ext4: speed up truncate/unlink by not using bforget() unless needed
[firefly-linux-kernel-4.4.55.git] / fs / ext4 / extents.c
1 /*
2  * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
3  * Written by Alex Tomas <alex@clusterfs.com>
4  *
5  * Architecture independence:
6  *   Copyright (c) 2005, Bull S.A.
7  *   Written by Pierre Peiffer <pierre.peiffer@bull.net>
8  *
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.
12  *
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.
17  *
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-
21  */
22
23 /*
24  * Extents support for EXT4
25  *
26  * TODO:
27  *   - ext4*_error() should be used in some situations
28  *   - analyze all BUG()/BUG_ON(), use -EIO where appropriate
29  *   - smart tree reduction
30  */
31
32 #include <linux/fs.h>
33 #include <linux/time.h>
34 #include <linux/jbd2.h>
35 #include <linux/highuid.h>
36 #include <linux/pagemap.h>
37 #include <linux/quotaops.h>
38 #include <linux/string.h>
39 #include <linux/slab.h>
40 #include <linux/falloc.h>
41 #include <asm/uaccess.h>
42 #include <linux/fiemap.h>
43 #include "ext4_jbd2.h"
44
45 #include <trace/events/ext4.h>
46
47 /*
48  * used by extent splitting.
49  */
50 #define EXT4_EXT_MAY_ZEROOUT    0x1  /* safe to zeroout if split fails \
51                                         due to ENOSPC */
52 #define EXT4_EXT_MARK_UNINIT1   0x2  /* mark first half uninitialized */
53 #define EXT4_EXT_MARK_UNINIT2   0x4  /* mark second half uninitialized */
54
55 static __le32 ext4_extent_block_csum(struct inode *inode,
56                                      struct ext4_extent_header *eh)
57 {
58         struct ext4_inode_info *ei = EXT4_I(inode);
59         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
60         __u32 csum;
61
62         csum = ext4_chksum(sbi, ei->i_csum_seed, (__u8 *)eh,
63                            EXT4_EXTENT_TAIL_OFFSET(eh));
64         return cpu_to_le32(csum);
65 }
66
67 static int ext4_extent_block_csum_verify(struct inode *inode,
68                                          struct ext4_extent_header *eh)
69 {
70         struct ext4_extent_tail *et;
71
72         if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb,
73                 EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
74                 return 1;
75
76         et = find_ext4_extent_tail(eh);
77         if (et->et_checksum != ext4_extent_block_csum(inode, eh))
78                 return 0;
79         return 1;
80 }
81
82 static void ext4_extent_block_csum_set(struct inode *inode,
83                                        struct ext4_extent_header *eh)
84 {
85         struct ext4_extent_tail *et;
86
87         if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb,
88                 EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
89                 return;
90
91         et = find_ext4_extent_tail(eh);
92         et->et_checksum = ext4_extent_block_csum(inode, eh);
93 }
94
95 static int ext4_split_extent(handle_t *handle,
96                                 struct inode *inode,
97                                 struct ext4_ext_path *path,
98                                 struct ext4_map_blocks *map,
99                                 int split_flag,
100                                 int flags);
101
102 static int ext4_split_extent_at(handle_t *handle,
103                              struct inode *inode,
104                              struct ext4_ext_path *path,
105                              ext4_lblk_t split,
106                              int split_flag,
107                              int flags);
108
109 static int ext4_ext_truncate_extend_restart(handle_t *handle,
110                                             struct inode *inode,
111                                             int needed)
112 {
113         int err;
114
115         if (!ext4_handle_valid(handle))
116                 return 0;
117         if (handle->h_buffer_credits > needed)
118                 return 0;
119         err = ext4_journal_extend(handle, needed);
120         if (err <= 0)
121                 return err;
122         err = ext4_truncate_restart_trans(handle, inode, needed);
123         if (err == 0)
124                 err = -EAGAIN;
125
126         return err;
127 }
128
129 /*
130  * could return:
131  *  - EROFS
132  *  - ENOMEM
133  */
134 static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
135                                 struct ext4_ext_path *path)
136 {
137         if (path->p_bh) {
138                 /* path points to block */
139                 return ext4_journal_get_write_access(handle, path->p_bh);
140         }
141         /* path points to leaf/index in inode body */
142         /* we use in-core data, no need to protect them */
143         return 0;
144 }
145
146 /*
147  * could return:
148  *  - EROFS
149  *  - ENOMEM
150  *  - EIO
151  */
152 #define ext4_ext_dirty(handle, inode, path) \
153                 __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path))
154 static int __ext4_ext_dirty(const char *where, unsigned int line,
155                             handle_t *handle, struct inode *inode,
156                             struct ext4_ext_path *path)
157 {
158         int err;
159         if (path->p_bh) {
160                 ext4_extent_block_csum_set(inode, ext_block_hdr(path->p_bh));
161                 /* path points to block */
162                 err = __ext4_handle_dirty_metadata(where, line, handle,
163                                                    inode, path->p_bh);
164         } else {
165                 /* path points to leaf/index in inode body */
166                 err = ext4_mark_inode_dirty(handle, inode);
167         }
168         return err;
169 }
170
171 static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
172                               struct ext4_ext_path *path,
173                               ext4_lblk_t block)
174 {
175         if (path) {
176                 int depth = path->p_depth;
177                 struct ext4_extent *ex;
178
179                 /*
180                  * Try to predict block placement assuming that we are
181                  * filling in a file which will eventually be
182                  * non-sparse --- i.e., in the case of libbfd writing
183                  * an ELF object sections out-of-order but in a way
184                  * the eventually results in a contiguous object or
185                  * executable file, or some database extending a table
186                  * space file.  However, this is actually somewhat
187                  * non-ideal if we are writing a sparse file such as
188                  * qemu or KVM writing a raw image file that is going
189                  * to stay fairly sparse, since it will end up
190                  * fragmenting the file system's free space.  Maybe we
191                  * should have some hueristics or some way to allow
192                  * userspace to pass a hint to file system,
193                  * especially if the latter case turns out to be
194                  * common.
195                  */
196                 ex = path[depth].p_ext;
197                 if (ex) {
198                         ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
199                         ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block);
200
201                         if (block > ext_block)
202                                 return ext_pblk + (block - ext_block);
203                         else
204                                 return ext_pblk - (ext_block - block);
205                 }
206
207                 /* it looks like index is empty;
208                  * try to find starting block from index itself */
209                 if (path[depth].p_bh)
210                         return path[depth].p_bh->b_blocknr;
211         }
212
213         /* OK. use inode's group */
214         return ext4_inode_to_goal_block(inode);
215 }
216
217 /*
218  * Allocation for a meta data block
219  */
220 static ext4_fsblk_t
221 ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
222                         struct ext4_ext_path *path,
223                         struct ext4_extent *ex, int *err, unsigned int flags)
224 {
225         ext4_fsblk_t goal, newblock;
226
227         goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
228         newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
229                                         NULL, err);
230         return newblock;
231 }
232
233 static inline int ext4_ext_space_block(struct inode *inode, int check)
234 {
235         int size;
236
237         size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
238                         / sizeof(struct ext4_extent);
239 #ifdef AGGRESSIVE_TEST
240         if (!check && size > 6)
241                 size = 6;
242 #endif
243         return size;
244 }
245
246 static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
247 {
248         int size;
249
250         size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
251                         / sizeof(struct ext4_extent_idx);
252 #ifdef AGGRESSIVE_TEST
253         if (!check && size > 5)
254                 size = 5;
255 #endif
256         return size;
257 }
258
259 static inline int ext4_ext_space_root(struct inode *inode, int check)
260 {
261         int size;
262
263         size = sizeof(EXT4_I(inode)->i_data);
264         size -= sizeof(struct ext4_extent_header);
265         size /= sizeof(struct ext4_extent);
266 #ifdef AGGRESSIVE_TEST
267         if (!check && size > 3)
268                 size = 3;
269 #endif
270         return size;
271 }
272
273 static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
274 {
275         int size;
276
277         size = sizeof(EXT4_I(inode)->i_data);
278         size -= sizeof(struct ext4_extent_header);
279         size /= sizeof(struct ext4_extent_idx);
280 #ifdef AGGRESSIVE_TEST
281         if (!check && size > 4)
282                 size = 4;
283 #endif
284         return size;
285 }
286
287 /*
288  * Calculate the number of metadata blocks needed
289  * to allocate @blocks
290  * Worse case is one block per extent
291  */
292 int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
293 {
294         struct ext4_inode_info *ei = EXT4_I(inode);
295         int idxs;
296
297         idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
298                 / sizeof(struct ext4_extent_idx));
299
300         /*
301          * If the new delayed allocation block is contiguous with the
302          * previous da block, it can share index blocks with the
303          * previous block, so we only need to allocate a new index
304          * block every idxs leaf blocks.  At ldxs**2 blocks, we need
305          * an additional index block, and at ldxs**3 blocks, yet
306          * another index blocks.
307          */
308         if (ei->i_da_metadata_calc_len &&
309             ei->i_da_metadata_calc_last_lblock+1 == lblock) {
310                 int num = 0;
311
312                 if ((ei->i_da_metadata_calc_len % idxs) == 0)
313                         num++;
314                 if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
315                         num++;
316                 if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
317                         num++;
318                         ei->i_da_metadata_calc_len = 0;
319                 } else
320                         ei->i_da_metadata_calc_len++;
321                 ei->i_da_metadata_calc_last_lblock++;
322                 return num;
323         }
324
325         /*
326          * In the worst case we need a new set of index blocks at
327          * every level of the inode's extent tree.
328          */
329         ei->i_da_metadata_calc_len = 1;
330         ei->i_da_metadata_calc_last_lblock = lblock;
331         return ext_depth(inode) + 1;
332 }
333
334 static int
335 ext4_ext_max_entries(struct inode *inode, int depth)
336 {
337         int max;
338
339         if (depth == ext_depth(inode)) {
340                 if (depth == 0)
341                         max = ext4_ext_space_root(inode, 1);
342                 else
343                         max = ext4_ext_space_root_idx(inode, 1);
344         } else {
345                 if (depth == 0)
346                         max = ext4_ext_space_block(inode, 1);
347                 else
348                         max = ext4_ext_space_block_idx(inode, 1);
349         }
350
351         return max;
352 }
353
354 static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
355 {
356         ext4_fsblk_t block = ext4_ext_pblock(ext);
357         int len = ext4_ext_get_actual_len(ext);
358
359         if (len == 0)
360                 return 0;
361         return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
362 }
363
364 static int ext4_valid_extent_idx(struct inode *inode,
365                                 struct ext4_extent_idx *ext_idx)
366 {
367         ext4_fsblk_t block = ext4_idx_pblock(ext_idx);
368
369         return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
370 }
371
372 static int ext4_valid_extent_entries(struct inode *inode,
373                                 struct ext4_extent_header *eh,
374                                 int depth)
375 {
376         unsigned short entries;
377         if (eh->eh_entries == 0)
378                 return 1;
379
380         entries = le16_to_cpu(eh->eh_entries);
381
382         if (depth == 0) {
383                 /* leaf entries */
384                 struct ext4_extent *ext = EXT_FIRST_EXTENT(eh);
385                 while (entries) {
386                         if (!ext4_valid_extent(inode, ext))
387                                 return 0;
388                         ext++;
389                         entries--;
390                 }
391         } else {
392                 struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh);
393                 while (entries) {
394                         if (!ext4_valid_extent_idx(inode, ext_idx))
395                                 return 0;
396                         ext_idx++;
397                         entries--;
398                 }
399         }
400         return 1;
401 }
402
403 static int __ext4_ext_check(const char *function, unsigned int line,
404                             struct inode *inode, struct ext4_extent_header *eh,
405                             int depth)
406 {
407         const char *error_msg;
408         int max = 0;
409
410         if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
411                 error_msg = "invalid magic";
412                 goto corrupted;
413         }
414         if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
415                 error_msg = "unexpected eh_depth";
416                 goto corrupted;
417         }
418         if (unlikely(eh->eh_max == 0)) {
419                 error_msg = "invalid eh_max";
420                 goto corrupted;
421         }
422         max = ext4_ext_max_entries(inode, depth);
423         if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
424                 error_msg = "too large eh_max";
425                 goto corrupted;
426         }
427         if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
428                 error_msg = "invalid eh_entries";
429                 goto corrupted;
430         }
431         if (!ext4_valid_extent_entries(inode, eh, depth)) {
432                 error_msg = "invalid extent entries";
433                 goto corrupted;
434         }
435         /* Verify checksum on non-root extent tree nodes */
436         if (ext_depth(inode) != depth &&
437             !ext4_extent_block_csum_verify(inode, eh)) {
438                 error_msg = "extent tree corrupted";
439                 goto corrupted;
440         }
441         return 0;
442
443 corrupted:
444         ext4_error_inode(inode, function, line, 0,
445                         "bad header/extent: %s - magic %x, "
446                         "entries %u, max %u(%u), depth %u(%u)",
447                         error_msg, le16_to_cpu(eh->eh_magic),
448                         le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
449                         max, le16_to_cpu(eh->eh_depth), depth);
450
451         return -EIO;
452 }
453
454 #define ext4_ext_check(inode, eh, depth)        \
455         __ext4_ext_check(__func__, __LINE__, inode, eh, depth)
456
457 int ext4_ext_check_inode(struct inode *inode)
458 {
459         return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
460 }
461
462 static int __ext4_ext_check_block(const char *function, unsigned int line,
463                                   struct inode *inode,
464                                   struct ext4_extent_header *eh,
465                                   int depth,
466                                   struct buffer_head *bh)
467 {
468         int ret;
469
470         if (buffer_verified(bh))
471                 return 0;
472         ret = ext4_ext_check(inode, eh, depth);
473         if (ret)
474                 return ret;
475         set_buffer_verified(bh);
476         return ret;
477 }
478
479 #define ext4_ext_check_block(inode, eh, depth, bh)      \
480         __ext4_ext_check_block(__func__, __LINE__, inode, eh, depth, bh)
481
482 #ifdef EXT_DEBUG
483 static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
484 {
485         int k, l = path->p_depth;
486
487         ext_debug("path:");
488         for (k = 0; k <= l; k++, path++) {
489                 if (path->p_idx) {
490                   ext_debug("  %d->%llu", le32_to_cpu(path->p_idx->ei_block),
491                             ext4_idx_pblock(path->p_idx));
492                 } else if (path->p_ext) {
493                         ext_debug("  %d:[%d]%d:%llu ",
494                                   le32_to_cpu(path->p_ext->ee_block),
495                                   ext4_ext_is_uninitialized(path->p_ext),
496                                   ext4_ext_get_actual_len(path->p_ext),
497                                   ext4_ext_pblock(path->p_ext));
498                 } else
499                         ext_debug("  []");
500         }
501         ext_debug("\n");
502 }
503
504 static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
505 {
506         int depth = ext_depth(inode);
507         struct ext4_extent_header *eh;
508         struct ext4_extent *ex;
509         int i;
510
511         if (!path)
512                 return;
513
514         eh = path[depth].p_hdr;
515         ex = EXT_FIRST_EXTENT(eh);
516
517         ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
518
519         for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
520                 ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
521                           ext4_ext_is_uninitialized(ex),
522                           ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
523         }
524         ext_debug("\n");
525 }
526
527 static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path,
528                         ext4_fsblk_t newblock, int level)
529 {
530         int depth = ext_depth(inode);
531         struct ext4_extent *ex;
532
533         if (depth != level) {
534                 struct ext4_extent_idx *idx;
535                 idx = path[level].p_idx;
536                 while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
537                         ext_debug("%d: move %d:%llu in new index %llu\n", level,
538                                         le32_to_cpu(idx->ei_block),
539                                         ext4_idx_pblock(idx),
540                                         newblock);
541                         idx++;
542                 }
543
544                 return;
545         }
546
547         ex = path[depth].p_ext;
548         while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
549                 ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
550                                 le32_to_cpu(ex->ee_block),
551                                 ext4_ext_pblock(ex),
552                                 ext4_ext_is_uninitialized(ex),
553                                 ext4_ext_get_actual_len(ex),
554                                 newblock);
555                 ex++;
556         }
557 }
558
559 #else
560 #define ext4_ext_show_path(inode, path)
561 #define ext4_ext_show_leaf(inode, path)
562 #define ext4_ext_show_move(inode, path, newblock, level)
563 #endif
564
565 void ext4_ext_drop_refs(struct ext4_ext_path *path)
566 {
567         int depth = path->p_depth;
568         int i;
569
570         for (i = 0; i <= depth; i++, path++)
571                 if (path->p_bh) {
572                         brelse(path->p_bh);
573                         path->p_bh = NULL;
574                 }
575 }
576
577 /*
578  * ext4_ext_binsearch_idx:
579  * binary search for the closest index of the given block
580  * the header must be checked before calling this
581  */
582 static void
583 ext4_ext_binsearch_idx(struct inode *inode,
584                         struct ext4_ext_path *path, ext4_lblk_t block)
585 {
586         struct ext4_extent_header *eh = path->p_hdr;
587         struct ext4_extent_idx *r, *l, *m;
588
589
590         ext_debug("binsearch for %u(idx):  ", block);
591
592         l = EXT_FIRST_INDEX(eh) + 1;
593         r = EXT_LAST_INDEX(eh);
594         while (l <= r) {
595                 m = l + (r - l) / 2;
596                 if (block < le32_to_cpu(m->ei_block))
597                         r = m - 1;
598                 else
599                         l = m + 1;
600                 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
601                                 m, le32_to_cpu(m->ei_block),
602                                 r, le32_to_cpu(r->ei_block));
603         }
604
605         path->p_idx = l - 1;
606         ext_debug("  -> %u->%lld ", le32_to_cpu(path->p_idx->ei_block),
607                   ext4_idx_pblock(path->p_idx));
608
609 #ifdef CHECK_BINSEARCH
610         {
611                 struct ext4_extent_idx *chix, *ix;
612                 int k;
613
614                 chix = ix = EXT_FIRST_INDEX(eh);
615                 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
616                   if (k != 0 &&
617                       le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
618                                 printk(KERN_DEBUG "k=%d, ix=0x%p, "
619                                        "first=0x%p\n", k,
620                                        ix, EXT_FIRST_INDEX(eh));
621                                 printk(KERN_DEBUG "%u <= %u\n",
622                                        le32_to_cpu(ix->ei_block),
623                                        le32_to_cpu(ix[-1].ei_block));
624                         }
625                         BUG_ON(k && le32_to_cpu(ix->ei_block)
626                                            <= le32_to_cpu(ix[-1].ei_block));
627                         if (block < le32_to_cpu(ix->ei_block))
628                                 break;
629                         chix = ix;
630                 }
631                 BUG_ON(chix != path->p_idx);
632         }
633 #endif
634
635 }
636
637 /*
638  * ext4_ext_binsearch:
639  * binary search for closest extent of the given block
640  * the header must be checked before calling this
641  */
642 static void
643 ext4_ext_binsearch(struct inode *inode,
644                 struct ext4_ext_path *path, ext4_lblk_t block)
645 {
646         struct ext4_extent_header *eh = path->p_hdr;
647         struct ext4_extent *r, *l, *m;
648
649         if (eh->eh_entries == 0) {
650                 /*
651                  * this leaf is empty:
652                  * we get such a leaf in split/add case
653                  */
654                 return;
655         }
656
657         ext_debug("binsearch for %u:  ", block);
658
659         l = EXT_FIRST_EXTENT(eh) + 1;
660         r = EXT_LAST_EXTENT(eh);
661
662         while (l <= r) {
663                 m = l + (r - l) / 2;
664                 if (block < le32_to_cpu(m->ee_block))
665                         r = m - 1;
666                 else
667                         l = m + 1;
668                 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
669                                 m, le32_to_cpu(m->ee_block),
670                                 r, le32_to_cpu(r->ee_block));
671         }
672
673         path->p_ext = l - 1;
674         ext_debug("  -> %d:%llu:[%d]%d ",
675                         le32_to_cpu(path->p_ext->ee_block),
676                         ext4_ext_pblock(path->p_ext),
677                         ext4_ext_is_uninitialized(path->p_ext),
678                         ext4_ext_get_actual_len(path->p_ext));
679
680 #ifdef CHECK_BINSEARCH
681         {
682                 struct ext4_extent *chex, *ex;
683                 int k;
684
685                 chex = ex = EXT_FIRST_EXTENT(eh);
686                 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
687                         BUG_ON(k && le32_to_cpu(ex->ee_block)
688                                           <= le32_to_cpu(ex[-1].ee_block));
689                         if (block < le32_to_cpu(ex->ee_block))
690                                 break;
691                         chex = ex;
692                 }
693                 BUG_ON(chex != path->p_ext);
694         }
695 #endif
696
697 }
698
699 int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
700 {
701         struct ext4_extent_header *eh;
702
703         eh = ext_inode_hdr(inode);
704         eh->eh_depth = 0;
705         eh->eh_entries = 0;
706         eh->eh_magic = EXT4_EXT_MAGIC;
707         eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
708         ext4_mark_inode_dirty(handle, inode);
709         ext4_ext_invalidate_cache(inode);
710         return 0;
711 }
712
713 struct ext4_ext_path *
714 ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
715                                         struct ext4_ext_path *path)
716 {
717         struct ext4_extent_header *eh;
718         struct buffer_head *bh;
719         short int depth, i, ppos = 0, alloc = 0;
720
721         eh = ext_inode_hdr(inode);
722         depth = ext_depth(inode);
723
724         /* account possible depth increase */
725         if (!path) {
726                 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
727                                 GFP_NOFS);
728                 if (!path)
729                         return ERR_PTR(-ENOMEM);
730                 alloc = 1;
731         }
732         path[0].p_hdr = eh;
733         path[0].p_bh = NULL;
734
735         i = depth;
736         /* walk through the tree */
737         while (i) {
738                 ext_debug("depth %d: num %d, max %d\n",
739                           ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
740
741                 ext4_ext_binsearch_idx(inode, path + ppos, block);
742                 path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
743                 path[ppos].p_depth = i;
744                 path[ppos].p_ext = NULL;
745
746                 bh = sb_getblk(inode->i_sb, path[ppos].p_block);
747                 if (unlikely(!bh))
748                         goto err;
749                 if (!bh_uptodate_or_lock(bh)) {
750                         trace_ext4_ext_load_extent(inode, block,
751                                                 path[ppos].p_block);
752                         if (bh_submit_read(bh) < 0) {
753                                 put_bh(bh);
754                                 goto err;
755                         }
756                 }
757                 eh = ext_block_hdr(bh);
758                 ppos++;
759                 if (unlikely(ppos > depth)) {
760                         put_bh(bh);
761                         EXT4_ERROR_INODE(inode,
762                                          "ppos %d > depth %d", ppos, depth);
763                         goto err;
764                 }
765                 path[ppos].p_bh = bh;
766                 path[ppos].p_hdr = eh;
767                 i--;
768
769                 if (ext4_ext_check_block(inode, eh, i, bh))
770                         goto err;
771         }
772
773         path[ppos].p_depth = i;
774         path[ppos].p_ext = NULL;
775         path[ppos].p_idx = NULL;
776
777         /* find extent */
778         ext4_ext_binsearch(inode, path + ppos, block);
779         /* if not an empty leaf */
780         if (path[ppos].p_ext)
781                 path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);
782
783         ext4_ext_show_path(inode, path);
784
785         return path;
786
787 err:
788         ext4_ext_drop_refs(path);
789         if (alloc)
790                 kfree(path);
791         return ERR_PTR(-EIO);
792 }
793
794 /*
795  * ext4_ext_insert_index:
796  * insert new index [@logical;@ptr] into the block at @curp;
797  * check where to insert: before @curp or after @curp
798  */
799 static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
800                                  struct ext4_ext_path *curp,
801                                  int logical, ext4_fsblk_t ptr)
802 {
803         struct ext4_extent_idx *ix;
804         int len, err;
805
806         err = ext4_ext_get_access(handle, inode, curp);
807         if (err)
808                 return err;
809
810         if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
811                 EXT4_ERROR_INODE(inode,
812                                  "logical %d == ei_block %d!",
813                                  logical, le32_to_cpu(curp->p_idx->ei_block));
814                 return -EIO;
815         }
816
817         if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
818                              >= le16_to_cpu(curp->p_hdr->eh_max))) {
819                 EXT4_ERROR_INODE(inode,
820                                  "eh_entries %d >= eh_max %d!",
821                                  le16_to_cpu(curp->p_hdr->eh_entries),
822                                  le16_to_cpu(curp->p_hdr->eh_max));
823                 return -EIO;
824         }
825
826         if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
827                 /* insert after */
828                 ext_debug("insert new index %d after: %llu\n", logical, ptr);
829                 ix = curp->p_idx + 1;
830         } else {
831                 /* insert before */
832                 ext_debug("insert new index %d before: %llu\n", logical, ptr);
833                 ix = curp->p_idx;
834         }
835
836         len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1;
837         BUG_ON(len < 0);
838         if (len > 0) {
839                 ext_debug("insert new index %d: "
840                                 "move %d indices from 0x%p to 0x%p\n",
841                                 logical, len, ix, ix + 1);
842                 memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx));
843         }
844
845         if (unlikely(ix > EXT_MAX_INDEX(curp->p_hdr))) {
846                 EXT4_ERROR_INODE(inode, "ix > EXT_MAX_INDEX!");
847                 return -EIO;
848         }
849
850         ix->ei_block = cpu_to_le32(logical);
851         ext4_idx_store_pblock(ix, ptr);
852         le16_add_cpu(&curp->p_hdr->eh_entries, 1);
853
854         if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
855                 EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
856                 return -EIO;
857         }
858
859         err = ext4_ext_dirty(handle, inode, curp);
860         ext4_std_error(inode->i_sb, err);
861
862         return err;
863 }
864
865 /*
866  * ext4_ext_split:
867  * inserts new subtree into the path, using free index entry
868  * at depth @at:
869  * - allocates all needed blocks (new leaf and all intermediate index blocks)
870  * - makes decision where to split
871  * - moves remaining extents and index entries (right to the split point)
872  *   into the newly allocated blocks
873  * - initializes subtree
874  */
875 static int ext4_ext_split(handle_t *handle, struct inode *inode,
876                           unsigned int flags,
877                           struct ext4_ext_path *path,
878                           struct ext4_extent *newext, int at)
879 {
880         struct buffer_head *bh = NULL;
881         int depth = ext_depth(inode);
882         struct ext4_extent_header *neh;
883         struct ext4_extent_idx *fidx;
884         int i = at, k, m, a;
885         ext4_fsblk_t newblock, oldblock;
886         __le32 border;
887         ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
888         int err = 0;
889
890         /* make decision: where to split? */
891         /* FIXME: now decision is simplest: at current extent */
892
893         /* if current leaf will be split, then we should use
894          * border from split point */
895         if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
896                 EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
897                 return -EIO;
898         }
899         if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
900                 border = path[depth].p_ext[1].ee_block;
901                 ext_debug("leaf will be split."
902                                 " next leaf starts at %d\n",
903                                   le32_to_cpu(border));
904         } else {
905                 border = newext->ee_block;
906                 ext_debug("leaf will be added."
907                                 " next leaf starts at %d\n",
908                                 le32_to_cpu(border));
909         }
910
911         /*
912          * If error occurs, then we break processing
913          * and mark filesystem read-only. index won't
914          * be inserted and tree will be in consistent
915          * state. Next mount will repair buffers too.
916          */
917
918         /*
919          * Get array to track all allocated blocks.
920          * We need this to handle errors and free blocks
921          * upon them.
922          */
923         ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
924         if (!ablocks)
925                 return -ENOMEM;
926
927         /* allocate all needed blocks */
928         ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
929         for (a = 0; a < depth - at; a++) {
930                 newblock = ext4_ext_new_meta_block(handle, inode, path,
931                                                    newext, &err, flags);
932                 if (newblock == 0)
933                         goto cleanup;
934                 ablocks[a] = newblock;
935         }
936
937         /* initialize new leaf */
938         newblock = ablocks[--a];
939         if (unlikely(newblock == 0)) {
940                 EXT4_ERROR_INODE(inode, "newblock == 0!");
941                 err = -EIO;
942                 goto cleanup;
943         }
944         bh = sb_getblk(inode->i_sb, newblock);
945         if (!bh) {
946                 err = -EIO;
947                 goto cleanup;
948         }
949         lock_buffer(bh);
950
951         err = ext4_journal_get_create_access(handle, bh);
952         if (err)
953                 goto cleanup;
954
955         neh = ext_block_hdr(bh);
956         neh->eh_entries = 0;
957         neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
958         neh->eh_magic = EXT4_EXT_MAGIC;
959         neh->eh_depth = 0;
960
961         /* move remainder of path[depth] to the new leaf */
962         if (unlikely(path[depth].p_hdr->eh_entries !=
963                      path[depth].p_hdr->eh_max)) {
964                 EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
965                                  path[depth].p_hdr->eh_entries,
966                                  path[depth].p_hdr->eh_max);
967                 err = -EIO;
968                 goto cleanup;
969         }
970         /* start copy from next extent */
971         m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++;
972         ext4_ext_show_move(inode, path, newblock, depth);
973         if (m) {
974                 struct ext4_extent *ex;
975                 ex = EXT_FIRST_EXTENT(neh);
976                 memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m);
977                 le16_add_cpu(&neh->eh_entries, m);
978         }
979
980         ext4_extent_block_csum_set(inode, neh);
981         set_buffer_uptodate(bh);
982         unlock_buffer(bh);
983
984         err = ext4_handle_dirty_metadata(handle, inode, bh);
985         if (err)
986                 goto cleanup;
987         brelse(bh);
988         bh = NULL;
989
990         /* correct old leaf */
991         if (m) {
992                 err = ext4_ext_get_access(handle, inode, path + depth);
993                 if (err)
994                         goto cleanup;
995                 le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
996                 err = ext4_ext_dirty(handle, inode, path + depth);
997                 if (err)
998                         goto cleanup;
999
1000         }
1001
1002         /* create intermediate indexes */
1003         k = depth - at - 1;
1004         if (unlikely(k < 0)) {
1005                 EXT4_ERROR_INODE(inode, "k %d < 0!", k);
1006                 err = -EIO;
1007                 goto cleanup;
1008         }
1009         if (k)
1010                 ext_debug("create %d intermediate indices\n", k);
1011         /* insert new index into current index block */
1012         /* current depth stored in i var */
1013         i = depth - 1;
1014         while (k--) {
1015                 oldblock = newblock;
1016                 newblock = ablocks[--a];
1017                 bh = sb_getblk(inode->i_sb, newblock);
1018                 if (!bh) {
1019                         err = -EIO;
1020                         goto cleanup;
1021                 }
1022                 lock_buffer(bh);
1023
1024                 err = ext4_journal_get_create_access(handle, bh);
1025                 if (err)
1026                         goto cleanup;
1027
1028                 neh = ext_block_hdr(bh);
1029                 neh->eh_entries = cpu_to_le16(1);
1030                 neh->eh_magic = EXT4_EXT_MAGIC;
1031                 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
1032                 neh->eh_depth = cpu_to_le16(depth - i);
1033                 fidx = EXT_FIRST_INDEX(neh);
1034                 fidx->ei_block = border;
1035                 ext4_idx_store_pblock(fidx, oldblock);
1036
1037                 ext_debug("int.index at %d (block %llu): %u -> %llu\n",
1038                                 i, newblock, le32_to_cpu(border), oldblock);
1039
1040                 /* move remainder of path[i] to the new index block */
1041                 if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
1042                                         EXT_LAST_INDEX(path[i].p_hdr))) {
1043                         EXT4_ERROR_INODE(inode,
1044                                          "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
1045                                          le32_to_cpu(path[i].p_ext->ee_block));
1046                         err = -EIO;
1047                         goto cleanup;
1048                 }
1049                 /* start copy indexes */
1050                 m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
1051                 ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
1052                                 EXT_MAX_INDEX(path[i].p_hdr));
1053                 ext4_ext_show_move(inode, path, newblock, i);
1054                 if (m) {
1055                         memmove(++fidx, path[i].p_idx,
1056                                 sizeof(struct ext4_extent_idx) * m);
1057                         le16_add_cpu(&neh->eh_entries, m);
1058                 }
1059                 ext4_extent_block_csum_set(inode, neh);
1060                 set_buffer_uptodate(bh);
1061                 unlock_buffer(bh);
1062
1063                 err = ext4_handle_dirty_metadata(handle, inode, bh);
1064                 if (err)
1065                         goto cleanup;
1066                 brelse(bh);
1067                 bh = NULL;
1068
1069                 /* correct old index */
1070                 if (m) {
1071                         err = ext4_ext_get_access(handle, inode, path + i);
1072                         if (err)
1073                                 goto cleanup;
1074                         le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
1075                         err = ext4_ext_dirty(handle, inode, path + i);
1076                         if (err)
1077                                 goto cleanup;
1078                 }
1079
1080                 i--;
1081         }
1082
1083         /* insert new index */
1084         err = ext4_ext_insert_index(handle, inode, path + at,
1085                                     le32_to_cpu(border), newblock);
1086
1087 cleanup:
1088         if (bh) {
1089                 if (buffer_locked(bh))
1090                         unlock_buffer(bh);
1091                 brelse(bh);
1092         }
1093
1094         if (err) {
1095                 /* free all allocated blocks in error case */
1096                 for (i = 0; i < depth; i++) {
1097                         if (!ablocks[i])
1098                                 continue;
1099                         ext4_free_blocks(handle, inode, NULL, ablocks[i], 1,
1100                                          EXT4_FREE_BLOCKS_METADATA);
1101                 }
1102         }
1103         kfree(ablocks);
1104
1105         return err;
1106 }
1107
1108 /*
1109  * ext4_ext_grow_indepth:
1110  * implements tree growing procedure:
1111  * - allocates new block
1112  * - moves top-level data (index block or leaf) into the new block
1113  * - initializes new top-level, creating index that points to the
1114  *   just created block
1115  */
1116 static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
1117                                  unsigned int flags,
1118                                  struct ext4_extent *newext)
1119 {
1120         struct ext4_extent_header *neh;
1121         struct buffer_head *bh;
1122         ext4_fsblk_t newblock;
1123         int err = 0;
1124
1125         newblock = ext4_ext_new_meta_block(handle, inode, NULL,
1126                 newext, &err, flags);
1127         if (newblock == 0)
1128                 return err;
1129
1130         bh = sb_getblk(inode->i_sb, newblock);
1131         if (!bh) {
1132                 err = -EIO;
1133                 ext4_std_error(inode->i_sb, err);
1134                 return err;
1135         }
1136         lock_buffer(bh);
1137
1138         err = ext4_journal_get_create_access(handle, bh);
1139         if (err) {
1140                 unlock_buffer(bh);
1141                 goto out;
1142         }
1143
1144         /* move top-level index/leaf into new block */
1145         memmove(bh->b_data, EXT4_I(inode)->i_data,
1146                 sizeof(EXT4_I(inode)->i_data));
1147
1148         /* set size of new block */
1149         neh = ext_block_hdr(bh);
1150         /* old root could have indexes or leaves
1151          * so calculate e_max right way */
1152         if (ext_depth(inode))
1153                 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
1154         else
1155                 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
1156         neh->eh_magic = EXT4_EXT_MAGIC;
1157         ext4_extent_block_csum_set(inode, neh);
1158         set_buffer_uptodate(bh);
1159         unlock_buffer(bh);
1160
1161         err = ext4_handle_dirty_metadata(handle, inode, bh);
1162         if (err)
1163                 goto out;
1164
1165         /* Update top-level index: num,max,pointer */
1166         neh = ext_inode_hdr(inode);
1167         neh->eh_entries = cpu_to_le16(1);
1168         ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock);
1169         if (neh->eh_depth == 0) {
1170                 /* Root extent block becomes index block */
1171                 neh->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
1172                 EXT_FIRST_INDEX(neh)->ei_block =
1173                         EXT_FIRST_EXTENT(neh)->ee_block;
1174         }
1175         ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
1176                   le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
1177                   le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
1178                   ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
1179
1180         neh->eh_depth = cpu_to_le16(le16_to_cpu(neh->eh_depth) + 1);
1181         ext4_mark_inode_dirty(handle, inode);
1182 out:
1183         brelse(bh);
1184
1185         return err;
1186 }
1187
1188 /*
1189  * ext4_ext_create_new_leaf:
1190  * finds empty index and adds new leaf.
1191  * if no free index is found, then it requests in-depth growing.
1192  */
1193 static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
1194                                     unsigned int flags,
1195                                     struct ext4_ext_path *path,
1196                                     struct ext4_extent *newext)
1197 {
1198         struct ext4_ext_path *curp;
1199         int depth, i, err = 0;
1200
1201 repeat:
1202         i = depth = ext_depth(inode);
1203
1204         /* walk up to the tree and look for free index entry */
1205         curp = path + depth;
1206         while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
1207                 i--;
1208                 curp--;
1209         }
1210
1211         /* we use already allocated block for index block,
1212          * so subsequent data blocks should be contiguous */
1213         if (EXT_HAS_FREE_INDEX(curp)) {
1214                 /* if we found index with free entry, then use that
1215                  * entry: create all needed subtree and add new leaf */
1216                 err = ext4_ext_split(handle, inode, flags, path, newext, i);
1217                 if (err)
1218                         goto out;
1219
1220                 /* refill path */
1221                 ext4_ext_drop_refs(path);
1222                 path = ext4_ext_find_extent(inode,
1223                                     (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1224                                     path);
1225                 if (IS_ERR(path))
1226                         err = PTR_ERR(path);
1227         } else {
1228                 /* tree is full, time to grow in depth */
1229                 err = ext4_ext_grow_indepth(handle, inode, flags, newext);
1230                 if (err)
1231                         goto out;
1232
1233                 /* refill path */
1234                 ext4_ext_drop_refs(path);
1235                 path = ext4_ext_find_extent(inode,
1236                                    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1237                                     path);
1238                 if (IS_ERR(path)) {
1239                         err = PTR_ERR(path);
1240                         goto out;
1241                 }
1242
1243                 /*
1244                  * only first (depth 0 -> 1) produces free space;
1245                  * in all other cases we have to split the grown tree
1246                  */
1247                 depth = ext_depth(inode);
1248                 if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
1249                         /* now we need to split */
1250                         goto repeat;
1251                 }
1252         }
1253
1254 out:
1255         return err;
1256 }
1257
1258 /*
1259  * search the closest allocated block to the left for *logical
1260  * and returns it at @logical + it's physical address at @phys
1261  * if *logical is the smallest allocated block, the function
1262  * returns 0 at @phys
1263  * return value contains 0 (success) or error code
1264  */
1265 static int ext4_ext_search_left(struct inode *inode,
1266                                 struct ext4_ext_path *path,
1267                                 ext4_lblk_t *logical, ext4_fsblk_t *phys)
1268 {
1269         struct ext4_extent_idx *ix;
1270         struct ext4_extent *ex;
1271         int depth, ee_len;
1272
1273         if (unlikely(path == NULL)) {
1274                 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1275                 return -EIO;
1276         }
1277         depth = path->p_depth;
1278         *phys = 0;
1279
1280         if (depth == 0 && path->p_ext == NULL)
1281                 return 0;
1282
1283         /* usually extent in the path covers blocks smaller
1284          * then *logical, but it can be that extent is the
1285          * first one in the file */
1286
1287         ex = path[depth].p_ext;
1288         ee_len = ext4_ext_get_actual_len(ex);
1289         if (*logical < le32_to_cpu(ex->ee_block)) {
1290                 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1291                         EXT4_ERROR_INODE(inode,
1292                                          "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
1293                                          *logical, le32_to_cpu(ex->ee_block));
1294                         return -EIO;
1295                 }
1296                 while (--depth >= 0) {
1297                         ix = path[depth].p_idx;
1298                         if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1299                                 EXT4_ERROR_INODE(inode,
1300                                   "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
1301                                   ix != NULL ? le32_to_cpu(ix->ei_block) : 0,
1302                                   EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
1303                 le32_to_cpu(EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block) : 0,
1304                                   depth);
1305                                 return -EIO;
1306                         }
1307                 }
1308                 return 0;
1309         }
1310
1311         if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1312                 EXT4_ERROR_INODE(inode,
1313                                  "logical %d < ee_block %d + ee_len %d!",
1314                                  *logical, le32_to_cpu(ex->ee_block), ee_len);
1315                 return -EIO;
1316         }
1317
1318         *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
1319         *phys = ext4_ext_pblock(ex) + ee_len - 1;
1320         return 0;
1321 }
1322
1323 /*
1324  * search the closest allocated block to the right for *logical
1325  * and returns it at @logical + it's physical address at @phys
1326  * if *logical is the largest allocated block, the function
1327  * returns 0 at @phys
1328  * return value contains 0 (success) or error code
1329  */
1330 static int ext4_ext_search_right(struct inode *inode,
1331                                  struct ext4_ext_path *path,
1332                                  ext4_lblk_t *logical, ext4_fsblk_t *phys,
1333                                  struct ext4_extent **ret_ex)
1334 {
1335         struct buffer_head *bh = NULL;
1336         struct ext4_extent_header *eh;
1337         struct ext4_extent_idx *ix;
1338         struct ext4_extent *ex;
1339         ext4_fsblk_t block;
1340         int depth;      /* Note, NOT eh_depth; depth from top of tree */
1341         int ee_len;
1342
1343         if (unlikely(path == NULL)) {
1344                 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1345                 return -EIO;
1346         }
1347         depth = path->p_depth;
1348         *phys = 0;
1349
1350         if (depth == 0 && path->p_ext == NULL)
1351                 return 0;
1352
1353         /* usually extent in the path covers blocks smaller
1354          * then *logical, but it can be that extent is the
1355          * first one in the file */
1356
1357         ex = path[depth].p_ext;
1358         ee_len = ext4_ext_get_actual_len(ex);
1359         if (*logical < le32_to_cpu(ex->ee_block)) {
1360                 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1361                         EXT4_ERROR_INODE(inode,
1362                                          "first_extent(path[%d].p_hdr) != ex",
1363                                          depth);
1364                         return -EIO;
1365                 }
1366                 while (--depth >= 0) {
1367                         ix = path[depth].p_idx;
1368                         if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1369                                 EXT4_ERROR_INODE(inode,
1370                                                  "ix != EXT_FIRST_INDEX *logical %d!",
1371                                                  *logical);
1372                                 return -EIO;
1373                         }
1374                 }
1375                 goto found_extent;
1376         }
1377
1378         if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1379                 EXT4_ERROR_INODE(inode,
1380                                  "logical %d < ee_block %d + ee_len %d!",
1381                                  *logical, le32_to_cpu(ex->ee_block), ee_len);
1382                 return -EIO;
1383         }
1384
1385         if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
1386                 /* next allocated block in this leaf */
1387                 ex++;
1388                 goto found_extent;
1389         }
1390
1391         /* go up and search for index to the right */
1392         while (--depth >= 0) {
1393                 ix = path[depth].p_idx;
1394                 if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
1395                         goto got_index;
1396         }
1397
1398         /* we've gone up to the root and found no index to the right */
1399         return 0;
1400
1401 got_index:
1402         /* we've found index to the right, let's
1403          * follow it and find the closest allocated
1404          * block to the right */
1405         ix++;
1406         block = ext4_idx_pblock(ix);
1407         while (++depth < path->p_depth) {
1408                 bh = sb_bread(inode->i_sb, block);
1409                 if (bh == NULL)
1410                         return -EIO;
1411                 eh = ext_block_hdr(bh);
1412                 /* subtract from p_depth to get proper eh_depth */
1413                 if (ext4_ext_check_block(inode, eh,
1414                                          path->p_depth - depth, bh)) {
1415                         put_bh(bh);
1416                         return -EIO;
1417                 }
1418                 ix = EXT_FIRST_INDEX(eh);
1419                 block = ext4_idx_pblock(ix);
1420                 put_bh(bh);
1421         }
1422
1423         bh = sb_bread(inode->i_sb, block);
1424         if (bh == NULL)
1425                 return -EIO;
1426         eh = ext_block_hdr(bh);
1427         if (ext4_ext_check_block(inode, eh, path->p_depth - depth, bh)) {
1428                 put_bh(bh);
1429                 return -EIO;
1430         }
1431         ex = EXT_FIRST_EXTENT(eh);
1432 found_extent:
1433         *logical = le32_to_cpu(ex->ee_block);
1434         *phys = ext4_ext_pblock(ex);
1435         *ret_ex = ex;
1436         if (bh)
1437                 put_bh(bh);
1438         return 0;
1439 }
1440
1441 /*
1442  * ext4_ext_next_allocated_block:
1443  * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
1444  * NOTE: it considers block number from index entry as
1445  * allocated block. Thus, index entries have to be consistent
1446  * with leaves.
1447  */
1448 static ext4_lblk_t
1449 ext4_ext_next_allocated_block(struct ext4_ext_path *path)
1450 {
1451         int depth;
1452
1453         BUG_ON(path == NULL);
1454         depth = path->p_depth;
1455
1456         if (depth == 0 && path->p_ext == NULL)
1457                 return EXT_MAX_BLOCKS;
1458
1459         while (depth >= 0) {
1460                 if (depth == path->p_depth) {
1461                         /* leaf */
1462                         if (path[depth].p_ext &&
1463                                 path[depth].p_ext !=
1464                                         EXT_LAST_EXTENT(path[depth].p_hdr))
1465                           return le32_to_cpu(path[depth].p_ext[1].ee_block);
1466                 } else {
1467                         /* index */
1468                         if (path[depth].p_idx !=
1469                                         EXT_LAST_INDEX(path[depth].p_hdr))
1470                           return le32_to_cpu(path[depth].p_idx[1].ei_block);
1471                 }
1472                 depth--;
1473         }
1474
1475         return EXT_MAX_BLOCKS;
1476 }
1477
1478 /*
1479  * ext4_ext_next_leaf_block:
1480  * returns first allocated block from next leaf or EXT_MAX_BLOCKS
1481  */
1482 static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path)
1483 {
1484         int depth;
1485
1486         BUG_ON(path == NULL);
1487         depth = path->p_depth;
1488
1489         /* zero-tree has no leaf blocks at all */
1490         if (depth == 0)
1491                 return EXT_MAX_BLOCKS;
1492
1493         /* go to index block */
1494         depth--;
1495
1496         while (depth >= 0) {
1497                 if (path[depth].p_idx !=
1498                                 EXT_LAST_INDEX(path[depth].p_hdr))
1499                         return (ext4_lblk_t)
1500                                 le32_to_cpu(path[depth].p_idx[1].ei_block);
1501                 depth--;
1502         }
1503
1504         return EXT_MAX_BLOCKS;
1505 }
1506
1507 /*
1508  * ext4_ext_correct_indexes:
1509  * if leaf gets modified and modified extent is first in the leaf,
1510  * then we have to correct all indexes above.
1511  * TODO: do we need to correct tree in all cases?
1512  */
1513 static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
1514                                 struct ext4_ext_path *path)
1515 {
1516         struct ext4_extent_header *eh;
1517         int depth = ext_depth(inode);
1518         struct ext4_extent *ex;
1519         __le32 border;
1520         int k, err = 0;
1521
1522         eh = path[depth].p_hdr;
1523         ex = path[depth].p_ext;
1524
1525         if (unlikely(ex == NULL || eh == NULL)) {
1526                 EXT4_ERROR_INODE(inode,
1527                                  "ex %p == NULL or eh %p == NULL", ex, eh);
1528                 return -EIO;
1529         }
1530
1531         if (depth == 0) {
1532                 /* there is no tree at all */
1533                 return 0;
1534         }
1535
1536         if (ex != EXT_FIRST_EXTENT(eh)) {
1537                 /* we correct tree if first leaf got modified only */
1538                 return 0;
1539         }
1540
1541         /*
1542          * TODO: we need correction if border is smaller than current one
1543          */
1544         k = depth - 1;
1545         border = path[depth].p_ext->ee_block;
1546         err = ext4_ext_get_access(handle, inode, path + k);
1547         if (err)
1548                 return err;
1549         path[k].p_idx->ei_block = border;
1550         err = ext4_ext_dirty(handle, inode, path + k);
1551         if (err)
1552                 return err;
1553
1554         while (k--) {
1555                 /* change all left-side indexes */
1556                 if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
1557                         break;
1558                 err = ext4_ext_get_access(handle, inode, path + k);
1559                 if (err)
1560                         break;
1561                 path[k].p_idx->ei_block = border;
1562                 err = ext4_ext_dirty(handle, inode, path + k);
1563                 if (err)
1564                         break;
1565         }
1566
1567         return err;
1568 }
1569
1570 int
1571 ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
1572                                 struct ext4_extent *ex2)
1573 {
1574         unsigned short ext1_ee_len, ext2_ee_len, max_len;
1575
1576         /*
1577          * Make sure that either both extents are uninitialized, or
1578          * both are _not_.
1579          */
1580         if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
1581                 return 0;
1582
1583         if (ext4_ext_is_uninitialized(ex1))
1584                 max_len = EXT_UNINIT_MAX_LEN;
1585         else
1586                 max_len = EXT_INIT_MAX_LEN;
1587
1588         ext1_ee_len = ext4_ext_get_actual_len(ex1);
1589         ext2_ee_len = ext4_ext_get_actual_len(ex2);
1590
1591         if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
1592                         le32_to_cpu(ex2->ee_block))
1593                 return 0;
1594
1595         /*
1596          * To allow future support for preallocated extents to be added
1597          * as an RO_COMPAT feature, refuse to merge to extents if
1598          * this can result in the top bit of ee_len being set.
1599          */
1600         if (ext1_ee_len + ext2_ee_len > max_len)
1601                 return 0;
1602 #ifdef AGGRESSIVE_TEST
1603         if (ext1_ee_len >= 4)
1604                 return 0;
1605 #endif
1606
1607         if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
1608                 return 1;
1609         return 0;
1610 }
1611
1612 /*
1613  * This function tries to merge the "ex" extent to the next extent in the tree.
1614  * It always tries to merge towards right. If you want to merge towards
1615  * left, pass "ex - 1" as argument instead of "ex".
1616  * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
1617  * 1 if they got merged.
1618  */
1619 static int ext4_ext_try_to_merge_right(struct inode *inode,
1620                                  struct ext4_ext_path *path,
1621                                  struct ext4_extent *ex)
1622 {
1623         struct ext4_extent_header *eh;
1624         unsigned int depth, len;
1625         int merge_done = 0;
1626         int uninitialized = 0;
1627
1628         depth = ext_depth(inode);
1629         BUG_ON(path[depth].p_hdr == NULL);
1630         eh = path[depth].p_hdr;
1631
1632         while (ex < EXT_LAST_EXTENT(eh)) {
1633                 if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
1634                         break;
1635                 /* merge with next extent! */
1636                 if (ext4_ext_is_uninitialized(ex))
1637                         uninitialized = 1;
1638                 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1639                                 + ext4_ext_get_actual_len(ex + 1));
1640                 if (uninitialized)
1641                         ext4_ext_mark_uninitialized(ex);
1642
1643                 if (ex + 1 < EXT_LAST_EXTENT(eh)) {
1644                         len = (EXT_LAST_EXTENT(eh) - ex - 1)
1645                                 * sizeof(struct ext4_extent);
1646                         memmove(ex + 1, ex + 2, len);
1647                 }
1648                 le16_add_cpu(&eh->eh_entries, -1);
1649                 merge_done = 1;
1650                 WARN_ON(eh->eh_entries == 0);
1651                 if (!eh->eh_entries)
1652                         EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
1653         }
1654
1655         return merge_done;
1656 }
1657
1658 /*
1659  * This function does a very simple check to see if we can collapse
1660  * an extent tree with a single extent tree leaf block into the inode.
1661  */
1662 static void ext4_ext_try_to_merge_up(handle_t *handle,
1663                                      struct inode *inode,
1664                                      struct ext4_ext_path *path)
1665 {
1666         size_t s;
1667         unsigned max_root = ext4_ext_space_root(inode, 0);
1668         ext4_fsblk_t blk;
1669
1670         if ((path[0].p_depth != 1) ||
1671             (le16_to_cpu(path[0].p_hdr->eh_entries) != 1) ||
1672             (le16_to_cpu(path[1].p_hdr->eh_entries) > max_root))
1673                 return;
1674
1675         /*
1676          * We need to modify the block allocation bitmap and the block
1677          * group descriptor to release the extent tree block.  If we
1678          * can't get the journal credits, give up.
1679          */
1680         if (ext4_journal_extend(handle, 2))
1681                 return;
1682
1683         /*
1684          * Copy the extent data up to the inode
1685          */
1686         blk = ext4_idx_pblock(path[0].p_idx);
1687         s = le16_to_cpu(path[1].p_hdr->eh_entries) *
1688                 sizeof(struct ext4_extent_idx);
1689         s += sizeof(struct ext4_extent_header);
1690
1691         memcpy(path[0].p_hdr, path[1].p_hdr, s);
1692         path[0].p_depth = 0;
1693         path[0].p_ext = EXT_FIRST_EXTENT(path[0].p_hdr) +
1694                 (path[1].p_ext - EXT_FIRST_EXTENT(path[1].p_hdr));
1695         path[0].p_hdr->eh_max = cpu_to_le16(max_root);
1696
1697         brelse(path[1].p_bh);
1698         ext4_free_blocks(handle, inode, NULL, blk, 1,
1699                          EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
1700 }
1701
1702 /*
1703  * This function tries to merge the @ex extent to neighbours in the tree.
1704  * return 1 if merge left else 0.
1705  */
1706 static void ext4_ext_try_to_merge(handle_t *handle,
1707                                   struct inode *inode,
1708                                   struct ext4_ext_path *path,
1709                                   struct ext4_extent *ex) {
1710         struct ext4_extent_header *eh;
1711         unsigned int depth;
1712         int merge_done = 0;
1713
1714         depth = ext_depth(inode);
1715         BUG_ON(path[depth].p_hdr == NULL);
1716         eh = path[depth].p_hdr;
1717
1718         if (ex > EXT_FIRST_EXTENT(eh))
1719                 merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1);
1720
1721         if (!merge_done)
1722                 (void) ext4_ext_try_to_merge_right(inode, path, ex);
1723
1724         ext4_ext_try_to_merge_up(handle, inode, path);
1725 }
1726
1727 /*
1728  * check if a portion of the "newext" extent overlaps with an
1729  * existing extent.
1730  *
1731  * If there is an overlap discovered, it updates the length of the newext
1732  * such that there will be no overlap, and then returns 1.
1733  * If there is no overlap found, it returns 0.
1734  */
1735 static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi,
1736                                            struct inode *inode,
1737                                            struct ext4_extent *newext,
1738                                            struct ext4_ext_path *path)
1739 {
1740         ext4_lblk_t b1, b2;
1741         unsigned int depth, len1;
1742         unsigned int ret = 0;
1743
1744         b1 = le32_to_cpu(newext->ee_block);
1745         len1 = ext4_ext_get_actual_len(newext);
1746         depth = ext_depth(inode);
1747         if (!path[depth].p_ext)
1748                 goto out;
1749         b2 = le32_to_cpu(path[depth].p_ext->ee_block);
1750         b2 &= ~(sbi->s_cluster_ratio - 1);
1751
1752         /*
1753          * get the next allocated block if the extent in the path
1754          * is before the requested block(s)
1755          */
1756         if (b2 < b1) {
1757                 b2 = ext4_ext_next_allocated_block(path);
1758                 if (b2 == EXT_MAX_BLOCKS)
1759                         goto out;
1760                 b2 &= ~(sbi->s_cluster_ratio - 1);
1761         }
1762
1763         /* check for wrap through zero on extent logical start block*/
1764         if (b1 + len1 < b1) {
1765                 len1 = EXT_MAX_BLOCKS - b1;
1766                 newext->ee_len = cpu_to_le16(len1);
1767                 ret = 1;
1768         }
1769
1770         /* check for overlap */
1771         if (b1 + len1 > b2) {
1772                 newext->ee_len = cpu_to_le16(b2 - b1);
1773                 ret = 1;
1774         }
1775 out:
1776         return ret;
1777 }
1778
1779 /*
1780  * ext4_ext_insert_extent:
1781  * tries to merge requsted extent into the existing extent or
1782  * inserts requested extent as new one into the tree,
1783  * creating new leaf in the no-space case.
1784  */
1785 int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
1786                                 struct ext4_ext_path *path,
1787                                 struct ext4_extent *newext, int flag)
1788 {
1789         struct ext4_extent_header *eh;
1790         struct ext4_extent *ex, *fex;
1791         struct ext4_extent *nearex; /* nearest extent */
1792         struct ext4_ext_path *npath = NULL;
1793         int depth, len, err;
1794         ext4_lblk_t next;
1795         unsigned uninitialized = 0;
1796         int flags = 0;
1797
1798         if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
1799                 EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
1800                 return -EIO;
1801         }
1802         depth = ext_depth(inode);
1803         ex = path[depth].p_ext;
1804         if (unlikely(path[depth].p_hdr == NULL)) {
1805                 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1806                 return -EIO;
1807         }
1808
1809         /* try to insert block into found extent and return */
1810         if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
1811                 && ext4_can_extents_be_merged(inode, ex, newext)) {
1812                 ext_debug("append [%d]%d block to %u:[%d]%d (from %llu)\n",
1813                           ext4_ext_is_uninitialized(newext),
1814                           ext4_ext_get_actual_len(newext),
1815                           le32_to_cpu(ex->ee_block),
1816                           ext4_ext_is_uninitialized(ex),
1817                           ext4_ext_get_actual_len(ex),
1818                           ext4_ext_pblock(ex));
1819                 err = ext4_ext_get_access(handle, inode, path + depth);
1820                 if (err)
1821                         return err;
1822
1823                 /*
1824                  * ext4_can_extents_be_merged should have checked that either
1825                  * both extents are uninitialized, or both aren't. Thus we
1826                  * need to check only one of them here.
1827                  */
1828                 if (ext4_ext_is_uninitialized(ex))
1829                         uninitialized = 1;
1830                 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1831                                         + ext4_ext_get_actual_len(newext));
1832                 if (uninitialized)
1833                         ext4_ext_mark_uninitialized(ex);
1834                 eh = path[depth].p_hdr;
1835                 nearex = ex;
1836                 goto merge;
1837         }
1838
1839         depth = ext_depth(inode);
1840         eh = path[depth].p_hdr;
1841         if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
1842                 goto has_space;
1843
1844         /* probably next leaf has space for us? */
1845         fex = EXT_LAST_EXTENT(eh);
1846         next = EXT_MAX_BLOCKS;
1847         if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block))
1848                 next = ext4_ext_next_leaf_block(path);
1849         if (next != EXT_MAX_BLOCKS) {
1850                 ext_debug("next leaf block - %u\n", next);
1851                 BUG_ON(npath != NULL);
1852                 npath = ext4_ext_find_extent(inode, next, NULL);
1853                 if (IS_ERR(npath))
1854                         return PTR_ERR(npath);
1855                 BUG_ON(npath->p_depth != path->p_depth);
1856                 eh = npath[depth].p_hdr;
1857                 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
1858                         ext_debug("next leaf isn't full(%d)\n",
1859                                   le16_to_cpu(eh->eh_entries));
1860                         path = npath;
1861                         goto has_space;
1862                 }
1863                 ext_debug("next leaf has no free space(%d,%d)\n",
1864                           le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
1865         }
1866
1867         /*
1868          * There is no free space in the found leaf.
1869          * We're gonna add a new leaf in the tree.
1870          */
1871         if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT)
1872                 flags = EXT4_MB_USE_ROOT_BLOCKS;
1873         err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
1874         if (err)
1875                 goto cleanup;
1876         depth = ext_depth(inode);
1877         eh = path[depth].p_hdr;
1878
1879 has_space:
1880         nearex = path[depth].p_ext;
1881
1882         err = ext4_ext_get_access(handle, inode, path + depth);
1883         if (err)
1884                 goto cleanup;
1885
1886         if (!nearex) {
1887                 /* there is no extent in this leaf, create first one */
1888                 ext_debug("first extent in the leaf: %u:%llu:[%d]%d\n",
1889                                 le32_to_cpu(newext->ee_block),
1890                                 ext4_ext_pblock(newext),
1891                                 ext4_ext_is_uninitialized(newext),
1892                                 ext4_ext_get_actual_len(newext));
1893                 nearex = EXT_FIRST_EXTENT(eh);
1894         } else {
1895                 if (le32_to_cpu(newext->ee_block)
1896                            > le32_to_cpu(nearex->ee_block)) {
1897                         /* Insert after */
1898                         ext_debug("insert %u:%llu:[%d]%d before: "
1899                                         "nearest %p\n",
1900                                         le32_to_cpu(newext->ee_block),
1901                                         ext4_ext_pblock(newext),
1902                                         ext4_ext_is_uninitialized(newext),
1903                                         ext4_ext_get_actual_len(newext),
1904                                         nearex);
1905                         nearex++;
1906                 } else {
1907                         /* Insert before */
1908                         BUG_ON(newext->ee_block == nearex->ee_block);
1909                         ext_debug("insert %u:%llu:[%d]%d after: "
1910                                         "nearest %p\n",
1911                                         le32_to_cpu(newext->ee_block),
1912                                         ext4_ext_pblock(newext),
1913                                         ext4_ext_is_uninitialized(newext),
1914                                         ext4_ext_get_actual_len(newext),
1915                                         nearex);
1916                 }
1917                 len = EXT_LAST_EXTENT(eh) - nearex + 1;
1918                 if (len > 0) {
1919                         ext_debug("insert %u:%llu:[%d]%d: "
1920                                         "move %d extents from 0x%p to 0x%p\n",
1921                                         le32_to_cpu(newext->ee_block),
1922                                         ext4_ext_pblock(newext),
1923                                         ext4_ext_is_uninitialized(newext),
1924                                         ext4_ext_get_actual_len(newext),
1925                                         len, nearex, nearex + 1);
1926                         memmove(nearex + 1, nearex,
1927                                 len * sizeof(struct ext4_extent));
1928                 }
1929         }
1930
1931         le16_add_cpu(&eh->eh_entries, 1);
1932         path[depth].p_ext = nearex;
1933         nearex->ee_block = newext->ee_block;
1934         ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
1935         nearex->ee_len = newext->ee_len;
1936
1937 merge:
1938         /* try to merge extents */
1939         if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
1940                 ext4_ext_try_to_merge(handle, inode, path, nearex);
1941
1942
1943         /* time to correct all indexes above */
1944         err = ext4_ext_correct_indexes(handle, inode, path);
1945         if (err)
1946                 goto cleanup;
1947
1948         err = ext4_ext_dirty(handle, inode, path + path->p_depth);
1949
1950 cleanup:
1951         if (npath) {
1952                 ext4_ext_drop_refs(npath);
1953                 kfree(npath);
1954         }
1955         ext4_ext_invalidate_cache(inode);
1956         return err;
1957 }
1958
1959 static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
1960                                ext4_lblk_t num, ext_prepare_callback func,
1961                                void *cbdata)
1962 {
1963         struct ext4_ext_path *path = NULL;
1964         struct ext4_ext_cache cbex;
1965         struct ext4_extent *ex;
1966         ext4_lblk_t next, start = 0, end = 0;
1967         ext4_lblk_t last = block + num;
1968         int depth, exists, err = 0;
1969
1970         BUG_ON(func == NULL);
1971         BUG_ON(inode == NULL);
1972
1973         while (block < last && block != EXT_MAX_BLOCKS) {
1974                 num = last - block;
1975                 /* find extent for this block */
1976                 down_read(&EXT4_I(inode)->i_data_sem);
1977                 path = ext4_ext_find_extent(inode, block, path);
1978                 up_read(&EXT4_I(inode)->i_data_sem);
1979                 if (IS_ERR(path)) {
1980                         err = PTR_ERR(path);
1981                         path = NULL;
1982                         break;
1983                 }
1984
1985                 depth = ext_depth(inode);
1986                 if (unlikely(path[depth].p_hdr == NULL)) {
1987                         EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1988                         err = -EIO;
1989                         break;
1990                 }
1991                 ex = path[depth].p_ext;
1992                 next = ext4_ext_next_allocated_block(path);
1993
1994                 exists = 0;
1995                 if (!ex) {
1996                         /* there is no extent yet, so try to allocate
1997                          * all requested space */
1998                         start = block;
1999                         end = block + num;
2000                 } else if (le32_to_cpu(ex->ee_block) > block) {
2001                         /* need to allocate space before found extent */
2002                         start = block;
2003                         end = le32_to_cpu(ex->ee_block);
2004                         if (block + num < end)
2005                                 end = block + num;
2006                 } else if (block >= le32_to_cpu(ex->ee_block)
2007                                         + ext4_ext_get_actual_len(ex)) {
2008                         /* need to allocate space after found extent */
2009                         start = block;
2010                         end = block + num;
2011                         if (end >= next)
2012                                 end = next;
2013                 } else if (block >= le32_to_cpu(ex->ee_block)) {
2014                         /*
2015                          * some part of requested space is covered
2016                          * by found extent
2017                          */
2018                         start = block;
2019                         end = le32_to_cpu(ex->ee_block)
2020                                 + ext4_ext_get_actual_len(ex);
2021                         if (block + num < end)
2022                                 end = block + num;
2023                         exists = 1;
2024                 } else {
2025                         BUG();
2026                 }
2027                 BUG_ON(end <= start);
2028
2029                 if (!exists) {
2030                         cbex.ec_block = start;
2031                         cbex.ec_len = end - start;
2032                         cbex.ec_start = 0;
2033                 } else {
2034                         cbex.ec_block = le32_to_cpu(ex->ee_block);
2035                         cbex.ec_len = ext4_ext_get_actual_len(ex);
2036                         cbex.ec_start = ext4_ext_pblock(ex);
2037                 }
2038
2039                 if (unlikely(cbex.ec_len == 0)) {
2040                         EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
2041                         err = -EIO;
2042                         break;
2043                 }
2044                 err = func(inode, next, &cbex, ex, cbdata);
2045                 ext4_ext_drop_refs(path);
2046
2047                 if (err < 0)
2048                         break;
2049
2050                 if (err == EXT_REPEAT)
2051                         continue;
2052                 else if (err == EXT_BREAK) {
2053                         err = 0;
2054                         break;
2055                 }
2056
2057                 if (ext_depth(inode) != depth) {
2058                         /* depth was changed. we have to realloc path */
2059                         kfree(path);
2060                         path = NULL;
2061                 }
2062
2063                 block = cbex.ec_block + cbex.ec_len;
2064         }
2065
2066         if (path) {
2067                 ext4_ext_drop_refs(path);
2068                 kfree(path);
2069         }
2070
2071         return err;
2072 }
2073
2074 static void
2075 ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
2076                         __u32 len, ext4_fsblk_t start)
2077 {
2078         struct ext4_ext_cache *cex;
2079         BUG_ON(len == 0);
2080         spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
2081         trace_ext4_ext_put_in_cache(inode, block, len, start);
2082         cex = &EXT4_I(inode)->i_cached_extent;
2083         cex->ec_block = block;
2084         cex->ec_len = len;
2085         cex->ec_start = start;
2086         spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
2087 }
2088
2089 /*
2090  * ext4_ext_put_gap_in_cache:
2091  * calculate boundaries of the gap that the requested block fits into
2092  * and cache this gap
2093  */
2094 static void
2095 ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
2096                                 ext4_lblk_t block)
2097 {
2098         int depth = ext_depth(inode);
2099         unsigned long len;
2100         ext4_lblk_t lblock;
2101         struct ext4_extent *ex;
2102
2103         ex = path[depth].p_ext;
2104         if (ex == NULL) {
2105                 /* there is no extent yet, so gap is [0;-] */
2106                 lblock = 0;
2107                 len = EXT_MAX_BLOCKS;
2108                 ext_debug("cache gap(whole file):");
2109         } else if (block < le32_to_cpu(ex->ee_block)) {
2110                 lblock = block;
2111                 len = le32_to_cpu(ex->ee_block) - block;
2112                 ext_debug("cache gap(before): %u [%u:%u]",
2113                                 block,
2114                                 le32_to_cpu(ex->ee_block),
2115                                  ext4_ext_get_actual_len(ex));
2116         } else if (block >= le32_to_cpu(ex->ee_block)
2117                         + ext4_ext_get_actual_len(ex)) {
2118                 ext4_lblk_t next;
2119                 lblock = le32_to_cpu(ex->ee_block)
2120                         + ext4_ext_get_actual_len(ex);
2121
2122                 next = ext4_ext_next_allocated_block(path);
2123                 ext_debug("cache gap(after): [%u:%u] %u",
2124                                 le32_to_cpu(ex->ee_block),
2125                                 ext4_ext_get_actual_len(ex),
2126                                 block);
2127                 BUG_ON(next == lblock);
2128                 len = next - lblock;
2129         } else {
2130                 lblock = len = 0;
2131                 BUG();
2132         }
2133
2134         ext_debug(" -> %u:%lu\n", lblock, len);
2135         ext4_ext_put_in_cache(inode, lblock, len, 0);
2136 }
2137
2138 /*
2139  * ext4_ext_check_cache()
2140  * Checks to see if the given block is in the cache.
2141  * If it is, the cached extent is stored in the given
2142  * cache extent pointer.  If the cached extent is a hole,
2143  * this routine should be used instead of
2144  * ext4_ext_in_cache if the calling function needs to
2145  * know the size of the hole.
2146  *
2147  * @inode: The files inode
2148  * @block: The block to look for in the cache
2149  * @ex:    Pointer where the cached extent will be stored
2150  *         if it contains block
2151  *
2152  * Return 0 if cache is invalid; 1 if the cache is valid
2153  */
2154 static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
2155         struct ext4_ext_cache *ex){
2156         struct ext4_ext_cache *cex;
2157         struct ext4_sb_info *sbi;
2158         int ret = 0;
2159
2160         /*
2161          * We borrow i_block_reservation_lock to protect i_cached_extent
2162          */
2163         spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
2164         cex = &EXT4_I(inode)->i_cached_extent;
2165         sbi = EXT4_SB(inode->i_sb);
2166
2167         /* has cache valid data? */
2168         if (cex->ec_len == 0)
2169                 goto errout;
2170
2171         if (in_range(block, cex->ec_block, cex->ec_len)) {
2172                 memcpy(ex, cex, sizeof(struct ext4_ext_cache));
2173                 ext_debug("%u cached by %u:%u:%llu\n",
2174                                 block,
2175                                 cex->ec_block, cex->ec_len, cex->ec_start);
2176                 ret = 1;
2177         }
2178 errout:
2179         trace_ext4_ext_in_cache(inode, block, ret);
2180         spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
2181         return ret;
2182 }
2183
2184 /*
2185  * ext4_ext_in_cache()
2186  * Checks to see if the given block is in the cache.
2187  * If it is, the cached extent is stored in the given
2188  * extent pointer.
2189  *
2190  * @inode: The files inode
2191  * @block: The block to look for in the cache
2192  * @ex:    Pointer where the cached extent will be stored
2193  *         if it contains block
2194  *
2195  * Return 0 if cache is invalid; 1 if the cache is valid
2196  */
2197 static int
2198 ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
2199                         struct ext4_extent *ex)
2200 {
2201         struct ext4_ext_cache cex;
2202         int ret = 0;
2203
2204         if (ext4_ext_check_cache(inode, block, &cex)) {
2205                 ex->ee_block = cpu_to_le32(cex.ec_block);
2206                 ext4_ext_store_pblock(ex, cex.ec_start);
2207                 ex->ee_len = cpu_to_le16(cex.ec_len);
2208                 ret = 1;
2209         }
2210
2211         return ret;
2212 }
2213
2214
2215 /*
2216  * ext4_ext_rm_idx:
2217  * removes index from the index block.
2218  */
2219 static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
2220                         struct ext4_ext_path *path)
2221 {
2222         int err;
2223         ext4_fsblk_t leaf;
2224
2225         /* free index block */
2226         path--;
2227         leaf = ext4_idx_pblock(path->p_idx);
2228         if (unlikely(path->p_hdr->eh_entries == 0)) {
2229                 EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
2230                 return -EIO;
2231         }
2232         err = ext4_ext_get_access(handle, inode, path);
2233         if (err)
2234                 return err;
2235
2236         if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) {
2237                 int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx;
2238                 len *= sizeof(struct ext4_extent_idx);
2239                 memmove(path->p_idx, path->p_idx + 1, len);
2240         }
2241
2242         le16_add_cpu(&path->p_hdr->eh_entries, -1);
2243         err = ext4_ext_dirty(handle, inode, path);
2244         if (err)
2245                 return err;
2246         ext_debug("index is empty, remove it, free block %llu\n", leaf);
2247         trace_ext4_ext_rm_idx(inode, leaf);
2248
2249         ext4_free_blocks(handle, inode, NULL, leaf, 1,
2250                          EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
2251         return err;
2252 }
2253
2254 /*
2255  * ext4_ext_calc_credits_for_single_extent:
2256  * This routine returns max. credits that needed to insert an extent
2257  * to the extent tree.
2258  * When pass the actual path, the caller should calculate credits
2259  * under i_data_sem.
2260  */
2261 int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
2262                                                 struct ext4_ext_path *path)
2263 {
2264         if (path) {
2265                 int depth = ext_depth(inode);
2266                 int ret = 0;
2267
2268                 /* probably there is space in leaf? */
2269                 if (le16_to_cpu(path[depth].p_hdr->eh_entries)
2270                                 < le16_to_cpu(path[depth].p_hdr->eh_max)) {
2271
2272                         /*
2273                          *  There are some space in the leaf tree, no
2274                          *  need to account for leaf block credit
2275                          *
2276                          *  bitmaps and block group descriptor blocks
2277                          *  and other metadata blocks still need to be
2278                          *  accounted.
2279                          */
2280                         /* 1 bitmap, 1 block group descriptor */
2281                         ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
2282                         return ret;
2283                 }
2284         }
2285
2286         return ext4_chunk_trans_blocks(inode, nrblocks);
2287 }
2288
2289 /*
2290  * How many index/leaf blocks need to change/allocate to modify nrblocks?
2291  *
2292  * if nrblocks are fit in a single extent (chunk flag is 1), then
2293  * in the worse case, each tree level index/leaf need to be changed
2294  * if the tree split due to insert a new extent, then the old tree
2295  * index/leaf need to be updated too
2296  *
2297  * If the nrblocks are discontiguous, they could cause
2298  * the whole tree split more than once, but this is really rare.
2299  */
2300 int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
2301 {
2302         int index;
2303         int depth = ext_depth(inode);
2304
2305         if (chunk)
2306                 index = depth * 2;
2307         else
2308                 index = depth * 3;
2309
2310         return index;
2311 }
2312
2313 static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
2314                               struct ext4_extent *ex,
2315                               ext4_fsblk_t *partial_cluster,
2316                               ext4_lblk_t from, ext4_lblk_t to)
2317 {
2318         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2319         unsigned short ee_len =  ext4_ext_get_actual_len(ex);
2320         ext4_fsblk_t pblk;
2321         int flags = 0;
2322
2323         if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2324                 flags |= EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET;
2325         else if (ext4_should_journal_data(inode))
2326                 flags |= EXT4_FREE_BLOCKS_FORGET;
2327
2328         /*
2329          * For bigalloc file systems, we never free a partial cluster
2330          * at the beginning of the extent.  Instead, we make a note
2331          * that we tried freeing the cluster, and check to see if we
2332          * need to free it on a subsequent call to ext4_remove_blocks,
2333          * or at the end of the ext4_truncate() operation.
2334          */
2335         flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;
2336
2337         trace_ext4_remove_blocks(inode, ex, from, to, *partial_cluster);
2338         /*
2339          * If we have a partial cluster, and it's different from the
2340          * cluster of the last block, we need to explicitly free the
2341          * partial cluster here.
2342          */
2343         pblk = ext4_ext_pblock(ex) + ee_len - 1;
2344         if (*partial_cluster && (EXT4_B2C(sbi, pblk) != *partial_cluster)) {
2345                 ext4_free_blocks(handle, inode, NULL,
2346                                  EXT4_C2B(sbi, *partial_cluster),
2347                                  sbi->s_cluster_ratio, flags);
2348                 *partial_cluster = 0;
2349         }
2350
2351 #ifdef EXTENTS_STATS
2352         {
2353                 struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2354                 spin_lock(&sbi->s_ext_stats_lock);
2355                 sbi->s_ext_blocks += ee_len;
2356                 sbi->s_ext_extents++;
2357                 if (ee_len < sbi->s_ext_min)
2358                         sbi->s_ext_min = ee_len;
2359                 if (ee_len > sbi->s_ext_max)
2360                         sbi->s_ext_max = ee_len;
2361                 if (ext_depth(inode) > sbi->s_depth_max)
2362                         sbi->s_depth_max = ext_depth(inode);
2363                 spin_unlock(&sbi->s_ext_stats_lock);
2364         }
2365 #endif
2366         if (from >= le32_to_cpu(ex->ee_block)
2367             && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
2368                 /* tail removal */
2369                 ext4_lblk_t num;
2370
2371                 num = le32_to_cpu(ex->ee_block) + ee_len - from;
2372                 pblk = ext4_ext_pblock(ex) + ee_len - num;
2373                 ext_debug("free last %u blocks starting %llu\n", num, pblk);
2374                 ext4_free_blocks(handle, inode, NULL, pblk, num, flags);
2375                 /*
2376                  * If the block range to be freed didn't start at the
2377                  * beginning of a cluster, and we removed the entire
2378                  * extent, save the partial cluster here, since we
2379                  * might need to delete if we determine that the
2380                  * truncate operation has removed all of the blocks in
2381                  * the cluster.
2382                  */
2383                 if (pblk & (sbi->s_cluster_ratio - 1) &&
2384                     (ee_len == num))
2385                         *partial_cluster = EXT4_B2C(sbi, pblk);
2386                 else
2387                         *partial_cluster = 0;
2388         } else if (from == le32_to_cpu(ex->ee_block)
2389                    && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
2390                 /* head removal */
2391                 ext4_lblk_t num;
2392                 ext4_fsblk_t start;
2393
2394                 num = to - from;
2395                 start = ext4_ext_pblock(ex);
2396
2397                 ext_debug("free first %u blocks starting %llu\n", num, start);
2398                 ext4_free_blocks(handle, inode, NULL, start, num, flags);
2399
2400         } else {
2401                 printk(KERN_INFO "strange request: removal(2) "
2402                                 "%u-%u from %u:%u\n",
2403                                 from, to, le32_to_cpu(ex->ee_block), ee_len);
2404         }
2405         return 0;
2406 }
2407
2408
2409 /*
2410  * ext4_ext_rm_leaf() Removes the extents associated with the
2411  * blocks appearing between "start" and "end", and splits the extents
2412  * if "start" and "end" appear in the same extent
2413  *
2414  * @handle: The journal handle
2415  * @inode:  The files inode
2416  * @path:   The path to the leaf
2417  * @start:  The first block to remove
2418  * @end:   The last block to remove
2419  */
2420 static int
2421 ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
2422                  struct ext4_ext_path *path, ext4_fsblk_t *partial_cluster,
2423                  ext4_lblk_t start, ext4_lblk_t end)
2424 {
2425         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2426         int err = 0, correct_index = 0;
2427         int depth = ext_depth(inode), credits;
2428         struct ext4_extent_header *eh;
2429         ext4_lblk_t a, b;
2430         unsigned num;
2431         ext4_lblk_t ex_ee_block;
2432         unsigned short ex_ee_len;
2433         unsigned uninitialized = 0;
2434         struct ext4_extent *ex;
2435
2436         /* the header must be checked already in ext4_ext_remove_space() */
2437         ext_debug("truncate since %u in leaf to %u\n", start, end);
2438         if (!path[depth].p_hdr)
2439                 path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
2440         eh = path[depth].p_hdr;
2441         if (unlikely(path[depth].p_hdr == NULL)) {
2442                 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
2443                 return -EIO;
2444         }
2445         /* find where to start removing */
2446         ex = EXT_LAST_EXTENT(eh);
2447
2448         ex_ee_block = le32_to_cpu(ex->ee_block);
2449         ex_ee_len = ext4_ext_get_actual_len(ex);
2450
2451         trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster);
2452
2453         while (ex >= EXT_FIRST_EXTENT(eh) &&
2454                         ex_ee_block + ex_ee_len > start) {
2455
2456                 if (ext4_ext_is_uninitialized(ex))
2457                         uninitialized = 1;
2458                 else
2459                         uninitialized = 0;
2460
2461                 ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
2462                          uninitialized, ex_ee_len);
2463                 path[depth].p_ext = ex;
2464
2465                 a = ex_ee_block > start ? ex_ee_block : start;
2466                 b = ex_ee_block+ex_ee_len - 1 < end ?
2467                         ex_ee_block+ex_ee_len - 1 : end;
2468
2469                 ext_debug("  border %u:%u\n", a, b);
2470
2471                 /* If this extent is beyond the end of the hole, skip it */
2472                 if (end < ex_ee_block) {
2473                         ex--;
2474                         ex_ee_block = le32_to_cpu(ex->ee_block);
2475                         ex_ee_len = ext4_ext_get_actual_len(ex);
2476                         continue;
2477                 } else if (b != ex_ee_block + ex_ee_len - 1) {
2478                         EXT4_ERROR_INODE(inode,
2479                                          "can not handle truncate %u:%u "
2480                                          "on extent %u:%u",
2481                                          start, end, ex_ee_block,
2482                                          ex_ee_block + ex_ee_len - 1);
2483                         err = -EIO;
2484                         goto out;
2485                 } else if (a != ex_ee_block) {
2486                         /* remove tail of the extent */
2487                         num = a - ex_ee_block;
2488                 } else {
2489                         /* remove whole extent: excellent! */
2490                         num = 0;
2491                 }
2492                 /*
2493                  * 3 for leaf, sb, and inode plus 2 (bmap and group
2494                  * descriptor) for each block group; assume two block
2495                  * groups plus ex_ee_len/blocks_per_block_group for
2496                  * the worst case
2497                  */
2498                 credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb));
2499                 if (ex == EXT_FIRST_EXTENT(eh)) {
2500                         correct_index = 1;
2501                         credits += (ext_depth(inode)) + 1;
2502                 }
2503                 credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
2504
2505                 err = ext4_ext_truncate_extend_restart(handle, inode, credits);
2506                 if (err)
2507                         goto out;
2508
2509                 err = ext4_ext_get_access(handle, inode, path + depth);
2510                 if (err)
2511                         goto out;
2512
2513                 err = ext4_remove_blocks(handle, inode, ex, partial_cluster,
2514                                          a, b);
2515                 if (err)
2516                         goto out;
2517
2518                 if (num == 0)
2519                         /* this extent is removed; mark slot entirely unused */
2520                         ext4_ext_store_pblock(ex, 0);
2521
2522                 ex->ee_len = cpu_to_le16(num);
2523                 /*
2524                  * Do not mark uninitialized if all the blocks in the
2525                  * extent have been removed.
2526                  */
2527                 if (uninitialized && num)
2528                         ext4_ext_mark_uninitialized(ex);
2529                 /*
2530                  * If the extent was completely released,
2531                  * we need to remove it from the leaf
2532                  */
2533                 if (num == 0) {
2534                         if (end != EXT_MAX_BLOCKS - 1) {
2535                                 /*
2536                                  * For hole punching, we need to scoot all the
2537                                  * extents up when an extent is removed so that
2538                                  * we dont have blank extents in the middle
2539                                  */
2540                                 memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) *
2541                                         sizeof(struct ext4_extent));
2542
2543                                 /* Now get rid of the one at the end */
2544                                 memset(EXT_LAST_EXTENT(eh), 0,
2545                                         sizeof(struct ext4_extent));
2546                         }
2547                         le16_add_cpu(&eh->eh_entries, -1);
2548                 } else
2549                         *partial_cluster = 0;
2550
2551                 err = ext4_ext_dirty(handle, inode, path + depth);
2552                 if (err)
2553                         goto out;
2554
2555                 ext_debug("new extent: %u:%u:%llu\n", ex_ee_block, num,
2556                                 ext4_ext_pblock(ex));
2557                 ex--;
2558                 ex_ee_block = le32_to_cpu(ex->ee_block);
2559                 ex_ee_len = ext4_ext_get_actual_len(ex);
2560         }
2561
2562         if (correct_index && eh->eh_entries)
2563                 err = ext4_ext_correct_indexes(handle, inode, path);
2564
2565         /*
2566          * If there is still a entry in the leaf node, check to see if
2567          * it references the partial cluster.  This is the only place
2568          * where it could; if it doesn't, we can free the cluster.
2569          */
2570         if (*partial_cluster && ex >= EXT_FIRST_EXTENT(eh) &&
2571             (EXT4_B2C(sbi, ext4_ext_pblock(ex) + ex_ee_len - 1) !=
2572              *partial_cluster)) {
2573                 int flags = EXT4_FREE_BLOCKS_FORGET;
2574
2575                 if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2576                         flags |= EXT4_FREE_BLOCKS_METADATA;
2577
2578                 ext4_free_blocks(handle, inode, NULL,
2579                                  EXT4_C2B(sbi, *partial_cluster),
2580                                  sbi->s_cluster_ratio, flags);
2581                 *partial_cluster = 0;
2582         }
2583
2584         /* if this leaf is free, then we should
2585          * remove it from index block above */
2586         if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
2587                 err = ext4_ext_rm_idx(handle, inode, path + depth);
2588
2589 out:
2590         return err;
2591 }
2592
2593 /*
2594  * ext4_ext_more_to_rm:
2595  * returns 1 if current index has to be freed (even partial)
2596  */
2597 static int
2598 ext4_ext_more_to_rm(struct ext4_ext_path *path)
2599 {
2600         BUG_ON(path->p_idx == NULL);
2601
2602         if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
2603                 return 0;
2604
2605         /*
2606          * if truncate on deeper level happened, it wasn't partial,
2607          * so we have to consider current index for truncation
2608          */
2609         if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
2610                 return 0;
2611         return 1;
2612 }
2613
2614 static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
2615                                  ext4_lblk_t end)
2616 {
2617         struct super_block *sb = inode->i_sb;
2618         int depth = ext_depth(inode);
2619         struct ext4_ext_path *path = NULL;
2620         ext4_fsblk_t partial_cluster = 0;
2621         handle_t *handle;
2622         int i = 0, err;
2623
2624         ext_debug("truncate since %u to %u\n", start, end);
2625
2626         /* probably first extent we're gonna free will be last in block */
2627         handle = ext4_journal_start(inode, depth + 1);
2628         if (IS_ERR(handle))
2629                 return PTR_ERR(handle);
2630
2631 again:
2632         ext4_ext_invalidate_cache(inode);
2633
2634         trace_ext4_ext_remove_space(inode, start, depth);
2635
2636         /*
2637          * Check if we are removing extents inside the extent tree. If that
2638          * is the case, we are going to punch a hole inside the extent tree
2639          * so we have to check whether we need to split the extent covering
2640          * the last block to remove so we can easily remove the part of it
2641          * in ext4_ext_rm_leaf().
2642          */
2643         if (end < EXT_MAX_BLOCKS - 1) {
2644                 struct ext4_extent *ex;
2645                 ext4_lblk_t ee_block;
2646
2647                 /* find extent for this block */
2648                 path = ext4_ext_find_extent(inode, end, NULL);
2649                 if (IS_ERR(path)) {
2650                         ext4_journal_stop(handle);
2651                         return PTR_ERR(path);
2652                 }
2653                 depth = ext_depth(inode);
2654                 ex = path[depth].p_ext;
2655                 if (!ex) {
2656                         ext4_ext_drop_refs(path);
2657                         kfree(path);
2658                         path = NULL;
2659                         goto cont;
2660                 }
2661
2662                 ee_block = le32_to_cpu(ex->ee_block);
2663
2664                 /*
2665                  * See if the last block is inside the extent, if so split
2666                  * the extent at 'end' block so we can easily remove the
2667                  * tail of the first part of the split extent in
2668                  * ext4_ext_rm_leaf().
2669                  */
2670                 if (end >= ee_block &&
2671                     end < ee_block + ext4_ext_get_actual_len(ex) - 1) {
2672                         int split_flag = 0;
2673
2674                         if (ext4_ext_is_uninitialized(ex))
2675                                 split_flag = EXT4_EXT_MARK_UNINIT1 |
2676                                              EXT4_EXT_MARK_UNINIT2;
2677
2678                         /*
2679                          * Split the extent in two so that 'end' is the last
2680                          * block in the first new extent
2681                          */
2682                         err = ext4_split_extent_at(handle, inode, path,
2683                                                 end + 1, split_flag,
2684                                                 EXT4_GET_BLOCKS_PRE_IO |
2685                                                 EXT4_GET_BLOCKS_PUNCH_OUT_EXT);
2686
2687                         if (err < 0)
2688                                 goto out;
2689                 }
2690         }
2691 cont:
2692
2693         /*
2694          * We start scanning from right side, freeing all the blocks
2695          * after i_size and walking into the tree depth-wise.
2696          */
2697         depth = ext_depth(inode);
2698         if (path) {
2699                 int k = i = depth;
2700                 while (--k > 0)
2701                         path[k].p_block =
2702                                 le16_to_cpu(path[k].p_hdr->eh_entries)+1;
2703         } else {
2704                 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1),
2705                                GFP_NOFS);
2706                 if (path == NULL) {
2707                         ext4_journal_stop(handle);
2708                         return -ENOMEM;
2709                 }
2710                 path[0].p_depth = depth;
2711                 path[0].p_hdr = ext_inode_hdr(inode);
2712                 i = 0;
2713
2714                 if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
2715                         err = -EIO;
2716                         goto out;
2717                 }
2718         }
2719         err = 0;
2720
2721         while (i >= 0 && err == 0) {
2722                 if (i == depth) {
2723                         /* this is leaf block */
2724                         err = ext4_ext_rm_leaf(handle, inode, path,
2725                                                &partial_cluster, start,
2726                                                end);
2727                         /* root level has p_bh == NULL, brelse() eats this */
2728                         brelse(path[i].p_bh);
2729                         path[i].p_bh = NULL;
2730                         i--;
2731                         continue;
2732                 }
2733
2734                 /* this is index block */
2735                 if (!path[i].p_hdr) {
2736                         ext_debug("initialize header\n");
2737                         path[i].p_hdr = ext_block_hdr(path[i].p_bh);
2738                 }
2739
2740                 if (!path[i].p_idx) {
2741                         /* this level hasn't been touched yet */
2742                         path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2743                         path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
2744                         ext_debug("init index ptr: hdr 0x%p, num %d\n",
2745                                   path[i].p_hdr,
2746                                   le16_to_cpu(path[i].p_hdr->eh_entries));
2747                 } else {
2748                         /* we were already here, see at next index */
2749                         path[i].p_idx--;
2750                 }
2751
2752                 ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
2753                                 i, EXT_FIRST_INDEX(path[i].p_hdr),
2754                                 path[i].p_idx);
2755                 if (ext4_ext_more_to_rm(path + i)) {
2756                         struct buffer_head *bh;
2757                         /* go to the next level */
2758                         ext_debug("move to level %d (block %llu)\n",
2759                                   i + 1, ext4_idx_pblock(path[i].p_idx));
2760                         memset(path + i + 1, 0, sizeof(*path));
2761                         bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
2762                         if (!bh) {
2763                                 /* should we reset i_size? */
2764                                 err = -EIO;
2765                                 break;
2766                         }
2767                         if (WARN_ON(i + 1 > depth)) {
2768                                 err = -EIO;
2769                                 break;
2770                         }
2771                         if (ext4_ext_check_block(inode, ext_block_hdr(bh),
2772                                                         depth - i - 1, bh)) {
2773                                 err = -EIO;
2774                                 break;
2775                         }
2776                         path[i + 1].p_bh = bh;
2777
2778                         /* save actual number of indexes since this
2779                          * number is changed at the next iteration */
2780                         path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
2781                         i++;
2782                 } else {
2783                         /* we finished processing this index, go up */
2784                         if (path[i].p_hdr->eh_entries == 0 && i > 0) {
2785                                 /* index is empty, remove it;
2786                                  * handle must be already prepared by the
2787                                  * truncatei_leaf() */
2788                                 err = ext4_ext_rm_idx(handle, inode, path + i);
2789                         }
2790                         /* root level has p_bh == NULL, brelse() eats this */
2791                         brelse(path[i].p_bh);
2792                         path[i].p_bh = NULL;
2793                         i--;
2794                         ext_debug("return to level %d\n", i);
2795                 }
2796         }
2797
2798         trace_ext4_ext_remove_space_done(inode, start, depth, partial_cluster,
2799                         path->p_hdr->eh_entries);
2800
2801         /* If we still have something in the partial cluster and we have removed
2802          * even the first extent, then we should free the blocks in the partial
2803          * cluster as well. */
2804         if (partial_cluster && path->p_hdr->eh_entries == 0) {
2805                 int flags = EXT4_FREE_BLOCKS_FORGET;
2806
2807                 if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2808                         flags |= EXT4_FREE_BLOCKS_METADATA;
2809
2810                 ext4_free_blocks(handle, inode, NULL,
2811                                  EXT4_C2B(EXT4_SB(sb), partial_cluster),
2812                                  EXT4_SB(sb)->s_cluster_ratio, flags);
2813                 partial_cluster = 0;
2814         }
2815
2816         /* TODO: flexible tree reduction should be here */
2817         if (path->p_hdr->eh_entries == 0) {
2818                 /*
2819                  * truncate to zero freed all the tree,
2820                  * so we need to correct eh_depth
2821                  */
2822                 err = ext4_ext_get_access(handle, inode, path);
2823                 if (err == 0) {
2824                         ext_inode_hdr(inode)->eh_depth = 0;
2825                         ext_inode_hdr(inode)->eh_max =
2826                                 cpu_to_le16(ext4_ext_space_root(inode, 0));
2827                         err = ext4_ext_dirty(handle, inode, path);
2828                 }
2829         }
2830 out:
2831         ext4_ext_drop_refs(path);
2832         kfree(path);
2833         if (err == -EAGAIN) {
2834                 path = NULL;
2835                 goto again;
2836         }
2837         ext4_journal_stop(handle);
2838
2839         return err;
2840 }
2841
2842 /*
2843  * called at mount time
2844  */
2845 void ext4_ext_init(struct super_block *sb)
2846 {
2847         /*
2848          * possible initialization would be here
2849          */
2850
2851         if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
2852 #if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
2853                 printk(KERN_INFO "EXT4-fs: file extents enabled"
2854 #ifdef AGGRESSIVE_TEST
2855                        ", aggressive tests"
2856 #endif
2857 #ifdef CHECK_BINSEARCH
2858                        ", check binsearch"
2859 #endif
2860 #ifdef EXTENTS_STATS
2861                        ", stats"
2862 #endif
2863                        "\n");
2864 #endif
2865 #ifdef EXTENTS_STATS
2866                 spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
2867                 EXT4_SB(sb)->s_ext_min = 1 << 30;
2868                 EXT4_SB(sb)->s_ext_max = 0;
2869 #endif
2870         }
2871 }
2872
2873 /*
2874  * called at umount time
2875  */
2876 void ext4_ext_release(struct super_block *sb)
2877 {
2878         if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
2879                 return;
2880
2881 #ifdef EXTENTS_STATS
2882         if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
2883                 struct ext4_sb_info *sbi = EXT4_SB(sb);
2884                 printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
2885                         sbi->s_ext_blocks, sbi->s_ext_extents,
2886                         sbi->s_ext_blocks / sbi->s_ext_extents);
2887                 printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
2888                         sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
2889         }
2890 #endif
2891 }
2892
2893 /* FIXME!! we need to try to merge to left or right after zero-out  */
2894 static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
2895 {
2896         ext4_fsblk_t ee_pblock;
2897         unsigned int ee_len;
2898         int ret;
2899
2900         ee_len    = ext4_ext_get_actual_len(ex);
2901         ee_pblock = ext4_ext_pblock(ex);
2902
2903         ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
2904         if (ret > 0)
2905                 ret = 0;
2906
2907         return ret;
2908 }
2909
2910 /*
2911  * ext4_split_extent_at() splits an extent at given block.
2912  *
2913  * @handle: the journal handle
2914  * @inode: the file inode
2915  * @path: the path to the extent
2916  * @split: the logical block where the extent is splitted.
2917  * @split_flags: indicates if the extent could be zeroout if split fails, and
2918  *               the states(init or uninit) of new extents.
2919  * @flags: flags used to insert new extent to extent tree.
2920  *
2921  *
2922  * Splits extent [a, b] into two extents [a, @split) and [@split, b], states
2923  * of which are deterimined by split_flag.
2924  *
2925  * There are two cases:
2926  *  a> the extent are splitted into two extent.
2927  *  b> split is not needed, and just mark the extent.
2928  *
2929  * return 0 on success.
2930  */
2931 static int ext4_split_extent_at(handle_t *handle,
2932                              struct inode *inode,
2933                              struct ext4_ext_path *path,
2934                              ext4_lblk_t split,
2935                              int split_flag,
2936                              int flags)
2937 {
2938         ext4_fsblk_t newblock;
2939         ext4_lblk_t ee_block;
2940         struct ext4_extent *ex, newex, orig_ex;
2941         struct ext4_extent *ex2 = NULL;
2942         unsigned int ee_len, depth;
2943         int err = 0;
2944
2945         ext_debug("ext4_split_extents_at: inode %lu, logical"
2946                 "block %llu\n", inode->i_ino, (unsigned long long)split);
2947
2948         ext4_ext_show_leaf(inode, path);
2949
2950         depth = ext_depth(inode);
2951         ex = path[depth].p_ext;
2952         ee_block = le32_to_cpu(ex->ee_block);
2953         ee_len = ext4_ext_get_actual_len(ex);
2954         newblock = split - ee_block + ext4_ext_pblock(ex);
2955
2956         BUG_ON(split < ee_block || split >= (ee_block + ee_len));
2957
2958         err = ext4_ext_get_access(handle, inode, path + depth);
2959         if (err)
2960                 goto out;
2961
2962         if (split == ee_block) {
2963                 /*
2964                  * case b: block @split is the block that the extent begins with
2965                  * then we just change the state of the extent, and splitting
2966                  * is not needed.
2967                  */
2968                 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2969                         ext4_ext_mark_uninitialized(ex);
2970                 else
2971                         ext4_ext_mark_initialized(ex);
2972
2973                 if (!(flags & EXT4_GET_BLOCKS_PRE_IO))
2974                         ext4_ext_try_to_merge(handle, inode, path, ex);
2975
2976                 err = ext4_ext_dirty(handle, inode, path + path->p_depth);
2977                 goto out;
2978         }
2979
2980         /* case a */
2981         memcpy(&orig_ex, ex, sizeof(orig_ex));
2982         ex->ee_len = cpu_to_le16(split - ee_block);
2983         if (split_flag & EXT4_EXT_MARK_UNINIT1)
2984                 ext4_ext_mark_uninitialized(ex);
2985
2986         /*
2987          * path may lead to new leaf, not to original leaf any more
2988          * after ext4_ext_insert_extent() returns,
2989          */
2990         err = ext4_ext_dirty(handle, inode, path + depth);
2991         if (err)
2992                 goto fix_extent_len;
2993
2994         ex2 = &newex;
2995         ex2->ee_block = cpu_to_le32(split);
2996         ex2->ee_len   = cpu_to_le16(ee_len - (split - ee_block));
2997         ext4_ext_store_pblock(ex2, newblock);
2998         if (split_flag & EXT4_EXT_MARK_UNINIT2)
2999                 ext4_ext_mark_uninitialized(ex2);
3000
3001         err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
3002         if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3003                 err = ext4_ext_zeroout(inode, &orig_ex);
3004                 if (err)
3005                         goto fix_extent_len;
3006                 /* update the extent length and mark as initialized */
3007                 ex->ee_len = cpu_to_le16(ee_len);
3008                 ext4_ext_try_to_merge(handle, inode, path, ex);
3009                 err = ext4_ext_dirty(handle, inode, path + path->p_depth);
3010                 goto out;
3011         } else if (err)
3012                 goto fix_extent_len;
3013
3014 out:
3015         ext4_ext_show_leaf(inode, path);
3016         return err;
3017
3018 fix_extent_len:
3019         ex->ee_len = orig_ex.ee_len;
3020         ext4_ext_dirty(handle, inode, path + depth);
3021         return err;
3022 }
3023
3024 /*
3025  * ext4_split_extents() splits an extent and mark extent which is covered
3026  * by @map as split_flags indicates
3027  *
3028  * It may result in splitting the extent into multiple extents (upto three)
3029  * There are three possibilities:
3030  *   a> There is no split required
3031  *   b> Splits in two extents: Split is happening at either end of the extent
3032  *   c> Splits in three extents: Somone is splitting in middle of the extent
3033  *
3034  */
3035 static int ext4_split_extent(handle_t *handle,
3036                               struct inode *inode,
3037                               struct ext4_ext_path *path,
3038                               struct ext4_map_blocks *map,
3039                               int split_flag,
3040                               int flags)
3041 {
3042         ext4_lblk_t ee_block;
3043         struct ext4_extent *ex;
3044         unsigned int ee_len, depth;
3045         int err = 0;
3046         int uninitialized;
3047         int split_flag1, flags1;
3048
3049         depth = ext_depth(inode);
3050         ex = path[depth].p_ext;
3051         ee_block = le32_to_cpu(ex->ee_block);
3052         ee_len = ext4_ext_get_actual_len(ex);
3053         uninitialized = ext4_ext_is_uninitialized(ex);
3054
3055         if (map->m_lblk + map->m_len < ee_block + ee_len) {
3056                 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
3057                               EXT4_EXT_MAY_ZEROOUT : 0;
3058                 flags1 = flags | EXT4_GET_BLOCKS_PRE_IO;
3059                 if (uninitialized)
3060                         split_flag1 |= EXT4_EXT_MARK_UNINIT1 |
3061                                        EXT4_EXT_MARK_UNINIT2;
3062                 err = ext4_split_extent_at(handle, inode, path,
3063                                 map->m_lblk + map->m_len, split_flag1, flags1);
3064                 if (err)
3065                         goto out;
3066         }
3067
3068         ext4_ext_drop_refs(path);
3069         path = ext4_ext_find_extent(inode, map->m_lblk, path);
3070         if (IS_ERR(path))
3071                 return PTR_ERR(path);
3072
3073         if (map->m_lblk >= ee_block) {
3074                 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
3075                               EXT4_EXT_MAY_ZEROOUT : 0;
3076                 if (uninitialized)
3077                         split_flag1 |= EXT4_EXT_MARK_UNINIT1;
3078                 if (split_flag & EXT4_EXT_MARK_UNINIT2)
3079                         split_flag1 |= EXT4_EXT_MARK_UNINIT2;
3080                 err = ext4_split_extent_at(handle, inode, path,
3081                                 map->m_lblk, split_flag1, flags);
3082                 if (err)
3083                         goto out;
3084         }
3085
3086         ext4_ext_show_leaf(inode, path);
3087 out:
3088         return err ? err : map->m_len;
3089 }
3090
3091 /*
3092  * This function is called by ext4_ext_map_blocks() if someone tries to write
3093  * to an uninitialized extent. It may result in splitting the uninitialized
3094  * extent into multiple extents (up to three - one initialized and two
3095  * uninitialized).
3096  * There are three possibilities:
3097  *   a> There is no split required: Entire extent should be initialized
3098  *   b> Splits in two extents: Write is happening at either end of the extent
3099  *   c> Splits in three extents: Somone is writing in middle of the extent
3100  *
3101  * Pre-conditions:
3102  *  - The extent pointed to by 'path' is uninitialized.
3103  *  - The extent pointed to by 'path' contains a superset
3104  *    of the logical span [map->m_lblk, map->m_lblk + map->m_len).
3105  *
3106  * Post-conditions on success:
3107  *  - the returned value is the number of blocks beyond map->l_lblk
3108  *    that are allocated and initialized.
3109  *    It is guaranteed to be >= map->m_len.
3110  */
3111 static int ext4_ext_convert_to_initialized(handle_t *handle,
3112                                            struct inode *inode,
3113                                            struct ext4_map_blocks *map,
3114                                            struct ext4_ext_path *path)
3115 {
3116         struct ext4_sb_info *sbi;
3117         struct ext4_extent_header *eh;
3118         struct ext4_map_blocks split_map;
3119         struct ext4_extent zero_ex;
3120         struct ext4_extent *ex;
3121         ext4_lblk_t ee_block, eof_block;
3122         unsigned int ee_len, depth;
3123         int allocated, max_zeroout = 0;
3124         int err = 0;
3125         int split_flag = 0;
3126
3127         ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
3128                 "block %llu, max_blocks %u\n", inode->i_ino,
3129                 (unsigned long long)map->m_lblk, map->m_len);
3130
3131         sbi = EXT4_SB(inode->i_sb);
3132         eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
3133                 inode->i_sb->s_blocksize_bits;
3134         if (eof_block < map->m_lblk + map->m_len)
3135                 eof_block = map->m_lblk + map->m_len;
3136
3137         depth = ext_depth(inode);
3138         eh = path[depth].p_hdr;
3139         ex = path[depth].p_ext;
3140         ee_block = le32_to_cpu(ex->ee_block);
3141         ee_len = ext4_ext_get_actual_len(ex);
3142         allocated = ee_len - (map->m_lblk - ee_block);
3143
3144         trace_ext4_ext_convert_to_initialized_enter(inode, map, ex);
3145
3146         /* Pre-conditions */
3147         BUG_ON(!ext4_ext_is_uninitialized(ex));
3148         BUG_ON(!in_range(map->m_lblk, ee_block, ee_len));
3149
3150         /*
3151          * Attempt to transfer newly initialized blocks from the currently
3152          * uninitialized extent to its left neighbor. This is much cheaper
3153          * than an insertion followed by a merge as those involve costly
3154          * memmove() calls. This is the common case in steady state for
3155          * workloads doing fallocate(FALLOC_FL_KEEP_SIZE) followed by append
3156          * writes.
3157          *
3158          * Limitations of the current logic:
3159          *  - L1: we only deal with writes at the start of the extent.
3160          *    The approach could be extended to writes at the end
3161          *    of the extent but this scenario was deemed less common.
3162          *  - L2: we do not deal with writes covering the whole extent.
3163          *    This would require removing the extent if the transfer
3164          *    is possible.
3165          *  - L3: we only attempt to merge with an extent stored in the
3166          *    same extent tree node.
3167          */
3168         if ((map->m_lblk == ee_block) &&        /*L1*/
3169                 (map->m_len < ee_len) &&        /*L2*/
3170                 (ex > EXT_FIRST_EXTENT(eh))) {  /*L3*/
3171                 struct ext4_extent *prev_ex;
3172                 ext4_lblk_t prev_lblk;
3173                 ext4_fsblk_t prev_pblk, ee_pblk;
3174                 unsigned int prev_len, write_len;
3175
3176                 prev_ex = ex - 1;
3177                 prev_lblk = le32_to_cpu(prev_ex->ee_block);
3178                 prev_len = ext4_ext_get_actual_len(prev_ex);
3179                 prev_pblk = ext4_ext_pblock(prev_ex);
3180                 ee_pblk = ext4_ext_pblock(ex);
3181                 write_len = map->m_len;
3182
3183                 /*
3184                  * A transfer of blocks from 'ex' to 'prev_ex' is allowed
3185                  * upon those conditions:
3186                  * - C1: prev_ex is initialized,
3187                  * - C2: prev_ex is logically abutting ex,
3188                  * - C3: prev_ex is physically abutting ex,
3189                  * - C4: prev_ex can receive the additional blocks without
3190                  *   overflowing the (initialized) length limit.
3191                  */
3192                 if ((!ext4_ext_is_uninitialized(prev_ex)) &&            /*C1*/
3193                         ((prev_lblk + prev_len) == ee_block) &&         /*C2*/
3194                         ((prev_pblk + prev_len) == ee_pblk) &&          /*C3*/
3195                         (prev_len < (EXT_INIT_MAX_LEN - write_len))) {  /*C4*/
3196                         err = ext4_ext_get_access(handle, inode, path + depth);
3197                         if (err)
3198                                 goto out;
3199
3200                         trace_ext4_ext_convert_to_initialized_fastpath(inode,
3201                                 map, ex, prev_ex);
3202
3203                         /* Shift the start of ex by 'write_len' blocks */
3204                         ex->ee_block = cpu_to_le32(ee_block + write_len);
3205                         ext4_ext_store_pblock(ex, ee_pblk + write_len);
3206                         ex->ee_len = cpu_to_le16(ee_len - write_len);
3207                         ext4_ext_mark_uninitialized(ex); /* Restore the flag */
3208
3209                         /* Extend prev_ex by 'write_len' blocks */
3210                         prev_ex->ee_len = cpu_to_le16(prev_len + write_len);
3211
3212                         /* Mark the block containing both extents as dirty */
3213                         ext4_ext_dirty(handle, inode, path + depth);
3214
3215                         /* Update path to point to the right extent */
3216                         path[depth].p_ext = prev_ex;
3217
3218                         /* Result: number of initialized blocks past m_lblk */
3219                         allocated = write_len;
3220                         goto out;
3221                 }
3222         }
3223
3224         WARN_ON(map->m_lblk < ee_block);
3225         /*
3226          * It is safe to convert extent to initialized via explicit
3227          * zeroout only if extent is fully insde i_size or new_size.
3228          */
3229         split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3230
3231         if (EXT4_EXT_MAY_ZEROOUT & split_flag)
3232                 max_zeroout = sbi->s_extent_max_zeroout_kb >>
3233                         inode->i_sb->s_blocksize_bits;
3234
3235         /* If extent is less than s_max_zeroout_kb, zeroout directly */
3236         if (max_zeroout && (ee_len <= max_zeroout)) {
3237                 err = ext4_ext_zeroout(inode, ex);
3238                 if (err)
3239                         goto out;
3240
3241                 err = ext4_ext_get_access(handle, inode, path + depth);
3242                 if (err)
3243                         goto out;
3244                 ext4_ext_mark_initialized(ex);
3245                 ext4_ext_try_to_merge(handle, inode, path, ex);
3246                 err = ext4_ext_dirty(handle, inode, path + path->p_depth);
3247                 goto out;
3248         }
3249
3250         /*
3251          * four cases:
3252          * 1. split the extent into three extents.
3253          * 2. split the extent into two extents, zeroout the first half.
3254          * 3. split the extent into two extents, zeroout the second half.
3255          * 4. split the extent into two extents with out zeroout.
3256          */
3257         split_map.m_lblk = map->m_lblk;
3258         split_map.m_len = map->m_len;
3259
3260         if (max_zeroout && (allocated > map->m_len)) {
3261                 if (allocated <= max_zeroout) {
3262                         /* case 3 */
3263                         zero_ex.ee_block =
3264                                          cpu_to_le32(map->m_lblk);
3265                         zero_ex.ee_len = cpu_to_le16(allocated);
3266                         ext4_ext_store_pblock(&zero_ex,
3267                                 ext4_ext_pblock(ex) + map->m_lblk - ee_block);
3268                         err = ext4_ext_zeroout(inode, &zero_ex);
3269                         if (err)
3270                                 goto out;
3271                         split_map.m_lblk = map->m_lblk;
3272                         split_map.m_len = allocated;
3273                 } else if (map->m_lblk - ee_block + map->m_len < max_zeroout) {
3274                         /* case 2 */
3275                         if (map->m_lblk != ee_block) {
3276                                 zero_ex.ee_block = ex->ee_block;
3277                                 zero_ex.ee_len = cpu_to_le16(map->m_lblk -
3278                                                         ee_block);
3279                                 ext4_ext_store_pblock(&zero_ex,
3280                                                       ext4_ext_pblock(ex));
3281                                 err = ext4_ext_zeroout(inode, &zero_ex);
3282                                 if (err)
3283                                         goto out;
3284                         }
3285
3286                         split_map.m_lblk = ee_block;
3287                         split_map.m_len = map->m_lblk - ee_block + map->m_len;
3288                         allocated = map->m_len;
3289                 }
3290         }
3291
3292         allocated = ext4_split_extent(handle, inode, path,
3293                                       &split_map, split_flag, 0);
3294         if (allocated < 0)
3295                 err = allocated;
3296
3297 out:
3298         return err ? err : allocated;
3299 }
3300
3301 /*
3302  * This function is called by ext4_ext_map_blocks() from
3303  * ext4_get_blocks_dio_write() when DIO to write
3304  * to an uninitialized extent.
3305  *
3306  * Writing to an uninitialized extent may result in splitting the uninitialized
3307  * extent into multiple initialized/uninitialized extents (up to three)
3308  * There are three possibilities:
3309  *   a> There is no split required: Entire extent should be uninitialized
3310  *   b> Splits in two extents: Write is happening at either end of the extent
3311  *   c> Splits in three extents: Somone is writing in middle of the extent
3312  *
3313  * One of more index blocks maybe needed if the extent tree grow after
3314  * the uninitialized extent split. To prevent ENOSPC occur at the IO
3315  * complete, we need to split the uninitialized extent before DIO submit
3316  * the IO. The uninitialized extent called at this time will be split
3317  * into three uninitialized extent(at most). After IO complete, the part
3318  * being filled will be convert to initialized by the end_io callback function
3319  * via ext4_convert_unwritten_extents().
3320  *
3321  * Returns the size of uninitialized extent to be written on success.
3322  */
3323 static int ext4_split_unwritten_extents(handle_t *handle,
3324                                         struct inode *inode,
3325                                         struct ext4_map_blocks *map,
3326                                         struct ext4_ext_path *path,
3327                                         int flags)
3328 {
3329         ext4_lblk_t eof_block;
3330         ext4_lblk_t ee_block;
3331         struct ext4_extent *ex;
3332         unsigned int ee_len;
3333         int split_flag = 0, depth;
3334
3335         ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
3336                 "block %llu, max_blocks %u\n", inode->i_ino,
3337                 (unsigned long long)map->m_lblk, map->m_len);
3338
3339         eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
3340                 inode->i_sb->s_blocksize_bits;
3341         if (eof_block < map->m_lblk + map->m_len)
3342                 eof_block = map->m_lblk + map->m_len;
3343         /*
3344          * It is safe to convert extent to initialized via explicit
3345          * zeroout only if extent is fully insde i_size or new_size.
3346          */
3347         depth = ext_depth(inode);
3348         ex = path[depth].p_ext;
3349         ee_block = le32_to_cpu(ex->ee_block);
3350         ee_len = ext4_ext_get_actual_len(ex);
3351
3352         split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3353         split_flag |= EXT4_EXT_MARK_UNINIT2;
3354
3355         flags |= EXT4_GET_BLOCKS_PRE_IO;
3356         return ext4_split_extent(handle, inode, path, map, split_flag, flags);
3357 }
3358
3359 static int ext4_convert_unwritten_extents_endio(handle_t *handle,
3360                                               struct inode *inode,
3361                                               struct ext4_ext_path *path)
3362 {
3363         struct ext4_extent *ex;
3364         int depth;
3365         int err = 0;
3366
3367         depth = ext_depth(inode);
3368         ex = path[depth].p_ext;
3369
3370         ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
3371                 "block %llu, max_blocks %u\n", inode->i_ino,
3372                 (unsigned long long)le32_to_cpu(ex->ee_block),
3373                 ext4_ext_get_actual_len(ex));
3374
3375         err = ext4_ext_get_access(handle, inode, path + depth);
3376         if (err)
3377                 goto out;
3378         /* first mark the extent as initialized */
3379         ext4_ext_mark_initialized(ex);
3380
3381         /* note: ext4_ext_correct_indexes() isn't needed here because
3382          * borders are not changed
3383          */
3384         ext4_ext_try_to_merge(handle, inode, path, ex);
3385
3386         /* Mark modified extent as dirty */
3387         err = ext4_ext_dirty(handle, inode, path + path->p_depth);
3388 out:
3389         ext4_ext_show_leaf(inode, path);
3390         return err;
3391 }
3392
3393 static void unmap_underlying_metadata_blocks(struct block_device *bdev,
3394                         sector_t block, int count)
3395 {
3396         int i;
3397         for (i = 0; i < count; i++)
3398                 unmap_underlying_metadata(bdev, block + i);
3399 }
3400
3401 /*
3402  * Handle EOFBLOCKS_FL flag, clearing it if necessary
3403  */
3404 static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
3405                               ext4_lblk_t lblk,
3406                               struct ext4_ext_path *path,
3407                               unsigned int len)
3408 {
3409         int i, depth;
3410         struct ext4_extent_header *eh;
3411         struct ext4_extent *last_ex;
3412
3413         if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
3414                 return 0;
3415
3416         depth = ext_depth(inode);
3417         eh = path[depth].p_hdr;
3418
3419         /*
3420          * We're going to remove EOFBLOCKS_FL entirely in future so we
3421          * do not care for this case anymore. Simply remove the flag
3422          * if there are no extents.
3423          */
3424         if (unlikely(!eh->eh_entries))
3425                 goto out;
3426         last_ex = EXT_LAST_EXTENT(eh);
3427         /*
3428          * We should clear the EOFBLOCKS_FL flag if we are writing the
3429          * last block in the last extent in the file.  We test this by
3430          * first checking to see if the caller to
3431          * ext4_ext_get_blocks() was interested in the last block (or
3432          * a block beyond the last block) in the current extent.  If
3433          * this turns out to be false, we can bail out from this
3434          * function immediately.
3435          */
3436         if (lblk + len < le32_to_cpu(last_ex->ee_block) +
3437             ext4_ext_get_actual_len(last_ex))
3438                 return 0;
3439         /*
3440          * If the caller does appear to be planning to write at or
3441          * beyond the end of the current extent, we then test to see
3442          * if the current extent is the last extent in the file, by
3443          * checking to make sure it was reached via the rightmost node
3444          * at each level of the tree.
3445          */
3446         for (i = depth-1; i >= 0; i--)
3447                 if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
3448                         return 0;
3449 out:
3450         ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3451         return ext4_mark_inode_dirty(handle, inode);
3452 }
3453
3454 /**
3455  * ext4_find_delalloc_range: find delayed allocated block in the given range.
3456  *
3457  * Goes through the buffer heads in the range [lblk_start, lblk_end] and returns
3458  * whether there are any buffers marked for delayed allocation. It returns '1'
3459  * on the first delalloc'ed buffer head found. If no buffer head in the given
3460  * range is marked for delalloc, it returns 0.
3461  * lblk_start should always be <= lblk_end.
3462  * search_hint_reverse is to indicate that searching in reverse from lblk_end to
3463  * lblk_start might be more efficient (i.e., we will likely hit the delalloc'ed
3464  * block sooner). This is useful when blocks are truncated sequentially from
3465  * lblk_start towards lblk_end.
3466  */
3467 static int ext4_find_delalloc_range(struct inode *inode,
3468                                     ext4_lblk_t lblk_start,
3469                                     ext4_lblk_t lblk_end,
3470                                     int search_hint_reverse)
3471 {
3472         struct address_space *mapping = inode->i_mapping;
3473         struct buffer_head *head, *bh = NULL;
3474         struct page *page;
3475         ext4_lblk_t i, pg_lblk;
3476         pgoff_t index;
3477
3478         if (!test_opt(inode->i_sb, DELALLOC))
3479                 return 0;
3480
3481         /* reverse search wont work if fs block size is less than page size */
3482         if (inode->i_blkbits < PAGE_CACHE_SHIFT)
3483                 search_hint_reverse = 0;
3484
3485         if (search_hint_reverse)
3486                 i = lblk_end;
3487         else
3488                 i = lblk_start;
3489
3490         index = i >> (PAGE_CACHE_SHIFT - inode->i_blkbits);
3491
3492         while ((i >= lblk_start) && (i <= lblk_end)) {
3493                 page = find_get_page(mapping, index);
3494                 if (!page)
3495                         goto nextpage;
3496
3497                 if (!page_has_buffers(page))
3498                         goto nextpage;
3499
3500                 head = page_buffers(page);
3501                 if (!head)
3502                         goto nextpage;
3503
3504                 bh = head;
3505                 pg_lblk = index << (PAGE_CACHE_SHIFT -
3506                                                 inode->i_blkbits);
3507                 do {
3508                         if (unlikely(pg_lblk < lblk_start)) {
3509                                 /*
3510                                  * This is possible when fs block size is less
3511                                  * than page size and our cluster starts/ends in
3512                                  * middle of the page. So we need to skip the
3513                                  * initial few blocks till we reach the 'lblk'
3514                                  */
3515                                 pg_lblk++;
3516                                 continue;
3517                         }
3518
3519                         /* Check if the buffer is delayed allocated and that it
3520                          * is not yet mapped. (when da-buffers are mapped during
3521                          * their writeout, their da_mapped bit is set.)
3522                          */
3523                         if (buffer_delay(bh) && !buffer_da_mapped(bh)) {
3524                                 page_cache_release(page);
3525                                 trace_ext4_find_delalloc_range(inode,
3526                                                 lblk_start, lblk_end,
3527                                                 search_hint_reverse,
3528                                                 1, i);
3529                                 return 1;
3530                         }
3531                         if (search_hint_reverse)
3532                                 i--;
3533                         else
3534                                 i++;
3535                 } while ((i >= lblk_start) && (i <= lblk_end) &&
3536                                 ((bh = bh->b_this_page) != head));
3537 nextpage:
3538                 if (page)
3539                         page_cache_release(page);
3540                 /*
3541                  * Move to next page. 'i' will be the first lblk in the next
3542                  * page.
3543                  */
3544                 if (search_hint_reverse)
3545                         index--;
3546                 else
3547                         index++;
3548                 i = index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
3549         }
3550
3551         trace_ext4_find_delalloc_range(inode, lblk_start, lblk_end,
3552                                         search_hint_reverse, 0, 0);
3553         return 0;
3554 }
3555
3556 int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk,
3557                                int search_hint_reverse)
3558 {
3559         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3560         ext4_lblk_t lblk_start, lblk_end;
3561         lblk_start = lblk & (~(sbi->s_cluster_ratio - 1));
3562         lblk_end = lblk_start + sbi->s_cluster_ratio - 1;
3563
3564         return ext4_find_delalloc_range(inode, lblk_start, lblk_end,
3565                                         search_hint_reverse);
3566 }
3567
3568 /**
3569  * Determines how many complete clusters (out of those specified by the 'map')
3570  * are under delalloc and were reserved quota for.
3571  * This function is called when we are writing out the blocks that were
3572  * originally written with their allocation delayed, but then the space was
3573  * allocated using fallocate() before the delayed allocation could be resolved.
3574  * The cases to look for are:
3575  * ('=' indicated delayed allocated blocks
3576  *  '-' indicates non-delayed allocated blocks)
3577  * (a) partial clusters towards beginning and/or end outside of allocated range
3578  *     are not delalloc'ed.
3579  *      Ex:
3580  *      |----c---=|====c====|====c====|===-c----|
3581  *               |++++++ allocated ++++++|
3582  *      ==> 4 complete clusters in above example
3583  *
3584  * (b) partial cluster (outside of allocated range) towards either end is
3585  *     marked for delayed allocation. In this case, we will exclude that
3586  *     cluster.
3587  *      Ex:
3588  *      |----====c========|========c========|
3589  *           |++++++ allocated ++++++|
3590  *      ==> 1 complete clusters in above example
3591  *
3592  *      Ex:
3593  *      |================c================|
3594  *            |++++++ allocated ++++++|
3595  *      ==> 0 complete clusters in above example
3596  *
3597  * The ext4_da_update_reserve_space will be called only if we
3598  * determine here that there were some "entire" clusters that span
3599  * this 'allocated' range.
3600  * In the non-bigalloc case, this function will just end up returning num_blks
3601  * without ever calling ext4_find_delalloc_range.
3602  */
3603 static unsigned int
3604 get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start,
3605                            unsigned int num_blks)
3606 {
3607         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3608         ext4_lblk_t alloc_cluster_start, alloc_cluster_end;
3609         ext4_lblk_t lblk_from, lblk_to, c_offset;
3610         unsigned int allocated_clusters = 0;
3611
3612         alloc_cluster_start = EXT4_B2C(sbi, lblk_start);
3613         alloc_cluster_end = EXT4_B2C(sbi, lblk_start + num_blks - 1);
3614
3615         /* max possible clusters for this allocation */
3616         allocated_clusters = alloc_cluster_end - alloc_cluster_start + 1;
3617
3618         trace_ext4_get_reserved_cluster_alloc(inode, lblk_start, num_blks);
3619
3620         /* Check towards left side */
3621         c_offset = lblk_start & (sbi->s_cluster_ratio - 1);
3622         if (c_offset) {
3623                 lblk_from = lblk_start & (~(sbi->s_cluster_ratio - 1));
3624                 lblk_to = lblk_from + c_offset - 1;
3625
3626                 if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0))
3627                         allocated_clusters--;
3628         }
3629
3630         /* Now check towards right. */
3631         c_offset = (lblk_start + num_blks) & (sbi->s_cluster_ratio - 1);
3632         if (allocated_clusters && c_offset) {
3633                 lblk_from = lblk_start + num_blks;
3634                 lblk_to = lblk_from + (sbi->s_cluster_ratio - c_offset) - 1;
3635
3636                 if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0))
3637                         allocated_clusters--;
3638         }
3639
3640         return allocated_clusters;
3641 }
3642
3643 static int
3644 ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
3645                         struct ext4_map_blocks *map,
3646                         struct ext4_ext_path *path, int flags,
3647                         unsigned int allocated, ext4_fsblk_t newblock)
3648 {
3649         int ret = 0;
3650         int err = 0;
3651         ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3652
3653         ext_debug("ext4_ext_handle_uninitialized_extents: inode %lu, logical "
3654                   "block %llu, max_blocks %u, flags %x, allocated %u\n",
3655                   inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
3656                   flags, allocated);
3657         ext4_ext_show_leaf(inode, path);
3658
3659         trace_ext4_ext_handle_uninitialized_extents(inode, map, allocated,
3660                                                     newblock);
3661
3662         /* get_block() before submit the IO, split the extent */
3663         if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
3664                 ret = ext4_split_unwritten_extents(handle, inode, map,
3665                                                    path, flags);
3666                 /*
3667                  * Flag the inode(non aio case) or end_io struct (aio case)
3668                  * that this IO needs to conversion to written when IO is
3669                  * completed
3670                  */
3671                 if (io)
3672                         ext4_set_io_unwritten_flag(inode, io);
3673                 else
3674                         ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
3675                 if (ext4_should_dioread_nolock(inode))
3676                         map->m_flags |= EXT4_MAP_UNINIT;
3677                 goto out;
3678         }
3679         /* IO end_io complete, convert the filled extent to written */
3680         if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
3681                 ret = ext4_convert_unwritten_extents_endio(handle, inode,
3682                                                         path);
3683                 if (ret >= 0) {
3684                         ext4_update_inode_fsync_trans(handle, inode, 1);
3685                         err = check_eofblocks_fl(handle, inode, map->m_lblk,
3686                                                  path, map->m_len);
3687                 } else
3688                         err = ret;
3689                 goto out2;
3690         }
3691         /* buffered IO case */
3692         /*
3693          * repeat fallocate creation request
3694          * we already have an unwritten extent
3695          */
3696         if (flags & EXT4_GET_BLOCKS_UNINIT_EXT)
3697                 goto map_out;
3698
3699         /* buffered READ or buffered write_begin() lookup */
3700         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3701                 /*
3702                  * We have blocks reserved already.  We
3703                  * return allocated blocks so that delalloc
3704                  * won't do block reservation for us.  But
3705                  * the buffer head will be unmapped so that
3706                  * a read from the block returns 0s.
3707                  */
3708                 map->m_flags |= EXT4_MAP_UNWRITTEN;
3709                 goto out1;
3710         }
3711
3712         /* buffered write, writepage time, convert*/
3713         ret = ext4_ext_convert_to_initialized(handle, inode, map, path);
3714         if (ret >= 0)
3715                 ext4_update_inode_fsync_trans(handle, inode, 1);
3716 out:
3717         if (ret <= 0) {
3718                 err = ret;
3719                 goto out2;
3720         } else
3721                 allocated = ret;
3722         map->m_flags |= EXT4_MAP_NEW;
3723         /*
3724          * if we allocated more blocks than requested
3725          * we need to make sure we unmap the extra block
3726          * allocated. The actual needed block will get
3727          * unmapped later when we find the buffer_head marked
3728          * new.
3729          */
3730         if (allocated > map->m_len) {
3731                 unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
3732                                         newblock + map->m_len,
3733                                         allocated - map->m_len);
3734                 allocated = map->m_len;
3735         }
3736
3737         /*
3738          * If we have done fallocate with the offset that is already
3739          * delayed allocated, we would have block reservation
3740          * and quota reservation done in the delayed write path.
3741          * But fallocate would have already updated quota and block
3742          * count for this offset. So cancel these reservation
3743          */
3744         if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
3745                 unsigned int reserved_clusters;
3746                 reserved_clusters = get_reserved_cluster_alloc(inode,
3747                                 map->m_lblk, map->m_len);
3748                 if (reserved_clusters)
3749                         ext4_da_update_reserve_space(inode,
3750                                                      reserved_clusters,
3751                                                      0);
3752         }
3753
3754 map_out:
3755         map->m_flags |= EXT4_MAP_MAPPED;
3756         if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0) {
3757                 err = check_eofblocks_fl(handle, inode, map->m_lblk, path,
3758                                          map->m_len);
3759                 if (err < 0)
3760                         goto out2;
3761         }
3762 out1:
3763         if (allocated > map->m_len)
3764                 allocated = map->m_len;
3765         ext4_ext_show_leaf(inode, path);
3766         map->m_pblk = newblock;
3767         map->m_len = allocated;
3768 out2:
3769         if (path) {
3770                 ext4_ext_drop_refs(path);
3771                 kfree(path);
3772         }
3773         return err ? err : allocated;
3774 }
3775
3776 /*
3777  * get_implied_cluster_alloc - check to see if the requested
3778  * allocation (in the map structure) overlaps with a cluster already
3779  * allocated in an extent.
3780  *      @sb     The filesystem superblock structure
3781  *      @map    The requested lblk->pblk mapping
3782  *      @ex     The extent structure which might contain an implied
3783  *                      cluster allocation
3784  *
3785  * This function is called by ext4_ext_map_blocks() after we failed to
3786  * find blocks that were already in the inode's extent tree.  Hence,
3787  * we know that the beginning of the requested region cannot overlap
3788  * the extent from the inode's extent tree.  There are three cases we
3789  * want to catch.  The first is this case:
3790  *
3791  *               |--- cluster # N--|
3792  *    |--- extent ---|  |---- requested region ---|
3793  *                      |==========|
3794  *
3795  * The second case that we need to test for is this one:
3796  *
3797  *   |--------- cluster # N ----------------|
3798  *         |--- requested region --|   |------- extent ----|
3799  *         |=======================|
3800  *
3801  * The third case is when the requested region lies between two extents
3802  * within the same cluster:
3803  *          |------------- cluster # N-------------|
3804  * |----- ex -----|                  |---- ex_right ----|
3805  *                  |------ requested region ------|
3806  *                  |================|
3807  *
3808  * In each of the above cases, we need to set the map->m_pblk and
3809  * map->m_len so it corresponds to the return the extent labelled as
3810  * "|====|" from cluster #N, since it is already in use for data in
3811  * cluster EXT4_B2C(sbi, map->m_lblk).  We will then return 1 to
3812  * signal to ext4_ext_map_blocks() that map->m_pblk should be treated
3813  * as a new "allocated" block region.  Otherwise, we will return 0 and
3814  * ext4_ext_map_blocks() will then allocate one or more new clusters
3815  * by calling ext4_mb_new_blocks().
3816  */
3817 static int get_implied_cluster_alloc(struct super_block *sb,
3818                                      struct ext4_map_blocks *map,
3819                                      struct ext4_extent *ex,
3820                                      struct ext4_ext_path *path)
3821 {
3822         struct ext4_sb_info *sbi = EXT4_SB(sb);
3823         ext4_lblk_t c_offset = map->m_lblk & (sbi->s_cluster_ratio-1);
3824         ext4_lblk_t ex_cluster_start, ex_cluster_end;
3825         ext4_lblk_t rr_cluster_start;
3826         ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
3827         ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
3828         unsigned short ee_len = ext4_ext_get_actual_len(ex);
3829
3830         /* The extent passed in that we are trying to match */
3831         ex_cluster_start = EXT4_B2C(sbi, ee_block);
3832         ex_cluster_end = EXT4_B2C(sbi, ee_block + ee_len - 1);
3833
3834         /* The requested region passed into ext4_map_blocks() */
3835         rr_cluster_start = EXT4_B2C(sbi, map->m_lblk);
3836
3837         if ((rr_cluster_start == ex_cluster_end) ||
3838             (rr_cluster_start == ex_cluster_start)) {
3839                 if (rr_cluster_start == ex_cluster_end)
3840                         ee_start += ee_len - 1;
3841                 map->m_pblk = (ee_start & ~(sbi->s_cluster_ratio - 1)) +
3842                         c_offset;
3843                 map->m_len = min(map->m_len,
3844                                  (unsigned) sbi->s_cluster_ratio - c_offset);
3845                 /*
3846                  * Check for and handle this case:
3847                  *
3848                  *   |--------- cluster # N-------------|
3849                  *                     |------- extent ----|
3850                  *         |--- requested region ---|
3851                  *         |===========|
3852                  */
3853
3854                 if (map->m_lblk < ee_block)
3855                         map->m_len = min(map->m_len, ee_block - map->m_lblk);
3856
3857                 /*
3858                  * Check for the case where there is already another allocated
3859                  * block to the right of 'ex' but before the end of the cluster.
3860                  *
3861                  *          |------------- cluster # N-------------|
3862                  * |----- ex -----|                  |---- ex_right ----|
3863                  *                  |------ requested region ------|
3864                  *                  |================|
3865                  */
3866                 if (map->m_lblk > ee_block) {
3867                         ext4_lblk_t next = ext4_ext_next_allocated_block(path);
3868                         map->m_len = min(map->m_len, next - map->m_lblk);
3869                 }
3870
3871                 trace_ext4_get_implied_cluster_alloc_exit(sb, map, 1);
3872                 return 1;
3873         }
3874
3875         trace_ext4_get_implied_cluster_alloc_exit(sb, map, 0);
3876         return 0;
3877 }
3878
3879
3880 /*
3881  * Block allocation/map/preallocation routine for extents based files
3882  *
3883  *
3884  * Need to be called with
3885  * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
3886  * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
3887  *
3888  * return > 0, number of of blocks already mapped/allocated
3889  *          if create == 0 and these are pre-allocated blocks
3890  *              buffer head is unmapped
3891  *          otherwise blocks are mapped
3892  *
3893  * return = 0, if plain look up failed (blocks have not been allocated)
3894  *          buffer head is unmapped
3895  *
3896  * return < 0, error case.
3897  */
3898 int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
3899                         struct ext4_map_blocks *map, int flags)
3900 {
3901         struct ext4_ext_path *path = NULL;
3902         struct ext4_extent newex, *ex, *ex2;
3903         struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3904         ext4_fsblk_t newblock = 0;
3905         int free_on_err = 0, err = 0, depth, ret;
3906         unsigned int allocated = 0, offset = 0;
3907         unsigned int allocated_clusters = 0;
3908         struct ext4_allocation_request ar;
3909         ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3910         ext4_lblk_t cluster_offset;
3911
3912         ext_debug("blocks %u/%u requested for inode %lu\n",
3913                   map->m_lblk, map->m_len, inode->i_ino);
3914         trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
3915
3916         /* check in cache */
3917         if (ext4_ext_in_cache(inode, map->m_lblk, &newex)) {
3918                 if (!newex.ee_start_lo && !newex.ee_start_hi) {
3919                         if ((sbi->s_cluster_ratio > 1) &&
3920                             ext4_find_delalloc_cluster(inode, map->m_lblk, 0))
3921                                 map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3922
3923                         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3924                                 /*
3925                                  * block isn't allocated yet and
3926                                  * user doesn't want to allocate it
3927                                  */
3928                                 goto out2;
3929                         }
3930                         /* we should allocate requested block */
3931                 } else {
3932                         /* block is already allocated */
3933                         if (sbi->s_cluster_ratio > 1)
3934                                 map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3935                         newblock = map->m_lblk
3936                                    - le32_to_cpu(newex.ee_block)
3937                                    + ext4_ext_pblock(&newex);
3938                         /* number of remaining blocks in the extent */
3939                         allocated = ext4_ext_get_actual_len(&newex) -
3940                                 (map->m_lblk - le32_to_cpu(newex.ee_block));
3941                         goto out;
3942                 }
3943         }
3944
3945         /* find extent for this block */
3946         path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
3947         if (IS_ERR(path)) {
3948                 err = PTR_ERR(path);
3949                 path = NULL;
3950                 goto out2;
3951         }
3952
3953         depth = ext_depth(inode);
3954
3955         /*
3956          * consistent leaf must not be empty;
3957          * this situation is possible, though, _during_ tree modification;
3958          * this is why assert can't be put in ext4_ext_find_extent()
3959          */
3960         if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
3961                 EXT4_ERROR_INODE(inode, "bad extent address "
3962                                  "lblock: %lu, depth: %d pblock %lld",
3963                                  (unsigned long) map->m_lblk, depth,
3964                                  path[depth].p_block);
3965                 err = -EIO;
3966                 goto out2;
3967         }
3968
3969         ex = path[depth].p_ext;
3970         if (ex) {
3971                 ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
3972                 ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
3973                 unsigned short ee_len;
3974
3975                 /*
3976                  * Uninitialized extents are treated as holes, except that
3977                  * we split out initialized portions during a write.
3978                  */
3979                 ee_len = ext4_ext_get_actual_len(ex);
3980
3981                 trace_ext4_ext_show_extent(inode, ee_block, ee_start, ee_len);
3982
3983                 /* if found extent covers block, simply return it */
3984                 if (in_range(map->m_lblk, ee_block, ee_len)) {
3985                         newblock = map->m_lblk - ee_block + ee_start;
3986                         /* number of remaining blocks in the extent */
3987                         allocated = ee_len - (map->m_lblk - ee_block);
3988                         ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
3989                                   ee_block, ee_len, newblock);
3990
3991                         /*
3992                          * Do not put uninitialized extent
3993                          * in the cache
3994                          */
3995                         if (!ext4_ext_is_uninitialized(ex)) {
3996                                 ext4_ext_put_in_cache(inode, ee_block,
3997                                         ee_len, ee_start);
3998                                 goto out;
3999                         }
4000                         ret = ext4_ext_handle_uninitialized_extents(
4001                                 handle, inode, map, path, flags,
4002                                 allocated, newblock);
4003                         return ret;
4004                 }
4005         }
4006
4007         if ((sbi->s_cluster_ratio > 1) &&
4008             ext4_find_delalloc_cluster(inode, map->m_lblk, 0))
4009                 map->m_flags |= EXT4_MAP_FROM_CLUSTER;
4010
4011         /*
4012          * requested block isn't allocated yet;
4013          * we couldn't try to create block if create flag is zero
4014          */
4015         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
4016                 /*
4017                  * put just found gap into cache to speed up
4018                  * subsequent requests
4019                  */
4020                 ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
4021                 goto out2;
4022         }
4023
4024         /*
4025          * Okay, we need to do block allocation.
4026          */
4027         map->m_flags &= ~EXT4_MAP_FROM_CLUSTER;
4028         newex.ee_block = cpu_to_le32(map->m_lblk);
4029         cluster_offset = map->m_lblk & (sbi->s_cluster_ratio-1);
4030
4031         /*
4032          * If we are doing bigalloc, check to see if the extent returned
4033          * by ext4_ext_find_extent() implies a cluster we can use.
4034          */
4035         if (cluster_offset && ex &&
4036             get_implied_cluster_alloc(inode->i_sb, map, ex, path)) {
4037                 ar.len = allocated = map->m_len;
4038                 newblock = map->m_pblk;
4039                 map->m_flags |= EXT4_MAP_FROM_CLUSTER;
4040                 goto got_allocated_blocks;
4041         }
4042
4043         /* find neighbour allocated blocks */
4044         ar.lleft = map->m_lblk;
4045         err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
4046         if (err)
4047                 goto out2;
4048         ar.lright = map->m_lblk;
4049         ex2 = NULL;
4050         err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright, &ex2);
4051         if (err)
4052                 goto out2;
4053
4054         /* Check if the extent after searching to the right implies a
4055          * cluster we can use. */
4056         if ((sbi->s_cluster_ratio > 1) && ex2 &&
4057             get_implied_cluster_alloc(inode->i_sb, map, ex2, path)) {
4058                 ar.len = allocated = map->m_len;
4059                 newblock = map->m_pblk;
4060                 map->m_flags |= EXT4_MAP_FROM_CLUSTER;
4061                 goto got_allocated_blocks;
4062         }
4063
4064         /*
4065          * See if request is beyond maximum number of blocks we can have in
4066          * a single extent. For an initialized extent this limit is
4067          * EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
4068          * EXT_UNINIT_MAX_LEN.
4069          */
4070         if (map->m_len > EXT_INIT_MAX_LEN &&
4071             !(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
4072                 map->m_len = EXT_INIT_MAX_LEN;
4073         else if (map->m_len > EXT_UNINIT_MAX_LEN &&
4074                  (flags & EXT4_GET_BLOCKS_UNINIT_EXT))
4075                 map->m_len = EXT_UNINIT_MAX_LEN;
4076
4077         /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */
4078         newex.ee_len = cpu_to_le16(map->m_len);
4079         err = ext4_ext_check_overlap(sbi, inode, &newex, path);
4080         if (err)
4081                 allocated = ext4_ext_get_actual_len(&newex);
4082         else
4083                 allocated = map->m_len;
4084
4085         /* allocate new block */
4086         ar.inode = inode;
4087         ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk);
4088         ar.logical = map->m_lblk;
4089         /*
4090          * We calculate the offset from the beginning of the cluster
4091          * for the logical block number, since when we allocate a
4092          * physical cluster, the physical block should start at the
4093          * same offset from the beginning of the cluster.  This is
4094          * needed so that future calls to get_implied_cluster_alloc()
4095          * work correctly.
4096          */
4097         offset = map->m_lblk & (sbi->s_cluster_ratio - 1);
4098         ar.len = EXT4_NUM_B2C(sbi, offset+allocated);
4099         ar.goal -= offset;
4100         ar.logical -= offset;
4101         if (S_ISREG(inode->i_mode))
4102                 ar.flags = EXT4_MB_HINT_DATA;
4103         else
4104                 /* disable in-core preallocation for non-regular files */
4105                 ar.flags = 0;
4106         if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE)
4107                 ar.flags |= EXT4_MB_HINT_NOPREALLOC;
4108         newblock = ext4_mb_new_blocks(handle, &ar, &err);
4109         if (!newblock)
4110                 goto out2;
4111         ext_debug("allocate new block: goal %llu, found %llu/%u\n",
4112                   ar.goal, newblock, allocated);
4113         free_on_err = 1;
4114         allocated_clusters = ar.len;
4115         ar.len = EXT4_C2B(sbi, ar.len) - offset;
4116         if (ar.len > allocated)
4117                 ar.len = allocated;
4118
4119 got_allocated_blocks:
4120         /* try to insert new extent into found leaf and return */
4121         ext4_ext_store_pblock(&newex, newblock + offset);
4122         newex.ee_len = cpu_to_le16(ar.len);
4123         /* Mark uninitialized */
4124         if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
4125                 ext4_ext_mark_uninitialized(&newex);
4126                 /*
4127                  * io_end structure was created for every IO write to an
4128                  * uninitialized extent. To avoid unnecessary conversion,
4129                  * here we flag the IO that really needs the conversion.
4130                  * For non asycn direct IO case, flag the inode state
4131                  * that we need to perform conversion when IO is done.
4132                  */
4133                 if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
4134                         if (io)
4135                                 ext4_set_io_unwritten_flag(inode, io);
4136                         else
4137                                 ext4_set_inode_state(inode,
4138                                                      EXT4_STATE_DIO_UNWRITTEN);
4139                 }
4140                 if (ext4_should_dioread_nolock(inode))
4141                         map->m_flags |= EXT4_MAP_UNINIT;
4142         }
4143
4144         err = 0;
4145         if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0)
4146                 err = check_eofblocks_fl(handle, inode, map->m_lblk,
4147                                          path, ar.len);
4148         if (!err)
4149                 err = ext4_ext_insert_extent(handle, inode, path,
4150                                              &newex, flags);
4151         if (err && free_on_err) {
4152                 int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ?
4153                         EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0;
4154                 /* free data blocks we just allocated */
4155                 /* not a good idea to call discard here directly,
4156                  * but otherwise we'd need to call it every free() */
4157                 ext4_discard_preallocations(inode);
4158                 ext4_free_blocks(handle, inode, NULL, ext4_ext_pblock(&newex),
4159                                  ext4_ext_get_actual_len(&newex), fb_flags);
4160                 goto out2;
4161         }
4162
4163         /* previous routine could use block we allocated */
4164         newblock = ext4_ext_pblock(&newex);
4165         allocated = ext4_ext_get_actual_len(&newex);
4166         if (allocated > map->m_len)
4167                 allocated = map->m_len;
4168         map->m_flags |= EXT4_MAP_NEW;
4169
4170         /*
4171          * Update reserved blocks/metadata blocks after successful
4172          * block allocation which had been deferred till now.
4173          */
4174         if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
4175                 unsigned int reserved_clusters;
4176                 /*
4177                  * Check how many clusters we had reserved this allocated range
4178                  */
4179                 reserved_clusters = get_reserved_cluster_alloc(inode,
4180                                                 map->m_lblk, allocated);
4181                 if (map->m_flags & EXT4_MAP_FROM_CLUSTER) {
4182                         if (reserved_clusters) {
4183                                 /*
4184                                  * We have clusters reserved for this range.
4185                                  * But since we are not doing actual allocation
4186                                  * and are simply using blocks from previously
4187                                  * allocated cluster, we should release the
4188                                  * reservation and not claim quota.
4189                                  */
4190                                 ext4_da_update_reserve_space(inode,
4191                                                 reserved_clusters, 0);
4192                         }
4193                 } else {
4194                         BUG_ON(allocated_clusters < reserved_clusters);
4195                         /* We will claim quota for all newly allocated blocks.*/
4196                         ext4_da_update_reserve_space(inode, allocated_clusters,
4197                                                         1);
4198                         if (reserved_clusters < allocated_clusters) {
4199                                 struct ext4_inode_info *ei = EXT4_I(inode);
4200                                 int reservation = allocated_clusters -
4201                                                   reserved_clusters;
4202                                 /*
4203                                  * It seems we claimed few clusters outside of
4204                                  * the range of this allocation. We should give
4205                                  * it back to the reservation pool. This can
4206                                  * happen in the following case:
4207                                  *
4208                                  * * Suppose s_cluster_ratio is 4 (i.e., each
4209                                  *   cluster has 4 blocks. Thus, the clusters
4210                                  *   are [0-3],[4-7],[8-11]...
4211                                  * * First comes delayed allocation write for
4212                                  *   logical blocks 10 & 11. Since there were no
4213                                  *   previous delayed allocated blocks in the
4214                                  *   range [8-11], we would reserve 1 cluster
4215                                  *   for this write.
4216                                  * * Next comes write for logical blocks 3 to 8.
4217                                  *   In this case, we will reserve 2 clusters
4218                                  *   (for [0-3] and [4-7]; and not for [8-11] as
4219                                  *   that range has a delayed allocated blocks.
4220                                  *   Thus total reserved clusters now becomes 3.
4221                                  * * Now, during the delayed allocation writeout
4222                                  *   time, we will first write blocks [3-8] and
4223                                  *   allocate 3 clusters for writing these
4224                                  *   blocks. Also, we would claim all these
4225                                  *   three clusters above.
4226                                  * * Now when we come here to writeout the
4227                                  *   blocks [10-11], we would expect to claim
4228                                  *   the reservation of 1 cluster we had made
4229                                  *   (and we would claim it since there are no
4230                                  *   more delayed allocated blocks in the range
4231                                  *   [8-11]. But our reserved cluster count had
4232                                  *   already gone to 0.
4233                                  *
4234                                  *   Thus, at the step 4 above when we determine
4235                                  *   that there are still some unwritten delayed
4236                                  *   allocated blocks outside of our current
4237                                  *   block range, we should increment the
4238                                  *   reserved clusters count so that when the
4239                                  *   remaining blocks finally gets written, we
4240                                  *   could claim them.
4241                                  */
4242                                 dquot_reserve_block(inode,
4243                                                 EXT4_C2B(sbi, reservation));
4244                                 spin_lock(&ei->i_block_reservation_lock);
4245                                 ei->i_reserved_data_blocks += reservation;
4246                                 spin_unlock(&ei->i_block_reservation_lock);
4247                         }
4248                 }
4249         }
4250
4251         /*
4252          * Cache the extent and update transaction to commit on fdatasync only
4253          * when it is _not_ an uninitialized extent.
4254          */
4255         if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) {
4256                 ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock);
4257                 ext4_update_inode_fsync_trans(handle, inode, 1);
4258         } else
4259                 ext4_update_inode_fsync_trans(handle, inode, 0);
4260 out:
4261         if (allocated > map->m_len)
4262                 allocated = map->m_len;
4263         ext4_ext_show_leaf(inode, path);
4264         map->m_flags |= EXT4_MAP_MAPPED;
4265         map->m_pblk = newblock;
4266         map->m_len = allocated;
4267 out2:
4268         if (path) {
4269                 ext4_ext_drop_refs(path);
4270                 kfree(path);
4271         }
4272
4273         trace_ext4_ext_map_blocks_exit(inode, map->m_lblk,
4274                 newblock, map->m_len, err ? err : allocated);
4275
4276         return err ? err : allocated;
4277 }
4278
4279 void ext4_ext_truncate(struct inode *inode)
4280 {
4281         struct address_space *mapping = inode->i_mapping;
4282         struct super_block *sb = inode->i_sb;
4283         ext4_lblk_t last_block;
4284         handle_t *handle;
4285         loff_t page_len;
4286         int err = 0;
4287
4288         /*
4289          * finish any pending end_io work so we won't run the risk of
4290          * converting any truncated blocks to initialized later
4291          */
4292         ext4_flush_completed_IO(inode);
4293
4294         /*
4295          * probably first extent we're gonna free will be last in block
4296          */
4297         err = ext4_writepage_trans_blocks(inode);
4298         handle = ext4_journal_start(inode, err);
4299         if (IS_ERR(handle))
4300                 return;
4301
4302         if (inode->i_size % PAGE_CACHE_SIZE != 0) {
4303                 page_len = PAGE_CACHE_SIZE -
4304                         (inode->i_size & (PAGE_CACHE_SIZE - 1));
4305
4306                 err = ext4_discard_partial_page_buffers(handle,
4307                         mapping, inode->i_size, page_len, 0);
4308
4309                 if (err)
4310                         goto out_stop;
4311         }
4312
4313         if (ext4_orphan_add(handle, inode))
4314                 goto out_stop;
4315
4316         down_write(&EXT4_I(inode)->i_data_sem);
4317         ext4_ext_invalidate_cache(inode);
4318
4319         ext4_discard_preallocations(inode);
4320
4321         /*
4322          * TODO: optimization is possible here.
4323          * Probably we need not scan at all,
4324          * because page truncation is enough.
4325          */
4326
4327         /* we have to know where to truncate from in crash case */
4328         EXT4_I(inode)->i_disksize = inode->i_size;
4329         ext4_mark_inode_dirty(handle, inode);
4330
4331         last_block = (inode->i_size + sb->s_blocksize - 1)
4332                         >> EXT4_BLOCK_SIZE_BITS(sb);
4333         err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1);
4334
4335         /* In a multi-transaction truncate, we only make the final
4336          * transaction synchronous.
4337          */
4338         if (IS_SYNC(inode))
4339                 ext4_handle_sync(handle);
4340
4341         up_write(&EXT4_I(inode)->i_data_sem);
4342
4343 out_stop:
4344         /*
4345          * If this was a simple ftruncate() and the file will remain alive,
4346          * then we need to clear up the orphan record which we created above.
4347          * However, if this was a real unlink then we were called by
4348          * ext4_delete_inode(), and we allow that function to clean up the
4349          * orphan info for us.
4350          */
4351         if (inode->i_nlink)
4352                 ext4_orphan_del(handle, inode);
4353
4354         inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4355         ext4_mark_inode_dirty(handle, inode);
4356         ext4_journal_stop(handle);
4357 }
4358
4359 static void ext4_falloc_update_inode(struct inode *inode,
4360                                 int mode, loff_t new_size, int update_ctime)
4361 {
4362         struct timespec now;
4363
4364         if (update_ctime) {
4365                 now = current_fs_time(inode->i_sb);
4366                 if (!timespec_equal(&inode->i_ctime, &now))
4367                         inode->i_ctime = now;
4368         }
4369         /*
4370          * Update only when preallocation was requested beyond
4371          * the file size.
4372          */
4373         if (!(mode & FALLOC_FL_KEEP_SIZE)) {
4374                 if (new_size > i_size_read(inode))
4375                         i_size_write(inode, new_size);
4376                 if (new_size > EXT4_I(inode)->i_disksize)
4377                         ext4_update_i_disksize(inode, new_size);
4378         } else {
4379                 /*
4380                  * Mark that we allocate beyond EOF so the subsequent truncate
4381                  * can proceed even if the new size is the same as i_size.
4382                  */
4383                 if (new_size > i_size_read(inode))
4384                         ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
4385         }
4386
4387 }
4388
4389 /*
4390  * preallocate space for a file. This implements ext4's fallocate file
4391  * operation, which gets called from sys_fallocate system call.
4392  * For block-mapped files, posix_fallocate should fall back to the method
4393  * of writing zeroes to the required new blocks (the same behavior which is
4394  * expected for file systems which do not support fallocate() system call).
4395  */
4396 long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
4397 {
4398         struct inode *inode = file->f_path.dentry->d_inode;
4399         handle_t *handle;
4400         loff_t new_size;
4401         unsigned int max_blocks;
4402         int ret = 0;
4403         int ret2 = 0;
4404         int retries = 0;
4405         int flags;
4406         struct ext4_map_blocks map;
4407         unsigned int credits, blkbits = inode->i_blkbits;
4408
4409         /*
4410          * currently supporting (pre)allocate mode for extent-based
4411          * files _only_
4412          */
4413         if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
4414                 return -EOPNOTSUPP;
4415
4416         /* Return error if mode is not supported */
4417         if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
4418                 return -EOPNOTSUPP;
4419
4420         if (mode & FALLOC_FL_PUNCH_HOLE)
4421                 return ext4_punch_hole(file, offset, len);
4422
4423         trace_ext4_fallocate_enter(inode, offset, len, mode);
4424         map.m_lblk = offset >> blkbits;
4425         /*
4426          * We can't just convert len to max_blocks because
4427          * If blocksize = 4096 offset = 3072 and len = 2048
4428          */
4429         max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
4430                 - map.m_lblk;
4431         /*
4432          * credits to insert 1 extent into extent tree
4433          */
4434         credits = ext4_chunk_trans_blocks(inode, max_blocks);
4435         mutex_lock(&inode->i_mutex);
4436         ret = inode_newsize_ok(inode, (len + offset));
4437         if (ret) {
4438                 mutex_unlock(&inode->i_mutex);
4439                 trace_ext4_fallocate_exit(inode, offset, max_blocks, ret);
4440                 return ret;
4441         }
4442         flags = EXT4_GET_BLOCKS_CREATE_UNINIT_EXT;
4443         if (mode & FALLOC_FL_KEEP_SIZE)
4444                 flags |= EXT4_GET_BLOCKS_KEEP_SIZE;
4445         /*
4446          * Don't normalize the request if it can fit in one extent so
4447          * that it doesn't get unnecessarily split into multiple
4448          * extents.
4449          */
4450         if (len <= EXT_UNINIT_MAX_LEN << blkbits)
4451                 flags |= EXT4_GET_BLOCKS_NO_NORMALIZE;
4452 retry:
4453         while (ret >= 0 && ret < max_blocks) {
4454                 map.m_lblk = map.m_lblk + ret;
4455                 map.m_len = max_blocks = max_blocks - ret;
4456                 handle = ext4_journal_start(inode, credits);
4457                 if (IS_ERR(handle)) {
4458                         ret = PTR_ERR(handle);
4459                         break;
4460                 }
4461                 ret = ext4_map_blocks(handle, inode, &map, flags);
4462                 if (ret <= 0) {
4463 #ifdef EXT4FS_DEBUG
4464                         WARN_ON(ret <= 0);
4465                         printk(KERN_ERR "%s: ext4_ext_map_blocks "
4466                                     "returned error inode#%lu, block=%u, "
4467                                     "max_blocks=%u", __func__,
4468                                     inode->i_ino, map.m_lblk, max_blocks);
4469 #endif
4470                         ext4_mark_inode_dirty(handle, inode);
4471                         ret2 = ext4_journal_stop(handle);
4472                         break;
4473                 }
4474                 if ((map.m_lblk + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
4475                                                 blkbits) >> blkbits))
4476                         new_size = offset + len;
4477                 else
4478                         new_size = ((loff_t) map.m_lblk + ret) << blkbits;
4479
4480                 ext4_falloc_update_inode(inode, mode, new_size,
4481                                          (map.m_flags & EXT4_MAP_NEW));
4482                 ext4_mark_inode_dirty(handle, inode);
4483                 if ((file->f_flags & O_SYNC) && ret >= max_blocks)
4484                         ext4_handle_sync(handle);
4485                 ret2 = ext4_journal_stop(handle);
4486                 if (ret2)
4487                         break;
4488         }
4489         if (ret == -ENOSPC &&
4490                         ext4_should_retry_alloc(inode->i_sb, &retries)) {
4491                 ret = 0;
4492                 goto retry;
4493         }
4494         mutex_unlock(&inode->i_mutex);
4495         trace_ext4_fallocate_exit(inode, offset, max_blocks,
4496                                 ret > 0 ? ret2 : ret);
4497         return ret > 0 ? ret2 : ret;
4498 }
4499
4500 /*
4501  * This function convert a range of blocks to written extents
4502  * The caller of this function will pass the start offset and the size.
4503  * all unwritten extents within this range will be converted to
4504  * written extents.
4505  *
4506  * This function is called from the direct IO end io call back
4507  * function, to convert the fallocated extents after IO is completed.
4508  * Returns 0 on success.
4509  */
4510 int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
4511                                     ssize_t len)
4512 {
4513         handle_t *handle;
4514         unsigned int max_blocks;
4515         int ret = 0;
4516         int ret2 = 0;
4517         struct ext4_map_blocks map;
4518         unsigned int credits, blkbits = inode->i_blkbits;
4519
4520         map.m_lblk = offset >> blkbits;
4521         /*
4522          * We can't just convert len to max_blocks because
4523          * If blocksize = 4096 offset = 3072 and len = 2048
4524          */
4525         max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) -
4526                       map.m_lblk);
4527         /*
4528          * credits to insert 1 extent into extent tree
4529          */
4530         credits = ext4_chunk_trans_blocks(inode, max_blocks);
4531         while (ret >= 0 && ret < max_blocks) {
4532                 map.m_lblk += ret;
4533                 map.m_len = (max_blocks -= ret);
4534                 handle = ext4_journal_start(inode, credits);
4535                 if (IS_ERR(handle)) {
4536                         ret = PTR_ERR(handle);
4537                         break;
4538                 }
4539                 ret = ext4_map_blocks(handle, inode, &map,
4540                                       EXT4_GET_BLOCKS_IO_CONVERT_EXT);
4541                 if (ret <= 0) {
4542                         WARN_ON(ret <= 0);
4543                         ext4_msg(inode->i_sb, KERN_ERR,
4544                                  "%s:%d: inode #%lu: block %u: len %u: "
4545                                  "ext4_ext_map_blocks returned %d",
4546                                  __func__, __LINE__, inode->i_ino, map.m_lblk,
4547                                  map.m_len, ret);
4548                 }
4549                 ext4_mark_inode_dirty(handle, inode);
4550                 ret2 = ext4_journal_stop(handle);
4551                 if (ret <= 0 || ret2 )
4552                         break;
4553         }
4554         return ret > 0 ? ret2 : ret;
4555 }
4556
4557 /*
4558  * Callback function called for each extent to gather FIEMAP information.
4559  */
4560 static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next,
4561                        struct ext4_ext_cache *newex, struct ext4_extent *ex,
4562                        void *data)
4563 {
4564         __u64   logical;
4565         __u64   physical;
4566         __u64   length;
4567         __u32   flags = 0;
4568         int             ret = 0;
4569         struct fiemap_extent_info *fieinfo = data;
4570         unsigned char blksize_bits;
4571
4572         blksize_bits = inode->i_sb->s_blocksize_bits;
4573         logical = (__u64)newex->ec_block << blksize_bits;
4574
4575         if (newex->ec_start == 0) {
4576                 /*
4577                  * No extent in extent-tree contains block @newex->ec_start,
4578                  * then the block may stay in 1)a hole or 2)delayed-extent.
4579                  *
4580                  * Holes or delayed-extents are processed as follows.
4581                  * 1. lookup dirty pages with specified range in pagecache.
4582                  *    If no page is got, then there is no delayed-extent and
4583                  *    return with EXT_CONTINUE.
4584                  * 2. find the 1st mapped buffer,
4585                  * 3. check if the mapped buffer is both in the request range
4586                  *    and a delayed buffer. If not, there is no delayed-extent,
4587                  *    then return.
4588                  * 4. a delayed-extent is found, the extent will be collected.
4589                  */
4590                 ext4_lblk_t     end = 0;
4591                 pgoff_t         last_offset;
4592                 pgoff_t         offset;
4593                 pgoff_t         index;
4594                 pgoff_t         start_index = 0;
4595                 struct page     **pages = NULL;
4596                 struct buffer_head *bh = NULL;
4597                 struct buffer_head *head = NULL;
4598                 unsigned int nr_pages = PAGE_SIZE / sizeof(struct page *);
4599
4600                 pages = kmalloc(PAGE_SIZE, GFP_KERNEL);
4601                 if (pages == NULL)
4602                         return -ENOMEM;
4603
4604                 offset = logical >> PAGE_SHIFT;
4605 repeat:
4606                 last_offset = offset;
4607                 head = NULL;
4608                 ret = find_get_pages_tag(inode->i_mapping, &offset,
4609                                         PAGECACHE_TAG_DIRTY, nr_pages, pages);
4610
4611                 if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4612                         /* First time, try to find a mapped buffer. */
4613                         if (ret == 0) {
4614 out:
4615                                 for (index = 0; index < ret; index++)
4616                                         page_cache_release(pages[index]);
4617                                 /* just a hole. */
4618                                 kfree(pages);
4619                                 return EXT_CONTINUE;
4620                         }
4621                         index = 0;
4622
4623 next_page:
4624                         /* Try to find the 1st mapped buffer. */
4625                         end = ((__u64)pages[index]->index << PAGE_SHIFT) >>
4626                                   blksize_bits;
4627                         if (!page_has_buffers(pages[index]))
4628                                 goto out;
4629                         head = page_buffers(pages[index]);
4630                         if (!head)
4631                                 goto out;
4632
4633                         index++;
4634                         bh = head;
4635                         do {
4636                                 if (end >= newex->ec_block +
4637                                         newex->ec_len)
4638                                         /* The buffer is out of
4639                                          * the request range.
4640                                          */
4641                                         goto out;
4642
4643                                 if (buffer_mapped(bh) &&
4644                                     end >= newex->ec_block) {
4645                                         start_index = index - 1;
4646                                         /* get the 1st mapped buffer. */
4647                                         goto found_mapped_buffer;
4648                                 }
4649
4650                                 bh = bh->b_this_page;
4651                                 end++;
4652                         } while (bh != head);
4653
4654                         /* No mapped buffer in the range found in this page,
4655                          * We need to look up next page.
4656                          */
4657                         if (index >= ret) {
4658                                 /* There is no page left, but we need to limit
4659                                  * newex->ec_len.
4660                                  */
4661                                 newex->ec_len = end - newex->ec_block;
4662                                 goto out;
4663                         }
4664                         goto next_page;
4665                 } else {
4666                         /*Find contiguous delayed buffers. */
4667                         if (ret > 0 && pages[0]->index == last_offset)
4668                                 head = page_buffers(pages[0]);
4669                         bh = head;
4670                         index = 1;
4671                         start_index = 0;
4672                 }
4673
4674 found_mapped_buffer:
4675                 if (bh != NULL && buffer_delay(bh)) {
4676                         /* 1st or contiguous delayed buffer found. */
4677                         if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4678                                 /*
4679                                  * 1st delayed buffer found, record
4680                                  * the start of extent.
4681                                  */
4682                                 flags |= FIEMAP_EXTENT_DELALLOC;
4683                                 newex->ec_block = end;
4684                                 logical = (__u64)end << blksize_bits;
4685                         }
4686                         /* Find contiguous delayed buffers. */
4687                         do {
4688                                 if (!buffer_delay(bh))
4689                                         goto found_delayed_extent;
4690                                 bh = bh->b_this_page;
4691                                 end++;
4692                         } while (bh != head);
4693
4694                         for (; index < ret; index++) {
4695                                 if (!page_has_buffers(pages[index])) {
4696                                         bh = NULL;
4697                                         break;
4698                                 }
4699                                 head = page_buffers(pages[index]);
4700                                 if (!head) {
4701                                         bh = NULL;
4702                                         break;
4703                                 }
4704
4705                                 if (pages[index]->index !=
4706                                     pages[start_index]->index + index
4707                                     - start_index) {
4708                                         /* Blocks are not contiguous. */
4709                                         bh = NULL;
4710                                         break;
4711                                 }
4712                                 bh = head;
4713                                 do {
4714                                         if (!buffer_delay(bh))
4715                                                 /* Delayed-extent ends. */
4716                                                 goto found_delayed_extent;
4717                                         bh = bh->b_this_page;
4718                                         end++;
4719                                 } while (bh != head);
4720                         }
4721                 } else if (!(flags & FIEMAP_EXTENT_DELALLOC))
4722                         /* a hole found. */
4723                         goto out;
4724
4725 found_delayed_extent:
4726                 newex->ec_len = min(end - newex->ec_block,
4727                                                 (ext4_lblk_t)EXT_INIT_MAX_LEN);
4728                 if (ret == nr_pages && bh != NULL &&
4729                         newex->ec_len < EXT_INIT_MAX_LEN &&
4730                         buffer_delay(bh)) {
4731                         /* Have not collected an extent and continue. */
4732                         for (index = 0; index < ret; index++)
4733                                 page_cache_release(pages[index]);
4734                         goto repeat;
4735                 }
4736
4737                 for (index = 0; index < ret; index++)
4738                         page_cache_release(pages[index]);
4739                 kfree(pages);
4740         }
4741
4742         physical = (__u64)newex->ec_start << blksize_bits;
4743         length =   (__u64)newex->ec_len << blksize_bits;
4744
4745         if (ex && ext4_ext_is_uninitialized(ex))
4746                 flags |= FIEMAP_EXTENT_UNWRITTEN;
4747
4748         if (next == EXT_MAX_BLOCKS)
4749                 flags |= FIEMAP_EXTENT_LAST;
4750
4751         ret = fiemap_fill_next_extent(fieinfo, logical, physical,
4752                                         length, flags);
4753         if (ret < 0)
4754                 return ret;
4755         if (ret == 1)
4756                 return EXT_BREAK;
4757         return EXT_CONTINUE;
4758 }
4759 /* fiemap flags we can handle specified here */
4760 #define EXT4_FIEMAP_FLAGS       (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
4761
4762 static int ext4_xattr_fiemap(struct inode *inode,
4763                                 struct fiemap_extent_info *fieinfo)
4764 {
4765         __u64 physical = 0;
4766         __u64 length;
4767         __u32 flags = FIEMAP_EXTENT_LAST;
4768         int blockbits = inode->i_sb->s_blocksize_bits;
4769         int error = 0;
4770
4771         /* in-inode? */
4772         if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
4773                 struct ext4_iloc iloc;
4774                 int offset;     /* offset of xattr in inode */
4775
4776                 error = ext4_get_inode_loc(inode, &iloc);
4777                 if (error)
4778                         return error;
4779                 physical = iloc.bh->b_blocknr << blockbits;
4780                 offset = EXT4_GOOD_OLD_INODE_SIZE +
4781                                 EXT4_I(inode)->i_extra_isize;
4782                 physical += offset;
4783                 length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
4784                 flags |= FIEMAP_EXTENT_DATA_INLINE;
4785                 brelse(iloc.bh);
4786         } else { /* external block */
4787                 physical = EXT4_I(inode)->i_file_acl << blockbits;
4788                 length = inode->i_sb->s_blocksize;
4789         }
4790
4791         if (physical)
4792                 error = fiemap_fill_next_extent(fieinfo, 0, physical,
4793                                                 length, flags);
4794         return (error < 0 ? error : 0);
4795 }
4796
4797 /*
4798  * ext4_ext_punch_hole
4799  *
4800  * Punches a hole of "length" bytes in a file starting
4801  * at byte "offset"
4802  *
4803  * @inode:  The inode of the file to punch a hole in
4804  * @offset: The starting byte offset of the hole
4805  * @length: The length of the hole
4806  *
4807  * Returns the number of blocks removed or negative on err
4808  */
4809 int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length)
4810 {
4811         struct inode *inode = file->f_path.dentry->d_inode;
4812         struct super_block *sb = inode->i_sb;
4813         ext4_lblk_t first_block, stop_block;
4814         struct address_space *mapping = inode->i_mapping;
4815         handle_t *handle;
4816         loff_t first_page, last_page, page_len;
4817         loff_t first_page_offset, last_page_offset;
4818         int credits, err = 0;
4819
4820         /* No need to punch hole beyond i_size */
4821         if (offset >= inode->i_size)
4822                 return 0;
4823
4824         /*
4825          * If the hole extends beyond i_size, set the hole
4826          * to end after the page that contains i_size
4827          */
4828         if (offset + length > inode->i_size) {
4829                 length = inode->i_size +
4830                    PAGE_CACHE_SIZE - (inode->i_size & (PAGE_CACHE_SIZE - 1)) -
4831                    offset;
4832         }
4833
4834         first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
4835         last_page = (offset + length) >> PAGE_CACHE_SHIFT;
4836
4837         first_page_offset = first_page << PAGE_CACHE_SHIFT;
4838         last_page_offset = last_page << PAGE_CACHE_SHIFT;
4839
4840         /*
4841          * Write out all dirty pages to avoid race conditions
4842          * Then release them.
4843          */
4844         if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
4845                 err = filemap_write_and_wait_range(mapping,
4846                         offset, offset + length - 1);
4847
4848                 if (err)
4849                         return err;
4850         }
4851
4852         /* Now release the pages */
4853         if (last_page_offset > first_page_offset) {
4854                 truncate_pagecache_range(inode, first_page_offset,
4855                                          last_page_offset - 1);
4856         }
4857
4858         /* finish any pending end_io work */
4859         ext4_flush_completed_IO(inode);
4860
4861         credits = ext4_writepage_trans_blocks(inode);
4862         handle = ext4_journal_start(inode, credits);
4863         if (IS_ERR(handle))
4864                 return PTR_ERR(handle);
4865
4866
4867         /*
4868          * Now we need to zero out the non-page-aligned data in the
4869          * pages at the start and tail of the hole, and unmap the buffer
4870          * heads for the block aligned regions of the page that were
4871          * completely zeroed.
4872          */
4873         if (first_page > last_page) {
4874                 /*
4875                  * If the file space being truncated is contained within a page
4876                  * just zero out and unmap the middle of that page
4877                  */
4878                 err = ext4_discard_partial_page_buffers(handle,
4879                         mapping, offset, length, 0);
4880
4881                 if (err)
4882                         goto out;
4883         } else {
4884                 /*
4885                  * zero out and unmap the partial page that contains
4886                  * the start of the hole
4887                  */
4888                 page_len  = first_page_offset - offset;
4889                 if (page_len > 0) {
4890                         err = ext4_discard_partial_page_buffers(handle, mapping,
4891                                                    offset, page_len, 0);
4892                         if (err)
4893                                 goto out;
4894                 }
4895
4896                 /*
4897                  * zero out and unmap the partial page that contains
4898                  * the end of the hole
4899                  */
4900                 page_len = offset + length - last_page_offset;
4901                 if (page_len > 0) {
4902                         err = ext4_discard_partial_page_buffers(handle, mapping,
4903                                         last_page_offset, page_len, 0);
4904                         if (err)
4905                                 goto out;
4906                 }
4907         }
4908
4909         /*
4910          * If i_size is contained in the last page, we need to
4911          * unmap and zero the partial page after i_size
4912          */
4913         if (inode->i_size >> PAGE_CACHE_SHIFT == last_page &&
4914            inode->i_size % PAGE_CACHE_SIZE != 0) {
4915
4916                 page_len = PAGE_CACHE_SIZE -
4917                         (inode->i_size & (PAGE_CACHE_SIZE - 1));
4918
4919                 if (page_len > 0) {
4920                         err = ext4_discard_partial_page_buffers(handle,
4921                           mapping, inode->i_size, page_len, 0);
4922
4923                         if (err)
4924                                 goto out;
4925                 }
4926         }
4927
4928         first_block = (offset + sb->s_blocksize - 1) >>
4929                 EXT4_BLOCK_SIZE_BITS(sb);
4930         stop_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb);
4931
4932         /* If there are no blocks to remove, return now */
4933         if (first_block >= stop_block)
4934                 goto out;
4935
4936         down_write(&EXT4_I(inode)->i_data_sem);
4937         ext4_ext_invalidate_cache(inode);
4938         ext4_discard_preallocations(inode);
4939
4940         err = ext4_ext_remove_space(inode, first_block, stop_block - 1);
4941
4942         ext4_ext_invalidate_cache(inode);
4943         ext4_discard_preallocations(inode);
4944
4945         if (IS_SYNC(inode))
4946                 ext4_handle_sync(handle);
4947
4948         up_write(&EXT4_I(inode)->i_data_sem);
4949
4950 out:
4951         inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4952         ext4_mark_inode_dirty(handle, inode);
4953         ext4_journal_stop(handle);
4954         return err;
4955 }
4956 int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
4957                 __u64 start, __u64 len)
4958 {
4959         ext4_lblk_t start_blk;
4960         int error = 0;
4961
4962         /* fallback to generic here if not in extents fmt */
4963         if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
4964                 return generic_block_fiemap(inode, fieinfo, start, len,
4965                         ext4_get_block);
4966
4967         if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
4968                 return -EBADR;
4969
4970         if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
4971                 error = ext4_xattr_fiemap(inode, fieinfo);
4972         } else {
4973                 ext4_lblk_t len_blks;
4974                 __u64 last_blk;
4975
4976                 start_blk = start >> inode->i_sb->s_blocksize_bits;
4977                 last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
4978                 if (last_blk >= EXT_MAX_BLOCKS)
4979                         last_blk = EXT_MAX_BLOCKS-1;
4980                 len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
4981
4982                 /*
4983                  * Walk the extent tree gathering extent information.
4984                  * ext4_ext_fiemap_cb will push extents back to user.
4985                  */
4986                 error = ext4_ext_walk_space(inode, start_blk, len_blks,
4987                                           ext4_ext_fiemap_cb, fieinfo);
4988         }
4989
4990         return error;
4991 }