]>
Commit | Line | Data |
---|---|---|
a86c6181 AT |
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 | ||
a86c6181 AT |
32 | #include <linux/fs.h> |
33 | #include <linux/time.h> | |
cd02ff0b | 34 | #include <linux/jbd2.h> |
a86c6181 AT |
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> | |
a86c6181 | 40 | #include <asm/uaccess.h> |
6873fa0d | 41 | #include <linux/fiemap.h> |
3dcf5451 | 42 | #include "ext4_jbd2.h" |
4a092d73 | 43 | #include "ext4_extents.h" |
f19d5870 | 44 | #include "xattr.h" |
a86c6181 | 45 | |
0562e0ba JZ |
46 | #include <trace/events/ext4.h> |
47 | ||
5f95d21f LC |
48 | /* |
49 | * used by extent splitting. | |
50 | */ | |
51 | #define EXT4_EXT_MAY_ZEROOUT 0x1 /* safe to zeroout if split fails \ | |
52 | due to ENOSPC */ | |
556615dc LC |
53 | #define EXT4_EXT_MARK_UNWRIT1 0x2 /* mark first half unwritten */ |
54 | #define EXT4_EXT_MARK_UNWRIT2 0x4 /* mark second half unwritten */ | |
5f95d21f | 55 | |
dee1f973 DM |
56 | #define EXT4_EXT_DATA_VALID1 0x8 /* first half contains valid data */ |
57 | #define EXT4_EXT_DATA_VALID2 0x10 /* second half contains valid data */ | |
58 | ||
7ac5990d DW |
59 | static __le32 ext4_extent_block_csum(struct inode *inode, |
60 | struct ext4_extent_header *eh) | |
61 | { | |
62 | struct ext4_inode_info *ei = EXT4_I(inode); | |
63 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); | |
64 | __u32 csum; | |
65 | ||
66 | csum = ext4_chksum(sbi, ei->i_csum_seed, (__u8 *)eh, | |
67 | EXT4_EXTENT_TAIL_OFFSET(eh)); | |
68 | return cpu_to_le32(csum); | |
69 | } | |
70 | ||
71 | static int ext4_extent_block_csum_verify(struct inode *inode, | |
72 | struct ext4_extent_header *eh) | |
73 | { | |
74 | struct ext4_extent_tail *et; | |
75 | ||
76 | if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, | |
77 | EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) | |
78 | return 1; | |
79 | ||
80 | et = find_ext4_extent_tail(eh); | |
81 | if (et->et_checksum != ext4_extent_block_csum(inode, eh)) | |
82 | return 0; | |
83 | return 1; | |
84 | } | |
85 | ||
86 | static void ext4_extent_block_csum_set(struct inode *inode, | |
87 | struct ext4_extent_header *eh) | |
88 | { | |
89 | struct ext4_extent_tail *et; | |
90 | ||
91 | if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, | |
92 | EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) | |
93 | return; | |
94 | ||
95 | et = find_ext4_extent_tail(eh); | |
96 | et->et_checksum = ext4_extent_block_csum(inode, eh); | |
97 | } | |
98 | ||
d583fb87 AH |
99 | static int ext4_split_extent(handle_t *handle, |
100 | struct inode *inode, | |
101 | struct ext4_ext_path *path, | |
102 | struct ext4_map_blocks *map, | |
103 | int split_flag, | |
104 | int flags); | |
105 | ||
5f95d21f LC |
106 | static int ext4_split_extent_at(handle_t *handle, |
107 | struct inode *inode, | |
108 | struct ext4_ext_path *path, | |
109 | ext4_lblk_t split, | |
110 | int split_flag, | |
111 | int flags); | |
112 | ||
91dd8c11 | 113 | static int ext4_find_delayed_extent(struct inode *inode, |
69eb33dc | 114 | struct extent_status *newes); |
91dd8c11 | 115 | |
487caeef JK |
116 | static int ext4_ext_truncate_extend_restart(handle_t *handle, |
117 | struct inode *inode, | |
118 | int needed) | |
a86c6181 AT |
119 | { |
120 | int err; | |
121 | ||
0390131b FM |
122 | if (!ext4_handle_valid(handle)) |
123 | return 0; | |
a86c6181 | 124 | if (handle->h_buffer_credits > needed) |
9102e4fa SF |
125 | return 0; |
126 | err = ext4_journal_extend(handle, needed); | |
0123c939 | 127 | if (err <= 0) |
9102e4fa | 128 | return err; |
487caeef | 129 | err = ext4_truncate_restart_trans(handle, inode, needed); |
0617b83f DM |
130 | if (err == 0) |
131 | err = -EAGAIN; | |
487caeef JK |
132 | |
133 | return err; | |
a86c6181 AT |
134 | } |
135 | ||
136 | /* | |
137 | * could return: | |
138 | * - EROFS | |
139 | * - ENOMEM | |
140 | */ | |
141 | static int ext4_ext_get_access(handle_t *handle, struct inode *inode, | |
142 | struct ext4_ext_path *path) | |
143 | { | |
144 | if (path->p_bh) { | |
145 | /* path points to block */ | |
5d601255 | 146 | BUFFER_TRACE(path->p_bh, "get_write_access"); |
a86c6181 AT |
147 | return ext4_journal_get_write_access(handle, path->p_bh); |
148 | } | |
149 | /* path points to leaf/index in inode body */ | |
150 | /* we use in-core data, no need to protect them */ | |
151 | return 0; | |
152 | } | |
153 | ||
154 | /* | |
155 | * could return: | |
156 | * - EROFS | |
157 | * - ENOMEM | |
158 | * - EIO | |
159 | */ | |
2656497b DW |
160 | int __ext4_ext_dirty(const char *where, unsigned int line, handle_t *handle, |
161 | struct inode *inode, struct ext4_ext_path *path) | |
a86c6181 AT |
162 | { |
163 | int err; | |
4b1f1660 DM |
164 | |
165 | WARN_ON(!rwsem_is_locked(&EXT4_I(inode)->i_data_sem)); | |
a86c6181 | 166 | if (path->p_bh) { |
7ac5990d | 167 | ext4_extent_block_csum_set(inode, ext_block_hdr(path->p_bh)); |
a86c6181 | 168 | /* path points to block */ |
9ea7a0df TT |
169 | err = __ext4_handle_dirty_metadata(where, line, handle, |
170 | inode, path->p_bh); | |
a86c6181 AT |
171 | } else { |
172 | /* path points to leaf/index in inode body */ | |
173 | err = ext4_mark_inode_dirty(handle, inode); | |
174 | } | |
175 | return err; | |
176 | } | |
177 | ||
f65e6fba | 178 | static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, |
a86c6181 | 179 | struct ext4_ext_path *path, |
725d26d3 | 180 | ext4_lblk_t block) |
a86c6181 | 181 | { |
a86c6181 | 182 | if (path) { |
81fdbb4a | 183 | int depth = path->p_depth; |
a86c6181 | 184 | struct ext4_extent *ex; |
a86c6181 | 185 | |
ad4fb9ca KM |
186 | /* |
187 | * Try to predict block placement assuming that we are | |
188 | * filling in a file which will eventually be | |
189 | * non-sparse --- i.e., in the case of libbfd writing | |
190 | * an ELF object sections out-of-order but in a way | |
191 | * the eventually results in a contiguous object or | |
192 | * executable file, or some database extending a table | |
193 | * space file. However, this is actually somewhat | |
194 | * non-ideal if we are writing a sparse file such as | |
195 | * qemu or KVM writing a raw image file that is going | |
196 | * to stay fairly sparse, since it will end up | |
197 | * fragmenting the file system's free space. Maybe we | |
198 | * should have some hueristics or some way to allow | |
199 | * userspace to pass a hint to file system, | |
b8d6568a | 200 | * especially if the latter case turns out to be |
ad4fb9ca KM |
201 | * common. |
202 | */ | |
7e028976 | 203 | ex = path[depth].p_ext; |
ad4fb9ca KM |
204 | if (ex) { |
205 | ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex); | |
206 | ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block); | |
207 | ||
208 | if (block > ext_block) | |
209 | return ext_pblk + (block - ext_block); | |
210 | else | |
211 | return ext_pblk - (ext_block - block); | |
212 | } | |
a86c6181 | 213 | |
d0d856e8 RD |
214 | /* it looks like index is empty; |
215 | * try to find starting block from index itself */ | |
a86c6181 AT |
216 | if (path[depth].p_bh) |
217 | return path[depth].p_bh->b_blocknr; | |
218 | } | |
219 | ||
220 | /* OK. use inode's group */ | |
f86186b4 | 221 | return ext4_inode_to_goal_block(inode); |
a86c6181 AT |
222 | } |
223 | ||
654b4908 AK |
224 | /* |
225 | * Allocation for a meta data block | |
226 | */ | |
f65e6fba | 227 | static ext4_fsblk_t |
654b4908 | 228 | ext4_ext_new_meta_block(handle_t *handle, struct inode *inode, |
a86c6181 | 229 | struct ext4_ext_path *path, |
55f020db | 230 | struct ext4_extent *ex, int *err, unsigned int flags) |
a86c6181 | 231 | { |
f65e6fba | 232 | ext4_fsblk_t goal, newblock; |
a86c6181 AT |
233 | |
234 | goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block)); | |
55f020db AH |
235 | newblock = ext4_new_meta_blocks(handle, inode, goal, flags, |
236 | NULL, err); | |
a86c6181 AT |
237 | return newblock; |
238 | } | |
239 | ||
55ad63bf | 240 | static inline int ext4_ext_space_block(struct inode *inode, int check) |
a86c6181 AT |
241 | { |
242 | int size; | |
243 | ||
244 | size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header)) | |
245 | / sizeof(struct ext4_extent); | |
bbf2f9fb | 246 | #ifdef AGGRESSIVE_TEST |
02dc62fb YY |
247 | if (!check && size > 6) |
248 | size = 6; | |
a86c6181 AT |
249 | #endif |
250 | return size; | |
251 | } | |
252 | ||
55ad63bf | 253 | static inline int ext4_ext_space_block_idx(struct inode *inode, int check) |
a86c6181 AT |
254 | { |
255 | int size; | |
256 | ||
257 | size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header)) | |
258 | / sizeof(struct ext4_extent_idx); | |
bbf2f9fb | 259 | #ifdef AGGRESSIVE_TEST |
02dc62fb YY |
260 | if (!check && size > 5) |
261 | size = 5; | |
a86c6181 AT |
262 | #endif |
263 | return size; | |
264 | } | |
265 | ||
55ad63bf | 266 | static inline int ext4_ext_space_root(struct inode *inode, int check) |
a86c6181 AT |
267 | { |
268 | int size; | |
269 | ||
270 | size = sizeof(EXT4_I(inode)->i_data); | |
271 | size -= sizeof(struct ext4_extent_header); | |
272 | size /= sizeof(struct ext4_extent); | |
bbf2f9fb | 273 | #ifdef AGGRESSIVE_TEST |
02dc62fb YY |
274 | if (!check && size > 3) |
275 | size = 3; | |
a86c6181 AT |
276 | #endif |
277 | return size; | |
278 | } | |
279 | ||
55ad63bf | 280 | static inline int ext4_ext_space_root_idx(struct inode *inode, int check) |
a86c6181 AT |
281 | { |
282 | int size; | |
283 | ||
284 | size = sizeof(EXT4_I(inode)->i_data); | |
285 | size -= sizeof(struct ext4_extent_header); | |
286 | size /= sizeof(struct ext4_extent_idx); | |
bbf2f9fb | 287 | #ifdef AGGRESSIVE_TEST |
02dc62fb YY |
288 | if (!check && size > 4) |
289 | size = 4; | |
a86c6181 AT |
290 | #endif |
291 | return size; | |
292 | } | |
293 | ||
d2a17637 MC |
294 | /* |
295 | * Calculate the number of metadata blocks needed | |
296 | * to allocate @blocks | |
297 | * Worse case is one block per extent | |
298 | */ | |
01f49d0b | 299 | int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock) |
d2a17637 | 300 | { |
9d0be502 | 301 | struct ext4_inode_info *ei = EXT4_I(inode); |
81fdbb4a | 302 | int idxs; |
d2a17637 | 303 | |
9d0be502 TT |
304 | idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header)) |
305 | / sizeof(struct ext4_extent_idx)); | |
d2a17637 MC |
306 | |
307 | /* | |
9d0be502 TT |
308 | * If the new delayed allocation block is contiguous with the |
309 | * previous da block, it can share index blocks with the | |
310 | * previous block, so we only need to allocate a new index | |
311 | * block every idxs leaf blocks. At ldxs**2 blocks, we need | |
312 | * an additional index block, and at ldxs**3 blocks, yet | |
313 | * another index blocks. | |
d2a17637 | 314 | */ |
9d0be502 TT |
315 | if (ei->i_da_metadata_calc_len && |
316 | ei->i_da_metadata_calc_last_lblock+1 == lblock) { | |
81fdbb4a YY |
317 | int num = 0; |
318 | ||
9d0be502 TT |
319 | if ((ei->i_da_metadata_calc_len % idxs) == 0) |
320 | num++; | |
321 | if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0) | |
322 | num++; | |
323 | if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) { | |
324 | num++; | |
325 | ei->i_da_metadata_calc_len = 0; | |
326 | } else | |
327 | ei->i_da_metadata_calc_len++; | |
328 | ei->i_da_metadata_calc_last_lblock++; | |
329 | return num; | |
330 | } | |
d2a17637 | 331 | |
9d0be502 TT |
332 | /* |
333 | * In the worst case we need a new set of index blocks at | |
334 | * every level of the inode's extent tree. | |
335 | */ | |
336 | ei->i_da_metadata_calc_len = 1; | |
337 | ei->i_da_metadata_calc_last_lblock = lblock; | |
338 | return ext_depth(inode) + 1; | |
d2a17637 MC |
339 | } |
340 | ||
c29c0ae7 AT |
341 | static int |
342 | ext4_ext_max_entries(struct inode *inode, int depth) | |
343 | { | |
344 | int max; | |
345 | ||
346 | if (depth == ext_depth(inode)) { | |
347 | if (depth == 0) | |
55ad63bf | 348 | max = ext4_ext_space_root(inode, 1); |
c29c0ae7 | 349 | else |
55ad63bf | 350 | max = ext4_ext_space_root_idx(inode, 1); |
c29c0ae7 AT |
351 | } else { |
352 | if (depth == 0) | |
55ad63bf | 353 | max = ext4_ext_space_block(inode, 1); |
c29c0ae7 | 354 | else |
55ad63bf | 355 | max = ext4_ext_space_block_idx(inode, 1); |
c29c0ae7 AT |
356 | } |
357 | ||
358 | return max; | |
359 | } | |
360 | ||
56b19868 AK |
361 | static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext) |
362 | { | |
bf89d16f | 363 | ext4_fsblk_t block = ext4_ext_pblock(ext); |
56b19868 | 364 | int len = ext4_ext_get_actual_len(ext); |
5946d089 EG |
365 | ext4_lblk_t lblock = le32_to_cpu(ext->ee_block); |
366 | ext4_lblk_t last = lblock + len - 1; | |
e84a26ce | 367 | |
5946d089 | 368 | if (lblock > last) |
31d4f3a2 | 369 | return 0; |
6fd058f7 | 370 | return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len); |
56b19868 AK |
371 | } |
372 | ||
373 | static int ext4_valid_extent_idx(struct inode *inode, | |
374 | struct ext4_extent_idx *ext_idx) | |
375 | { | |
bf89d16f | 376 | ext4_fsblk_t block = ext4_idx_pblock(ext_idx); |
e84a26ce | 377 | |
6fd058f7 | 378 | return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1); |
56b19868 AK |
379 | } |
380 | ||
381 | static int ext4_valid_extent_entries(struct inode *inode, | |
382 | struct ext4_extent_header *eh, | |
383 | int depth) | |
384 | { | |
56b19868 AK |
385 | unsigned short entries; |
386 | if (eh->eh_entries == 0) | |
387 | return 1; | |
388 | ||
389 | entries = le16_to_cpu(eh->eh_entries); | |
390 | ||
391 | if (depth == 0) { | |
392 | /* leaf entries */ | |
81fdbb4a | 393 | struct ext4_extent *ext = EXT_FIRST_EXTENT(eh); |
5946d089 EG |
394 | struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; |
395 | ext4_fsblk_t pblock = 0; | |
396 | ext4_lblk_t lblock = 0; | |
397 | ext4_lblk_t prev = 0; | |
398 | int len = 0; | |
56b19868 AK |
399 | while (entries) { |
400 | if (!ext4_valid_extent(inode, ext)) | |
401 | return 0; | |
5946d089 EG |
402 | |
403 | /* Check for overlapping extents */ | |
404 | lblock = le32_to_cpu(ext->ee_block); | |
405 | len = ext4_ext_get_actual_len(ext); | |
406 | if ((lblock <= prev) && prev) { | |
407 | pblock = ext4_ext_pblock(ext); | |
408 | es->s_last_error_block = cpu_to_le64(pblock); | |
409 | return 0; | |
410 | } | |
56b19868 AK |
411 | ext++; |
412 | entries--; | |
5946d089 | 413 | prev = lblock + len - 1; |
56b19868 AK |
414 | } |
415 | } else { | |
81fdbb4a | 416 | struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh); |
56b19868 AK |
417 | while (entries) { |
418 | if (!ext4_valid_extent_idx(inode, ext_idx)) | |
419 | return 0; | |
420 | ext_idx++; | |
421 | entries--; | |
422 | } | |
423 | } | |
424 | return 1; | |
425 | } | |
426 | ||
c398eda0 TT |
427 | static int __ext4_ext_check(const char *function, unsigned int line, |
428 | struct inode *inode, struct ext4_extent_header *eh, | |
c349179b | 429 | int depth, ext4_fsblk_t pblk) |
c29c0ae7 AT |
430 | { |
431 | const char *error_msg; | |
432 | int max = 0; | |
433 | ||
434 | if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) { | |
435 | error_msg = "invalid magic"; | |
436 | goto corrupted; | |
437 | } | |
438 | if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) { | |
439 | error_msg = "unexpected eh_depth"; | |
440 | goto corrupted; | |
441 | } | |
442 | if (unlikely(eh->eh_max == 0)) { | |
443 | error_msg = "invalid eh_max"; | |
444 | goto corrupted; | |
445 | } | |
446 | max = ext4_ext_max_entries(inode, depth); | |
447 | if (unlikely(le16_to_cpu(eh->eh_max) > max)) { | |
448 | error_msg = "too large eh_max"; | |
449 | goto corrupted; | |
450 | } | |
451 | if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) { | |
452 | error_msg = "invalid eh_entries"; | |
453 | goto corrupted; | |
454 | } | |
56b19868 AK |
455 | if (!ext4_valid_extent_entries(inode, eh, depth)) { |
456 | error_msg = "invalid extent entries"; | |
457 | goto corrupted; | |
458 | } | |
7ac5990d DW |
459 | /* Verify checksum on non-root extent tree nodes */ |
460 | if (ext_depth(inode) != depth && | |
461 | !ext4_extent_block_csum_verify(inode, eh)) { | |
462 | error_msg = "extent tree corrupted"; | |
463 | goto corrupted; | |
464 | } | |
c29c0ae7 AT |
465 | return 0; |
466 | ||
467 | corrupted: | |
c398eda0 | 468 | ext4_error_inode(inode, function, line, 0, |
c349179b TT |
469 | "pblk %llu bad header/extent: %s - magic %x, " |
470 | "entries %u, max %u(%u), depth %u(%u)", | |
471 | (unsigned long long) pblk, error_msg, | |
472 | le16_to_cpu(eh->eh_magic), | |
473 | le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max), | |
474 | max, le16_to_cpu(eh->eh_depth), depth); | |
c29c0ae7 AT |
475 | return -EIO; |
476 | } | |
477 | ||
c349179b TT |
478 | #define ext4_ext_check(inode, eh, depth, pblk) \ |
479 | __ext4_ext_check(__func__, __LINE__, (inode), (eh), (depth), (pblk)) | |
c29c0ae7 | 480 | |
7a262f7c AK |
481 | int ext4_ext_check_inode(struct inode *inode) |
482 | { | |
c349179b | 483 | return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode), 0); |
7a262f7c AK |
484 | } |
485 | ||
7d7ea89e TT |
486 | static struct buffer_head * |
487 | __read_extent_tree_block(const char *function, unsigned int line, | |
107a7bd3 TT |
488 | struct inode *inode, ext4_fsblk_t pblk, int depth, |
489 | int flags) | |
f8489128 | 490 | { |
7d7ea89e TT |
491 | struct buffer_head *bh; |
492 | int err; | |
493 | ||
494 | bh = sb_getblk(inode->i_sb, pblk); | |
495 | if (unlikely(!bh)) | |
496 | return ERR_PTR(-ENOMEM); | |
f8489128 | 497 | |
7d7ea89e TT |
498 | if (!bh_uptodate_or_lock(bh)) { |
499 | trace_ext4_ext_load_extent(inode, pblk, _RET_IP_); | |
500 | err = bh_submit_read(bh); | |
501 | if (err < 0) | |
502 | goto errout; | |
503 | } | |
7869a4a6 | 504 | if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE)) |
7d7ea89e TT |
505 | return bh; |
506 | err = __ext4_ext_check(function, line, inode, | |
c349179b | 507 | ext_block_hdr(bh), depth, pblk); |
7d7ea89e TT |
508 | if (err) |
509 | goto errout; | |
f8489128 | 510 | set_buffer_verified(bh); |
107a7bd3 TT |
511 | /* |
512 | * If this is a leaf block, cache all of its entries | |
513 | */ | |
514 | if (!(flags & EXT4_EX_NOCACHE) && depth == 0) { | |
515 | struct ext4_extent_header *eh = ext_block_hdr(bh); | |
516 | struct ext4_extent *ex = EXT_FIRST_EXTENT(eh); | |
517 | ext4_lblk_t prev = 0; | |
518 | int i; | |
519 | ||
520 | for (i = le16_to_cpu(eh->eh_entries); i > 0; i--, ex++) { | |
521 | unsigned int status = EXTENT_STATUS_WRITTEN; | |
522 | ext4_lblk_t lblk = le32_to_cpu(ex->ee_block); | |
523 | int len = ext4_ext_get_actual_len(ex); | |
524 | ||
525 | if (prev && (prev != lblk)) | |
526 | ext4_es_cache_extent(inode, prev, | |
527 | lblk - prev, ~0, | |
528 | EXTENT_STATUS_HOLE); | |
529 | ||
556615dc | 530 | if (ext4_ext_is_unwritten(ex)) |
107a7bd3 TT |
531 | status = EXTENT_STATUS_UNWRITTEN; |
532 | ext4_es_cache_extent(inode, lblk, len, | |
533 | ext4_ext_pblock(ex), status); | |
534 | prev = lblk + len; | |
535 | } | |
536 | } | |
7d7ea89e TT |
537 | return bh; |
538 | errout: | |
539 | put_bh(bh); | |
540 | return ERR_PTR(err); | |
541 | ||
f8489128 DW |
542 | } |
543 | ||
107a7bd3 TT |
544 | #define read_extent_tree_block(inode, pblk, depth, flags) \ |
545 | __read_extent_tree_block(__func__, __LINE__, (inode), (pblk), \ | |
546 | (depth), (flags)) | |
f8489128 | 547 | |
7869a4a6 TT |
548 | /* |
549 | * This function is called to cache a file's extent information in the | |
550 | * extent status tree | |
551 | */ | |
552 | int ext4_ext_precache(struct inode *inode) | |
553 | { | |
554 | struct ext4_inode_info *ei = EXT4_I(inode); | |
555 | struct ext4_ext_path *path = NULL; | |
556 | struct buffer_head *bh; | |
557 | int i = 0, depth, ret = 0; | |
558 | ||
559 | if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) | |
560 | return 0; /* not an extent-mapped inode */ | |
561 | ||
562 | down_read(&ei->i_data_sem); | |
563 | depth = ext_depth(inode); | |
564 | ||
565 | path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), | |
566 | GFP_NOFS); | |
567 | if (path == NULL) { | |
568 | up_read(&ei->i_data_sem); | |
569 | return -ENOMEM; | |
570 | } | |
571 | ||
572 | /* Don't cache anything if there are no external extent blocks */ | |
573 | if (depth == 0) | |
574 | goto out; | |
575 | path[0].p_hdr = ext_inode_hdr(inode); | |
576 | ret = ext4_ext_check(inode, path[0].p_hdr, depth, 0); | |
577 | if (ret) | |
578 | goto out; | |
579 | path[0].p_idx = EXT_FIRST_INDEX(path[0].p_hdr); | |
580 | while (i >= 0) { | |
581 | /* | |
582 | * If this is a leaf block or we've reached the end of | |
583 | * the index block, go up | |
584 | */ | |
585 | if ((i == depth) || | |
586 | path[i].p_idx > EXT_LAST_INDEX(path[i].p_hdr)) { | |
587 | brelse(path[i].p_bh); | |
588 | path[i].p_bh = NULL; | |
589 | i--; | |
590 | continue; | |
591 | } | |
592 | bh = read_extent_tree_block(inode, | |
593 | ext4_idx_pblock(path[i].p_idx++), | |
594 | depth - i - 1, | |
595 | EXT4_EX_FORCE_CACHE); | |
596 | if (IS_ERR(bh)) { | |
597 | ret = PTR_ERR(bh); | |
598 | break; | |
599 | } | |
600 | i++; | |
601 | path[i].p_bh = bh; | |
602 | path[i].p_hdr = ext_block_hdr(bh); | |
603 | path[i].p_idx = EXT_FIRST_INDEX(path[i].p_hdr); | |
604 | } | |
605 | ext4_set_inode_state(inode, EXT4_STATE_EXT_PRECACHED); | |
606 | out: | |
607 | up_read(&ei->i_data_sem); | |
608 | ext4_ext_drop_refs(path); | |
609 | kfree(path); | |
610 | return ret; | |
611 | } | |
612 | ||
a86c6181 AT |
613 | #ifdef EXT_DEBUG |
614 | static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path) | |
615 | { | |
616 | int k, l = path->p_depth; | |
617 | ||
618 | ext_debug("path:"); | |
619 | for (k = 0; k <= l; k++, path++) { | |
620 | if (path->p_idx) { | |
2ae02107 | 621 | ext_debug(" %d->%llu", le32_to_cpu(path->p_idx->ei_block), |
bf89d16f | 622 | ext4_idx_pblock(path->p_idx)); |
a86c6181 | 623 | } else if (path->p_ext) { |
553f9008 | 624 | ext_debug(" %d:[%d]%d:%llu ", |
a86c6181 | 625 | le32_to_cpu(path->p_ext->ee_block), |
556615dc | 626 | ext4_ext_is_unwritten(path->p_ext), |
a2df2a63 | 627 | ext4_ext_get_actual_len(path->p_ext), |
bf89d16f | 628 | ext4_ext_pblock(path->p_ext)); |
a86c6181 AT |
629 | } else |
630 | ext_debug(" []"); | |
631 | } | |
632 | ext_debug("\n"); | |
633 | } | |
634 | ||
635 | static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path) | |
636 | { | |
637 | int depth = ext_depth(inode); | |
638 | struct ext4_extent_header *eh; | |
639 | struct ext4_extent *ex; | |
640 | int i; | |
641 | ||
642 | if (!path) | |
643 | return; | |
644 | ||
645 | eh = path[depth].p_hdr; | |
646 | ex = EXT_FIRST_EXTENT(eh); | |
647 | ||
553f9008 M |
648 | ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino); |
649 | ||
a86c6181 | 650 | for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) { |
553f9008 | 651 | ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block), |
556615dc | 652 | ext4_ext_is_unwritten(ex), |
bf89d16f | 653 | ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex)); |
a86c6181 AT |
654 | } |
655 | ext_debug("\n"); | |
656 | } | |
1b16da77 YY |
657 | |
658 | static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path, | |
659 | ext4_fsblk_t newblock, int level) | |
660 | { | |
661 | int depth = ext_depth(inode); | |
662 | struct ext4_extent *ex; | |
663 | ||
664 | if (depth != level) { | |
665 | struct ext4_extent_idx *idx; | |
666 | idx = path[level].p_idx; | |
667 | while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) { | |
668 | ext_debug("%d: move %d:%llu in new index %llu\n", level, | |
669 | le32_to_cpu(idx->ei_block), | |
670 | ext4_idx_pblock(idx), | |
671 | newblock); | |
672 | idx++; | |
673 | } | |
674 | ||
675 | return; | |
676 | } | |
677 | ||
678 | ex = path[depth].p_ext; | |
679 | while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) { | |
680 | ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n", | |
681 | le32_to_cpu(ex->ee_block), | |
682 | ext4_ext_pblock(ex), | |
556615dc | 683 | ext4_ext_is_unwritten(ex), |
1b16da77 YY |
684 | ext4_ext_get_actual_len(ex), |
685 | newblock); | |
686 | ex++; | |
687 | } | |
688 | } | |
689 | ||
a86c6181 | 690 | #else |
af5bc92d TT |
691 | #define ext4_ext_show_path(inode, path) |
692 | #define ext4_ext_show_leaf(inode, path) | |
1b16da77 | 693 | #define ext4_ext_show_move(inode, path, newblock, level) |
a86c6181 AT |
694 | #endif |
695 | ||
b35905c1 | 696 | void ext4_ext_drop_refs(struct ext4_ext_path *path) |
a86c6181 AT |
697 | { |
698 | int depth = path->p_depth; | |
699 | int i; | |
700 | ||
701 | for (i = 0; i <= depth; i++, path++) | |
702 | if (path->p_bh) { | |
703 | brelse(path->p_bh); | |
704 | path->p_bh = NULL; | |
705 | } | |
706 | } | |
707 | ||
708 | /* | |
d0d856e8 RD |
709 | * ext4_ext_binsearch_idx: |
710 | * binary search for the closest index of the given block | |
c29c0ae7 | 711 | * the header must be checked before calling this |
a86c6181 AT |
712 | */ |
713 | static void | |
725d26d3 AK |
714 | ext4_ext_binsearch_idx(struct inode *inode, |
715 | struct ext4_ext_path *path, ext4_lblk_t block) | |
a86c6181 AT |
716 | { |
717 | struct ext4_extent_header *eh = path->p_hdr; | |
718 | struct ext4_extent_idx *r, *l, *m; | |
719 | ||
a86c6181 | 720 | |
bba90743 | 721 | ext_debug("binsearch for %u(idx): ", block); |
a86c6181 AT |
722 | |
723 | l = EXT_FIRST_INDEX(eh) + 1; | |
e9f410b1 | 724 | r = EXT_LAST_INDEX(eh); |
a86c6181 AT |
725 | while (l <= r) { |
726 | m = l + (r - l) / 2; | |
727 | if (block < le32_to_cpu(m->ei_block)) | |
728 | r = m - 1; | |
729 | else | |
730 | l = m + 1; | |
26d535ed DM |
731 | ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block), |
732 | m, le32_to_cpu(m->ei_block), | |
733 | r, le32_to_cpu(r->ei_block)); | |
a86c6181 AT |
734 | } |
735 | ||
736 | path->p_idx = l - 1; | |
4a3c3a51 | 737 | ext_debug(" -> %u->%lld ", le32_to_cpu(path->p_idx->ei_block), |
bf89d16f | 738 | ext4_idx_pblock(path->p_idx)); |
a86c6181 AT |
739 | |
740 | #ifdef CHECK_BINSEARCH | |
741 | { | |
742 | struct ext4_extent_idx *chix, *ix; | |
743 | int k; | |
744 | ||
745 | chix = ix = EXT_FIRST_INDEX(eh); | |
746 | for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) { | |
747 | if (k != 0 && | |
748 | le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) { | |
4776004f TT |
749 | printk(KERN_DEBUG "k=%d, ix=0x%p, " |
750 | "first=0x%p\n", k, | |
751 | ix, EXT_FIRST_INDEX(eh)); | |
752 | printk(KERN_DEBUG "%u <= %u\n", | |
a86c6181 AT |
753 | le32_to_cpu(ix->ei_block), |
754 | le32_to_cpu(ix[-1].ei_block)); | |
755 | } | |
756 | BUG_ON(k && le32_to_cpu(ix->ei_block) | |
8c55e204 | 757 | <= le32_to_cpu(ix[-1].ei_block)); |
a86c6181 AT |
758 | if (block < le32_to_cpu(ix->ei_block)) |
759 | break; | |
760 | chix = ix; | |
761 | } | |
762 | BUG_ON(chix != path->p_idx); | |
763 | } | |
764 | #endif | |
765 | ||
766 | } | |
767 | ||
768 | /* | |
d0d856e8 RD |
769 | * ext4_ext_binsearch: |
770 | * binary search for closest extent of the given block | |
c29c0ae7 | 771 | * the header must be checked before calling this |
a86c6181 AT |
772 | */ |
773 | static void | |
725d26d3 AK |
774 | ext4_ext_binsearch(struct inode *inode, |
775 | struct ext4_ext_path *path, ext4_lblk_t block) | |
a86c6181 AT |
776 | { |
777 | struct ext4_extent_header *eh = path->p_hdr; | |
778 | struct ext4_extent *r, *l, *m; | |
779 | ||
a86c6181 AT |
780 | if (eh->eh_entries == 0) { |
781 | /* | |
d0d856e8 RD |
782 | * this leaf is empty: |
783 | * we get such a leaf in split/add case | |
a86c6181 AT |
784 | */ |
785 | return; | |
786 | } | |
787 | ||
bba90743 | 788 | ext_debug("binsearch for %u: ", block); |
a86c6181 AT |
789 | |
790 | l = EXT_FIRST_EXTENT(eh) + 1; | |
e9f410b1 | 791 | r = EXT_LAST_EXTENT(eh); |
a86c6181 AT |
792 | |
793 | while (l <= r) { | |
794 | m = l + (r - l) / 2; | |
795 | if (block < le32_to_cpu(m->ee_block)) | |
796 | r = m - 1; | |
797 | else | |
798 | l = m + 1; | |
26d535ed DM |
799 | ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block), |
800 | m, le32_to_cpu(m->ee_block), | |
801 | r, le32_to_cpu(r->ee_block)); | |
a86c6181 AT |
802 | } |
803 | ||
804 | path->p_ext = l - 1; | |
553f9008 | 805 | ext_debug(" -> %d:%llu:[%d]%d ", |
8c55e204 | 806 | le32_to_cpu(path->p_ext->ee_block), |
bf89d16f | 807 | ext4_ext_pblock(path->p_ext), |
556615dc | 808 | ext4_ext_is_unwritten(path->p_ext), |
a2df2a63 | 809 | ext4_ext_get_actual_len(path->p_ext)); |
a86c6181 AT |
810 | |
811 | #ifdef CHECK_BINSEARCH | |
812 | { | |
813 | struct ext4_extent *chex, *ex; | |
814 | int k; | |
815 | ||
816 | chex = ex = EXT_FIRST_EXTENT(eh); | |
817 | for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) { | |
818 | BUG_ON(k && le32_to_cpu(ex->ee_block) | |
8c55e204 | 819 | <= le32_to_cpu(ex[-1].ee_block)); |
a86c6181 AT |
820 | if (block < le32_to_cpu(ex->ee_block)) |
821 | break; | |
822 | chex = ex; | |
823 | } | |
824 | BUG_ON(chex != path->p_ext); | |
825 | } | |
826 | #endif | |
827 | ||
828 | } | |
829 | ||
830 | int ext4_ext_tree_init(handle_t *handle, struct inode *inode) | |
831 | { | |
832 | struct ext4_extent_header *eh; | |
833 | ||
834 | eh = ext_inode_hdr(inode); | |
835 | eh->eh_depth = 0; | |
836 | eh->eh_entries = 0; | |
837 | eh->eh_magic = EXT4_EXT_MAGIC; | |
55ad63bf | 838 | eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0)); |
a86c6181 | 839 | ext4_mark_inode_dirty(handle, inode); |
a86c6181 AT |
840 | return 0; |
841 | } | |
842 | ||
843 | struct ext4_ext_path * | |
725d26d3 | 844 | ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block, |
107a7bd3 | 845 | struct ext4_ext_path *path, int flags) |
a86c6181 AT |
846 | { |
847 | struct ext4_extent_header *eh; | |
848 | struct buffer_head *bh; | |
849 | short int depth, i, ppos = 0, alloc = 0; | |
860d21e2 | 850 | int ret; |
a86c6181 AT |
851 | |
852 | eh = ext_inode_hdr(inode); | |
c29c0ae7 | 853 | depth = ext_depth(inode); |
a86c6181 AT |
854 | |
855 | /* account possible depth increase */ | |
856 | if (!path) { | |
5d4958f9 | 857 | path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2), |
a86c6181 AT |
858 | GFP_NOFS); |
859 | if (!path) | |
860 | return ERR_PTR(-ENOMEM); | |
861 | alloc = 1; | |
862 | } | |
a86c6181 | 863 | path[0].p_hdr = eh; |
1973adcb | 864 | path[0].p_bh = NULL; |
a86c6181 | 865 | |
c29c0ae7 | 866 | i = depth; |
a86c6181 AT |
867 | /* walk through the tree */ |
868 | while (i) { | |
869 | ext_debug("depth %d: num %d, max %d\n", | |
870 | ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max)); | |
c29c0ae7 | 871 | |
a86c6181 | 872 | ext4_ext_binsearch_idx(inode, path + ppos, block); |
bf89d16f | 873 | path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx); |
a86c6181 AT |
874 | path[ppos].p_depth = i; |
875 | path[ppos].p_ext = NULL; | |
876 | ||
107a7bd3 TT |
877 | bh = read_extent_tree_block(inode, path[ppos].p_block, --i, |
878 | flags); | |
7d7ea89e TT |
879 | if (IS_ERR(bh)) { |
880 | ret = PTR_ERR(bh); | |
a86c6181 | 881 | goto err; |
860d21e2 | 882 | } |
7d7ea89e | 883 | |
a86c6181 AT |
884 | eh = ext_block_hdr(bh); |
885 | ppos++; | |
273df556 FM |
886 | if (unlikely(ppos > depth)) { |
887 | put_bh(bh); | |
888 | EXT4_ERROR_INODE(inode, | |
889 | "ppos %d > depth %d", ppos, depth); | |
860d21e2 | 890 | ret = -EIO; |
273df556 FM |
891 | goto err; |
892 | } | |
a86c6181 AT |
893 | path[ppos].p_bh = bh; |
894 | path[ppos].p_hdr = eh; | |
a86c6181 AT |
895 | } |
896 | ||
897 | path[ppos].p_depth = i; | |
a86c6181 AT |
898 | path[ppos].p_ext = NULL; |
899 | path[ppos].p_idx = NULL; | |
900 | ||
a86c6181 AT |
901 | /* find extent */ |
902 | ext4_ext_binsearch(inode, path + ppos, block); | |
1973adcb SF |
903 | /* if not an empty leaf */ |
904 | if (path[ppos].p_ext) | |
bf89d16f | 905 | path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext); |
a86c6181 AT |
906 | |
907 | ext4_ext_show_path(inode, path); | |
908 | ||
909 | return path; | |
910 | ||
911 | err: | |
912 | ext4_ext_drop_refs(path); | |
913 | if (alloc) | |
914 | kfree(path); | |
860d21e2 | 915 | return ERR_PTR(ret); |
a86c6181 AT |
916 | } |
917 | ||
918 | /* | |
d0d856e8 RD |
919 | * ext4_ext_insert_index: |
920 | * insert new index [@logical;@ptr] into the block at @curp; | |
921 | * check where to insert: before @curp or after @curp | |
a86c6181 | 922 | */ |
1f109d5a TT |
923 | static int ext4_ext_insert_index(handle_t *handle, struct inode *inode, |
924 | struct ext4_ext_path *curp, | |
925 | int logical, ext4_fsblk_t ptr) | |
a86c6181 AT |
926 | { |
927 | struct ext4_extent_idx *ix; | |
928 | int len, err; | |
929 | ||
7e028976 AM |
930 | err = ext4_ext_get_access(handle, inode, curp); |
931 | if (err) | |
a86c6181 AT |
932 | return err; |
933 | ||
273df556 FM |
934 | if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) { |
935 | EXT4_ERROR_INODE(inode, | |
936 | "logical %d == ei_block %d!", | |
937 | logical, le32_to_cpu(curp->p_idx->ei_block)); | |
938 | return -EIO; | |
939 | } | |
d4620315 RD |
940 | |
941 | if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries) | |
942 | >= le16_to_cpu(curp->p_hdr->eh_max))) { | |
943 | EXT4_ERROR_INODE(inode, | |
944 | "eh_entries %d >= eh_max %d!", | |
945 | le16_to_cpu(curp->p_hdr->eh_entries), | |
946 | le16_to_cpu(curp->p_hdr->eh_max)); | |
947 | return -EIO; | |
948 | } | |
949 | ||
a86c6181 AT |
950 | if (logical > le32_to_cpu(curp->p_idx->ei_block)) { |
951 | /* insert after */ | |
80e675f9 | 952 | ext_debug("insert new index %d after: %llu\n", logical, ptr); |
a86c6181 AT |
953 | ix = curp->p_idx + 1; |
954 | } else { | |
955 | /* insert before */ | |
80e675f9 | 956 | ext_debug("insert new index %d before: %llu\n", logical, ptr); |
a86c6181 AT |
957 | ix = curp->p_idx; |
958 | } | |
959 | ||
80e675f9 EG |
960 | len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1; |
961 | BUG_ON(len < 0); | |
962 | if (len > 0) { | |
963 | ext_debug("insert new index %d: " | |
964 | "move %d indices from 0x%p to 0x%p\n", | |
965 | logical, len, ix, ix + 1); | |
966 | memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx)); | |
967 | } | |
968 | ||
f472e026 TM |
969 | if (unlikely(ix > EXT_MAX_INDEX(curp->p_hdr))) { |
970 | EXT4_ERROR_INODE(inode, "ix > EXT_MAX_INDEX!"); | |
971 | return -EIO; | |
972 | } | |
973 | ||
a86c6181 | 974 | ix->ei_block = cpu_to_le32(logical); |
f65e6fba | 975 | ext4_idx_store_pblock(ix, ptr); |
e8546d06 | 976 | le16_add_cpu(&curp->p_hdr->eh_entries, 1); |
a86c6181 | 977 | |
273df556 FM |
978 | if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) { |
979 | EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!"); | |
980 | return -EIO; | |
981 | } | |
a86c6181 AT |
982 | |
983 | err = ext4_ext_dirty(handle, inode, curp); | |
984 | ext4_std_error(inode->i_sb, err); | |
985 | ||
986 | return err; | |
987 | } | |
988 | ||
989 | /* | |
d0d856e8 RD |
990 | * ext4_ext_split: |
991 | * inserts new subtree into the path, using free index entry | |
992 | * at depth @at: | |
993 | * - allocates all needed blocks (new leaf and all intermediate index blocks) | |
994 | * - makes decision where to split | |
995 | * - moves remaining extents and index entries (right to the split point) | |
996 | * into the newly allocated blocks | |
997 | * - initializes subtree | |
a86c6181 AT |
998 | */ |
999 | static int ext4_ext_split(handle_t *handle, struct inode *inode, | |
55f020db AH |
1000 | unsigned int flags, |
1001 | struct ext4_ext_path *path, | |
1002 | struct ext4_extent *newext, int at) | |
a86c6181 AT |
1003 | { |
1004 | struct buffer_head *bh = NULL; | |
1005 | int depth = ext_depth(inode); | |
1006 | struct ext4_extent_header *neh; | |
1007 | struct ext4_extent_idx *fidx; | |
a86c6181 | 1008 | int i = at, k, m, a; |
f65e6fba | 1009 | ext4_fsblk_t newblock, oldblock; |
a86c6181 | 1010 | __le32 border; |
f65e6fba | 1011 | ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */ |
a86c6181 AT |
1012 | int err = 0; |
1013 | ||
1014 | /* make decision: where to split? */ | |
d0d856e8 | 1015 | /* FIXME: now decision is simplest: at current extent */ |
a86c6181 | 1016 | |
d0d856e8 | 1017 | /* if current leaf will be split, then we should use |
a86c6181 | 1018 | * border from split point */ |
273df556 FM |
1019 | if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) { |
1020 | EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!"); | |
1021 | return -EIO; | |
1022 | } | |
a86c6181 AT |
1023 | if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) { |
1024 | border = path[depth].p_ext[1].ee_block; | |
d0d856e8 | 1025 | ext_debug("leaf will be split." |
a86c6181 | 1026 | " next leaf starts at %d\n", |
8c55e204 | 1027 | le32_to_cpu(border)); |
a86c6181 AT |
1028 | } else { |
1029 | border = newext->ee_block; | |
1030 | ext_debug("leaf will be added." | |
1031 | " next leaf starts at %d\n", | |
8c55e204 | 1032 | le32_to_cpu(border)); |
a86c6181 AT |
1033 | } |
1034 | ||
1035 | /* | |
d0d856e8 RD |
1036 | * If error occurs, then we break processing |
1037 | * and mark filesystem read-only. index won't | |
a86c6181 | 1038 | * be inserted and tree will be in consistent |
d0d856e8 | 1039 | * state. Next mount will repair buffers too. |
a86c6181 AT |
1040 | */ |
1041 | ||
1042 | /* | |
d0d856e8 RD |
1043 | * Get array to track all allocated blocks. |
1044 | * We need this to handle errors and free blocks | |
1045 | * upon them. | |
a86c6181 | 1046 | */ |
5d4958f9 | 1047 | ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS); |
a86c6181 AT |
1048 | if (!ablocks) |
1049 | return -ENOMEM; | |
a86c6181 AT |
1050 | |
1051 | /* allocate all needed blocks */ | |
1052 | ext_debug("allocate %d blocks for indexes/leaf\n", depth - at); | |
1053 | for (a = 0; a < depth - at; a++) { | |
654b4908 | 1054 | newblock = ext4_ext_new_meta_block(handle, inode, path, |
55f020db | 1055 | newext, &err, flags); |
a86c6181 AT |
1056 | if (newblock == 0) |
1057 | goto cleanup; | |
1058 | ablocks[a] = newblock; | |
1059 | } | |
1060 | ||
1061 | /* initialize new leaf */ | |
1062 | newblock = ablocks[--a]; | |
273df556 FM |
1063 | if (unlikely(newblock == 0)) { |
1064 | EXT4_ERROR_INODE(inode, "newblock == 0!"); | |
1065 | err = -EIO; | |
1066 | goto cleanup; | |
1067 | } | |
a86c6181 | 1068 | bh = sb_getblk(inode->i_sb, newblock); |
aebf0243 | 1069 | if (unlikely(!bh)) { |
860d21e2 | 1070 | err = -ENOMEM; |
a86c6181 AT |
1071 | goto cleanup; |
1072 | } | |
1073 | lock_buffer(bh); | |
1074 | ||
7e028976 AM |
1075 | err = ext4_journal_get_create_access(handle, bh); |
1076 | if (err) | |
a86c6181 AT |
1077 | goto cleanup; |
1078 | ||
1079 | neh = ext_block_hdr(bh); | |
1080 | neh->eh_entries = 0; | |
55ad63bf | 1081 | neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0)); |
a86c6181 AT |
1082 | neh->eh_magic = EXT4_EXT_MAGIC; |
1083 | neh->eh_depth = 0; | |
a86c6181 | 1084 | |
d0d856e8 | 1085 | /* move remainder of path[depth] to the new leaf */ |
273df556 FM |
1086 | if (unlikely(path[depth].p_hdr->eh_entries != |
1087 | path[depth].p_hdr->eh_max)) { | |
1088 | EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!", | |
1089 | path[depth].p_hdr->eh_entries, | |
1090 | path[depth].p_hdr->eh_max); | |
1091 | err = -EIO; | |
1092 | goto cleanup; | |
1093 | } | |
a86c6181 | 1094 | /* start copy from next extent */ |
1b16da77 YY |
1095 | m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++; |
1096 | ext4_ext_show_move(inode, path, newblock, depth); | |
a86c6181 | 1097 | if (m) { |
1b16da77 YY |
1098 | struct ext4_extent *ex; |
1099 | ex = EXT_FIRST_EXTENT(neh); | |
1100 | memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m); | |
e8546d06 | 1101 | le16_add_cpu(&neh->eh_entries, m); |
a86c6181 AT |
1102 | } |
1103 | ||
7ac5990d | 1104 | ext4_extent_block_csum_set(inode, neh); |
a86c6181 AT |
1105 | set_buffer_uptodate(bh); |
1106 | unlock_buffer(bh); | |
1107 | ||
0390131b | 1108 | err = ext4_handle_dirty_metadata(handle, inode, bh); |
7e028976 | 1109 | if (err) |
a86c6181 AT |
1110 | goto cleanup; |
1111 | brelse(bh); | |
1112 | bh = NULL; | |
1113 | ||
1114 | /* correct old leaf */ | |
1115 | if (m) { | |
7e028976 AM |
1116 | err = ext4_ext_get_access(handle, inode, path + depth); |
1117 | if (err) | |
a86c6181 | 1118 | goto cleanup; |
e8546d06 | 1119 | le16_add_cpu(&path[depth].p_hdr->eh_entries, -m); |
7e028976 AM |
1120 | err = ext4_ext_dirty(handle, inode, path + depth); |
1121 | if (err) | |
a86c6181 AT |
1122 | goto cleanup; |
1123 | ||
1124 | } | |
1125 | ||
1126 | /* create intermediate indexes */ | |
1127 | k = depth - at - 1; | |
273df556 FM |
1128 | if (unlikely(k < 0)) { |
1129 | EXT4_ERROR_INODE(inode, "k %d < 0!", k); | |
1130 | err = -EIO; | |
1131 | goto cleanup; | |
1132 | } | |
a86c6181 AT |
1133 | if (k) |
1134 | ext_debug("create %d intermediate indices\n", k); | |
1135 | /* insert new index into current index block */ | |
1136 | /* current depth stored in i var */ | |
1137 | i = depth - 1; | |
1138 | while (k--) { | |
1139 | oldblock = newblock; | |
1140 | newblock = ablocks[--a]; | |
bba90743 | 1141 | bh = sb_getblk(inode->i_sb, newblock); |
aebf0243 | 1142 | if (unlikely(!bh)) { |
860d21e2 | 1143 | err = -ENOMEM; |
a86c6181 AT |
1144 | goto cleanup; |
1145 | } | |
1146 | lock_buffer(bh); | |
1147 | ||
7e028976 AM |
1148 | err = ext4_journal_get_create_access(handle, bh); |
1149 | if (err) | |
a86c6181 AT |
1150 | goto cleanup; |
1151 | ||
1152 | neh = ext_block_hdr(bh); | |
1153 | neh->eh_entries = cpu_to_le16(1); | |
1154 | neh->eh_magic = EXT4_EXT_MAGIC; | |
55ad63bf | 1155 | neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0)); |
a86c6181 AT |
1156 | neh->eh_depth = cpu_to_le16(depth - i); |
1157 | fidx = EXT_FIRST_INDEX(neh); | |
1158 | fidx->ei_block = border; | |
f65e6fba | 1159 | ext4_idx_store_pblock(fidx, oldblock); |
a86c6181 | 1160 | |
bba90743 ES |
1161 | ext_debug("int.index at %d (block %llu): %u -> %llu\n", |
1162 | i, newblock, le32_to_cpu(border), oldblock); | |
a86c6181 | 1163 | |
1b16da77 | 1164 | /* move remainder of path[i] to the new index block */ |
273df556 FM |
1165 | if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) != |
1166 | EXT_LAST_INDEX(path[i].p_hdr))) { | |
1167 | EXT4_ERROR_INODE(inode, | |
1168 | "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!", | |
1169 | le32_to_cpu(path[i].p_ext->ee_block)); | |
1170 | err = -EIO; | |
1171 | goto cleanup; | |
1172 | } | |
1b16da77 YY |
1173 | /* start copy indexes */ |
1174 | m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++; | |
1175 | ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx, | |
1176 | EXT_MAX_INDEX(path[i].p_hdr)); | |
1177 | ext4_ext_show_move(inode, path, newblock, i); | |
a86c6181 | 1178 | if (m) { |
1b16da77 | 1179 | memmove(++fidx, path[i].p_idx, |
a86c6181 | 1180 | sizeof(struct ext4_extent_idx) * m); |
e8546d06 | 1181 | le16_add_cpu(&neh->eh_entries, m); |
a86c6181 | 1182 | } |
7ac5990d | 1183 | ext4_extent_block_csum_set(inode, neh); |
a86c6181 AT |
1184 | set_buffer_uptodate(bh); |
1185 | unlock_buffer(bh); | |
1186 | ||
0390131b | 1187 | err = ext4_handle_dirty_metadata(handle, inode, bh); |
7e028976 | 1188 | if (err) |
a86c6181 AT |
1189 | goto cleanup; |
1190 | brelse(bh); | |
1191 | bh = NULL; | |
1192 | ||
1193 | /* correct old index */ | |
1194 | if (m) { | |
1195 | err = ext4_ext_get_access(handle, inode, path + i); | |
1196 | if (err) | |
1197 | goto cleanup; | |
e8546d06 | 1198 | le16_add_cpu(&path[i].p_hdr->eh_entries, -m); |
a86c6181 AT |
1199 | err = ext4_ext_dirty(handle, inode, path + i); |
1200 | if (err) | |
1201 | goto cleanup; | |
1202 | } | |
1203 | ||
1204 | i--; | |
1205 | } | |
1206 | ||
1207 | /* insert new index */ | |
a86c6181 AT |
1208 | err = ext4_ext_insert_index(handle, inode, path + at, |
1209 | le32_to_cpu(border), newblock); | |
1210 | ||
1211 | cleanup: | |
1212 | if (bh) { | |
1213 | if (buffer_locked(bh)) | |
1214 | unlock_buffer(bh); | |
1215 | brelse(bh); | |
1216 | } | |
1217 | ||
1218 | if (err) { | |
1219 | /* free all allocated blocks in error case */ | |
1220 | for (i = 0; i < depth; i++) { | |
1221 | if (!ablocks[i]) | |
1222 | continue; | |
7dc57615 | 1223 | ext4_free_blocks(handle, inode, NULL, ablocks[i], 1, |
e6362609 | 1224 | EXT4_FREE_BLOCKS_METADATA); |
a86c6181 AT |
1225 | } |
1226 | } | |
1227 | kfree(ablocks); | |
1228 | ||
1229 | return err; | |
1230 | } | |
1231 | ||
1232 | /* | |
d0d856e8 RD |
1233 | * ext4_ext_grow_indepth: |
1234 | * implements tree growing procedure: | |
1235 | * - allocates new block | |
1236 | * - moves top-level data (index block or leaf) into the new block | |
1237 | * - initializes new top-level, creating index that points to the | |
1238 | * just created block | |
a86c6181 AT |
1239 | */ |
1240 | static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, | |
55f020db | 1241 | unsigned int flags, |
55f020db | 1242 | struct ext4_extent *newext) |
a86c6181 | 1243 | { |
a86c6181 | 1244 | struct ext4_extent_header *neh; |
a86c6181 | 1245 | struct buffer_head *bh; |
f65e6fba | 1246 | ext4_fsblk_t newblock; |
a86c6181 AT |
1247 | int err = 0; |
1248 | ||
1939dd84 | 1249 | newblock = ext4_ext_new_meta_block(handle, inode, NULL, |
55f020db | 1250 | newext, &err, flags); |
a86c6181 AT |
1251 | if (newblock == 0) |
1252 | return err; | |
1253 | ||
1254 | bh = sb_getblk(inode->i_sb, newblock); | |
aebf0243 | 1255 | if (unlikely(!bh)) |
860d21e2 | 1256 | return -ENOMEM; |
a86c6181 AT |
1257 | lock_buffer(bh); |
1258 | ||
7e028976 AM |
1259 | err = ext4_journal_get_create_access(handle, bh); |
1260 | if (err) { | |
a86c6181 AT |
1261 | unlock_buffer(bh); |
1262 | goto out; | |
1263 | } | |
1264 | ||
1265 | /* move top-level index/leaf into new block */ | |
1939dd84 DM |
1266 | memmove(bh->b_data, EXT4_I(inode)->i_data, |
1267 | sizeof(EXT4_I(inode)->i_data)); | |
a86c6181 AT |
1268 | |
1269 | /* set size of new block */ | |
1270 | neh = ext_block_hdr(bh); | |
1271 | /* old root could have indexes or leaves | |
1272 | * so calculate e_max right way */ | |
1273 | if (ext_depth(inode)) | |
55ad63bf | 1274 | neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0)); |
a86c6181 | 1275 | else |
55ad63bf | 1276 | neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0)); |
a86c6181 | 1277 | neh->eh_magic = EXT4_EXT_MAGIC; |
7ac5990d | 1278 | ext4_extent_block_csum_set(inode, neh); |
a86c6181 AT |
1279 | set_buffer_uptodate(bh); |
1280 | unlock_buffer(bh); | |
1281 | ||
0390131b | 1282 | err = ext4_handle_dirty_metadata(handle, inode, bh); |
7e028976 | 1283 | if (err) |
a86c6181 AT |
1284 | goto out; |
1285 | ||
1939dd84 | 1286 | /* Update top-level index: num,max,pointer */ |
a86c6181 | 1287 | neh = ext_inode_hdr(inode); |
1939dd84 DM |
1288 | neh->eh_entries = cpu_to_le16(1); |
1289 | ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock); | |
1290 | if (neh->eh_depth == 0) { | |
1291 | /* Root extent block becomes index block */ | |
1292 | neh->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0)); | |
1293 | EXT_FIRST_INDEX(neh)->ei_block = | |
1294 | EXT_FIRST_EXTENT(neh)->ee_block; | |
1295 | } | |
2ae02107 | 1296 | ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n", |
a86c6181 | 1297 | le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max), |
5a0790c2 | 1298 | le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block), |
bf89d16f | 1299 | ext4_idx_pblock(EXT_FIRST_INDEX(neh))); |
a86c6181 | 1300 | |
ba39ebb6 | 1301 | le16_add_cpu(&neh->eh_depth, 1); |
1939dd84 | 1302 | ext4_mark_inode_dirty(handle, inode); |
a86c6181 AT |
1303 | out: |
1304 | brelse(bh); | |
1305 | ||
1306 | return err; | |
1307 | } | |
1308 | ||
1309 | /* | |
d0d856e8 RD |
1310 | * ext4_ext_create_new_leaf: |
1311 | * finds empty index and adds new leaf. | |
1312 | * if no free index is found, then it requests in-depth growing. | |
a86c6181 AT |
1313 | */ |
1314 | static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode, | |
107a7bd3 TT |
1315 | unsigned int mb_flags, |
1316 | unsigned int gb_flags, | |
55f020db AH |
1317 | struct ext4_ext_path *path, |
1318 | struct ext4_extent *newext) | |
a86c6181 AT |
1319 | { |
1320 | struct ext4_ext_path *curp; | |
1321 | int depth, i, err = 0; | |
1322 | ||
1323 | repeat: | |
1324 | i = depth = ext_depth(inode); | |
1325 | ||
1326 | /* walk up to the tree and look for free index entry */ | |
1327 | curp = path + depth; | |
1328 | while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) { | |
1329 | i--; | |
1330 | curp--; | |
1331 | } | |
1332 | ||
d0d856e8 RD |
1333 | /* we use already allocated block for index block, |
1334 | * so subsequent data blocks should be contiguous */ | |
a86c6181 AT |
1335 | if (EXT_HAS_FREE_INDEX(curp)) { |
1336 | /* if we found index with free entry, then use that | |
1337 | * entry: create all needed subtree and add new leaf */ | |
107a7bd3 | 1338 | err = ext4_ext_split(handle, inode, mb_flags, path, newext, i); |
787e0981 SF |
1339 | if (err) |
1340 | goto out; | |
a86c6181 AT |
1341 | |
1342 | /* refill path */ | |
1343 | ext4_ext_drop_refs(path); | |
1344 | path = ext4_ext_find_extent(inode, | |
725d26d3 | 1345 | (ext4_lblk_t)le32_to_cpu(newext->ee_block), |
107a7bd3 | 1346 | path, gb_flags); |
a86c6181 AT |
1347 | if (IS_ERR(path)) |
1348 | err = PTR_ERR(path); | |
1349 | } else { | |
1350 | /* tree is full, time to grow in depth */ | |
107a7bd3 | 1351 | err = ext4_ext_grow_indepth(handle, inode, mb_flags, newext); |
a86c6181 AT |
1352 | if (err) |
1353 | goto out; | |
1354 | ||
1355 | /* refill path */ | |
1356 | ext4_ext_drop_refs(path); | |
1357 | path = ext4_ext_find_extent(inode, | |
725d26d3 | 1358 | (ext4_lblk_t)le32_to_cpu(newext->ee_block), |
107a7bd3 | 1359 | path, gb_flags); |
a86c6181 AT |
1360 | if (IS_ERR(path)) { |
1361 | err = PTR_ERR(path); | |
1362 | goto out; | |
1363 | } | |
1364 | ||
1365 | /* | |
d0d856e8 RD |
1366 | * only first (depth 0 -> 1) produces free space; |
1367 | * in all other cases we have to split the grown tree | |
a86c6181 AT |
1368 | */ |
1369 | depth = ext_depth(inode); | |
1370 | if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) { | |
d0d856e8 | 1371 | /* now we need to split */ |
a86c6181 AT |
1372 | goto repeat; |
1373 | } | |
1374 | } | |
1375 | ||
1376 | out: | |
1377 | return err; | |
1378 | } | |
1379 | ||
1988b51e AT |
1380 | /* |
1381 | * search the closest allocated block to the left for *logical | |
1382 | * and returns it at @logical + it's physical address at @phys | |
1383 | * if *logical is the smallest allocated block, the function | |
1384 | * returns 0 at @phys | |
1385 | * return value contains 0 (success) or error code | |
1386 | */ | |
1f109d5a TT |
1387 | static int ext4_ext_search_left(struct inode *inode, |
1388 | struct ext4_ext_path *path, | |
1389 | ext4_lblk_t *logical, ext4_fsblk_t *phys) | |
1988b51e AT |
1390 | { |
1391 | struct ext4_extent_idx *ix; | |
1392 | struct ext4_extent *ex; | |
b939e376 | 1393 | int depth, ee_len; |
1988b51e | 1394 | |
273df556 FM |
1395 | if (unlikely(path == NULL)) { |
1396 | EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical); | |
1397 | return -EIO; | |
1398 | } | |
1988b51e AT |
1399 | depth = path->p_depth; |
1400 | *phys = 0; | |
1401 | ||
1402 | if (depth == 0 && path->p_ext == NULL) | |
1403 | return 0; | |
1404 | ||
1405 | /* usually extent in the path covers blocks smaller | |
1406 | * then *logical, but it can be that extent is the | |
1407 | * first one in the file */ | |
1408 | ||
1409 | ex = path[depth].p_ext; | |
b939e376 | 1410 | ee_len = ext4_ext_get_actual_len(ex); |
1988b51e | 1411 | if (*logical < le32_to_cpu(ex->ee_block)) { |
273df556 FM |
1412 | if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) { |
1413 | EXT4_ERROR_INODE(inode, | |
1414 | "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!", | |
1415 | *logical, le32_to_cpu(ex->ee_block)); | |
1416 | return -EIO; | |
1417 | } | |
1988b51e AT |
1418 | while (--depth >= 0) { |
1419 | ix = path[depth].p_idx; | |
273df556 FM |
1420 | if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) { |
1421 | EXT4_ERROR_INODE(inode, | |
1422 | "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!", | |
6ee3b212 | 1423 | ix != NULL ? le32_to_cpu(ix->ei_block) : 0, |
273df556 | 1424 | EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ? |
6ee3b212 | 1425 | le32_to_cpu(EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block) : 0, |
273df556 FM |
1426 | depth); |
1427 | return -EIO; | |
1428 | } | |
1988b51e AT |
1429 | } |
1430 | return 0; | |
1431 | } | |
1432 | ||
273df556 FM |
1433 | if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) { |
1434 | EXT4_ERROR_INODE(inode, | |
1435 | "logical %d < ee_block %d + ee_len %d!", | |
1436 | *logical, le32_to_cpu(ex->ee_block), ee_len); | |
1437 | return -EIO; | |
1438 | } | |
1988b51e | 1439 | |
b939e376 | 1440 | *logical = le32_to_cpu(ex->ee_block) + ee_len - 1; |
bf89d16f | 1441 | *phys = ext4_ext_pblock(ex) + ee_len - 1; |
1988b51e AT |
1442 | return 0; |
1443 | } | |
1444 | ||
1445 | /* | |
1446 | * search the closest allocated block to the right for *logical | |
1447 | * and returns it at @logical + it's physical address at @phys | |
df3ab170 | 1448 | * if *logical is the largest allocated block, the function |
1988b51e AT |
1449 | * returns 0 at @phys |
1450 | * return value contains 0 (success) or error code | |
1451 | */ | |
1f109d5a TT |
1452 | static int ext4_ext_search_right(struct inode *inode, |
1453 | struct ext4_ext_path *path, | |
4d33b1ef TT |
1454 | ext4_lblk_t *logical, ext4_fsblk_t *phys, |
1455 | struct ext4_extent **ret_ex) | |
1988b51e AT |
1456 | { |
1457 | struct buffer_head *bh = NULL; | |
1458 | struct ext4_extent_header *eh; | |
1459 | struct ext4_extent_idx *ix; | |
1460 | struct ext4_extent *ex; | |
1461 | ext4_fsblk_t block; | |
395a87bf ES |
1462 | int depth; /* Note, NOT eh_depth; depth from top of tree */ |
1463 | int ee_len; | |
1988b51e | 1464 | |
273df556 FM |
1465 | if (unlikely(path == NULL)) { |
1466 | EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical); | |
1467 | return -EIO; | |
1468 | } | |
1988b51e AT |
1469 | depth = path->p_depth; |
1470 | *phys = 0; | |
1471 | ||
1472 | if (depth == 0 && path->p_ext == NULL) | |
1473 | return 0; | |
1474 | ||
1475 | /* usually extent in the path covers blocks smaller | |
1476 | * then *logical, but it can be that extent is the | |
1477 | * first one in the file */ | |
1478 | ||
1479 | ex = path[depth].p_ext; | |
b939e376 | 1480 | ee_len = ext4_ext_get_actual_len(ex); |
1988b51e | 1481 | if (*logical < le32_to_cpu(ex->ee_block)) { |
273df556 FM |
1482 | if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) { |
1483 | EXT4_ERROR_INODE(inode, | |
1484 | "first_extent(path[%d].p_hdr) != ex", | |
1485 | depth); | |
1486 | return -EIO; | |
1487 | } | |
1988b51e AT |
1488 | while (--depth >= 0) { |
1489 | ix = path[depth].p_idx; | |
273df556 FM |
1490 | if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) { |
1491 | EXT4_ERROR_INODE(inode, | |
1492 | "ix != EXT_FIRST_INDEX *logical %d!", | |
1493 | *logical); | |
1494 | return -EIO; | |
1495 | } | |
1988b51e | 1496 | } |
4d33b1ef | 1497 | goto found_extent; |
1988b51e AT |
1498 | } |
1499 | ||
273df556 FM |
1500 | if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) { |
1501 | EXT4_ERROR_INODE(inode, | |
1502 | "logical %d < ee_block %d + ee_len %d!", | |
1503 | *logical, le32_to_cpu(ex->ee_block), ee_len); | |
1504 | return -EIO; | |
1505 | } | |
1988b51e AT |
1506 | |
1507 | if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) { | |
1508 | /* next allocated block in this leaf */ | |
1509 | ex++; | |
4d33b1ef | 1510 | goto found_extent; |
1988b51e AT |
1511 | } |
1512 | ||
1513 | /* go up and search for index to the right */ | |
1514 | while (--depth >= 0) { | |
1515 | ix = path[depth].p_idx; | |
1516 | if (ix != EXT_LAST_INDEX(path[depth].p_hdr)) | |
25f1ee3a | 1517 | goto got_index; |
1988b51e AT |
1518 | } |
1519 | ||
25f1ee3a WF |
1520 | /* we've gone up to the root and found no index to the right */ |
1521 | return 0; | |
1988b51e | 1522 | |
25f1ee3a | 1523 | got_index: |
1988b51e AT |
1524 | /* we've found index to the right, let's |
1525 | * follow it and find the closest allocated | |
1526 | * block to the right */ | |
1527 | ix++; | |
bf89d16f | 1528 | block = ext4_idx_pblock(ix); |
1988b51e | 1529 | while (++depth < path->p_depth) { |
395a87bf | 1530 | /* subtract from p_depth to get proper eh_depth */ |
7d7ea89e | 1531 | bh = read_extent_tree_block(inode, block, |
107a7bd3 | 1532 | path->p_depth - depth, 0); |
7d7ea89e TT |
1533 | if (IS_ERR(bh)) |
1534 | return PTR_ERR(bh); | |
1535 | eh = ext_block_hdr(bh); | |
1988b51e | 1536 | ix = EXT_FIRST_INDEX(eh); |
bf89d16f | 1537 | block = ext4_idx_pblock(ix); |
1988b51e AT |
1538 | put_bh(bh); |
1539 | } | |
1540 | ||
107a7bd3 | 1541 | bh = read_extent_tree_block(inode, block, path->p_depth - depth, 0); |
7d7ea89e TT |
1542 | if (IS_ERR(bh)) |
1543 | return PTR_ERR(bh); | |
1988b51e | 1544 | eh = ext_block_hdr(bh); |
1988b51e | 1545 | ex = EXT_FIRST_EXTENT(eh); |
4d33b1ef | 1546 | found_extent: |
1988b51e | 1547 | *logical = le32_to_cpu(ex->ee_block); |
bf89d16f | 1548 | *phys = ext4_ext_pblock(ex); |
4d33b1ef TT |
1549 | *ret_ex = ex; |
1550 | if (bh) | |
1551 | put_bh(bh); | |
1988b51e | 1552 | return 0; |
1988b51e AT |
1553 | } |
1554 | ||
a86c6181 | 1555 | /* |
d0d856e8 | 1556 | * ext4_ext_next_allocated_block: |
f17722f9 | 1557 | * returns allocated block in subsequent extent or EXT_MAX_BLOCKS. |
d0d856e8 RD |
1558 | * NOTE: it considers block number from index entry as |
1559 | * allocated block. Thus, index entries have to be consistent | |
1560 | * with leaves. | |
a86c6181 | 1561 | */ |
725d26d3 | 1562 | static ext4_lblk_t |
a86c6181 AT |
1563 | ext4_ext_next_allocated_block(struct ext4_ext_path *path) |
1564 | { | |
1565 | int depth; | |
1566 | ||
1567 | BUG_ON(path == NULL); | |
1568 | depth = path->p_depth; | |
1569 | ||
1570 | if (depth == 0 && path->p_ext == NULL) | |
f17722f9 | 1571 | return EXT_MAX_BLOCKS; |
a86c6181 AT |
1572 | |
1573 | while (depth >= 0) { | |
1574 | if (depth == path->p_depth) { | |
1575 | /* leaf */ | |
6f8ff537 CW |
1576 | if (path[depth].p_ext && |
1577 | path[depth].p_ext != | |
a86c6181 AT |
1578 | EXT_LAST_EXTENT(path[depth].p_hdr)) |
1579 | return le32_to_cpu(path[depth].p_ext[1].ee_block); | |
1580 | } else { | |
1581 | /* index */ | |
1582 | if (path[depth].p_idx != | |
1583 | EXT_LAST_INDEX(path[depth].p_hdr)) | |
1584 | return le32_to_cpu(path[depth].p_idx[1].ei_block); | |
1585 | } | |
1586 | depth--; | |
1587 | } | |
1588 | ||
f17722f9 | 1589 | return EXT_MAX_BLOCKS; |
a86c6181 AT |
1590 | } |
1591 | ||
1592 | /* | |
d0d856e8 | 1593 | * ext4_ext_next_leaf_block: |
f17722f9 | 1594 | * returns first allocated block from next leaf or EXT_MAX_BLOCKS |
a86c6181 | 1595 | */ |
5718789d | 1596 | static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path) |
a86c6181 AT |
1597 | { |
1598 | int depth; | |
1599 | ||
1600 | BUG_ON(path == NULL); | |
1601 | depth = path->p_depth; | |
1602 | ||
1603 | /* zero-tree has no leaf blocks at all */ | |
1604 | if (depth == 0) | |
f17722f9 | 1605 | return EXT_MAX_BLOCKS; |
a86c6181 AT |
1606 | |
1607 | /* go to index block */ | |
1608 | depth--; | |
1609 | ||
1610 | while (depth >= 0) { | |
1611 | if (path[depth].p_idx != | |
1612 | EXT_LAST_INDEX(path[depth].p_hdr)) | |
725d26d3 AK |
1613 | return (ext4_lblk_t) |
1614 | le32_to_cpu(path[depth].p_idx[1].ei_block); | |
a86c6181 AT |
1615 | depth--; |
1616 | } | |
1617 | ||
f17722f9 | 1618 | return EXT_MAX_BLOCKS; |
a86c6181 AT |
1619 | } |
1620 | ||
1621 | /* | |
d0d856e8 RD |
1622 | * ext4_ext_correct_indexes: |
1623 | * if leaf gets modified and modified extent is first in the leaf, | |
1624 | * then we have to correct all indexes above. | |
a86c6181 AT |
1625 | * TODO: do we need to correct tree in all cases? |
1626 | */ | |
1d03ec98 | 1627 | static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode, |
a86c6181 AT |
1628 | struct ext4_ext_path *path) |
1629 | { | |
1630 | struct ext4_extent_header *eh; | |
1631 | int depth = ext_depth(inode); | |
1632 | struct ext4_extent *ex; | |
1633 | __le32 border; | |
1634 | int k, err = 0; | |
1635 | ||
1636 | eh = path[depth].p_hdr; | |
1637 | ex = path[depth].p_ext; | |
273df556 FM |
1638 | |
1639 | if (unlikely(ex == NULL || eh == NULL)) { | |
1640 | EXT4_ERROR_INODE(inode, | |
1641 | "ex %p == NULL or eh %p == NULL", ex, eh); | |
1642 | return -EIO; | |
1643 | } | |
a86c6181 AT |
1644 | |
1645 | if (depth == 0) { | |
1646 | /* there is no tree at all */ | |
1647 | return 0; | |
1648 | } | |
1649 | ||
1650 | if (ex != EXT_FIRST_EXTENT(eh)) { | |
1651 | /* we correct tree if first leaf got modified only */ | |
1652 | return 0; | |
1653 | } | |
1654 | ||
1655 | /* | |
d0d856e8 | 1656 | * TODO: we need correction if border is smaller than current one |
a86c6181 AT |
1657 | */ |
1658 | k = depth - 1; | |
1659 | border = path[depth].p_ext->ee_block; | |
7e028976 AM |
1660 | err = ext4_ext_get_access(handle, inode, path + k); |
1661 | if (err) | |
a86c6181 AT |
1662 | return err; |
1663 | path[k].p_idx->ei_block = border; | |
7e028976 AM |
1664 | err = ext4_ext_dirty(handle, inode, path + k); |
1665 | if (err) | |
a86c6181 AT |
1666 | return err; |
1667 | ||
1668 | while (k--) { | |
1669 | /* change all left-side indexes */ | |
1670 | if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr)) | |
1671 | break; | |
7e028976 AM |
1672 | err = ext4_ext_get_access(handle, inode, path + k); |
1673 | if (err) | |
a86c6181 AT |
1674 | break; |
1675 | path[k].p_idx->ei_block = border; | |
7e028976 AM |
1676 | err = ext4_ext_dirty(handle, inode, path + k); |
1677 | if (err) | |
a86c6181 AT |
1678 | break; |
1679 | } | |
1680 | ||
1681 | return err; | |
1682 | } | |
1683 | ||
748de673 | 1684 | int |
a86c6181 AT |
1685 | ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1, |
1686 | struct ext4_extent *ex2) | |
1687 | { | |
da0169b3 | 1688 | unsigned short ext1_ee_len, ext2_ee_len; |
a2df2a63 AA |
1689 | |
1690 | /* | |
ec22ba8e | 1691 | * Make sure that both extents are initialized. We don't merge |
556615dc | 1692 | * unwritten extents so that we can be sure that end_io code has |
ec22ba8e DM |
1693 | * the extent that was written properly split out and conversion to |
1694 | * initialized is trivial. | |
a2df2a63 | 1695 | */ |
556615dc | 1696 | if (ext4_ext_is_unwritten(ex1) != ext4_ext_is_unwritten(ex2)) |
a2df2a63 AA |
1697 | return 0; |
1698 | ||
1699 | ext1_ee_len = ext4_ext_get_actual_len(ex1); | |
1700 | ext2_ee_len = ext4_ext_get_actual_len(ex2); | |
1701 | ||
1702 | if (le32_to_cpu(ex1->ee_block) + ext1_ee_len != | |
63f57933 | 1703 | le32_to_cpu(ex2->ee_block)) |
a86c6181 AT |
1704 | return 0; |
1705 | ||
471d4011 SB |
1706 | /* |
1707 | * To allow future support for preallocated extents to be added | |
1708 | * as an RO_COMPAT feature, refuse to merge to extents if | |
d0d856e8 | 1709 | * this can result in the top bit of ee_len being set. |
471d4011 | 1710 | */ |
da0169b3 | 1711 | if (ext1_ee_len + ext2_ee_len > EXT_INIT_MAX_LEN) |
471d4011 | 1712 | return 0; |
556615dc | 1713 | if (ext4_ext_is_unwritten(ex1) && |
a9b82415 DW |
1714 | (ext4_test_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN) || |
1715 | atomic_read(&EXT4_I(inode)->i_unwritten) || | |
556615dc | 1716 | (ext1_ee_len + ext2_ee_len > EXT_UNWRITTEN_MAX_LEN))) |
a9b82415 | 1717 | return 0; |
bbf2f9fb | 1718 | #ifdef AGGRESSIVE_TEST |
b939e376 | 1719 | if (ext1_ee_len >= 4) |
a86c6181 AT |
1720 | return 0; |
1721 | #endif | |
1722 | ||
bf89d16f | 1723 | if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2)) |
a86c6181 AT |
1724 | return 1; |
1725 | return 0; | |
1726 | } | |
1727 | ||
56055d3a AA |
1728 | /* |
1729 | * This function tries to merge the "ex" extent to the next extent in the tree. | |
1730 | * It always tries to merge towards right. If you want to merge towards | |
1731 | * left, pass "ex - 1" as argument instead of "ex". | |
1732 | * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns | |
1733 | * 1 if they got merged. | |
1734 | */ | |
197217a5 | 1735 | static int ext4_ext_try_to_merge_right(struct inode *inode, |
1f109d5a TT |
1736 | struct ext4_ext_path *path, |
1737 | struct ext4_extent *ex) | |
56055d3a AA |
1738 | { |
1739 | struct ext4_extent_header *eh; | |
1740 | unsigned int depth, len; | |
556615dc | 1741 | int merge_done = 0, unwritten; |
56055d3a AA |
1742 | |
1743 | depth = ext_depth(inode); | |
1744 | BUG_ON(path[depth].p_hdr == NULL); | |
1745 | eh = path[depth].p_hdr; | |
1746 | ||
1747 | while (ex < EXT_LAST_EXTENT(eh)) { | |
1748 | if (!ext4_can_extents_be_merged(inode, ex, ex + 1)) | |
1749 | break; | |
1750 | /* merge with next extent! */ | |
556615dc | 1751 | unwritten = ext4_ext_is_unwritten(ex); |
56055d3a AA |
1752 | ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) |
1753 | + ext4_ext_get_actual_len(ex + 1)); | |
556615dc LC |
1754 | if (unwritten) |
1755 | ext4_ext_mark_unwritten(ex); | |
56055d3a AA |
1756 | |
1757 | if (ex + 1 < EXT_LAST_EXTENT(eh)) { | |
1758 | len = (EXT_LAST_EXTENT(eh) - ex - 1) | |
1759 | * sizeof(struct ext4_extent); | |
1760 | memmove(ex + 1, ex + 2, len); | |
1761 | } | |
e8546d06 | 1762 | le16_add_cpu(&eh->eh_entries, -1); |
56055d3a AA |
1763 | merge_done = 1; |
1764 | WARN_ON(eh->eh_entries == 0); | |
1765 | if (!eh->eh_entries) | |
24676da4 | 1766 | EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!"); |
56055d3a AA |
1767 | } |
1768 | ||
1769 | return merge_done; | |
1770 | } | |
1771 | ||
ecb94f5f TT |
1772 | /* |
1773 | * This function does a very simple check to see if we can collapse | |
1774 | * an extent tree with a single extent tree leaf block into the inode. | |
1775 | */ | |
1776 | static void ext4_ext_try_to_merge_up(handle_t *handle, | |
1777 | struct inode *inode, | |
1778 | struct ext4_ext_path *path) | |
1779 | { | |
1780 | size_t s; | |
1781 | unsigned max_root = ext4_ext_space_root(inode, 0); | |
1782 | ext4_fsblk_t blk; | |
1783 | ||
1784 | if ((path[0].p_depth != 1) || | |
1785 | (le16_to_cpu(path[0].p_hdr->eh_entries) != 1) || | |
1786 | (le16_to_cpu(path[1].p_hdr->eh_entries) > max_root)) | |
1787 | return; | |
1788 | ||
1789 | /* | |
1790 | * We need to modify the block allocation bitmap and the block | |
1791 | * group descriptor to release the extent tree block. If we | |
1792 | * can't get the journal credits, give up. | |
1793 | */ | |
1794 | if (ext4_journal_extend(handle, 2)) | |
1795 | return; | |
1796 | ||
1797 | /* | |
1798 | * Copy the extent data up to the inode | |
1799 | */ | |
1800 | blk = ext4_idx_pblock(path[0].p_idx); | |
1801 | s = le16_to_cpu(path[1].p_hdr->eh_entries) * | |
1802 | sizeof(struct ext4_extent_idx); | |
1803 | s += sizeof(struct ext4_extent_header); | |
1804 | ||
1805 | memcpy(path[0].p_hdr, path[1].p_hdr, s); | |
1806 | path[0].p_depth = 0; | |
1807 | path[0].p_ext = EXT_FIRST_EXTENT(path[0].p_hdr) + | |
1808 | (path[1].p_ext - EXT_FIRST_EXTENT(path[1].p_hdr)); | |
1809 | path[0].p_hdr->eh_max = cpu_to_le16(max_root); | |
1810 | ||
1811 | brelse(path[1].p_bh); | |
1812 | ext4_free_blocks(handle, inode, NULL, blk, 1, | |
71d4f7d0 | 1813 | EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET); |
ecb94f5f TT |
1814 | } |
1815 | ||
197217a5 YY |
1816 | /* |
1817 | * This function tries to merge the @ex extent to neighbours in the tree. | |
1818 | * return 1 if merge left else 0. | |
1819 | */ | |
ecb94f5f TT |
1820 | static void ext4_ext_try_to_merge(handle_t *handle, |
1821 | struct inode *inode, | |
197217a5 YY |
1822 | struct ext4_ext_path *path, |
1823 | struct ext4_extent *ex) { | |
1824 | struct ext4_extent_header *eh; | |
1825 | unsigned int depth; | |
1826 | int merge_done = 0; | |
197217a5 YY |
1827 | |
1828 | depth = ext_depth(inode); | |
1829 | BUG_ON(path[depth].p_hdr == NULL); | |
1830 | eh = path[depth].p_hdr; | |
1831 | ||
1832 | if (ex > EXT_FIRST_EXTENT(eh)) | |
1833 | merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1); | |
1834 | ||
1835 | if (!merge_done) | |
ecb94f5f | 1836 | (void) ext4_ext_try_to_merge_right(inode, path, ex); |
197217a5 | 1837 | |
ecb94f5f | 1838 | ext4_ext_try_to_merge_up(handle, inode, path); |
197217a5 YY |
1839 | } |
1840 | ||
25d14f98 AA |
1841 | /* |
1842 | * check if a portion of the "newext" extent overlaps with an | |
1843 | * existing extent. | |
1844 | * | |
1845 | * If there is an overlap discovered, it updates the length of the newext | |
1846 | * such that there will be no overlap, and then returns 1. | |
1847 | * If there is no overlap found, it returns 0. | |
1848 | */ | |
4d33b1ef TT |
1849 | static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi, |
1850 | struct inode *inode, | |
1f109d5a TT |
1851 | struct ext4_extent *newext, |
1852 | struct ext4_ext_path *path) | |
25d14f98 | 1853 | { |
725d26d3 | 1854 | ext4_lblk_t b1, b2; |
25d14f98 AA |
1855 | unsigned int depth, len1; |
1856 | unsigned int ret = 0; | |
1857 | ||
1858 | b1 = le32_to_cpu(newext->ee_block); | |
a2df2a63 | 1859 | len1 = ext4_ext_get_actual_len(newext); |
25d14f98 AA |
1860 | depth = ext_depth(inode); |
1861 | if (!path[depth].p_ext) | |
1862 | goto out; | |
f5a44db5 | 1863 | b2 = EXT4_LBLK_CMASK(sbi, le32_to_cpu(path[depth].p_ext->ee_block)); |
25d14f98 AA |
1864 | |
1865 | /* | |
1866 | * get the next allocated block if the extent in the path | |
2b2d6d01 | 1867 | * is before the requested block(s) |
25d14f98 AA |
1868 | */ |
1869 | if (b2 < b1) { | |
1870 | b2 = ext4_ext_next_allocated_block(path); | |
f17722f9 | 1871 | if (b2 == EXT_MAX_BLOCKS) |
25d14f98 | 1872 | goto out; |
f5a44db5 | 1873 | b2 = EXT4_LBLK_CMASK(sbi, b2); |
25d14f98 AA |
1874 | } |
1875 | ||
725d26d3 | 1876 | /* check for wrap through zero on extent logical start block*/ |
25d14f98 | 1877 | if (b1 + len1 < b1) { |
f17722f9 | 1878 | len1 = EXT_MAX_BLOCKS - b1; |
25d14f98 AA |
1879 | newext->ee_len = cpu_to_le16(len1); |
1880 | ret = 1; | |
1881 | } | |
1882 | ||
1883 | /* check for overlap */ | |
1884 | if (b1 + len1 > b2) { | |
1885 | newext->ee_len = cpu_to_le16(b2 - b1); | |
1886 | ret = 1; | |
1887 | } | |
1888 | out: | |
1889 | return ret; | |
1890 | } | |
1891 | ||
a86c6181 | 1892 | /* |
d0d856e8 RD |
1893 | * ext4_ext_insert_extent: |
1894 | * tries to merge requsted extent into the existing extent or | |
1895 | * inserts requested extent as new one into the tree, | |
1896 | * creating new leaf in the no-space case. | |
a86c6181 AT |
1897 | */ |
1898 | int ext4_ext_insert_extent(handle_t *handle, struct inode *inode, | |
1899 | struct ext4_ext_path *path, | |
107a7bd3 | 1900 | struct ext4_extent *newext, int gb_flags) |
a86c6181 | 1901 | { |
af5bc92d | 1902 | struct ext4_extent_header *eh; |
a86c6181 AT |
1903 | struct ext4_extent *ex, *fex; |
1904 | struct ext4_extent *nearex; /* nearest extent */ | |
1905 | struct ext4_ext_path *npath = NULL; | |
725d26d3 AK |
1906 | int depth, len, err; |
1907 | ext4_lblk_t next; | |
556615dc | 1908 | int mb_flags = 0, unwritten; |
a86c6181 | 1909 | |
273df556 FM |
1910 | if (unlikely(ext4_ext_get_actual_len(newext) == 0)) { |
1911 | EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0"); | |
1912 | return -EIO; | |
1913 | } | |
a86c6181 AT |
1914 | depth = ext_depth(inode); |
1915 | ex = path[depth].p_ext; | |
be8981be | 1916 | eh = path[depth].p_hdr; |
273df556 FM |
1917 | if (unlikely(path[depth].p_hdr == NULL)) { |
1918 | EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); | |
1919 | return -EIO; | |
1920 | } | |
a86c6181 AT |
1921 | |
1922 | /* try to insert block into found extent and return */ | |
107a7bd3 | 1923 | if (ex && !(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) { |
a2df2a63 AA |
1924 | |
1925 | /* | |
be8981be LC |
1926 | * Try to see whether we should rather test the extent on |
1927 | * right from ex, or from the left of ex. This is because | |
1928 | * ext4_ext_find_extent() can return either extent on the | |
1929 | * left, or on the right from the searched position. This | |
1930 | * will make merging more effective. | |
a2df2a63 | 1931 | */ |
be8981be LC |
1932 | if (ex < EXT_LAST_EXTENT(eh) && |
1933 | (le32_to_cpu(ex->ee_block) + | |
1934 | ext4_ext_get_actual_len(ex) < | |
1935 | le32_to_cpu(newext->ee_block))) { | |
1936 | ex += 1; | |
1937 | goto prepend; | |
1938 | } else if ((ex > EXT_FIRST_EXTENT(eh)) && | |
1939 | (le32_to_cpu(newext->ee_block) + | |
1940 | ext4_ext_get_actual_len(newext) < | |
1941 | le32_to_cpu(ex->ee_block))) | |
1942 | ex -= 1; | |
1943 | ||
1944 | /* Try to append newex to the ex */ | |
1945 | if (ext4_can_extents_be_merged(inode, ex, newext)) { | |
1946 | ext_debug("append [%d]%d block to %u:[%d]%d" | |
1947 | "(from %llu)\n", | |
556615dc | 1948 | ext4_ext_is_unwritten(newext), |
be8981be LC |
1949 | ext4_ext_get_actual_len(newext), |
1950 | le32_to_cpu(ex->ee_block), | |
556615dc | 1951 | ext4_ext_is_unwritten(ex), |
be8981be LC |
1952 | ext4_ext_get_actual_len(ex), |
1953 | ext4_ext_pblock(ex)); | |
1954 | err = ext4_ext_get_access(handle, inode, | |
1955 | path + depth); | |
1956 | if (err) | |
1957 | return err; | |
556615dc | 1958 | unwritten = ext4_ext_is_unwritten(ex); |
be8981be | 1959 | ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) |
a2df2a63 | 1960 | + ext4_ext_get_actual_len(newext)); |
556615dc LC |
1961 | if (unwritten) |
1962 | ext4_ext_mark_unwritten(ex); | |
be8981be LC |
1963 | eh = path[depth].p_hdr; |
1964 | nearex = ex; | |
1965 | goto merge; | |
1966 | } | |
1967 | ||
1968 | prepend: | |
1969 | /* Try to prepend newex to the ex */ | |
1970 | if (ext4_can_extents_be_merged(inode, newext, ex)) { | |
1971 | ext_debug("prepend %u[%d]%d block to %u:[%d]%d" | |
1972 | "(from %llu)\n", | |
1973 | le32_to_cpu(newext->ee_block), | |
556615dc | 1974 | ext4_ext_is_unwritten(newext), |
be8981be LC |
1975 | ext4_ext_get_actual_len(newext), |
1976 | le32_to_cpu(ex->ee_block), | |
556615dc | 1977 | ext4_ext_is_unwritten(ex), |
be8981be LC |
1978 | ext4_ext_get_actual_len(ex), |
1979 | ext4_ext_pblock(ex)); | |
1980 | err = ext4_ext_get_access(handle, inode, | |
1981 | path + depth); | |
1982 | if (err) | |
1983 | return err; | |
1984 | ||
556615dc | 1985 | unwritten = ext4_ext_is_unwritten(ex); |
be8981be LC |
1986 | ex->ee_block = newext->ee_block; |
1987 | ext4_ext_store_pblock(ex, ext4_ext_pblock(newext)); | |
1988 | ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) | |
1989 | + ext4_ext_get_actual_len(newext)); | |
556615dc LC |
1990 | if (unwritten) |
1991 | ext4_ext_mark_unwritten(ex); | |
be8981be LC |
1992 | eh = path[depth].p_hdr; |
1993 | nearex = ex; | |
1994 | goto merge; | |
1995 | } | |
a86c6181 AT |
1996 | } |
1997 | ||
a86c6181 AT |
1998 | depth = ext_depth(inode); |
1999 | eh = path[depth].p_hdr; | |
2000 | if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) | |
2001 | goto has_space; | |
2002 | ||
2003 | /* probably next leaf has space for us? */ | |
2004 | fex = EXT_LAST_EXTENT(eh); | |
598dbdf2 RD |
2005 | next = EXT_MAX_BLOCKS; |
2006 | if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)) | |
5718789d | 2007 | next = ext4_ext_next_leaf_block(path); |
598dbdf2 | 2008 | if (next != EXT_MAX_BLOCKS) { |
32de6756 | 2009 | ext_debug("next leaf block - %u\n", next); |
a86c6181 | 2010 | BUG_ON(npath != NULL); |
107a7bd3 | 2011 | npath = ext4_ext_find_extent(inode, next, NULL, 0); |
a86c6181 AT |
2012 | if (IS_ERR(npath)) |
2013 | return PTR_ERR(npath); | |
2014 | BUG_ON(npath->p_depth != path->p_depth); | |
2015 | eh = npath[depth].p_hdr; | |
2016 | if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) { | |
25985edc | 2017 | ext_debug("next leaf isn't full(%d)\n", |
a86c6181 AT |
2018 | le16_to_cpu(eh->eh_entries)); |
2019 | path = npath; | |
ffb505ff | 2020 | goto has_space; |
a86c6181 AT |
2021 | } |
2022 | ext_debug("next leaf has no free space(%d,%d)\n", | |
2023 | le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max)); | |
2024 | } | |
2025 | ||
2026 | /* | |
d0d856e8 RD |
2027 | * There is no free space in the found leaf. |
2028 | * We're gonna add a new leaf in the tree. | |
a86c6181 | 2029 | */ |
107a7bd3 TT |
2030 | if (gb_flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) |
2031 | mb_flags = EXT4_MB_USE_RESERVED; | |
2032 | err = ext4_ext_create_new_leaf(handle, inode, mb_flags, gb_flags, | |
2033 | path, newext); | |
a86c6181 AT |
2034 | if (err) |
2035 | goto cleanup; | |
2036 | depth = ext_depth(inode); | |
2037 | eh = path[depth].p_hdr; | |
2038 | ||
2039 | has_space: | |
2040 | nearex = path[depth].p_ext; | |
2041 | ||
7e028976 AM |
2042 | err = ext4_ext_get_access(handle, inode, path + depth); |
2043 | if (err) | |
a86c6181 AT |
2044 | goto cleanup; |
2045 | ||
2046 | if (!nearex) { | |
2047 | /* there is no extent in this leaf, create first one */ | |
32de6756 | 2048 | ext_debug("first extent in the leaf: %u:%llu:[%d]%d\n", |
8c55e204 | 2049 | le32_to_cpu(newext->ee_block), |
bf89d16f | 2050 | ext4_ext_pblock(newext), |
556615dc | 2051 | ext4_ext_is_unwritten(newext), |
a2df2a63 | 2052 | ext4_ext_get_actual_len(newext)); |
80e675f9 EG |
2053 | nearex = EXT_FIRST_EXTENT(eh); |
2054 | } else { | |
2055 | if (le32_to_cpu(newext->ee_block) | |
8c55e204 | 2056 | > le32_to_cpu(nearex->ee_block)) { |
80e675f9 | 2057 | /* Insert after */ |
32de6756 YY |
2058 | ext_debug("insert %u:%llu:[%d]%d before: " |
2059 | "nearest %p\n", | |
80e675f9 EG |
2060 | le32_to_cpu(newext->ee_block), |
2061 | ext4_ext_pblock(newext), | |
556615dc | 2062 | ext4_ext_is_unwritten(newext), |
80e675f9 EG |
2063 | ext4_ext_get_actual_len(newext), |
2064 | nearex); | |
2065 | nearex++; | |
2066 | } else { | |
2067 | /* Insert before */ | |
2068 | BUG_ON(newext->ee_block == nearex->ee_block); | |
32de6756 YY |
2069 | ext_debug("insert %u:%llu:[%d]%d after: " |
2070 | "nearest %p\n", | |
8c55e204 | 2071 | le32_to_cpu(newext->ee_block), |
bf89d16f | 2072 | ext4_ext_pblock(newext), |
556615dc | 2073 | ext4_ext_is_unwritten(newext), |
a2df2a63 | 2074 | ext4_ext_get_actual_len(newext), |
80e675f9 EG |
2075 | nearex); |
2076 | } | |
2077 | len = EXT_LAST_EXTENT(eh) - nearex + 1; | |
2078 | if (len > 0) { | |
32de6756 | 2079 | ext_debug("insert %u:%llu:[%d]%d: " |
80e675f9 EG |
2080 | "move %d extents from 0x%p to 0x%p\n", |
2081 | le32_to_cpu(newext->ee_block), | |
2082 | ext4_ext_pblock(newext), | |
556615dc | 2083 | ext4_ext_is_unwritten(newext), |
80e675f9 EG |
2084 | ext4_ext_get_actual_len(newext), |
2085 | len, nearex, nearex + 1); | |
2086 | memmove(nearex + 1, nearex, | |
2087 | len * sizeof(struct ext4_extent)); | |
a86c6181 | 2088 | } |
a86c6181 AT |
2089 | } |
2090 | ||
e8546d06 | 2091 | le16_add_cpu(&eh->eh_entries, 1); |
80e675f9 | 2092 | path[depth].p_ext = nearex; |
a86c6181 | 2093 | nearex->ee_block = newext->ee_block; |
bf89d16f | 2094 | ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext)); |
a86c6181 | 2095 | nearex->ee_len = newext->ee_len; |
a86c6181 AT |
2096 | |
2097 | merge: | |
e7bcf823 | 2098 | /* try to merge extents */ |
107a7bd3 | 2099 | if (!(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) |
ecb94f5f | 2100 | ext4_ext_try_to_merge(handle, inode, path, nearex); |
a86c6181 | 2101 | |
a86c6181 AT |
2102 | |
2103 | /* time to correct all indexes above */ | |
2104 | err = ext4_ext_correct_indexes(handle, inode, path); | |
2105 | if (err) | |
2106 | goto cleanup; | |
2107 | ||
ecb94f5f | 2108 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); |
a86c6181 AT |
2109 | |
2110 | cleanup: | |
2111 | if (npath) { | |
2112 | ext4_ext_drop_refs(npath); | |
2113 | kfree(npath); | |
2114 | } | |
a86c6181 AT |
2115 | return err; |
2116 | } | |
2117 | ||
91dd8c11 LC |
2118 | static int ext4_fill_fiemap_extents(struct inode *inode, |
2119 | ext4_lblk_t block, ext4_lblk_t num, | |
2120 | struct fiemap_extent_info *fieinfo) | |
6873fa0d ES |
2121 | { |
2122 | struct ext4_ext_path *path = NULL; | |
6873fa0d | 2123 | struct ext4_extent *ex; |
69eb33dc | 2124 | struct extent_status es; |
91dd8c11 | 2125 | ext4_lblk_t next, next_del, start = 0, end = 0; |
6873fa0d | 2126 | ext4_lblk_t last = block + num; |
91dd8c11 LC |
2127 | int exists, depth = 0, err = 0; |
2128 | unsigned int flags = 0; | |
2129 | unsigned char blksize_bits = inode->i_sb->s_blocksize_bits; | |
6873fa0d | 2130 | |
f17722f9 | 2131 | while (block < last && block != EXT_MAX_BLOCKS) { |
6873fa0d ES |
2132 | num = last - block; |
2133 | /* find extent for this block */ | |
fab3a549 | 2134 | down_read(&EXT4_I(inode)->i_data_sem); |
91dd8c11 LC |
2135 | |
2136 | if (path && ext_depth(inode) != depth) { | |
2137 | /* depth was changed. we have to realloc path */ | |
2138 | kfree(path); | |
2139 | path = NULL; | |
2140 | } | |
2141 | ||
107a7bd3 | 2142 | path = ext4_ext_find_extent(inode, block, path, 0); |
6873fa0d | 2143 | if (IS_ERR(path)) { |
91dd8c11 | 2144 | up_read(&EXT4_I(inode)->i_data_sem); |
6873fa0d ES |
2145 | err = PTR_ERR(path); |
2146 | path = NULL; | |
2147 | break; | |
2148 | } | |
2149 | ||
2150 | depth = ext_depth(inode); | |
273df556 | 2151 | if (unlikely(path[depth].p_hdr == NULL)) { |
91dd8c11 | 2152 | up_read(&EXT4_I(inode)->i_data_sem); |
273df556 FM |
2153 | EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); |
2154 | err = -EIO; | |
2155 | break; | |
2156 | } | |
6873fa0d ES |
2157 | ex = path[depth].p_ext; |
2158 | next = ext4_ext_next_allocated_block(path); | |
91dd8c11 | 2159 | ext4_ext_drop_refs(path); |
6873fa0d | 2160 | |
91dd8c11 | 2161 | flags = 0; |
6873fa0d ES |
2162 | exists = 0; |
2163 | if (!ex) { | |
2164 | /* there is no extent yet, so try to allocate | |
2165 | * all requested space */ | |
2166 | start = block; | |
2167 | end = block + num; | |
2168 | } else if (le32_to_cpu(ex->ee_block) > block) { | |
2169 | /* need to allocate space before found extent */ | |
2170 | start = block; | |
2171 | end = le32_to_cpu(ex->ee_block); | |
2172 | if (block + num < end) | |
2173 | end = block + num; | |
2174 | } else if (block >= le32_to_cpu(ex->ee_block) | |
2175 | + ext4_ext_get_actual_len(ex)) { | |
2176 | /* need to allocate space after found extent */ | |
2177 | start = block; | |
2178 | end = block + num; | |
2179 | if (end >= next) | |
2180 | end = next; | |
2181 | } else if (block >= le32_to_cpu(ex->ee_block)) { | |
2182 | /* | |
2183 | * some part of requested space is covered | |
2184 | * by found extent | |
2185 | */ | |
2186 | start = block; | |
2187 | end = le32_to_cpu(ex->ee_block) | |
2188 | + ext4_ext_get_actual_len(ex); | |
2189 | if (block + num < end) | |
2190 | end = block + num; | |
2191 | exists = 1; | |
2192 | } else { | |
2193 | BUG(); | |
2194 | } | |
2195 | BUG_ON(end <= start); | |
2196 | ||
2197 | if (!exists) { | |
69eb33dc ZL |
2198 | es.es_lblk = start; |
2199 | es.es_len = end - start; | |
2200 | es.es_pblk = 0; | |
6873fa0d | 2201 | } else { |
69eb33dc ZL |
2202 | es.es_lblk = le32_to_cpu(ex->ee_block); |
2203 | es.es_len = ext4_ext_get_actual_len(ex); | |
2204 | es.es_pblk = ext4_ext_pblock(ex); | |
556615dc | 2205 | if (ext4_ext_is_unwritten(ex)) |
91dd8c11 | 2206 | flags |= FIEMAP_EXTENT_UNWRITTEN; |
6873fa0d ES |
2207 | } |
2208 | ||
91dd8c11 | 2209 | /* |
69eb33dc ZL |
2210 | * Find delayed extent and update es accordingly. We call |
2211 | * it even in !exists case to find out whether es is the | |
91dd8c11 LC |
2212 | * last existing extent or not. |
2213 | */ | |
69eb33dc | 2214 | next_del = ext4_find_delayed_extent(inode, &es); |
91dd8c11 LC |
2215 | if (!exists && next_del) { |
2216 | exists = 1; | |
72dac95d JL |
2217 | flags |= (FIEMAP_EXTENT_DELALLOC | |
2218 | FIEMAP_EXTENT_UNKNOWN); | |
91dd8c11 LC |
2219 | } |
2220 | up_read(&EXT4_I(inode)->i_data_sem); | |
2221 | ||
69eb33dc ZL |
2222 | if (unlikely(es.es_len == 0)) { |
2223 | EXT4_ERROR_INODE(inode, "es.es_len == 0"); | |
273df556 FM |
2224 | err = -EIO; |
2225 | break; | |
2226 | } | |
6873fa0d | 2227 | |
f7fec032 ZL |
2228 | /* |
2229 | * This is possible iff next == next_del == EXT_MAX_BLOCKS. | |
2230 | * we need to check next == EXT_MAX_BLOCKS because it is | |
2231 | * possible that an extent is with unwritten and delayed | |
2232 | * status due to when an extent is delayed allocated and | |
2233 | * is allocated by fallocate status tree will track both of | |
2234 | * them in a extent. | |
2235 | * | |
2236 | * So we could return a unwritten and delayed extent, and | |
2237 | * its block is equal to 'next'. | |
2238 | */ | |
2239 | if (next == next_del && next == EXT_MAX_BLOCKS) { | |
91dd8c11 LC |
2240 | flags |= FIEMAP_EXTENT_LAST; |
2241 | if (unlikely(next_del != EXT_MAX_BLOCKS || | |
2242 | next != EXT_MAX_BLOCKS)) { | |
2243 | EXT4_ERROR_INODE(inode, | |
2244 | "next extent == %u, next " | |
2245 | "delalloc extent = %u", | |
2246 | next, next_del); | |
2247 | err = -EIO; | |
2248 | break; | |
2249 | } | |
6873fa0d ES |
2250 | } |
2251 | ||
91dd8c11 LC |
2252 | if (exists) { |
2253 | err = fiemap_fill_next_extent(fieinfo, | |
69eb33dc ZL |
2254 | (__u64)es.es_lblk << blksize_bits, |
2255 | (__u64)es.es_pblk << blksize_bits, | |
2256 | (__u64)es.es_len << blksize_bits, | |
91dd8c11 LC |
2257 | flags); |
2258 | if (err < 0) | |
2259 | break; | |
2260 | if (err == 1) { | |
2261 | err = 0; | |
2262 | break; | |
2263 | } | |
6873fa0d ES |
2264 | } |
2265 | ||
69eb33dc | 2266 | block = es.es_lblk + es.es_len; |
6873fa0d ES |
2267 | } |
2268 | ||
2269 | if (path) { | |
2270 | ext4_ext_drop_refs(path); | |
2271 | kfree(path); | |
2272 | } | |
2273 | ||
2274 | return err; | |
2275 | } | |
2276 | ||
a86c6181 | 2277 | /* |
d0d856e8 RD |
2278 | * ext4_ext_put_gap_in_cache: |
2279 | * calculate boundaries of the gap that the requested block fits into | |
a86c6181 AT |
2280 | * and cache this gap |
2281 | */ | |
09b88252 | 2282 | static void |
a86c6181 | 2283 | ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, |
725d26d3 | 2284 | ext4_lblk_t block) |
a86c6181 AT |
2285 | { |
2286 | int depth = ext_depth(inode); | |
27b1b228 AS |
2287 | unsigned long len = 0; |
2288 | ext4_lblk_t lblock = 0; | |
a86c6181 AT |
2289 | struct ext4_extent *ex; |
2290 | ||
2291 | ex = path[depth].p_ext; | |
2292 | if (ex == NULL) { | |
69eb33dc ZL |
2293 | /* |
2294 | * there is no extent yet, so gap is [0;-] and we | |
2295 | * don't cache it | |
2296 | */ | |
a86c6181 AT |
2297 | ext_debug("cache gap(whole file):"); |
2298 | } else if (block < le32_to_cpu(ex->ee_block)) { | |
2299 | lblock = block; | |
2300 | len = le32_to_cpu(ex->ee_block) - block; | |
bba90743 ES |
2301 | ext_debug("cache gap(before): %u [%u:%u]", |
2302 | block, | |
2303 | le32_to_cpu(ex->ee_block), | |
2304 | ext4_ext_get_actual_len(ex)); | |
d100eef2 ZL |
2305 | if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1)) |
2306 | ext4_es_insert_extent(inode, lblock, len, ~0, | |
2307 | EXTENT_STATUS_HOLE); | |
a86c6181 | 2308 | } else if (block >= le32_to_cpu(ex->ee_block) |
a2df2a63 | 2309 | + ext4_ext_get_actual_len(ex)) { |
725d26d3 | 2310 | ext4_lblk_t next; |
8c55e204 | 2311 | lblock = le32_to_cpu(ex->ee_block) |
a2df2a63 | 2312 | + ext4_ext_get_actual_len(ex); |
725d26d3 AK |
2313 | |
2314 | next = ext4_ext_next_allocated_block(path); | |
bba90743 ES |
2315 | ext_debug("cache gap(after): [%u:%u] %u", |
2316 | le32_to_cpu(ex->ee_block), | |
2317 | ext4_ext_get_actual_len(ex), | |
2318 | block); | |
725d26d3 AK |
2319 | BUG_ON(next == lblock); |
2320 | len = next - lblock; | |
d100eef2 ZL |
2321 | if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1)) |
2322 | ext4_es_insert_extent(inode, lblock, len, ~0, | |
2323 | EXTENT_STATUS_HOLE); | |
a86c6181 | 2324 | } else { |
a86c6181 AT |
2325 | BUG(); |
2326 | } | |
2327 | ||
bba90743 | 2328 | ext_debug(" -> %u:%lu\n", lblock, len); |
a86c6181 AT |
2329 | } |
2330 | ||
2331 | /* | |
d0d856e8 RD |
2332 | * ext4_ext_rm_idx: |
2333 | * removes index from the index block. | |
a86c6181 | 2334 | */ |
1d03ec98 | 2335 | static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode, |
c36575e6 | 2336 | struct ext4_ext_path *path, int depth) |
a86c6181 | 2337 | { |
a86c6181 | 2338 | int err; |
f65e6fba | 2339 | ext4_fsblk_t leaf; |
a86c6181 AT |
2340 | |
2341 | /* free index block */ | |
c36575e6 FL |
2342 | depth--; |
2343 | path = path + depth; | |
bf89d16f | 2344 | leaf = ext4_idx_pblock(path->p_idx); |
273df556 FM |
2345 | if (unlikely(path->p_hdr->eh_entries == 0)) { |
2346 | EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0"); | |
2347 | return -EIO; | |
2348 | } | |
7e028976 AM |
2349 | err = ext4_ext_get_access(handle, inode, path); |
2350 | if (err) | |
a86c6181 | 2351 | return err; |
0e1147b0 RD |
2352 | |
2353 | if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) { | |
2354 | int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx; | |
2355 | len *= sizeof(struct ext4_extent_idx); | |
2356 | memmove(path->p_idx, path->p_idx + 1, len); | |
2357 | } | |
2358 | ||
e8546d06 | 2359 | le16_add_cpu(&path->p_hdr->eh_entries, -1); |
7e028976 AM |
2360 | err = ext4_ext_dirty(handle, inode, path); |
2361 | if (err) | |
a86c6181 | 2362 | return err; |
2ae02107 | 2363 | ext_debug("index is empty, remove it, free block %llu\n", leaf); |
d8990240 AK |
2364 | trace_ext4_ext_rm_idx(inode, leaf); |
2365 | ||
7dc57615 | 2366 | ext4_free_blocks(handle, inode, NULL, leaf, 1, |
e6362609 | 2367 | EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET); |
c36575e6 FL |
2368 | |
2369 | while (--depth >= 0) { | |
2370 | if (path->p_idx != EXT_FIRST_INDEX(path->p_hdr)) | |
2371 | break; | |
2372 | path--; | |
2373 | err = ext4_ext_get_access(handle, inode, path); | |
2374 | if (err) | |
2375 | break; | |
2376 | path->p_idx->ei_block = (path+1)->p_idx->ei_block; | |
2377 | err = ext4_ext_dirty(handle, inode, path); | |
2378 | if (err) | |
2379 | break; | |
2380 | } | |
a86c6181 AT |
2381 | return err; |
2382 | } | |
2383 | ||
2384 | /* | |
ee12b630 MC |
2385 | * ext4_ext_calc_credits_for_single_extent: |
2386 | * This routine returns max. credits that needed to insert an extent | |
2387 | * to the extent tree. | |
2388 | * When pass the actual path, the caller should calculate credits | |
2389 | * under i_data_sem. | |
a86c6181 | 2390 | */ |
525f4ed8 | 2391 | int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks, |
a86c6181 AT |
2392 | struct ext4_ext_path *path) |
2393 | { | |
a86c6181 | 2394 | if (path) { |
ee12b630 | 2395 | int depth = ext_depth(inode); |
f3bd1f3f | 2396 | int ret = 0; |
ee12b630 | 2397 | |
a86c6181 | 2398 | /* probably there is space in leaf? */ |
a86c6181 | 2399 | if (le16_to_cpu(path[depth].p_hdr->eh_entries) |
ee12b630 | 2400 | < le16_to_cpu(path[depth].p_hdr->eh_max)) { |
a86c6181 | 2401 | |
ee12b630 MC |
2402 | /* |
2403 | * There are some space in the leaf tree, no | |
2404 | * need to account for leaf block credit | |
2405 | * | |
2406 | * bitmaps and block group descriptor blocks | |
df3ab170 | 2407 | * and other metadata blocks still need to be |
ee12b630 MC |
2408 | * accounted. |
2409 | */ | |
525f4ed8 | 2410 | /* 1 bitmap, 1 block group descriptor */ |
ee12b630 | 2411 | ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb); |
5887e98b | 2412 | return ret; |
ee12b630 MC |
2413 | } |
2414 | } | |
a86c6181 | 2415 | |
525f4ed8 | 2416 | return ext4_chunk_trans_blocks(inode, nrblocks); |
ee12b630 | 2417 | } |
a86c6181 | 2418 | |
ee12b630 | 2419 | /* |
fffb2739 | 2420 | * How many index/leaf blocks need to change/allocate to add @extents extents? |
ee12b630 | 2421 | * |
fffb2739 JK |
2422 | * If we add a single extent, then in the worse case, each tree level |
2423 | * index/leaf need to be changed in case of the tree split. | |
ee12b630 | 2424 | * |
fffb2739 JK |
2425 | * If more extents are inserted, they could cause the whole tree split more |
2426 | * than once, but this is really rare. | |
ee12b630 | 2427 | */ |
fffb2739 | 2428 | int ext4_ext_index_trans_blocks(struct inode *inode, int extents) |
ee12b630 MC |
2429 | { |
2430 | int index; | |
f19d5870 TM |
2431 | int depth; |
2432 | ||
2433 | /* If we are converting the inline data, only one is needed here. */ | |
2434 | if (ext4_has_inline_data(inode)) | |
2435 | return 1; | |
2436 | ||
2437 | depth = ext_depth(inode); | |
a86c6181 | 2438 | |
fffb2739 | 2439 | if (extents <= 1) |
ee12b630 MC |
2440 | index = depth * 2; |
2441 | else | |
2442 | index = depth * 3; | |
a86c6181 | 2443 | |
ee12b630 | 2444 | return index; |
a86c6181 AT |
2445 | } |
2446 | ||
981250ca TT |
2447 | static inline int get_default_free_blocks_flags(struct inode *inode) |
2448 | { | |
2449 | if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) | |
2450 | return EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET; | |
2451 | else if (ext4_should_journal_data(inode)) | |
2452 | return EXT4_FREE_BLOCKS_FORGET; | |
2453 | return 0; | |
2454 | } | |
2455 | ||
a86c6181 | 2456 | static int ext4_remove_blocks(handle_t *handle, struct inode *inode, |
0aa06000 | 2457 | struct ext4_extent *ex, |
d23142c6 | 2458 | long long *partial_cluster, |
0aa06000 | 2459 | ext4_lblk_t from, ext4_lblk_t to) |
a86c6181 | 2460 | { |
0aa06000 | 2461 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); |
a2df2a63 | 2462 | unsigned short ee_len = ext4_ext_get_actual_len(ex); |
0aa06000 | 2463 | ext4_fsblk_t pblk; |
981250ca | 2464 | int flags = get_default_free_blocks_flags(inode); |
18888cf0 | 2465 | |
0aa06000 TT |
2466 | /* |
2467 | * For bigalloc file systems, we never free a partial cluster | |
2468 | * at the beginning of the extent. Instead, we make a note | |
2469 | * that we tried freeing the cluster, and check to see if we | |
2470 | * need to free it on a subsequent call to ext4_remove_blocks, | |
2471 | * or at the end of the ext4_truncate() operation. | |
2472 | */ | |
2473 | flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER; | |
2474 | ||
d8990240 | 2475 | trace_ext4_remove_blocks(inode, ex, from, to, *partial_cluster); |
0aa06000 TT |
2476 | /* |
2477 | * If we have a partial cluster, and it's different from the | |
2478 | * cluster of the last block, we need to explicitly free the | |
2479 | * partial cluster here. | |
2480 | */ | |
2481 | pblk = ext4_ext_pblock(ex) + ee_len - 1; | |
d23142c6 LC |
2482 | if ((*partial_cluster > 0) && |
2483 | (EXT4_B2C(sbi, pblk) != *partial_cluster)) { | |
0aa06000 TT |
2484 | ext4_free_blocks(handle, inode, NULL, |
2485 | EXT4_C2B(sbi, *partial_cluster), | |
2486 | sbi->s_cluster_ratio, flags); | |
2487 | *partial_cluster = 0; | |
2488 | } | |
2489 | ||
a86c6181 AT |
2490 | #ifdef EXTENTS_STATS |
2491 | { | |
2492 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); | |
a86c6181 AT |
2493 | spin_lock(&sbi->s_ext_stats_lock); |
2494 | sbi->s_ext_blocks += ee_len; | |
2495 | sbi->s_ext_extents++; | |
2496 | if (ee_len < sbi->s_ext_min) | |
2497 | sbi->s_ext_min = ee_len; | |
2498 | if (ee_len > sbi->s_ext_max) | |
2499 | sbi->s_ext_max = ee_len; | |
2500 | if (ext_depth(inode) > sbi->s_depth_max) | |
2501 | sbi->s_depth_max = ext_depth(inode); | |
2502 | spin_unlock(&sbi->s_ext_stats_lock); | |
2503 | } | |
2504 | #endif | |
2505 | if (from >= le32_to_cpu(ex->ee_block) | |
a2df2a63 | 2506 | && to == le32_to_cpu(ex->ee_block) + ee_len - 1) { |
a86c6181 | 2507 | /* tail removal */ |
725d26d3 | 2508 | ext4_lblk_t num; |
d23142c6 | 2509 | unsigned int unaligned; |
725d26d3 | 2510 | |
a2df2a63 | 2511 | num = le32_to_cpu(ex->ee_block) + ee_len - from; |
0aa06000 | 2512 | pblk = ext4_ext_pblock(ex) + ee_len - num; |
d23142c6 LC |
2513 | /* |
2514 | * Usually we want to free partial cluster at the end of the | |
2515 | * extent, except for the situation when the cluster is still | |
2516 | * used by any other extent (partial_cluster is negative). | |
2517 | */ | |
2518 | if (*partial_cluster < 0 && | |
2519 | -(*partial_cluster) == EXT4_B2C(sbi, pblk + num - 1)) | |
2520 | flags |= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER; | |
2521 | ||
2522 | ext_debug("free last %u blocks starting %llu partial %lld\n", | |
2523 | num, pblk, *partial_cluster); | |
0aa06000 TT |
2524 | ext4_free_blocks(handle, inode, NULL, pblk, num, flags); |
2525 | /* | |
2526 | * If the block range to be freed didn't start at the | |
2527 | * beginning of a cluster, and we removed the entire | |
d23142c6 LC |
2528 | * extent and the cluster is not used by any other extent, |
2529 | * save the partial cluster here, since we might need to | |
2530 | * delete if we determine that the truncate operation has | |
2531 | * removed all of the blocks in the cluster. | |
2532 | * | |
2533 | * On the other hand, if we did not manage to free the whole | |
2534 | * extent, we have to mark the cluster as used (store negative | |
2535 | * cluster number in partial_cluster). | |
0aa06000 | 2536 | */ |
f5a44db5 | 2537 | unaligned = EXT4_PBLK_COFF(sbi, pblk); |
d23142c6 LC |
2538 | if (unaligned && (ee_len == num) && |
2539 | (*partial_cluster != -((long long)EXT4_B2C(sbi, pblk)))) | |
0aa06000 | 2540 | *partial_cluster = EXT4_B2C(sbi, pblk); |
d23142c6 LC |
2541 | else if (unaligned) |
2542 | *partial_cluster = -((long long)EXT4_B2C(sbi, pblk)); | |
2543 | else if (*partial_cluster > 0) | |
0aa06000 | 2544 | *partial_cluster = 0; |
78fb9cdf LC |
2545 | } else |
2546 | ext4_error(sbi->s_sb, "strange request: removal(2) " | |
2547 | "%u-%u from %u:%u\n", | |
2548 | from, to, le32_to_cpu(ex->ee_block), ee_len); | |
a86c6181 AT |
2549 | return 0; |
2550 | } | |
2551 | ||
d583fb87 AH |
2552 | |
2553 | /* | |
2554 | * ext4_ext_rm_leaf() Removes the extents associated with the | |
2555 | * blocks appearing between "start" and "end", and splits the extents | |
2556 | * if "start" and "end" appear in the same extent | |
2557 | * | |
2558 | * @handle: The journal handle | |
2559 | * @inode: The files inode | |
2560 | * @path: The path to the leaf | |
d23142c6 LC |
2561 | * @partial_cluster: The cluster which we'll have to free if all extents |
2562 | * has been released from it. It gets negative in case | |
2563 | * that the cluster is still used. | |
d583fb87 AH |
2564 | * @start: The first block to remove |
2565 | * @end: The last block to remove | |
2566 | */ | |
a86c6181 AT |
2567 | static int |
2568 | ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, | |
d23142c6 LC |
2569 | struct ext4_ext_path *path, |
2570 | long long *partial_cluster, | |
0aa06000 | 2571 | ext4_lblk_t start, ext4_lblk_t end) |
a86c6181 | 2572 | { |
0aa06000 | 2573 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); |
a86c6181 AT |
2574 | int err = 0, correct_index = 0; |
2575 | int depth = ext_depth(inode), credits; | |
2576 | struct ext4_extent_header *eh; | |
750c9c47 | 2577 | ext4_lblk_t a, b; |
725d26d3 AK |
2578 | unsigned num; |
2579 | ext4_lblk_t ex_ee_block; | |
a86c6181 | 2580 | unsigned short ex_ee_len; |
556615dc | 2581 | unsigned unwritten = 0; |
a86c6181 | 2582 | struct ext4_extent *ex; |
d23142c6 | 2583 | ext4_fsblk_t pblk; |
a86c6181 | 2584 | |
c29c0ae7 | 2585 | /* the header must be checked already in ext4_ext_remove_space() */ |
5f95d21f | 2586 | ext_debug("truncate since %u in leaf to %u\n", start, end); |
a86c6181 AT |
2587 | if (!path[depth].p_hdr) |
2588 | path[depth].p_hdr = ext_block_hdr(path[depth].p_bh); | |
2589 | eh = path[depth].p_hdr; | |
273df556 FM |
2590 | if (unlikely(path[depth].p_hdr == NULL)) { |
2591 | EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); | |
2592 | return -EIO; | |
2593 | } | |
a86c6181 | 2594 | /* find where to start removing */ |
6ae06ff5 AS |
2595 | ex = path[depth].p_ext; |
2596 | if (!ex) | |
2597 | ex = EXT_LAST_EXTENT(eh); | |
a86c6181 AT |
2598 | |
2599 | ex_ee_block = le32_to_cpu(ex->ee_block); | |
a2df2a63 | 2600 | ex_ee_len = ext4_ext_get_actual_len(ex); |
a86c6181 | 2601 | |
c0634493 EW |
2602 | /* |
2603 | * If we're starting with an extent other than the last one in the | |
2604 | * node, we need to see if it shares a cluster with the extent to | |
2605 | * the right (towards the end of the file). If its leftmost cluster | |
2606 | * is this extent's rightmost cluster and it is not cluster aligned, | |
2607 | * we'll mark it as a partial that is not to be deallocated. | |
2608 | */ | |
2609 | ||
2610 | if (ex != EXT_LAST_EXTENT(eh)) { | |
2611 | ext4_fsblk_t current_pblk, right_pblk; | |
2612 | long long current_cluster, right_cluster; | |
2613 | ||
2614 | current_pblk = ext4_ext_pblock(ex) + ex_ee_len - 1; | |
2615 | current_cluster = (long long)EXT4_B2C(sbi, current_pblk); | |
2616 | right_pblk = ext4_ext_pblock(ex + 1); | |
2617 | right_cluster = (long long)EXT4_B2C(sbi, right_pblk); | |
2618 | if (current_cluster == right_cluster && | |
2619 | EXT4_PBLK_COFF(sbi, right_pblk)) | |
2620 | *partial_cluster = -right_cluster; | |
2621 | } | |
2622 | ||
d8990240 AK |
2623 | trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster); |
2624 | ||
a86c6181 AT |
2625 | while (ex >= EXT_FIRST_EXTENT(eh) && |
2626 | ex_ee_block + ex_ee_len > start) { | |
a41f2071 | 2627 | |
556615dc LC |
2628 | if (ext4_ext_is_unwritten(ex)) |
2629 | unwritten = 1; | |
a41f2071 | 2630 | else |
556615dc | 2631 | unwritten = 0; |
a41f2071 | 2632 | |
553f9008 | 2633 | ext_debug("remove ext %u:[%d]%d\n", ex_ee_block, |
556615dc | 2634 | unwritten, ex_ee_len); |
a86c6181 AT |
2635 | path[depth].p_ext = ex; |
2636 | ||
2637 | a = ex_ee_block > start ? ex_ee_block : start; | |
d583fb87 AH |
2638 | b = ex_ee_block+ex_ee_len - 1 < end ? |
2639 | ex_ee_block+ex_ee_len - 1 : end; | |
a86c6181 AT |
2640 | |
2641 | ext_debug(" border %u:%u\n", a, b); | |
2642 | ||
d583fb87 | 2643 | /* If this extent is beyond the end of the hole, skip it */ |
5f95d21f | 2644 | if (end < ex_ee_block) { |
d23142c6 LC |
2645 | /* |
2646 | * We're going to skip this extent and move to another, | |
2647 | * so if this extent is not cluster aligned we have | |
2648 | * to mark the current cluster as used to avoid | |
2649 | * accidentally freeing it later on | |
2650 | */ | |
2651 | pblk = ext4_ext_pblock(ex); | |
f5a44db5 | 2652 | if (EXT4_PBLK_COFF(sbi, pblk)) |
d23142c6 LC |
2653 | *partial_cluster = |
2654 | -((long long)EXT4_B2C(sbi, pblk)); | |
d583fb87 AH |
2655 | ex--; |
2656 | ex_ee_block = le32_to_cpu(ex->ee_block); | |
2657 | ex_ee_len = ext4_ext_get_actual_len(ex); | |
2658 | continue; | |
750c9c47 | 2659 | } else if (b != ex_ee_block + ex_ee_len - 1) { |
dc1841d6 LC |
2660 | EXT4_ERROR_INODE(inode, |
2661 | "can not handle truncate %u:%u " | |
2662 | "on extent %u:%u", | |
2663 | start, end, ex_ee_block, | |
2664 | ex_ee_block + ex_ee_len - 1); | |
750c9c47 DM |
2665 | err = -EIO; |
2666 | goto out; | |
a86c6181 AT |
2667 | } else if (a != ex_ee_block) { |
2668 | /* remove tail of the extent */ | |
750c9c47 | 2669 | num = a - ex_ee_block; |
a86c6181 AT |
2670 | } else { |
2671 | /* remove whole extent: excellent! */ | |
a86c6181 | 2672 | num = 0; |
a86c6181 | 2673 | } |
34071da7 TT |
2674 | /* |
2675 | * 3 for leaf, sb, and inode plus 2 (bmap and group | |
2676 | * descriptor) for each block group; assume two block | |
2677 | * groups plus ex_ee_len/blocks_per_block_group for | |
2678 | * the worst case | |
2679 | */ | |
2680 | credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb)); | |
a86c6181 AT |
2681 | if (ex == EXT_FIRST_EXTENT(eh)) { |
2682 | correct_index = 1; | |
2683 | credits += (ext_depth(inode)) + 1; | |
2684 | } | |
5aca07eb | 2685 | credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb); |
a86c6181 | 2686 | |
487caeef | 2687 | err = ext4_ext_truncate_extend_restart(handle, inode, credits); |
9102e4fa | 2688 | if (err) |
a86c6181 | 2689 | goto out; |
a86c6181 AT |
2690 | |
2691 | err = ext4_ext_get_access(handle, inode, path + depth); | |
2692 | if (err) | |
2693 | goto out; | |
2694 | ||
0aa06000 TT |
2695 | err = ext4_remove_blocks(handle, inode, ex, partial_cluster, |
2696 | a, b); | |
a86c6181 AT |
2697 | if (err) |
2698 | goto out; | |
2699 | ||
750c9c47 | 2700 | if (num == 0) |
d0d856e8 | 2701 | /* this extent is removed; mark slot entirely unused */ |
f65e6fba | 2702 | ext4_ext_store_pblock(ex, 0); |
a86c6181 | 2703 | |
a86c6181 | 2704 | ex->ee_len = cpu_to_le16(num); |
749269fa | 2705 | /* |
556615dc | 2706 | * Do not mark unwritten if all the blocks in the |
749269fa AA |
2707 | * extent have been removed. |
2708 | */ | |
556615dc LC |
2709 | if (unwritten && num) |
2710 | ext4_ext_mark_unwritten(ex); | |
d583fb87 AH |
2711 | /* |
2712 | * If the extent was completely released, | |
2713 | * we need to remove it from the leaf | |
2714 | */ | |
2715 | if (num == 0) { | |
f17722f9 | 2716 | if (end != EXT_MAX_BLOCKS - 1) { |
d583fb87 AH |
2717 | /* |
2718 | * For hole punching, we need to scoot all the | |
2719 | * extents up when an extent is removed so that | |
2720 | * we dont have blank extents in the middle | |
2721 | */ | |
2722 | memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) * | |
2723 | sizeof(struct ext4_extent)); | |
2724 | ||
2725 | /* Now get rid of the one at the end */ | |
2726 | memset(EXT_LAST_EXTENT(eh), 0, | |
2727 | sizeof(struct ext4_extent)); | |
2728 | } | |
2729 | le16_add_cpu(&eh->eh_entries, -1); | |
d23142c6 | 2730 | } else if (*partial_cluster > 0) |
0aa06000 | 2731 | *partial_cluster = 0; |
d583fb87 | 2732 | |
750c9c47 DM |
2733 | err = ext4_ext_dirty(handle, inode, path + depth); |
2734 | if (err) | |
2735 | goto out; | |
2736 | ||
bf52c6f7 | 2737 | ext_debug("new extent: %u:%u:%llu\n", ex_ee_block, num, |
bf89d16f | 2738 | ext4_ext_pblock(ex)); |
a86c6181 AT |
2739 | ex--; |
2740 | ex_ee_block = le32_to_cpu(ex->ee_block); | |
a2df2a63 | 2741 | ex_ee_len = ext4_ext_get_actual_len(ex); |
a86c6181 AT |
2742 | } |
2743 | ||
2744 | if (correct_index && eh->eh_entries) | |
2745 | err = ext4_ext_correct_indexes(handle, inode, path); | |
2746 | ||
0aa06000 | 2747 | /* |
ad6599ab EW |
2748 | * If there's a partial cluster and at least one extent remains in |
2749 | * the leaf, free the partial cluster if it isn't shared with the | |
2750 | * current extent. If there's a partial cluster and no extents | |
2751 | * remain in the leaf, it can't be freed here. It can only be | |
2752 | * freed when it's possible to determine if it's not shared with | |
2753 | * any other extent - when the next leaf is processed or when space | |
2754 | * removal is complete. | |
0aa06000 | 2755 | */ |
ad6599ab | 2756 | if (*partial_cluster > 0 && eh->eh_entries && |
0aa06000 TT |
2757 | (EXT4_B2C(sbi, ext4_ext_pblock(ex) + ex_ee_len - 1) != |
2758 | *partial_cluster)) { | |
981250ca | 2759 | int flags = get_default_free_blocks_flags(inode); |
0aa06000 TT |
2760 | |
2761 | ext4_free_blocks(handle, inode, NULL, | |
2762 | EXT4_C2B(sbi, *partial_cluster), | |
2763 | sbi->s_cluster_ratio, flags); | |
2764 | *partial_cluster = 0; | |
2765 | } | |
2766 | ||
a86c6181 AT |
2767 | /* if this leaf is free, then we should |
2768 | * remove it from index block above */ | |
2769 | if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL) | |
c36575e6 | 2770 | err = ext4_ext_rm_idx(handle, inode, path, depth); |
a86c6181 AT |
2771 | |
2772 | out: | |
2773 | return err; | |
2774 | } | |
2775 | ||
2776 | /* | |
d0d856e8 RD |
2777 | * ext4_ext_more_to_rm: |
2778 | * returns 1 if current index has to be freed (even partial) | |
a86c6181 | 2779 | */ |
09b88252 | 2780 | static int |
a86c6181 AT |
2781 | ext4_ext_more_to_rm(struct ext4_ext_path *path) |
2782 | { | |
2783 | BUG_ON(path->p_idx == NULL); | |
2784 | ||
2785 | if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr)) | |
2786 | return 0; | |
2787 | ||
2788 | /* | |
d0d856e8 | 2789 | * if truncate on deeper level happened, it wasn't partial, |
a86c6181 AT |
2790 | * so we have to consider current index for truncation |
2791 | */ | |
2792 | if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block) | |
2793 | return 0; | |
2794 | return 1; | |
2795 | } | |
2796 | ||
26a4c0c6 TT |
2797 | int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, |
2798 | ext4_lblk_t end) | |
a86c6181 AT |
2799 | { |
2800 | struct super_block *sb = inode->i_sb; | |
2801 | int depth = ext_depth(inode); | |
968dee77 | 2802 | struct ext4_ext_path *path = NULL; |
d23142c6 | 2803 | long long partial_cluster = 0; |
a86c6181 | 2804 | handle_t *handle; |
6f2080e6 | 2805 | int i = 0, err = 0; |
a86c6181 | 2806 | |
5f95d21f | 2807 | ext_debug("truncate since %u to %u\n", start, end); |
a86c6181 AT |
2808 | |
2809 | /* probably first extent we're gonna free will be last in block */ | |
9924a92a | 2810 | handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, depth + 1); |
a86c6181 AT |
2811 | if (IS_ERR(handle)) |
2812 | return PTR_ERR(handle); | |
2813 | ||
0617b83f | 2814 | again: |
61801325 | 2815 | trace_ext4_ext_remove_space(inode, start, end, depth); |
d8990240 | 2816 | |
5f95d21f LC |
2817 | /* |
2818 | * Check if we are removing extents inside the extent tree. If that | |
2819 | * is the case, we are going to punch a hole inside the extent tree | |
2820 | * so we have to check whether we need to split the extent covering | |
2821 | * the last block to remove so we can easily remove the part of it | |
2822 | * in ext4_ext_rm_leaf(). | |
2823 | */ | |
2824 | if (end < EXT_MAX_BLOCKS - 1) { | |
2825 | struct ext4_extent *ex; | |
2826 | ext4_lblk_t ee_block; | |
2827 | ||
2828 | /* find extent for this block */ | |
107a7bd3 | 2829 | path = ext4_ext_find_extent(inode, end, NULL, EXT4_EX_NOCACHE); |
5f95d21f LC |
2830 | if (IS_ERR(path)) { |
2831 | ext4_journal_stop(handle); | |
2832 | return PTR_ERR(path); | |
2833 | } | |
2834 | depth = ext_depth(inode); | |
6f2080e6 | 2835 | /* Leaf not may not exist only if inode has no blocks at all */ |
5f95d21f | 2836 | ex = path[depth].p_ext; |
968dee77 | 2837 | if (!ex) { |
6f2080e6 DM |
2838 | if (depth) { |
2839 | EXT4_ERROR_INODE(inode, | |
2840 | "path[%d].p_hdr == NULL", | |
2841 | depth); | |
2842 | err = -EIO; | |
2843 | } | |
2844 | goto out; | |
968dee77 | 2845 | } |
5f95d21f LC |
2846 | |
2847 | ee_block = le32_to_cpu(ex->ee_block); | |
2848 | ||
2849 | /* | |
2850 | * See if the last block is inside the extent, if so split | |
2851 | * the extent at 'end' block so we can easily remove the | |
2852 | * tail of the first part of the split extent in | |
2853 | * ext4_ext_rm_leaf(). | |
2854 | */ | |
2855 | if (end >= ee_block && | |
2856 | end < ee_block + ext4_ext_get_actual_len(ex) - 1) { | |
2857 | int split_flag = 0; | |
2858 | ||
556615dc LC |
2859 | if (ext4_ext_is_unwritten(ex)) |
2860 | split_flag = EXT4_EXT_MARK_UNWRIT1 | | |
2861 | EXT4_EXT_MARK_UNWRIT2; | |
5f95d21f LC |
2862 | |
2863 | /* | |
2864 | * Split the extent in two so that 'end' is the last | |
27dd4385 LC |
2865 | * block in the first new extent. Also we should not |
2866 | * fail removing space due to ENOSPC so try to use | |
2867 | * reserved block if that happens. | |
5f95d21f LC |
2868 | */ |
2869 | err = ext4_split_extent_at(handle, inode, path, | |
27dd4385 | 2870 | end + 1, split_flag, |
107a7bd3 | 2871 | EXT4_EX_NOCACHE | |
27dd4385 LC |
2872 | EXT4_GET_BLOCKS_PRE_IO | |
2873 | EXT4_GET_BLOCKS_METADATA_NOFAIL); | |
5f95d21f LC |
2874 | |
2875 | if (err < 0) | |
2876 | goto out; | |
2877 | } | |
5f95d21f | 2878 | } |
a86c6181 | 2879 | /* |
d0d856e8 RD |
2880 | * We start scanning from right side, freeing all the blocks |
2881 | * after i_size and walking into the tree depth-wise. | |
a86c6181 | 2882 | */ |
0617b83f | 2883 | depth = ext_depth(inode); |
968dee77 AS |
2884 | if (path) { |
2885 | int k = i = depth; | |
2886 | while (--k > 0) | |
2887 | path[k].p_block = | |
2888 | le16_to_cpu(path[k].p_hdr->eh_entries)+1; | |
2889 | } else { | |
2890 | path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), | |
2891 | GFP_NOFS); | |
2892 | if (path == NULL) { | |
2893 | ext4_journal_stop(handle); | |
2894 | return -ENOMEM; | |
2895 | } | |
2896 | path[0].p_depth = depth; | |
2897 | path[0].p_hdr = ext_inode_hdr(inode); | |
89a4e48f | 2898 | i = 0; |
5f95d21f | 2899 | |
c349179b | 2900 | if (ext4_ext_check(inode, path[0].p_hdr, depth, 0)) { |
968dee77 AS |
2901 | err = -EIO; |
2902 | goto out; | |
2903 | } | |
a86c6181 | 2904 | } |
968dee77 | 2905 | err = 0; |
a86c6181 AT |
2906 | |
2907 | while (i >= 0 && err == 0) { | |
2908 | if (i == depth) { | |
2909 | /* this is leaf block */ | |
d583fb87 | 2910 | err = ext4_ext_rm_leaf(handle, inode, path, |
0aa06000 | 2911 | &partial_cluster, start, |
5f95d21f | 2912 | end); |
d0d856e8 | 2913 | /* root level has p_bh == NULL, brelse() eats this */ |
a86c6181 AT |
2914 | brelse(path[i].p_bh); |
2915 | path[i].p_bh = NULL; | |
2916 | i--; | |
2917 | continue; | |
2918 | } | |
2919 | ||
2920 | /* this is index block */ | |
2921 | if (!path[i].p_hdr) { | |
2922 | ext_debug("initialize header\n"); | |
2923 | path[i].p_hdr = ext_block_hdr(path[i].p_bh); | |
a86c6181 AT |
2924 | } |
2925 | ||
a86c6181 | 2926 | if (!path[i].p_idx) { |
d0d856e8 | 2927 | /* this level hasn't been touched yet */ |
a86c6181 AT |
2928 | path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr); |
2929 | path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1; | |
2930 | ext_debug("init index ptr: hdr 0x%p, num %d\n", | |
2931 | path[i].p_hdr, | |
2932 | le16_to_cpu(path[i].p_hdr->eh_entries)); | |
2933 | } else { | |
d0d856e8 | 2934 | /* we were already here, see at next index */ |
a86c6181 AT |
2935 | path[i].p_idx--; |
2936 | } | |
2937 | ||
2938 | ext_debug("level %d - index, first 0x%p, cur 0x%p\n", | |
2939 | i, EXT_FIRST_INDEX(path[i].p_hdr), | |
2940 | path[i].p_idx); | |
2941 | if (ext4_ext_more_to_rm(path + i)) { | |
c29c0ae7 | 2942 | struct buffer_head *bh; |
a86c6181 | 2943 | /* go to the next level */ |
2ae02107 | 2944 | ext_debug("move to level %d (block %llu)\n", |
bf89d16f | 2945 | i + 1, ext4_idx_pblock(path[i].p_idx)); |
a86c6181 | 2946 | memset(path + i + 1, 0, sizeof(*path)); |
7d7ea89e | 2947 | bh = read_extent_tree_block(inode, |
107a7bd3 TT |
2948 | ext4_idx_pblock(path[i].p_idx), depth - i - 1, |
2949 | EXT4_EX_NOCACHE); | |
7d7ea89e | 2950 | if (IS_ERR(bh)) { |
a86c6181 | 2951 | /* should we reset i_size? */ |
7d7ea89e | 2952 | err = PTR_ERR(bh); |
a86c6181 AT |
2953 | break; |
2954 | } | |
76828c88 TT |
2955 | /* Yield here to deal with large extent trees. |
2956 | * Should be a no-op if we did IO above. */ | |
2957 | cond_resched(); | |
c29c0ae7 AT |
2958 | if (WARN_ON(i + 1 > depth)) { |
2959 | err = -EIO; | |
c29c0ae7 AT |
2960 | break; |
2961 | } | |
2962 | path[i + 1].p_bh = bh; | |
a86c6181 | 2963 | |
d0d856e8 RD |
2964 | /* save actual number of indexes since this |
2965 | * number is changed at the next iteration */ | |
a86c6181 AT |
2966 | path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries); |
2967 | i++; | |
2968 | } else { | |
d0d856e8 | 2969 | /* we finished processing this index, go up */ |
a86c6181 | 2970 | if (path[i].p_hdr->eh_entries == 0 && i > 0) { |
d0d856e8 | 2971 | /* index is empty, remove it; |
a86c6181 AT |
2972 | * handle must be already prepared by the |
2973 | * truncatei_leaf() */ | |
c36575e6 | 2974 | err = ext4_ext_rm_idx(handle, inode, path, i); |
a86c6181 | 2975 | } |
d0d856e8 | 2976 | /* root level has p_bh == NULL, brelse() eats this */ |
a86c6181 AT |
2977 | brelse(path[i].p_bh); |
2978 | path[i].p_bh = NULL; | |
2979 | i--; | |
2980 | ext_debug("return to level %d\n", i); | |
2981 | } | |
2982 | } | |
2983 | ||
61801325 LC |
2984 | trace_ext4_ext_remove_space_done(inode, start, end, depth, |
2985 | partial_cluster, path->p_hdr->eh_entries); | |
d8990240 | 2986 | |
7b415bf6 AK |
2987 | /* If we still have something in the partial cluster and we have removed |
2988 | * even the first extent, then we should free the blocks in the partial | |
2989 | * cluster as well. */ | |
d23142c6 | 2990 | if (partial_cluster > 0 && path->p_hdr->eh_entries == 0) { |
981250ca | 2991 | int flags = get_default_free_blocks_flags(inode); |
7b415bf6 AK |
2992 | |
2993 | ext4_free_blocks(handle, inode, NULL, | |
2994 | EXT4_C2B(EXT4_SB(sb), partial_cluster), | |
2995 | EXT4_SB(sb)->s_cluster_ratio, flags); | |
2996 | partial_cluster = 0; | |
2997 | } | |
2998 | ||
a86c6181 AT |
2999 | /* TODO: flexible tree reduction should be here */ |
3000 | if (path->p_hdr->eh_entries == 0) { | |
3001 | /* | |
d0d856e8 RD |
3002 | * truncate to zero freed all the tree, |
3003 | * so we need to correct eh_depth | |
a86c6181 AT |
3004 | */ |
3005 | err = ext4_ext_get_access(handle, inode, path); | |
3006 | if (err == 0) { | |
3007 | ext_inode_hdr(inode)->eh_depth = 0; | |
3008 | ext_inode_hdr(inode)->eh_max = | |
55ad63bf | 3009 | cpu_to_le16(ext4_ext_space_root(inode, 0)); |
a86c6181 AT |
3010 | err = ext4_ext_dirty(handle, inode, path); |
3011 | } | |
3012 | } | |
3013 | out: | |
a86c6181 AT |
3014 | ext4_ext_drop_refs(path); |
3015 | kfree(path); | |
968dee77 AS |
3016 | if (err == -EAGAIN) { |
3017 | path = NULL; | |
0617b83f | 3018 | goto again; |
968dee77 | 3019 | } |
a86c6181 AT |
3020 | ext4_journal_stop(handle); |
3021 | ||
3022 | return err; | |
3023 | } | |
3024 | ||
3025 | /* | |
3026 | * called at mount time | |
3027 | */ | |
3028 | void ext4_ext_init(struct super_block *sb) | |
3029 | { | |
3030 | /* | |
3031 | * possible initialization would be here | |
3032 | */ | |
3033 | ||
83982b6f | 3034 | if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) { |
90576c0b | 3035 | #if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS) |
92b97816 | 3036 | printk(KERN_INFO "EXT4-fs: file extents enabled" |
bbf2f9fb | 3037 | #ifdef AGGRESSIVE_TEST |
92b97816 | 3038 | ", aggressive tests" |
a86c6181 AT |
3039 | #endif |
3040 | #ifdef CHECK_BINSEARCH | |
92b97816 | 3041 | ", check binsearch" |
a86c6181 AT |
3042 | #endif |
3043 | #ifdef EXTENTS_STATS | |
92b97816 | 3044 | ", stats" |
a86c6181 | 3045 | #endif |
92b97816 | 3046 | "\n"); |
90576c0b | 3047 | #endif |
a86c6181 AT |
3048 | #ifdef EXTENTS_STATS |
3049 | spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock); | |
3050 | EXT4_SB(sb)->s_ext_min = 1 << 30; | |
3051 | EXT4_SB(sb)->s_ext_max = 0; | |
3052 | #endif | |
3053 | } | |
3054 | } | |
3055 | ||
3056 | /* | |
3057 | * called at umount time | |
3058 | */ | |
3059 | void ext4_ext_release(struct super_block *sb) | |
3060 | { | |
83982b6f | 3061 | if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) |
a86c6181 AT |
3062 | return; |
3063 | ||
3064 | #ifdef EXTENTS_STATS | |
3065 | if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) { | |
3066 | struct ext4_sb_info *sbi = EXT4_SB(sb); | |
3067 | printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n", | |
3068 | sbi->s_ext_blocks, sbi->s_ext_extents, | |
3069 | sbi->s_ext_blocks / sbi->s_ext_extents); | |
3070 | printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n", | |
3071 | sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max); | |
3072 | } | |
3073 | #endif | |
3074 | } | |
3075 | ||
d7b2a00c ZL |
3076 | static int ext4_zeroout_es(struct inode *inode, struct ext4_extent *ex) |
3077 | { | |
3078 | ext4_lblk_t ee_block; | |
3079 | ext4_fsblk_t ee_pblock; | |
3080 | unsigned int ee_len; | |
3081 | ||
3082 | ee_block = le32_to_cpu(ex->ee_block); | |
3083 | ee_len = ext4_ext_get_actual_len(ex); | |
3084 | ee_pblock = ext4_ext_pblock(ex); | |
3085 | ||
3086 | if (ee_len == 0) | |
3087 | return 0; | |
3088 | ||
3089 | return ext4_es_insert_extent(inode, ee_block, ee_len, ee_pblock, | |
3090 | EXTENT_STATUS_WRITTEN); | |
3091 | } | |
3092 | ||
093a088b AK |
3093 | /* FIXME!! we need to try to merge to left or right after zero-out */ |
3094 | static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex) | |
3095 | { | |
2407518d LC |
3096 | ext4_fsblk_t ee_pblock; |
3097 | unsigned int ee_len; | |
b720303d | 3098 | int ret; |
093a088b | 3099 | |
093a088b | 3100 | ee_len = ext4_ext_get_actual_len(ex); |
bf89d16f | 3101 | ee_pblock = ext4_ext_pblock(ex); |
b720303d | 3102 | |
a107e5a3 | 3103 | ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS); |
2407518d LC |
3104 | if (ret > 0) |
3105 | ret = 0; | |
093a088b | 3106 | |
2407518d | 3107 | return ret; |
093a088b AK |
3108 | } |
3109 | ||
47ea3bb5 YY |
3110 | /* |
3111 | * ext4_split_extent_at() splits an extent at given block. | |
3112 | * | |
3113 | * @handle: the journal handle | |
3114 | * @inode: the file inode | |
3115 | * @path: the path to the extent | |
3116 | * @split: the logical block where the extent is splitted. | |
3117 | * @split_flags: indicates if the extent could be zeroout if split fails, and | |
556615dc | 3118 | * the states(init or unwritten) of new extents. |
47ea3bb5 YY |
3119 | * @flags: flags used to insert new extent to extent tree. |
3120 | * | |
3121 | * | |
3122 | * Splits extent [a, b] into two extents [a, @split) and [@split, b], states | |
3123 | * of which are deterimined by split_flag. | |
3124 | * | |
3125 | * There are two cases: | |
3126 | * a> the extent are splitted into two extent. | |
3127 | * b> split is not needed, and just mark the extent. | |
3128 | * | |
3129 | * return 0 on success. | |
3130 | */ | |
3131 | static int ext4_split_extent_at(handle_t *handle, | |
3132 | struct inode *inode, | |
3133 | struct ext4_ext_path *path, | |
3134 | ext4_lblk_t split, | |
3135 | int split_flag, | |
3136 | int flags) | |
3137 | { | |
3138 | ext4_fsblk_t newblock; | |
3139 | ext4_lblk_t ee_block; | |
adb23551 | 3140 | struct ext4_extent *ex, newex, orig_ex, zero_ex; |
47ea3bb5 YY |
3141 | struct ext4_extent *ex2 = NULL; |
3142 | unsigned int ee_len, depth; | |
3143 | int err = 0; | |
3144 | ||
dee1f973 DM |
3145 | BUG_ON((split_flag & (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2)) == |
3146 | (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2)); | |
3147 | ||
47ea3bb5 YY |
3148 | ext_debug("ext4_split_extents_at: inode %lu, logical" |
3149 | "block %llu\n", inode->i_ino, (unsigned long long)split); | |
3150 | ||
3151 | ext4_ext_show_leaf(inode, path); | |
3152 | ||
3153 | depth = ext_depth(inode); | |
3154 | ex = path[depth].p_ext; | |
3155 | ee_block = le32_to_cpu(ex->ee_block); | |
3156 | ee_len = ext4_ext_get_actual_len(ex); | |
3157 | newblock = split - ee_block + ext4_ext_pblock(ex); | |
3158 | ||
3159 | BUG_ON(split < ee_block || split >= (ee_block + ee_len)); | |
556615dc | 3160 | BUG_ON(!ext4_ext_is_unwritten(ex) && |
357b66fd | 3161 | split_flag & (EXT4_EXT_MAY_ZEROOUT | |
556615dc LC |
3162 | EXT4_EXT_MARK_UNWRIT1 | |
3163 | EXT4_EXT_MARK_UNWRIT2)); | |
47ea3bb5 YY |
3164 | |
3165 | err = ext4_ext_get_access(handle, inode, path + depth); | |
3166 | if (err) | |
3167 | goto out; | |
3168 | ||
3169 | if (split == ee_block) { | |
3170 | /* | |
3171 | * case b: block @split is the block that the extent begins with | |
3172 | * then we just change the state of the extent, and splitting | |
3173 | * is not needed. | |
3174 | */ | |
556615dc LC |
3175 | if (split_flag & EXT4_EXT_MARK_UNWRIT2) |
3176 | ext4_ext_mark_unwritten(ex); | |
47ea3bb5 YY |
3177 | else |
3178 | ext4_ext_mark_initialized(ex); | |
3179 | ||
3180 | if (!(flags & EXT4_GET_BLOCKS_PRE_IO)) | |
ecb94f5f | 3181 | ext4_ext_try_to_merge(handle, inode, path, ex); |
47ea3bb5 | 3182 | |
ecb94f5f | 3183 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); |
47ea3bb5 YY |
3184 | goto out; |
3185 | } | |
3186 | ||
3187 | /* case a */ | |
3188 | memcpy(&orig_ex, ex, sizeof(orig_ex)); | |
3189 | ex->ee_len = cpu_to_le16(split - ee_block); | |
556615dc LC |
3190 | if (split_flag & EXT4_EXT_MARK_UNWRIT1) |
3191 | ext4_ext_mark_unwritten(ex); | |
47ea3bb5 YY |
3192 | |
3193 | /* | |
3194 | * path may lead to new leaf, not to original leaf any more | |
3195 | * after ext4_ext_insert_extent() returns, | |
3196 | */ | |
3197 | err = ext4_ext_dirty(handle, inode, path + depth); | |
3198 | if (err) | |
3199 | goto fix_extent_len; | |
3200 | ||
3201 | ex2 = &newex; | |
3202 | ex2->ee_block = cpu_to_le32(split); | |
3203 | ex2->ee_len = cpu_to_le16(ee_len - (split - ee_block)); | |
3204 | ext4_ext_store_pblock(ex2, newblock); | |
556615dc LC |
3205 | if (split_flag & EXT4_EXT_MARK_UNWRIT2) |
3206 | ext4_ext_mark_unwritten(ex2); | |
47ea3bb5 YY |
3207 | |
3208 | err = ext4_ext_insert_extent(handle, inode, path, &newex, flags); | |
3209 | if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) { | |
dee1f973 | 3210 | if (split_flag & (EXT4_EXT_DATA_VALID1|EXT4_EXT_DATA_VALID2)) { |
adb23551 | 3211 | if (split_flag & EXT4_EXT_DATA_VALID1) { |
dee1f973 | 3212 | err = ext4_ext_zeroout(inode, ex2); |
adb23551 | 3213 | zero_ex.ee_block = ex2->ee_block; |
8cde7ad1 ZL |
3214 | zero_ex.ee_len = cpu_to_le16( |
3215 | ext4_ext_get_actual_len(ex2)); | |
adb23551 ZL |
3216 | ext4_ext_store_pblock(&zero_ex, |
3217 | ext4_ext_pblock(ex2)); | |
3218 | } else { | |
dee1f973 | 3219 | err = ext4_ext_zeroout(inode, ex); |
adb23551 | 3220 | zero_ex.ee_block = ex->ee_block; |
8cde7ad1 ZL |
3221 | zero_ex.ee_len = cpu_to_le16( |
3222 | ext4_ext_get_actual_len(ex)); | |
adb23551 ZL |
3223 | ext4_ext_store_pblock(&zero_ex, |
3224 | ext4_ext_pblock(ex)); | |
3225 | } | |
3226 | } else { | |
dee1f973 | 3227 | err = ext4_ext_zeroout(inode, &orig_ex); |
adb23551 | 3228 | zero_ex.ee_block = orig_ex.ee_block; |
8cde7ad1 ZL |
3229 | zero_ex.ee_len = cpu_to_le16( |
3230 | ext4_ext_get_actual_len(&orig_ex)); | |
adb23551 ZL |
3231 | ext4_ext_store_pblock(&zero_ex, |
3232 | ext4_ext_pblock(&orig_ex)); | |
3233 | } | |
dee1f973 | 3234 | |
47ea3bb5 YY |
3235 | if (err) |
3236 | goto fix_extent_len; | |
3237 | /* update the extent length and mark as initialized */ | |
af1584f5 | 3238 | ex->ee_len = cpu_to_le16(ee_len); |
ecb94f5f TT |
3239 | ext4_ext_try_to_merge(handle, inode, path, ex); |
3240 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); | |
adb23551 ZL |
3241 | if (err) |
3242 | goto fix_extent_len; | |
3243 | ||
3244 | /* update extent status tree */ | |
d7b2a00c | 3245 | err = ext4_zeroout_es(inode, &zero_ex); |
adb23551 | 3246 | |
47ea3bb5 YY |
3247 | goto out; |
3248 | } else if (err) | |
3249 | goto fix_extent_len; | |
3250 | ||
3251 | out: | |
3252 | ext4_ext_show_leaf(inode, path); | |
3253 | return err; | |
3254 | ||
3255 | fix_extent_len: | |
3256 | ex->ee_len = orig_ex.ee_len; | |
29faed16 | 3257 | ext4_ext_dirty(handle, inode, path + path->p_depth); |
47ea3bb5 YY |
3258 | return err; |
3259 | } | |
3260 | ||
3261 | /* | |
3262 | * ext4_split_extents() splits an extent and mark extent which is covered | |
3263 | * by @map as split_flags indicates | |
3264 | * | |
70261f56 | 3265 | * It may result in splitting the extent into multiple extents (up to three) |
47ea3bb5 YY |
3266 | * There are three possibilities: |
3267 | * a> There is no split required | |
3268 | * b> Splits in two extents: Split is happening at either end of the extent | |
3269 | * c> Splits in three extents: Somone is splitting in middle of the extent | |
3270 | * | |
3271 | */ | |
3272 | static int ext4_split_extent(handle_t *handle, | |
3273 | struct inode *inode, | |
3274 | struct ext4_ext_path *path, | |
3275 | struct ext4_map_blocks *map, | |
3276 | int split_flag, | |
3277 | int flags) | |
3278 | { | |
3279 | ext4_lblk_t ee_block; | |
3280 | struct ext4_extent *ex; | |
3281 | unsigned int ee_len, depth; | |
3282 | int err = 0; | |
556615dc | 3283 | int unwritten; |
47ea3bb5 | 3284 | int split_flag1, flags1; |
3a225670 | 3285 | int allocated = map->m_len; |
47ea3bb5 YY |
3286 | |
3287 | depth = ext_depth(inode); | |
3288 | ex = path[depth].p_ext; | |
3289 | ee_block = le32_to_cpu(ex->ee_block); | |
3290 | ee_len = ext4_ext_get_actual_len(ex); | |
556615dc | 3291 | unwritten = ext4_ext_is_unwritten(ex); |
47ea3bb5 YY |
3292 | |
3293 | if (map->m_lblk + map->m_len < ee_block + ee_len) { | |
dee1f973 | 3294 | split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT; |
47ea3bb5 | 3295 | flags1 = flags | EXT4_GET_BLOCKS_PRE_IO; |
556615dc LC |
3296 | if (unwritten) |
3297 | split_flag1 |= EXT4_EXT_MARK_UNWRIT1 | | |
3298 | EXT4_EXT_MARK_UNWRIT2; | |
dee1f973 DM |
3299 | if (split_flag & EXT4_EXT_DATA_VALID2) |
3300 | split_flag1 |= EXT4_EXT_DATA_VALID1; | |
47ea3bb5 YY |
3301 | err = ext4_split_extent_at(handle, inode, path, |
3302 | map->m_lblk + map->m_len, split_flag1, flags1); | |
93917411 YY |
3303 | if (err) |
3304 | goto out; | |
3a225670 ZL |
3305 | } else { |
3306 | allocated = ee_len - (map->m_lblk - ee_block); | |
47ea3bb5 | 3307 | } |
357b66fd DM |
3308 | /* |
3309 | * Update path is required because previous ext4_split_extent_at() may | |
3310 | * result in split of original leaf or extent zeroout. | |
3311 | */ | |
47ea3bb5 | 3312 | ext4_ext_drop_refs(path); |
107a7bd3 | 3313 | path = ext4_ext_find_extent(inode, map->m_lblk, path, 0); |
47ea3bb5 YY |
3314 | if (IS_ERR(path)) |
3315 | return PTR_ERR(path); | |
357b66fd DM |
3316 | depth = ext_depth(inode); |
3317 | ex = path[depth].p_ext; | |
a18ed359 DM |
3318 | if (!ex) { |
3319 | EXT4_ERROR_INODE(inode, "unexpected hole at %lu", | |
3320 | (unsigned long) map->m_lblk); | |
3321 | return -EIO; | |
3322 | } | |
556615dc | 3323 | unwritten = ext4_ext_is_unwritten(ex); |
357b66fd | 3324 | split_flag1 = 0; |
47ea3bb5 YY |
3325 | |
3326 | if (map->m_lblk >= ee_block) { | |
357b66fd | 3327 | split_flag1 = split_flag & EXT4_EXT_DATA_VALID2; |
556615dc LC |
3328 | if (unwritten) { |
3329 | split_flag1 |= EXT4_EXT_MARK_UNWRIT1; | |
357b66fd | 3330 | split_flag1 |= split_flag & (EXT4_EXT_MAY_ZEROOUT | |
556615dc | 3331 | EXT4_EXT_MARK_UNWRIT2); |
357b66fd | 3332 | } |
47ea3bb5 YY |
3333 | err = ext4_split_extent_at(handle, inode, path, |
3334 | map->m_lblk, split_flag1, flags); | |
3335 | if (err) | |
3336 | goto out; | |
3337 | } | |
3338 | ||
3339 | ext4_ext_show_leaf(inode, path); | |
3340 | out: | |
3a225670 | 3341 | return err ? err : allocated; |
47ea3bb5 YY |
3342 | } |
3343 | ||
56055d3a | 3344 | /* |
e35fd660 | 3345 | * This function is called by ext4_ext_map_blocks() if someone tries to write |
556615dc | 3346 | * to an unwritten extent. It may result in splitting the unwritten |
25985edc | 3347 | * extent into multiple extents (up to three - one initialized and two |
556615dc | 3348 | * unwritten). |
56055d3a AA |
3349 | * There are three possibilities: |
3350 | * a> There is no split required: Entire extent should be initialized | |
3351 | * b> Splits in two extents: Write is happening at either end of the extent | |
3352 | * c> Splits in three extents: Somone is writing in middle of the extent | |
6f91bc5f EG |
3353 | * |
3354 | * Pre-conditions: | |
556615dc | 3355 | * - The extent pointed to by 'path' is unwritten. |
6f91bc5f EG |
3356 | * - The extent pointed to by 'path' contains a superset |
3357 | * of the logical span [map->m_lblk, map->m_lblk + map->m_len). | |
3358 | * | |
3359 | * Post-conditions on success: | |
3360 | * - the returned value is the number of blocks beyond map->l_lblk | |
3361 | * that are allocated and initialized. | |
3362 | * It is guaranteed to be >= map->m_len. | |
56055d3a | 3363 | */ |
725d26d3 | 3364 | static int ext4_ext_convert_to_initialized(handle_t *handle, |
e35fd660 TT |
3365 | struct inode *inode, |
3366 | struct ext4_map_blocks *map, | |
27dd4385 LC |
3367 | struct ext4_ext_path *path, |
3368 | int flags) | |
56055d3a | 3369 | { |
67a5da56 | 3370 | struct ext4_sb_info *sbi; |
6f91bc5f | 3371 | struct ext4_extent_header *eh; |
667eff35 YY |
3372 | struct ext4_map_blocks split_map; |
3373 | struct ext4_extent zero_ex; | |
bc2d9db4 | 3374 | struct ext4_extent *ex, *abut_ex; |
21ca087a | 3375 | ext4_lblk_t ee_block, eof_block; |
bc2d9db4 LC |
3376 | unsigned int ee_len, depth, map_len = map->m_len; |
3377 | int allocated = 0, max_zeroout = 0; | |
56055d3a | 3378 | int err = 0; |
667eff35 | 3379 | int split_flag = 0; |
21ca087a DM |
3380 | |
3381 | ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical" | |
3382 | "block %llu, max_blocks %u\n", inode->i_ino, | |
bc2d9db4 | 3383 | (unsigned long long)map->m_lblk, map_len); |
21ca087a | 3384 | |
67a5da56 | 3385 | sbi = EXT4_SB(inode->i_sb); |
21ca087a DM |
3386 | eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >> |
3387 | inode->i_sb->s_blocksize_bits; | |
bc2d9db4 LC |
3388 | if (eof_block < map->m_lblk + map_len) |
3389 | eof_block = map->m_lblk + map_len; | |
56055d3a AA |
3390 | |
3391 | depth = ext_depth(inode); | |
6f91bc5f | 3392 | eh = path[depth].p_hdr; |
56055d3a AA |
3393 | ex = path[depth].p_ext; |
3394 | ee_block = le32_to_cpu(ex->ee_block); | |
3395 | ee_len = ext4_ext_get_actual_len(ex); | |
adb23551 | 3396 | zero_ex.ee_len = 0; |
56055d3a | 3397 | |
6f91bc5f EG |
3398 | trace_ext4_ext_convert_to_initialized_enter(inode, map, ex); |
3399 | ||
3400 | /* Pre-conditions */ | |
556615dc | 3401 | BUG_ON(!ext4_ext_is_unwritten(ex)); |
6f91bc5f | 3402 | BUG_ON(!in_range(map->m_lblk, ee_block, ee_len)); |
6f91bc5f EG |
3403 | |
3404 | /* | |
3405 | * Attempt to transfer newly initialized blocks from the currently | |
556615dc | 3406 | * unwritten extent to its neighbor. This is much cheaper |
6f91bc5f | 3407 | * than an insertion followed by a merge as those involve costly |
bc2d9db4 LC |
3408 | * memmove() calls. Transferring to the left is the common case in |
3409 | * steady state for workloads doing fallocate(FALLOC_FL_KEEP_SIZE) | |
3410 | * followed by append writes. | |
6f91bc5f EG |
3411 | * |
3412 | * Limitations of the current logic: | |
bc2d9db4 | 3413 | * - L1: we do not deal with writes covering the whole extent. |
6f91bc5f EG |
3414 | * This would require removing the extent if the transfer |
3415 | * is possible. | |
bc2d9db4 | 3416 | * - L2: we only attempt to merge with an extent stored in the |
6f91bc5f EG |
3417 | * same extent tree node. |
3418 | */ | |
bc2d9db4 LC |
3419 | if ((map->m_lblk == ee_block) && |
3420 | /* See if we can merge left */ | |
3421 | (map_len < ee_len) && /*L1*/ | |
3422 | (ex > EXT_FIRST_EXTENT(eh))) { /*L2*/ | |
6f91bc5f EG |
3423 | ext4_lblk_t prev_lblk; |
3424 | ext4_fsblk_t prev_pblk, ee_pblk; | |
bc2d9db4 | 3425 | unsigned int prev_len; |
6f91bc5f | 3426 | |
bc2d9db4 LC |
3427 | abut_ex = ex - 1; |
3428 | prev_lblk = le32_to_cpu(abut_ex->ee_block); | |
3429 | prev_len = ext4_ext_get_actual_len(abut_ex); | |
3430 | prev_pblk = ext4_ext_pblock(abut_ex); | |
6f91bc5f | 3431 | ee_pblk = ext4_ext_pblock(ex); |
6f91bc5f EG |
3432 | |
3433 | /* | |
bc2d9db4 | 3434 | * A transfer of blocks from 'ex' to 'abut_ex' is allowed |
6f91bc5f | 3435 | * upon those conditions: |
bc2d9db4 LC |
3436 | * - C1: abut_ex is initialized, |
3437 | * - C2: abut_ex is logically abutting ex, | |
3438 | * - C3: abut_ex is physically abutting ex, | |
3439 | * - C4: abut_ex can receive the additional blocks without | |
6f91bc5f EG |
3440 | * overflowing the (initialized) length limit. |
3441 | */ | |
556615dc | 3442 | if ((!ext4_ext_is_unwritten(abut_ex)) && /*C1*/ |
6f91bc5f EG |
3443 | ((prev_lblk + prev_len) == ee_block) && /*C2*/ |
3444 | ((prev_pblk + prev_len) == ee_pblk) && /*C3*/ | |
bc2d9db4 | 3445 | (prev_len < (EXT_INIT_MAX_LEN - map_len))) { /*C4*/ |
6f91bc5f EG |
3446 | err = ext4_ext_get_access(handle, inode, path + depth); |
3447 | if (err) | |
3448 | goto out; | |
3449 | ||
3450 | trace_ext4_ext_convert_to_initialized_fastpath(inode, | |
bc2d9db4 | 3451 | map, ex, abut_ex); |
6f91bc5f | 3452 | |
bc2d9db4 LC |
3453 | /* Shift the start of ex by 'map_len' blocks */ |
3454 | ex->ee_block = cpu_to_le32(ee_block + map_len); | |
3455 | ext4_ext_store_pblock(ex, ee_pblk + map_len); | |
3456 | ex->ee_len = cpu_to_le16(ee_len - map_len); | |
556615dc | 3457 | ext4_ext_mark_unwritten(ex); /* Restore the flag */ |
6f91bc5f | 3458 | |
bc2d9db4 LC |
3459 | /* Extend abut_ex by 'map_len' blocks */ |
3460 | abut_ex->ee_len = cpu_to_le16(prev_len + map_len); | |
6f91bc5f | 3461 | |
bc2d9db4 LC |
3462 | /* Result: number of initialized blocks past m_lblk */ |
3463 | allocated = map_len; | |
3464 | } | |
3465 | } else if (((map->m_lblk + map_len) == (ee_block + ee_len)) && | |
3466 | (map_len < ee_len) && /*L1*/ | |
3467 | ex < EXT_LAST_EXTENT(eh)) { /*L2*/ | |
3468 | /* See if we can merge right */ | |
3469 | ext4_lblk_t next_lblk; | |
3470 | ext4_fsblk_t next_pblk, ee_pblk; | |
3471 | unsigned int next_len; | |
3472 | ||
3473 | abut_ex = ex + 1; | |
3474 | next_lblk = le32_to_cpu(abut_ex->ee_block); | |
3475 | next_len = ext4_ext_get_actual_len(abut_ex); | |
3476 | next_pblk = ext4_ext_pblock(abut_ex); | |
3477 | ee_pblk = ext4_ext_pblock(ex); | |
6f91bc5f | 3478 | |
bc2d9db4 LC |
3479 | /* |
3480 | * A transfer of blocks from 'ex' to 'abut_ex' is allowed | |
3481 | * upon those conditions: | |
3482 | * - C1: abut_ex is initialized, | |
3483 | * - C2: abut_ex is logically abutting ex, | |
3484 | * - C3: abut_ex is physically abutting ex, | |
3485 | * - C4: abut_ex can receive the additional blocks without | |
3486 | * overflowing the (initialized) length limit. | |
3487 | */ | |
556615dc | 3488 | if ((!ext4_ext_is_unwritten(abut_ex)) && /*C1*/ |
bc2d9db4 LC |
3489 | ((map->m_lblk + map_len) == next_lblk) && /*C2*/ |
3490 | ((ee_pblk + ee_len) == next_pblk) && /*C3*/ | |
3491 | (next_len < (EXT_INIT_MAX_LEN - map_len))) { /*C4*/ | |
3492 | err = ext4_ext_get_access(handle, inode, path + depth); | |
3493 | if (err) | |
3494 | goto out; | |
3495 | ||
3496 | trace_ext4_ext_convert_to_initialized_fastpath(inode, | |
3497 | map, ex, abut_ex); | |
3498 | ||
3499 | /* Shift the start of abut_ex by 'map_len' blocks */ | |
3500 | abut_ex->ee_block = cpu_to_le32(next_lblk - map_len); | |
3501 | ext4_ext_store_pblock(abut_ex, next_pblk - map_len); | |
3502 | ex->ee_len = cpu_to_le16(ee_len - map_len); | |
556615dc | 3503 | ext4_ext_mark_unwritten(ex); /* Restore the flag */ |
bc2d9db4 LC |
3504 | |
3505 | /* Extend abut_ex by 'map_len' blocks */ | |
3506 | abut_ex->ee_len = cpu_to_le16(next_len + map_len); | |
6f91bc5f EG |
3507 | |
3508 | /* Result: number of initialized blocks past m_lblk */ | |
bc2d9db4 | 3509 | allocated = map_len; |
6f91bc5f EG |
3510 | } |
3511 | } | |
bc2d9db4 LC |
3512 | if (allocated) { |
3513 | /* Mark the block containing both extents as dirty */ | |
3514 | ext4_ext_dirty(handle, inode, path + depth); | |
3515 | ||
3516 | /* Update path to point to the right extent */ | |
3517 | path[depth].p_ext = abut_ex; | |
3518 | goto out; | |
3519 | } else | |
3520 | allocated = ee_len - (map->m_lblk - ee_block); | |
6f91bc5f | 3521 | |
667eff35 | 3522 | WARN_ON(map->m_lblk < ee_block); |
21ca087a DM |
3523 | /* |
3524 | * It is safe to convert extent to initialized via explicit | |
9e740568 | 3525 | * zeroout only if extent is fully inside i_size or new_size. |
21ca087a | 3526 | */ |
667eff35 | 3527 | split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0; |
21ca087a | 3528 | |
67a5da56 ZL |
3529 | if (EXT4_EXT_MAY_ZEROOUT & split_flag) |
3530 | max_zeroout = sbi->s_extent_max_zeroout_kb >> | |
4f42f80a | 3531 | (inode->i_sb->s_blocksize_bits - 10); |
67a5da56 ZL |
3532 | |
3533 | /* If extent is less than s_max_zeroout_kb, zeroout directly */ | |
3534 | if (max_zeroout && (ee_len <= max_zeroout)) { | |
667eff35 | 3535 | err = ext4_ext_zeroout(inode, ex); |
3977c965 | 3536 | if (err) |
d03856bd | 3537 | goto out; |
adb23551 | 3538 | zero_ex.ee_block = ex->ee_block; |
8cde7ad1 | 3539 | zero_ex.ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)); |
adb23551 | 3540 | ext4_ext_store_pblock(&zero_ex, ext4_ext_pblock(ex)); |
d03856bd AK |
3541 | |
3542 | err = ext4_ext_get_access(handle, inode, path + depth); | |
3543 | if (err) | |
3544 | goto out; | |
667eff35 | 3545 | ext4_ext_mark_initialized(ex); |
ecb94f5f TT |
3546 | ext4_ext_try_to_merge(handle, inode, path, ex); |
3547 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); | |
667eff35 | 3548 | goto out; |
56055d3a | 3549 | } |
667eff35 | 3550 | |
56055d3a | 3551 | /* |
667eff35 YY |
3552 | * four cases: |
3553 | * 1. split the extent into three extents. | |
3554 | * 2. split the extent into two extents, zeroout the first half. | |
3555 | * 3. split the extent into two extents, zeroout the second half. | |
3556 | * 4. split the extent into two extents with out zeroout. | |
56055d3a | 3557 | */ |
667eff35 YY |
3558 | split_map.m_lblk = map->m_lblk; |
3559 | split_map.m_len = map->m_len; | |
3560 | ||
67a5da56 ZL |
3561 | if (max_zeroout && (allocated > map->m_len)) { |
3562 | if (allocated <= max_zeroout) { | |
667eff35 YY |
3563 | /* case 3 */ |
3564 | zero_ex.ee_block = | |
9b940f8e AH |
3565 | cpu_to_le32(map->m_lblk); |
3566 | zero_ex.ee_len = cpu_to_le16(allocated); | |
667eff35 YY |
3567 | ext4_ext_store_pblock(&zero_ex, |
3568 | ext4_ext_pblock(ex) + map->m_lblk - ee_block); | |
3569 | err = ext4_ext_zeroout(inode, &zero_ex); | |
56055d3a AA |
3570 | if (err) |
3571 | goto out; | |
667eff35 YY |
3572 | split_map.m_lblk = map->m_lblk; |
3573 | split_map.m_len = allocated; | |
67a5da56 | 3574 | } else if (map->m_lblk - ee_block + map->m_len < max_zeroout) { |
667eff35 YY |
3575 | /* case 2 */ |
3576 | if (map->m_lblk != ee_block) { | |
3577 | zero_ex.ee_block = ex->ee_block; | |
3578 | zero_ex.ee_len = cpu_to_le16(map->m_lblk - | |
3579 | ee_block); | |
3580 | ext4_ext_store_pblock(&zero_ex, | |
3581 | ext4_ext_pblock(ex)); | |
3582 | err = ext4_ext_zeroout(inode, &zero_ex); | |
3583 | if (err) | |
3584 | goto out; | |
3585 | } | |
3586 | ||
667eff35 | 3587 | split_map.m_lblk = ee_block; |
9b940f8e AH |
3588 | split_map.m_len = map->m_lblk - ee_block + map->m_len; |
3589 | allocated = map->m_len; | |
56055d3a AA |
3590 | } |
3591 | } | |
667eff35 YY |
3592 | |
3593 | allocated = ext4_split_extent(handle, inode, path, | |
27dd4385 | 3594 | &split_map, split_flag, flags); |
667eff35 YY |
3595 | if (allocated < 0) |
3596 | err = allocated; | |
3597 | ||
56055d3a | 3598 | out: |
adb23551 ZL |
3599 | /* If we have gotten a failure, don't zero out status tree */ |
3600 | if (!err) | |
d7b2a00c | 3601 | err = ext4_zeroout_es(inode, &zero_ex); |
56055d3a AA |
3602 | return err ? err : allocated; |
3603 | } | |
3604 | ||
0031462b | 3605 | /* |
e35fd660 | 3606 | * This function is called by ext4_ext_map_blocks() from |
0031462b | 3607 | * ext4_get_blocks_dio_write() when DIO to write |
556615dc | 3608 | * to an unwritten extent. |
0031462b | 3609 | * |
556615dc LC |
3610 | * Writing to an unwritten extent may result in splitting the unwritten |
3611 | * extent into multiple initialized/unwritten extents (up to three) | |
0031462b | 3612 | * There are three possibilities: |
556615dc | 3613 | * a> There is no split required: Entire extent should be unwritten |
0031462b MC |
3614 | * b> Splits in two extents: Write is happening at either end of the extent |
3615 | * c> Splits in three extents: Somone is writing in middle of the extent | |
3616 | * | |
b8a86845 LC |
3617 | * This works the same way in the case of initialized -> unwritten conversion. |
3618 | * | |
0031462b | 3619 | * One of more index blocks maybe needed if the extent tree grow after |
556615dc LC |
3620 | * the unwritten extent split. To prevent ENOSPC occur at the IO |
3621 | * complete, we need to split the unwritten extent before DIO submit | |
3622 | * the IO. The unwritten extent called at this time will be split | |
3623 | * into three unwritten extent(at most). After IO complete, the part | |
0031462b MC |
3624 | * being filled will be convert to initialized by the end_io callback function |
3625 | * via ext4_convert_unwritten_extents(). | |
ba230c3f | 3626 | * |
556615dc | 3627 | * Returns the size of unwritten extent to be written on success. |
0031462b | 3628 | */ |
b8a86845 | 3629 | static int ext4_split_convert_extents(handle_t *handle, |
0031462b | 3630 | struct inode *inode, |
e35fd660 | 3631 | struct ext4_map_blocks *map, |
0031462b | 3632 | struct ext4_ext_path *path, |
0031462b MC |
3633 | int flags) |
3634 | { | |
667eff35 YY |
3635 | ext4_lblk_t eof_block; |
3636 | ext4_lblk_t ee_block; | |
3637 | struct ext4_extent *ex; | |
3638 | unsigned int ee_len; | |
3639 | int split_flag = 0, depth; | |
21ca087a | 3640 | |
b8a86845 LC |
3641 | ext_debug("%s: inode %lu, logical block %llu, max_blocks %u\n", |
3642 | __func__, inode->i_ino, | |
3643 | (unsigned long long)map->m_lblk, map->m_len); | |
21ca087a DM |
3644 | |
3645 | eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >> | |
3646 | inode->i_sb->s_blocksize_bits; | |
e35fd660 TT |
3647 | if (eof_block < map->m_lblk + map->m_len) |
3648 | eof_block = map->m_lblk + map->m_len; | |
21ca087a DM |
3649 | /* |
3650 | * It is safe to convert extent to initialized via explicit | |
3651 | * zeroout only if extent is fully insde i_size or new_size. | |
3652 | */ | |
667eff35 YY |
3653 | depth = ext_depth(inode); |
3654 | ex = path[depth].p_ext; | |
3655 | ee_block = le32_to_cpu(ex->ee_block); | |
3656 | ee_len = ext4_ext_get_actual_len(ex); | |
0031462b | 3657 | |
b8a86845 LC |
3658 | /* Convert to unwritten */ |
3659 | if (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN) { | |
3660 | split_flag |= EXT4_EXT_DATA_VALID1; | |
3661 | /* Convert to initialized */ | |
3662 | } else if (flags & EXT4_GET_BLOCKS_CONVERT) { | |
3663 | split_flag |= ee_block + ee_len <= eof_block ? | |
3664 | EXT4_EXT_MAY_ZEROOUT : 0; | |
556615dc | 3665 | split_flag |= (EXT4_EXT_MARK_UNWRIT2 | EXT4_EXT_DATA_VALID2); |
b8a86845 | 3666 | } |
667eff35 YY |
3667 | flags |= EXT4_GET_BLOCKS_PRE_IO; |
3668 | return ext4_split_extent(handle, inode, path, map, split_flag, flags); | |
0031462b | 3669 | } |
197217a5 | 3670 | |
b8a86845 LC |
3671 | static int ext4_convert_initialized_extents(handle_t *handle, |
3672 | struct inode *inode, | |
3673 | struct ext4_map_blocks *map, | |
3674 | struct ext4_ext_path *path) | |
3675 | { | |
3676 | struct ext4_extent *ex; | |
3677 | ext4_lblk_t ee_block; | |
3678 | unsigned int ee_len; | |
3679 | int depth; | |
3680 | int err = 0; | |
3681 | ||
3682 | depth = ext_depth(inode); | |
3683 | ex = path[depth].p_ext; | |
3684 | ee_block = le32_to_cpu(ex->ee_block); | |
3685 | ee_len = ext4_ext_get_actual_len(ex); | |
3686 | ||
3687 | ext_debug("%s: inode %lu, logical" | |
3688 | "block %llu, max_blocks %u\n", __func__, inode->i_ino, | |
3689 | (unsigned long long)ee_block, ee_len); | |
3690 | ||
3691 | if (ee_block != map->m_lblk || ee_len > map->m_len) { | |
3692 | err = ext4_split_convert_extents(handle, inode, map, path, | |
3693 | EXT4_GET_BLOCKS_CONVERT_UNWRITTEN); | |
3694 | if (err < 0) | |
3695 | goto out; | |
3696 | ext4_ext_drop_refs(path); | |
3697 | path = ext4_ext_find_extent(inode, map->m_lblk, path, 0); | |
3698 | if (IS_ERR(path)) { | |
3699 | err = PTR_ERR(path); | |
3700 | goto out; | |
3701 | } | |
3702 | depth = ext_depth(inode); | |
3703 | ex = path[depth].p_ext; | |
a18ed359 DM |
3704 | if (!ex) { |
3705 | EXT4_ERROR_INODE(inode, "unexpected hole at %lu", | |
3706 | (unsigned long) map->m_lblk); | |
3707 | err = -EIO; | |
3708 | goto out; | |
3709 | } | |
b8a86845 LC |
3710 | } |
3711 | ||
3712 | err = ext4_ext_get_access(handle, inode, path + depth); | |
3713 | if (err) | |
3714 | goto out; | |
556615dc LC |
3715 | /* first mark the extent as unwritten */ |
3716 | ext4_ext_mark_unwritten(ex); | |
b8a86845 LC |
3717 | |
3718 | /* note: ext4_ext_correct_indexes() isn't needed here because | |
3719 | * borders are not changed | |
3720 | */ | |
3721 | ext4_ext_try_to_merge(handle, inode, path, ex); | |
3722 | ||
3723 | /* Mark modified extent as dirty */ | |
3724 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); | |
3725 | out: | |
3726 | ext4_ext_show_leaf(inode, path); | |
3727 | return err; | |
3728 | } | |
3729 | ||
3730 | ||
c7064ef1 | 3731 | static int ext4_convert_unwritten_extents_endio(handle_t *handle, |
dee1f973 DM |
3732 | struct inode *inode, |
3733 | struct ext4_map_blocks *map, | |
3734 | struct ext4_ext_path *path) | |
0031462b MC |
3735 | { |
3736 | struct ext4_extent *ex; | |
dee1f973 DM |
3737 | ext4_lblk_t ee_block; |
3738 | unsigned int ee_len; | |
0031462b MC |
3739 | int depth; |
3740 | int err = 0; | |
0031462b MC |
3741 | |
3742 | depth = ext_depth(inode); | |
0031462b | 3743 | ex = path[depth].p_ext; |
dee1f973 DM |
3744 | ee_block = le32_to_cpu(ex->ee_block); |
3745 | ee_len = ext4_ext_get_actual_len(ex); | |
0031462b | 3746 | |
197217a5 YY |
3747 | ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical" |
3748 | "block %llu, max_blocks %u\n", inode->i_ino, | |
dee1f973 DM |
3749 | (unsigned long long)ee_block, ee_len); |
3750 | ||
ff95ec22 DM |
3751 | /* If extent is larger than requested it is a clear sign that we still |
3752 | * have some extent state machine issues left. So extent_split is still | |
3753 | * required. | |
3754 | * TODO: Once all related issues will be fixed this situation should be | |
3755 | * illegal. | |
3756 | */ | |
dee1f973 | 3757 | if (ee_block != map->m_lblk || ee_len > map->m_len) { |
ff95ec22 DM |
3758 | #ifdef EXT4_DEBUG |
3759 | ext4_warning("Inode (%ld) finished: extent logical block %llu," | |
3760 | " len %u; IO logical block %llu, len %u\n", | |
3761 | inode->i_ino, (unsigned long long)ee_block, ee_len, | |
3762 | (unsigned long long)map->m_lblk, map->m_len); | |
3763 | #endif | |
b8a86845 LC |
3764 | err = ext4_split_convert_extents(handle, inode, map, path, |
3765 | EXT4_GET_BLOCKS_CONVERT); | |
dee1f973 DM |
3766 | if (err < 0) |
3767 | goto out; | |
3768 | ext4_ext_drop_refs(path); | |
107a7bd3 | 3769 | path = ext4_ext_find_extent(inode, map->m_lblk, path, 0); |
dee1f973 DM |
3770 | if (IS_ERR(path)) { |
3771 | err = PTR_ERR(path); | |
3772 | goto out; | |
3773 | } | |
3774 | depth = ext_depth(inode); | |
3775 | ex = path[depth].p_ext; | |
3776 | } | |
197217a5 | 3777 | |
0031462b MC |
3778 | err = ext4_ext_get_access(handle, inode, path + depth); |
3779 | if (err) | |
3780 | goto out; | |
3781 | /* first mark the extent as initialized */ | |
3782 | ext4_ext_mark_initialized(ex); | |
3783 | ||
197217a5 YY |
3784 | /* note: ext4_ext_correct_indexes() isn't needed here because |
3785 | * borders are not changed | |
0031462b | 3786 | */ |
ecb94f5f | 3787 | ext4_ext_try_to_merge(handle, inode, path, ex); |
197217a5 | 3788 | |
0031462b | 3789 | /* Mark modified extent as dirty */ |
ecb94f5f | 3790 | err = ext4_ext_dirty(handle, inode, path + path->p_depth); |
0031462b MC |
3791 | out: |
3792 | ext4_ext_show_leaf(inode, path); | |
3793 | return err; | |
3794 | } | |
3795 | ||
515f41c3 AK |
3796 | static void unmap_underlying_metadata_blocks(struct block_device *bdev, |
3797 | sector_t block, int count) | |
3798 | { | |
3799 | int i; | |
3800 | for (i = 0; i < count; i++) | |
3801 | unmap_underlying_metadata(bdev, block + i); | |
3802 | } | |
3803 | ||
58590b06 TT |
3804 | /* |
3805 | * Handle EOFBLOCKS_FL flag, clearing it if necessary | |
3806 | */ | |
3807 | static int check_eofblocks_fl(handle_t *handle, struct inode *inode, | |
d002ebf1 | 3808 | ext4_lblk_t lblk, |
58590b06 TT |
3809 | struct ext4_ext_path *path, |
3810 | unsigned int len) | |
3811 | { | |
3812 | int i, depth; | |
3813 | struct ext4_extent_header *eh; | |
65922cb5 | 3814 | struct ext4_extent *last_ex; |
58590b06 TT |
3815 | |
3816 | if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS)) | |
3817 | return 0; | |
3818 | ||
3819 | depth = ext_depth(inode); | |
3820 | eh = path[depth].p_hdr; | |
58590b06 | 3821 | |
afcff5d8 LC |
3822 | /* |
3823 | * We're going to remove EOFBLOCKS_FL entirely in future so we | |
3824 | * do not care for this case anymore. Simply remove the flag | |
3825 | * if there are no extents. | |
3826 | */ | |
3827 | if (unlikely(!eh->eh_entries)) | |
3828 | goto out; | |
58590b06 TT |
3829 | last_ex = EXT_LAST_EXTENT(eh); |
3830 | /* | |
3831 | * We should clear the EOFBLOCKS_FL flag if we are writing the | |
3832 | * last block in the last extent in the file. We test this by | |
3833 | * first checking to see if the caller to | |
3834 | * ext4_ext_get_blocks() was interested in the last block (or | |
3835 | * a block beyond the last block) in the current extent. If | |
3836 | * this turns out to be false, we can bail out from this | |
3837 | * function immediately. | |
3838 | */ | |
d002ebf1 | 3839 | if (lblk + len < le32_to_cpu(last_ex->ee_block) + |
58590b06 TT |
3840 | ext4_ext_get_actual_len(last_ex)) |
3841 | return 0; | |
3842 | /* | |
3843 | * If the caller does appear to be planning to write at or | |
3844 | * beyond the end of the current extent, we then test to see | |
3845 | * if the current extent is the last extent in the file, by | |
3846 | * checking to make sure it was reached via the rightmost node | |
3847 | * at each level of the tree. | |
3848 | */ | |
3849 | for (i = depth-1; i >= 0; i--) | |
3850 | if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr)) | |
3851 | return 0; | |
afcff5d8 | 3852 | out: |
58590b06 TT |
3853 | ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS); |
3854 | return ext4_mark_inode_dirty(handle, inode); | |
3855 | } | |
3856 | ||
7b415bf6 AK |
3857 | /** |
3858 | * ext4_find_delalloc_range: find delayed allocated block in the given range. | |
3859 | * | |
7d1b1fbc | 3860 | * Return 1 if there is a delalloc block in the range, otherwise 0. |
7b415bf6 | 3861 | */ |
f7fec032 ZL |
3862 | int ext4_find_delalloc_range(struct inode *inode, |
3863 | ext4_lblk_t lblk_start, | |
3864 | ext4_lblk_t lblk_end) | |
7b415bf6 | 3865 | { |
7d1b1fbc | 3866 | struct extent_status es; |
7b415bf6 | 3867 | |
e30b5dca | 3868 | ext4_es_find_delayed_extent_range(inode, lblk_start, lblk_end, &es); |
06b0c886 | 3869 | if (es.es_len == 0) |
7d1b1fbc | 3870 | return 0; /* there is no delay extent in this tree */ |
06b0c886 ZL |
3871 | else if (es.es_lblk <= lblk_start && |
3872 | lblk_start < es.es_lblk + es.es_len) | |
7d1b1fbc | 3873 | return 1; |
06b0c886 | 3874 | else if (lblk_start <= es.es_lblk && es.es_lblk <= lblk_end) |
7d1b1fbc | 3875 | return 1; |
7b415bf6 | 3876 | else |
7d1b1fbc | 3877 | return 0; |
7b415bf6 AK |
3878 | } |
3879 | ||
7d1b1fbc | 3880 | int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk) |
7b415bf6 AK |
3881 | { |
3882 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); | |
3883 | ext4_lblk_t lblk_start, lblk_end; | |
f5a44db5 | 3884 | lblk_start = EXT4_LBLK_CMASK(sbi, lblk); |
7b415bf6 AK |
3885 | lblk_end = lblk_start + sbi->s_cluster_ratio - 1; |
3886 | ||
7d1b1fbc | 3887 | return ext4_find_delalloc_range(inode, lblk_start, lblk_end); |
7b415bf6 AK |
3888 | } |
3889 | ||
3890 | /** | |
3891 | * Determines how many complete clusters (out of those specified by the 'map') | |
3892 | * are under delalloc and were reserved quota for. | |
3893 | * This function is called when we are writing out the blocks that were | |
3894 | * originally written with their allocation delayed, but then the space was | |
3895 | * allocated using fallocate() before the delayed allocation could be resolved. | |
3896 | * The cases to look for are: | |
3897 | * ('=' indicated delayed allocated blocks | |
3898 | * '-' indicates non-delayed allocated blocks) | |
3899 | * (a) partial clusters towards beginning and/or end outside of allocated range | |
3900 | * are not delalloc'ed. | |
3901 | * Ex: | |
3902 | * |----c---=|====c====|====c====|===-c----| | |
3903 | * |++++++ allocated ++++++| | |
3904 | * ==> 4 complete clusters in above example | |
3905 | * | |
3906 | * (b) partial cluster (outside of allocated range) towards either end is | |
3907 | * marked for delayed allocation. In this case, we will exclude that | |
3908 | * cluster. | |
3909 | * Ex: | |
3910 | * |----====c========|========c========| | |
3911 | * |++++++ allocated ++++++| | |
3912 | * ==> 1 complete clusters in above example | |
3913 | * | |
3914 | * Ex: | |
3915 | * |================c================| | |
3916 | * |++++++ allocated ++++++| | |
3917 | * ==> 0 complete clusters in above example | |
3918 | * | |
3919 | * The ext4_da_update_reserve_space will be called only if we | |
3920 | * determine here that there were some "entire" clusters that span | |
3921 | * this 'allocated' range. | |
3922 | * In the non-bigalloc case, this function will just end up returning num_blks | |
3923 | * without ever calling ext4_find_delalloc_range. | |
3924 | */ | |
3925 | static unsigned int | |
3926 | get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start, | |
3927 | unsigned int num_blks) | |
3928 | { | |
3929 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); | |
3930 | ext4_lblk_t alloc_cluster_start, alloc_cluster_end; | |
3931 | ext4_lblk_t lblk_from, lblk_to, c_offset; | |
3932 | unsigned int allocated_clusters = 0; | |
3933 | ||
3934 | alloc_cluster_start = EXT4_B2C(sbi, lblk_start); | |
3935 | alloc_cluster_end = EXT4_B2C(sbi, lblk_start + num_blks - 1); | |
3936 | ||
3937 | /* max possible clusters for this allocation */ | |
3938 | allocated_clusters = alloc_cluster_end - alloc_cluster_start + 1; | |
3939 | ||
d8990240 AK |
3940 | trace_ext4_get_reserved_cluster_alloc(inode, lblk_start, num_blks); |
3941 | ||
7b415bf6 | 3942 | /* Check towards left side */ |
f5a44db5 | 3943 | c_offset = EXT4_LBLK_COFF(sbi, lblk_start); |
7b415bf6 | 3944 | if (c_offset) { |
f5a44db5 | 3945 | lblk_from = EXT4_LBLK_CMASK(sbi, lblk_start); |
7b415bf6 AK |
3946 | lblk_to = lblk_from + c_offset - 1; |
3947 | ||
7d1b1fbc | 3948 | if (ext4_find_delalloc_range(inode, lblk_from, lblk_to)) |
7b415bf6 AK |
3949 | allocated_clusters--; |
3950 | } | |
3951 | ||
3952 | /* Now check towards right. */ | |
f5a44db5 | 3953 | c_offset = EXT4_LBLK_COFF(sbi, lblk_start + num_blks); |
7b415bf6 AK |
3954 | if (allocated_clusters && c_offset) { |
3955 | lblk_from = lblk_start + num_blks; | |
3956 | lblk_to = lblk_from + (sbi->s_cluster_ratio - c_offset) - 1; | |
3957 | ||
7d1b1fbc | 3958 | if (ext4_find_delalloc_range(inode, lblk_from, lblk_to)) |
7b415bf6 AK |
3959 | allocated_clusters--; |
3960 | } | |
3961 | ||
3962 | return allocated_clusters; | |
3963 | } | |
3964 | ||
b8a86845 LC |
3965 | static int |
3966 | ext4_ext_convert_initialized_extent(handle_t *handle, struct inode *inode, | |
3967 | struct ext4_map_blocks *map, | |
3968 | struct ext4_ext_path *path, int flags, | |
3969 | unsigned int allocated, ext4_fsblk_t newblock) | |
3970 | { | |
3971 | int ret = 0; | |
3972 | int err = 0; | |
3973 | ||
3974 | /* | |
3975 | * Make sure that the extent is no bigger than we support with | |
556615dc | 3976 | * unwritten extent |
b8a86845 | 3977 | */ |
556615dc LC |
3978 | if (map->m_len > EXT_UNWRITTEN_MAX_LEN) |
3979 | map->m_len = EXT_UNWRITTEN_MAX_LEN / 2; | |
b8a86845 LC |
3980 | |
3981 | ret = ext4_convert_initialized_extents(handle, inode, map, | |
3982 | path); | |
3983 | if (ret >= 0) { | |
3984 | ext4_update_inode_fsync_trans(handle, inode, 1); | |
3985 | err = check_eofblocks_fl(handle, inode, map->m_lblk, | |
3986 | path, map->m_len); | |
3987 | } else | |
3988 | err = ret; | |
3989 | map->m_flags |= EXT4_MAP_UNWRITTEN; | |
3990 | if (allocated > map->m_len) | |
3991 | allocated = map->m_len; | |
3992 | map->m_len = allocated; | |
3993 | ||
3994 | return err ? err : allocated; | |
3995 | } | |
3996 | ||
0031462b | 3997 | static int |
556615dc | 3998 | ext4_ext_handle_unwritten_extents(handle_t *handle, struct inode *inode, |
e35fd660 | 3999 | struct ext4_map_blocks *map, |
0031462b | 4000 | struct ext4_ext_path *path, int flags, |
e35fd660 | 4001 | unsigned int allocated, ext4_fsblk_t newblock) |
0031462b MC |
4002 | { |
4003 | int ret = 0; | |
4004 | int err = 0; | |
f45ee3a1 | 4005 | ext4_io_end_t *io = ext4_inode_aio(inode); |
0031462b | 4006 | |
556615dc | 4007 | ext_debug("ext4_ext_handle_unwritten_extents: inode %lu, logical " |
88635ca2 | 4008 | "block %llu, max_blocks %u, flags %x, allocated %u\n", |
e35fd660 | 4009 | inode->i_ino, (unsigned long long)map->m_lblk, map->m_len, |
0031462b MC |
4010 | flags, allocated); |
4011 | ext4_ext_show_leaf(inode, path); | |
4012 | ||
27dd4385 | 4013 | /* |
556615dc | 4014 | * When writing into unwritten space, we should not fail to |
27dd4385 LC |
4015 | * allocate metadata blocks for the new extent block if needed. |
4016 | */ | |
4017 | flags |= EXT4_GET_BLOCKS_METADATA_NOFAIL; | |
4018 | ||
556615dc | 4019 | trace_ext4_ext_handle_unwritten_extents(inode, map, flags, |
b5645534 | 4020 | allocated, newblock); |
d8990240 | 4021 | |
c7064ef1 | 4022 | /* get_block() before submit the IO, split the extent */ |
c8b459f4 | 4023 | if (flags & EXT4_GET_BLOCKS_PRE_IO) { |
b8a86845 LC |
4024 | ret = ext4_split_convert_extents(handle, inode, map, |
4025 | path, flags | EXT4_GET_BLOCKS_CONVERT); | |
82e54229 DM |
4026 | if (ret <= 0) |
4027 | goto out; | |
5f524950 M |
4028 | /* |
4029 | * Flag the inode(non aio case) or end_io struct (aio case) | |
25985edc | 4030 | * that this IO needs to conversion to written when IO is |
5f524950 M |
4031 | * completed |
4032 | */ | |
0edeb71d TM |
4033 | if (io) |
4034 | ext4_set_io_unwritten_flag(inode, io); | |
4035 | else | |
19f5fb7a | 4036 | ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); |
a25a4e1a | 4037 | map->m_flags |= EXT4_MAP_UNWRITTEN; |
0031462b MC |
4038 | goto out; |
4039 | } | |
c7064ef1 | 4040 | /* IO end_io complete, convert the filled extent to written */ |
c8b459f4 | 4041 | if (flags & EXT4_GET_BLOCKS_CONVERT) { |
dee1f973 | 4042 | ret = ext4_convert_unwritten_extents_endio(handle, inode, map, |
0031462b | 4043 | path); |
58590b06 | 4044 | if (ret >= 0) { |
b436b9be | 4045 | ext4_update_inode_fsync_trans(handle, inode, 1); |
d002ebf1 ES |
4046 | err = check_eofblocks_fl(handle, inode, map->m_lblk, |
4047 | path, map->m_len); | |
58590b06 TT |
4048 | } else |
4049 | err = ret; | |
cdee7843 | 4050 | map->m_flags |= EXT4_MAP_MAPPED; |
15cc1767 | 4051 | map->m_pblk = newblock; |
cdee7843 ZL |
4052 | if (allocated > map->m_len) |
4053 | allocated = map->m_len; | |
4054 | map->m_len = allocated; | |
0031462b MC |
4055 | goto out2; |
4056 | } | |
4057 | /* buffered IO case */ | |
4058 | /* | |
4059 | * repeat fallocate creation request | |
4060 | * we already have an unwritten extent | |
4061 | */ | |
556615dc | 4062 | if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT) { |
a25a4e1a | 4063 | map->m_flags |= EXT4_MAP_UNWRITTEN; |
0031462b | 4064 | goto map_out; |
a25a4e1a | 4065 | } |
0031462b MC |
4066 | |
4067 | /* buffered READ or buffered write_begin() lookup */ | |
4068 | if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { | |
4069 | /* | |
4070 | * We have blocks reserved already. We | |
4071 | * return allocated blocks so that delalloc | |
4072 | * won't do block reservation for us. But | |
4073 | * the buffer head will be unmapped so that | |
4074 | * a read from the block returns 0s. | |
4075 | */ | |
e35fd660 | 4076 | map->m_flags |= EXT4_MAP_UNWRITTEN; |
0031462b MC |
4077 | goto out1; |
4078 | } | |
4079 | ||
4080 | /* buffered write, writepage time, convert*/ | |
27dd4385 | 4081 | ret = ext4_ext_convert_to_initialized(handle, inode, map, path, flags); |
a4e5d88b | 4082 | if (ret >= 0) |
b436b9be | 4083 | ext4_update_inode_fsync_trans(handle, inode, 1); |
0031462b MC |
4084 | out: |
4085 | if (ret <= 0) { | |
4086 | err = ret; | |
4087 | goto out2; | |
4088 | } else | |
4089 | allocated = ret; | |
e35fd660 | 4090 | map->m_flags |= EXT4_MAP_NEW; |
515f41c3 AK |
4091 | /* |
4092 | * if we allocated more blocks than requested | |
4093 | * we need to make sure we unmap the extra block | |
4094 | * allocated. The actual needed block will get | |
4095 | * unmapped later when we find the buffer_head marked | |
4096 | * new. | |
4097 | */ | |
e35fd660 | 4098 | if (allocated > map->m_len) { |
515f41c3 | 4099 | unmap_underlying_metadata_blocks(inode->i_sb->s_bdev, |
e35fd660 TT |
4100 | newblock + map->m_len, |
4101 | allocated - map->m_len); | |
4102 | allocated = map->m_len; | |
515f41c3 | 4103 | } |
3a225670 | 4104 | map->m_len = allocated; |
5f634d06 AK |
4105 | |
4106 | /* | |
4107 | * If we have done fallocate with the offset that is already | |
4108 | * delayed allocated, we would have block reservation | |
4109 | * and quota reservation done in the delayed write path. | |
4110 | * But fallocate would have already updated quota and block | |
4111 | * count for this offset. So cancel these reservation | |
4112 | */ | |
7b415bf6 AK |
4113 | if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { |
4114 | unsigned int reserved_clusters; | |
4115 | reserved_clusters = get_reserved_cluster_alloc(inode, | |
4116 | map->m_lblk, map->m_len); | |
4117 | if (reserved_clusters) | |
4118 | ext4_da_update_reserve_space(inode, | |
4119 | reserved_clusters, | |
4120 | 0); | |
4121 | } | |
5f634d06 | 4122 | |
0031462b | 4123 | map_out: |
e35fd660 | 4124 | map->m_flags |= EXT4_MAP_MAPPED; |
a4e5d88b DM |
4125 | if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0) { |
4126 | err = check_eofblocks_fl(handle, inode, map->m_lblk, path, | |
4127 | map->m_len); | |
4128 | if (err < 0) | |
4129 | goto out2; | |
4130 | } | |
0031462b | 4131 | out1: |
e35fd660 TT |
4132 | if (allocated > map->m_len) |
4133 | allocated = map->m_len; | |
0031462b | 4134 | ext4_ext_show_leaf(inode, path); |
e35fd660 TT |
4135 | map->m_pblk = newblock; |
4136 | map->m_len = allocated; | |
0031462b | 4137 | out2: |
0031462b MC |
4138 | return err ? err : allocated; |
4139 | } | |
58590b06 | 4140 | |
4d33b1ef TT |
4141 | /* |
4142 | * get_implied_cluster_alloc - check to see if the requested | |
4143 | * allocation (in the map structure) overlaps with a cluster already | |
4144 | * allocated in an extent. | |
d8990240 | 4145 | * @sb The filesystem superblock structure |
4d33b1ef TT |
4146 | * @map The requested lblk->pblk mapping |
4147 | * @ex The extent structure which might contain an implied | |
4148 | * cluster allocation | |
4149 | * | |
4150 | * This function is called by ext4_ext_map_blocks() after we failed to | |
4151 | * find blocks that were already in the inode's extent tree. Hence, | |
4152 | * we know that the beginning of the requested region cannot overlap | |
4153 | * the extent from the inode's extent tree. There are three cases we | |
4154 | * want to catch. The first is this case: | |
4155 | * | |
4156 | * |--- cluster # N--| | |
4157 | * |--- extent ---| |---- requested region ---| | |
4158 | * |==========| | |
4159 | * | |
4160 | * The second case that we need to test for is this one: | |
4161 | * | |
4162 | * |--------- cluster # N ----------------| | |
4163 | * |--- requested region --| |------- extent ----| | |
4164 | * |=======================| | |
4165 | * | |
4166 | * The third case is when the requested region lies between two extents | |
4167 | * within the same cluster: | |
4168 | * |------------- cluster # N-------------| | |
4169 | * |----- ex -----| |---- ex_right ----| | |
4170 | * |------ requested region ------| | |
4171 | * |================| | |
4172 | * | |
4173 | * In each of the above cases, we need to set the map->m_pblk and | |
4174 | * map->m_len so it corresponds to the return the extent labelled as | |
4175 | * "|====|" from cluster #N, since it is already in use for data in | |
4176 | * cluster EXT4_B2C(sbi, map->m_lblk). We will then return 1 to | |
4177 | * signal to ext4_ext_map_blocks() that map->m_pblk should be treated | |
4178 | * as a new "allocated" block region. Otherwise, we will return 0 and | |
4179 | * ext4_ext_map_blocks() will then allocate one or more new clusters | |
4180 | * by calling ext4_mb_new_blocks(). | |
4181 | */ | |
d8990240 | 4182 | static int get_implied_cluster_alloc(struct super_block *sb, |
4d33b1ef TT |
4183 | struct ext4_map_blocks *map, |
4184 | struct ext4_extent *ex, | |
4185 | struct ext4_ext_path *path) | |
4186 | { | |
d8990240 | 4187 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
f5a44db5 | 4188 | ext4_lblk_t c_offset = EXT4_LBLK_COFF(sbi, map->m_lblk); |
4d33b1ef | 4189 | ext4_lblk_t ex_cluster_start, ex_cluster_end; |
14d7f3ef | 4190 | ext4_lblk_t rr_cluster_start; |
4d33b1ef TT |
4191 | ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block); |
4192 | ext4_fsblk_t ee_start = ext4_ext_pblock(ex); | |
4193 | unsigned short ee_len = ext4_ext_get_actual_len(ex); | |
4194 | ||
4195 | /* The extent passed in that we are trying to match */ | |
4196 | ex_cluster_start = EXT4_B2C(sbi, ee_block); | |
4197 | ex_cluster_end = EXT4_B2C(sbi, ee_block + ee_len - 1); | |
4198 | ||
4199 | /* The requested region passed into ext4_map_blocks() */ | |
4200 | rr_cluster_start = EXT4_B2C(sbi, map->m_lblk); | |
4d33b1ef TT |
4201 | |
4202 | if ((rr_cluster_start == ex_cluster_end) || | |
4203 | (rr_cluster_start == ex_cluster_start)) { | |
4204 | if (rr_cluster_start == ex_cluster_end) | |
4205 | ee_start += ee_len - 1; | |
f5a44db5 | 4206 | map->m_pblk = EXT4_PBLK_CMASK(sbi, ee_start) + c_offset; |
4d33b1ef TT |
4207 | map->m_len = min(map->m_len, |
4208 | (unsigned) sbi->s_cluster_ratio - c_offset); | |
4209 | /* | |
4210 | * Check for and handle this case: | |
4211 | * | |
4212 | * |--------- cluster # N-------------| | |
4213 | * |------- extent ----| | |
4214 | * |--- requested region ---| | |
4215 | * |===========| | |
4216 | */ | |
4217 | ||
4218 | if (map->m_lblk < ee_block) | |
4219 | map->m_len = min(map->m_len, ee_block - map->m_lblk); | |
4220 | ||
4221 | /* | |
4222 | * Check for the case where there is already another allocated | |
4223 | * block to the right of 'ex' but before the end of the cluster. | |
4224 | * | |
4225 | * |------------- cluster # N-------------| | |
4226 | * |----- ex -----| |---- ex_right ----| | |
4227 | * |------ requested region ------| | |
4228 | * |================| | |
4229 | */ | |
4230 | if (map->m_lblk > ee_block) { | |
4231 | ext4_lblk_t next = ext4_ext_next_allocated_block(path); | |
4232 | map->m_len = min(map->m_len, next - map->m_lblk); | |
4233 | } | |
d8990240 AK |
4234 | |
4235 | trace_ext4_get_implied_cluster_alloc_exit(sb, map, 1); | |
4d33b1ef TT |
4236 | return 1; |
4237 | } | |
d8990240 AK |
4238 | |
4239 | trace_ext4_get_implied_cluster_alloc_exit(sb, map, 0); | |
4d33b1ef TT |
4240 | return 0; |
4241 | } | |
4242 | ||
4243 | ||
c278bfec | 4244 | /* |
f5ab0d1f MC |
4245 | * Block allocation/map/preallocation routine for extents based files |
4246 | * | |
4247 | * | |
c278bfec | 4248 | * Need to be called with |
0e855ac8 AK |
4249 | * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block |
4250 | * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem) | |
f5ab0d1f MC |
4251 | * |
4252 | * return > 0, number of of blocks already mapped/allocated | |
4253 | * if create == 0 and these are pre-allocated blocks | |
4254 | * buffer head is unmapped | |
4255 | * otherwise blocks are mapped | |
4256 | * | |
4257 | * return = 0, if plain look up failed (blocks have not been allocated) | |
4258 | * buffer head is unmapped | |
4259 | * | |
4260 | * return < 0, error case. | |
c278bfec | 4261 | */ |
e35fd660 TT |
4262 | int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, |
4263 | struct ext4_map_blocks *map, int flags) | |
a86c6181 AT |
4264 | { |
4265 | struct ext4_ext_path *path = NULL; | |
4d33b1ef TT |
4266 | struct ext4_extent newex, *ex, *ex2; |
4267 | struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); | |
0562e0ba | 4268 | ext4_fsblk_t newblock = 0; |
ce37c429 | 4269 | int free_on_err = 0, err = 0, depth, ret; |
4d33b1ef | 4270 | unsigned int allocated = 0, offset = 0; |
81fdbb4a | 4271 | unsigned int allocated_clusters = 0; |
c9de560d | 4272 | struct ext4_allocation_request ar; |
f45ee3a1 | 4273 | ext4_io_end_t *io = ext4_inode_aio(inode); |
4d33b1ef | 4274 | ext4_lblk_t cluster_offset; |
82e54229 | 4275 | int set_unwritten = 0; |
a86c6181 | 4276 | |
84fe3bef | 4277 | ext_debug("blocks %u/%u requested for inode %lu\n", |
e35fd660 | 4278 | map->m_lblk, map->m_len, inode->i_ino); |
0562e0ba | 4279 | trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); |
a86c6181 | 4280 | |
a86c6181 | 4281 | /* find extent for this block */ |
107a7bd3 | 4282 | path = ext4_ext_find_extent(inode, map->m_lblk, NULL, 0); |
a86c6181 AT |
4283 | if (IS_ERR(path)) { |
4284 | err = PTR_ERR(path); | |
4285 | path = NULL; | |
4286 | goto out2; | |
4287 | } | |
4288 | ||
4289 | depth = ext_depth(inode); | |
4290 | ||
4291 | /* | |
d0d856e8 RD |
4292 | * consistent leaf must not be empty; |
4293 | * this situation is possible, though, _during_ tree modification; | |
a86c6181 AT |
4294 | * this is why assert can't be put in ext4_ext_find_extent() |
4295 | */ | |
273df556 FM |
4296 | if (unlikely(path[depth].p_ext == NULL && depth != 0)) { |
4297 | EXT4_ERROR_INODE(inode, "bad extent address " | |
f70f362b TT |
4298 | "lblock: %lu, depth: %d pblock %lld", |
4299 | (unsigned long) map->m_lblk, depth, | |
4300 | path[depth].p_block); | |
034fb4c9 SP |
4301 | err = -EIO; |
4302 | goto out2; | |
4303 | } | |
a86c6181 | 4304 | |
7e028976 AM |
4305 | ex = path[depth].p_ext; |
4306 | if (ex) { | |
725d26d3 | 4307 | ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block); |
bf89d16f | 4308 | ext4_fsblk_t ee_start = ext4_ext_pblock(ex); |
a2df2a63 | 4309 | unsigned short ee_len; |
471d4011 | 4310 | |
b8a86845 | 4311 | |
471d4011 | 4312 | /* |
556615dc | 4313 | * unwritten extents are treated as holes, except that |
56055d3a | 4314 | * we split out initialized portions during a write. |
471d4011 | 4315 | */ |
a2df2a63 | 4316 | ee_len = ext4_ext_get_actual_len(ex); |
d8990240 AK |
4317 | |
4318 | trace_ext4_ext_show_extent(inode, ee_block, ee_start, ee_len); | |
4319 | ||
d0d856e8 | 4320 | /* if found extent covers block, simply return it */ |
e35fd660 TT |
4321 | if (in_range(map->m_lblk, ee_block, ee_len)) { |
4322 | newblock = map->m_lblk - ee_block + ee_start; | |
d0d856e8 | 4323 | /* number of remaining blocks in the extent */ |
e35fd660 TT |
4324 | allocated = ee_len - (map->m_lblk - ee_block); |
4325 | ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk, | |
4326 | ee_block, ee_len, newblock); | |
56055d3a | 4327 | |
b8a86845 LC |
4328 | /* |
4329 | * If the extent is initialized check whether the | |
4330 | * caller wants to convert it to unwritten. | |
4331 | */ | |
556615dc | 4332 | if ((!ext4_ext_is_unwritten(ex)) && |
b8a86845 LC |
4333 | (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN)) { |
4334 | allocated = ext4_ext_convert_initialized_extent( | |
4335 | handle, inode, map, path, flags, | |
4336 | allocated, newblock); | |
4337 | goto out2; | |
556615dc | 4338 | } else if (!ext4_ext_is_unwritten(ex)) |
7877191c | 4339 | goto out; |
69eb33dc | 4340 | |
556615dc | 4341 | ret = ext4_ext_handle_unwritten_extents( |
7877191c LC |
4342 | handle, inode, map, path, flags, |
4343 | allocated, newblock); | |
ce37c429 EW |
4344 | if (ret < 0) |
4345 | err = ret; | |
4346 | else | |
4347 | allocated = ret; | |
31cf0f2c | 4348 | goto out2; |
a86c6181 AT |
4349 | } |
4350 | } | |
4351 | ||
7b415bf6 | 4352 | if ((sbi->s_cluster_ratio > 1) && |
7d1b1fbc | 4353 | ext4_find_delalloc_cluster(inode, map->m_lblk)) |
7b415bf6 AK |
4354 | map->m_flags |= EXT4_MAP_FROM_CLUSTER; |
4355 | ||
a86c6181 | 4356 | /* |
d0d856e8 | 4357 | * requested block isn't allocated yet; |
a86c6181 AT |
4358 | * we couldn't try to create block if create flag is zero |
4359 | */ | |
c2177057 | 4360 | if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { |
56055d3a AA |
4361 | /* |
4362 | * put just found gap into cache to speed up | |
4363 | * subsequent requests | |
4364 | */ | |
d100eef2 ZL |
4365 | if ((flags & EXT4_GET_BLOCKS_NO_PUT_HOLE) == 0) |
4366 | ext4_ext_put_gap_in_cache(inode, path, map->m_lblk); | |
a86c6181 AT |
4367 | goto out2; |
4368 | } | |
4d33b1ef | 4369 | |
a86c6181 | 4370 | /* |
c2ea3fde | 4371 | * Okay, we need to do block allocation. |
63f57933 | 4372 | */ |
7b415bf6 | 4373 | map->m_flags &= ~EXT4_MAP_FROM_CLUSTER; |
4d33b1ef | 4374 | newex.ee_block = cpu_to_le32(map->m_lblk); |
d0abafac | 4375 | cluster_offset = EXT4_LBLK_COFF(sbi, map->m_lblk); |
4d33b1ef TT |
4376 | |
4377 | /* | |
4378 | * If we are doing bigalloc, check to see if the extent returned | |
4379 | * by ext4_ext_find_extent() implies a cluster we can use. | |
4380 | */ | |
4381 | if (cluster_offset && ex && | |
d8990240 | 4382 | get_implied_cluster_alloc(inode->i_sb, map, ex, path)) { |
4d33b1ef TT |
4383 | ar.len = allocated = map->m_len; |
4384 | newblock = map->m_pblk; | |
7b415bf6 | 4385 | map->m_flags |= EXT4_MAP_FROM_CLUSTER; |
4d33b1ef TT |
4386 | goto got_allocated_blocks; |
4387 | } | |
a86c6181 | 4388 | |
c9de560d | 4389 | /* find neighbour allocated blocks */ |
e35fd660 | 4390 | ar.lleft = map->m_lblk; |
c9de560d AT |
4391 | err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft); |
4392 | if (err) | |
4393 | goto out2; | |
e35fd660 | 4394 | ar.lright = map->m_lblk; |
4d33b1ef TT |
4395 | ex2 = NULL; |
4396 | err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright, &ex2); | |
c9de560d AT |
4397 | if (err) |
4398 | goto out2; | |
25d14f98 | 4399 | |
4d33b1ef TT |
4400 | /* Check if the extent after searching to the right implies a |
4401 | * cluster we can use. */ | |
4402 | if ((sbi->s_cluster_ratio > 1) && ex2 && | |
d8990240 | 4403 | get_implied_cluster_alloc(inode->i_sb, map, ex2, path)) { |
4d33b1ef TT |
4404 | ar.len = allocated = map->m_len; |
4405 | newblock = map->m_pblk; | |
7b415bf6 | 4406 | map->m_flags |= EXT4_MAP_FROM_CLUSTER; |
4d33b1ef TT |
4407 | goto got_allocated_blocks; |
4408 | } | |
4409 | ||
749269fa AA |
4410 | /* |
4411 | * See if request is beyond maximum number of blocks we can have in | |
4412 | * a single extent. For an initialized extent this limit is | |
556615dc LC |
4413 | * EXT_INIT_MAX_LEN and for an unwritten extent this limit is |
4414 | * EXT_UNWRITTEN_MAX_LEN. | |
749269fa | 4415 | */ |
e35fd660 | 4416 | if (map->m_len > EXT_INIT_MAX_LEN && |
556615dc | 4417 | !(flags & EXT4_GET_BLOCKS_UNWRIT_EXT)) |
e35fd660 | 4418 | map->m_len = EXT_INIT_MAX_LEN; |
556615dc LC |
4419 | else if (map->m_len > EXT_UNWRITTEN_MAX_LEN && |
4420 | (flags & EXT4_GET_BLOCKS_UNWRIT_EXT)) | |
4421 | map->m_len = EXT_UNWRITTEN_MAX_LEN; | |
749269fa | 4422 | |
e35fd660 | 4423 | /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */ |
e35fd660 | 4424 | newex.ee_len = cpu_to_le16(map->m_len); |
4d33b1ef | 4425 | err = ext4_ext_check_overlap(sbi, inode, &newex, path); |
25d14f98 | 4426 | if (err) |
b939e376 | 4427 | allocated = ext4_ext_get_actual_len(&newex); |
25d14f98 | 4428 | else |
e35fd660 | 4429 | allocated = map->m_len; |
c9de560d AT |
4430 | |
4431 | /* allocate new block */ | |
4432 | ar.inode = inode; | |
e35fd660 TT |
4433 | ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk); |
4434 | ar.logical = map->m_lblk; | |
4d33b1ef TT |
4435 | /* |
4436 | * We calculate the offset from the beginning of the cluster | |
4437 | * for the logical block number, since when we allocate a | |
4438 | * physical cluster, the physical block should start at the | |
4439 | * same offset from the beginning of the cluster. This is | |
4440 | * needed so that future calls to get_implied_cluster_alloc() | |
4441 | * work correctly. | |
4442 | */ | |
f5a44db5 | 4443 | offset = EXT4_LBLK_COFF(sbi, map->m_lblk); |
4d33b1ef TT |
4444 | ar.len = EXT4_NUM_B2C(sbi, offset+allocated); |
4445 | ar.goal -= offset; | |
4446 | ar.logical -= offset; | |
c9de560d AT |
4447 | if (S_ISREG(inode->i_mode)) |
4448 | ar.flags = EXT4_MB_HINT_DATA; | |
4449 | else | |
4450 | /* disable in-core preallocation for non-regular files */ | |
4451 | ar.flags = 0; | |
556b27ab VH |
4452 | if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE) |
4453 | ar.flags |= EXT4_MB_HINT_NOPREALLOC; | |
c9de560d | 4454 | newblock = ext4_mb_new_blocks(handle, &ar, &err); |
a86c6181 AT |
4455 | if (!newblock) |
4456 | goto out2; | |
84fe3bef | 4457 | ext_debug("allocate new block: goal %llu, found %llu/%u\n", |
498e5f24 | 4458 | ar.goal, newblock, allocated); |
4d33b1ef | 4459 | free_on_err = 1; |
7b415bf6 | 4460 | allocated_clusters = ar.len; |
4d33b1ef TT |
4461 | ar.len = EXT4_C2B(sbi, ar.len) - offset; |
4462 | if (ar.len > allocated) | |
4463 | ar.len = allocated; | |
a86c6181 | 4464 | |
4d33b1ef | 4465 | got_allocated_blocks: |
a86c6181 | 4466 | /* try to insert new extent into found leaf and return */ |
4d33b1ef | 4467 | ext4_ext_store_pblock(&newex, newblock + offset); |
c9de560d | 4468 | newex.ee_len = cpu_to_le16(ar.len); |
556615dc LC |
4469 | /* Mark unwritten */ |
4470 | if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT){ | |
4471 | ext4_ext_mark_unwritten(&newex); | |
a25a4e1a | 4472 | map->m_flags |= EXT4_MAP_UNWRITTEN; |
8d5d02e6 | 4473 | /* |
744692dc | 4474 | * io_end structure was created for every IO write to an |
556615dc | 4475 | * unwritten extent. To avoid unnecessary conversion, |
744692dc | 4476 | * here we flag the IO that really needs the conversion. |
5f524950 | 4477 | * For non asycn direct IO case, flag the inode state |
25985edc | 4478 | * that we need to perform conversion when IO is done. |
8d5d02e6 | 4479 | */ |
c8b459f4 | 4480 | if (flags & EXT4_GET_BLOCKS_PRE_IO) |
82e54229 | 4481 | set_unwritten = 1; |
8d5d02e6 | 4482 | } |
c8d46e41 | 4483 | |
a4e5d88b DM |
4484 | err = 0; |
4485 | if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0) | |
4486 | err = check_eofblocks_fl(handle, inode, map->m_lblk, | |
4487 | path, ar.len); | |
575a1d4b JZ |
4488 | if (!err) |
4489 | err = ext4_ext_insert_extent(handle, inode, path, | |
4490 | &newex, flags); | |
82e54229 DM |
4491 | |
4492 | if (!err && set_unwritten) { | |
4493 | if (io) | |
4494 | ext4_set_io_unwritten_flag(inode, io); | |
4495 | else | |
4496 | ext4_set_inode_state(inode, | |
4497 | EXT4_STATE_DIO_UNWRITTEN); | |
4498 | } | |
4499 | ||
4d33b1ef | 4500 | if (err && free_on_err) { |
7132de74 MP |
4501 | int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ? |
4502 | EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0; | |
315054f0 | 4503 | /* free data blocks we just allocated */ |
c9de560d AT |
4504 | /* not a good idea to call discard here directly, |
4505 | * but otherwise we'd need to call it every free() */ | |
c2ea3fde | 4506 | ext4_discard_preallocations(inode); |
c8e15130 TT |
4507 | ext4_free_blocks(handle, inode, NULL, newblock, |
4508 | EXT4_C2B(sbi, allocated_clusters), fb_flags); | |
a86c6181 | 4509 | goto out2; |
315054f0 | 4510 | } |
a86c6181 | 4511 | |
a86c6181 | 4512 | /* previous routine could use block we allocated */ |
bf89d16f | 4513 | newblock = ext4_ext_pblock(&newex); |
b939e376 | 4514 | allocated = ext4_ext_get_actual_len(&newex); |
e35fd660 TT |
4515 | if (allocated > map->m_len) |
4516 | allocated = map->m_len; | |
4517 | map->m_flags |= EXT4_MAP_NEW; | |
a86c6181 | 4518 | |
5f634d06 AK |
4519 | /* |
4520 | * Update reserved blocks/metadata blocks after successful | |
4521 | * block allocation which had been deferred till now. | |
4522 | */ | |
7b415bf6 | 4523 | if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { |
81fdbb4a | 4524 | unsigned int reserved_clusters; |
7b415bf6 | 4525 | /* |
81fdbb4a | 4526 | * Check how many clusters we had reserved this allocated range |
7b415bf6 AK |
4527 | */ |
4528 | reserved_clusters = get_reserved_cluster_alloc(inode, | |
4529 | map->m_lblk, allocated); | |
4530 | if (map->m_flags & EXT4_MAP_FROM_CLUSTER) { | |
4531 | if (reserved_clusters) { | |
4532 | /* | |
4533 | * We have clusters reserved for this range. | |
4534 | * But since we are not doing actual allocation | |
4535 | * and are simply using blocks from previously | |
4536 | * allocated cluster, we should release the | |
4537 | * reservation and not claim quota. | |
4538 | */ | |
4539 | ext4_da_update_reserve_space(inode, | |
4540 | reserved_clusters, 0); | |
4541 | } | |
4542 | } else { | |
4543 | BUG_ON(allocated_clusters < reserved_clusters); | |
7b415bf6 | 4544 | if (reserved_clusters < allocated_clusters) { |
5356f261 | 4545 | struct ext4_inode_info *ei = EXT4_I(inode); |
7b415bf6 AK |
4546 | int reservation = allocated_clusters - |
4547 | reserved_clusters; | |
4548 | /* | |
4549 | * It seems we claimed few clusters outside of | |
4550 | * the range of this allocation. We should give | |
4551 | * it back to the reservation pool. This can | |
4552 | * happen in the following case: | |
4553 | * | |
4554 | * * Suppose s_cluster_ratio is 4 (i.e., each | |
4555 | * cluster has 4 blocks. Thus, the clusters | |
4556 | * are [0-3],[4-7],[8-11]... | |
4557 | * * First comes delayed allocation write for | |
4558 | * logical blocks 10 & 11. Since there were no | |
4559 | * previous delayed allocated blocks in the | |
4560 | * range [8-11], we would reserve 1 cluster | |
4561 | * for this write. | |
4562 | * * Next comes write for logical blocks 3 to 8. | |
4563 | * In this case, we will reserve 2 clusters | |
4564 | * (for [0-3] and [4-7]; and not for [8-11] as | |
4565 | * that range has a delayed allocated blocks. | |
4566 | * Thus total reserved clusters now becomes 3. | |
4567 | * * Now, during the delayed allocation writeout | |
4568 | * time, we will first write blocks [3-8] and | |
4569 | * allocate 3 clusters for writing these | |
4570 | * blocks. Also, we would claim all these | |
4571 | * three clusters above. | |
4572 | * * Now when we come here to writeout the | |
4573 | * blocks [10-11], we would expect to claim | |
4574 | * the reservation of 1 cluster we had made | |
4575 | * (and we would claim it since there are no | |
4576 | * more delayed allocated blocks in the range | |
4577 | * [8-11]. But our reserved cluster count had | |
4578 | * already gone to 0. | |
4579 | * | |
4580 | * Thus, at the step 4 above when we determine | |
4581 | * that there are still some unwritten delayed | |
4582 | * allocated blocks outside of our current | |
4583 | * block range, we should increment the | |
4584 | * reserved clusters count so that when the | |
4585 | * remaining blocks finally gets written, we | |
4586 | * could claim them. | |
4587 | */ | |
5356f261 AK |
4588 | dquot_reserve_block(inode, |
4589 | EXT4_C2B(sbi, reservation)); | |
4590 | spin_lock(&ei->i_block_reservation_lock); | |
4591 | ei->i_reserved_data_blocks += reservation; | |
4592 | spin_unlock(&ei->i_block_reservation_lock); | |
7b415bf6 | 4593 | } |
232ec872 LC |
4594 | /* |
4595 | * We will claim quota for all newly allocated blocks. | |
4596 | * We're updating the reserved space *after* the | |
4597 | * correction above so we do not accidentally free | |
4598 | * all the metadata reservation because we might | |
4599 | * actually need it later on. | |
4600 | */ | |
4601 | ext4_da_update_reserve_space(inode, allocated_clusters, | |
4602 | 1); | |
7b415bf6 AK |
4603 | } |
4604 | } | |
5f634d06 | 4605 | |
b436b9be JK |
4606 | /* |
4607 | * Cache the extent and update transaction to commit on fdatasync only | |
556615dc | 4608 | * when it is _not_ an unwritten extent. |
b436b9be | 4609 | */ |
556615dc | 4610 | if ((flags & EXT4_GET_BLOCKS_UNWRIT_EXT) == 0) |
b436b9be | 4611 | ext4_update_inode_fsync_trans(handle, inode, 1); |
69eb33dc | 4612 | else |
b436b9be | 4613 | ext4_update_inode_fsync_trans(handle, inode, 0); |
a86c6181 | 4614 | out: |
e35fd660 TT |
4615 | if (allocated > map->m_len) |
4616 | allocated = map->m_len; | |
a86c6181 | 4617 | ext4_ext_show_leaf(inode, path); |
e35fd660 TT |
4618 | map->m_flags |= EXT4_MAP_MAPPED; |
4619 | map->m_pblk = newblock; | |
4620 | map->m_len = allocated; | |
a86c6181 AT |
4621 | out2: |
4622 | if (path) { | |
4623 | ext4_ext_drop_refs(path); | |
4624 | kfree(path); | |
4625 | } | |
e861304b | 4626 | |
63b99968 TT |
4627 | trace_ext4_ext_map_blocks_exit(inode, flags, map, |
4628 | err ? err : allocated); | |
4629 | ext4_es_lru_add(inode); | |
7877191c | 4630 | return err ? err : allocated; |
a86c6181 AT |
4631 | } |
4632 | ||
819c4920 | 4633 | void ext4_ext_truncate(handle_t *handle, struct inode *inode) |
a86c6181 | 4634 | { |
a86c6181 | 4635 | struct super_block *sb = inode->i_sb; |
725d26d3 | 4636 | ext4_lblk_t last_block; |
a86c6181 AT |
4637 | int err = 0; |
4638 | ||
a86c6181 | 4639 | /* |
d0d856e8 RD |
4640 | * TODO: optimization is possible here. |
4641 | * Probably we need not scan at all, | |
4642 | * because page truncation is enough. | |
a86c6181 | 4643 | */ |
a86c6181 AT |
4644 | |
4645 | /* we have to know where to truncate from in crash case */ | |
4646 | EXT4_I(inode)->i_disksize = inode->i_size; | |
4647 | ext4_mark_inode_dirty(handle, inode); | |
4648 | ||
4649 | last_block = (inode->i_size + sb->s_blocksize - 1) | |
4650 | >> EXT4_BLOCK_SIZE_BITS(sb); | |
8acd5e9b | 4651 | retry: |
51865fda ZL |
4652 | err = ext4_es_remove_extent(inode, last_block, |
4653 | EXT_MAX_BLOCKS - last_block); | |
94eec0fc | 4654 | if (err == -ENOMEM) { |
8acd5e9b TT |
4655 | cond_resched(); |
4656 | congestion_wait(BLK_RW_ASYNC, HZ/50); | |
4657 | goto retry; | |
4658 | } | |
4659 | if (err) { | |
4660 | ext4_std_error(inode->i_sb, err); | |
4661 | return; | |
4662 | } | |
5f95d21f | 4663 | err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1); |
8acd5e9b | 4664 | ext4_std_error(inode->i_sb, err); |
a86c6181 AT |
4665 | } |
4666 | ||
0e8b6879 LC |
4667 | static int ext4_alloc_file_blocks(struct file *file, ext4_lblk_t offset, |
4668 | ext4_lblk_t len, int flags, int mode) | |
4669 | { | |
4670 | struct inode *inode = file_inode(file); | |
4671 | handle_t *handle; | |
4672 | int ret = 0; | |
4673 | int ret2 = 0; | |
4674 | int retries = 0; | |
4675 | struct ext4_map_blocks map; | |
4676 | unsigned int credits; | |
4677 | ||
4678 | map.m_lblk = offset; | |
4679 | /* | |
4680 | * Don't normalize the request if it can fit in one extent so | |
4681 | * that it doesn't get unnecessarily split into multiple | |
4682 | * extents. | |
4683 | */ | |
556615dc | 4684 | if (len <= EXT_UNWRITTEN_MAX_LEN) |
0e8b6879 LC |
4685 | flags |= EXT4_GET_BLOCKS_NO_NORMALIZE; |
4686 | ||
4687 | /* | |
4688 | * credits to insert 1 extent into extent tree | |
4689 | */ | |
4690 | credits = ext4_chunk_trans_blocks(inode, len); | |
4691 | ||
4692 | retry: | |
4693 | while (ret >= 0 && ret < len) { | |
4694 | map.m_lblk = map.m_lblk + ret; | |
4695 | map.m_len = len = len - ret; | |
4696 | handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, | |
4697 | credits); | |
4698 | if (IS_ERR(handle)) { | |
4699 | ret = PTR_ERR(handle); | |
4700 | break; | |
4701 | } | |
4702 | ret = ext4_map_blocks(handle, inode, &map, flags); | |
4703 | if (ret <= 0) { | |
4704 | ext4_debug("inode #%lu: block %u: len %u: " | |
4705 | "ext4_ext_map_blocks returned %d", | |
4706 | inode->i_ino, map.m_lblk, | |
4707 | map.m_len, ret); | |
4708 | ext4_mark_inode_dirty(handle, inode); | |
4709 | ret2 = ext4_journal_stop(handle); | |
4710 | break; | |
4711 | } | |
4712 | ret2 = ext4_journal_stop(handle); | |
4713 | if (ret2) | |
4714 | break; | |
4715 | } | |
4716 | if (ret == -ENOSPC && | |
4717 | ext4_should_retry_alloc(inode->i_sb, &retries)) { | |
4718 | ret = 0; | |
4719 | goto retry; | |
4720 | } | |
4721 | ||
4722 | return ret > 0 ? ret2 : ret; | |
4723 | } | |
4724 | ||
b8a86845 LC |
4725 | static long ext4_zero_range(struct file *file, loff_t offset, |
4726 | loff_t len, int mode) | |
4727 | { | |
4728 | struct inode *inode = file_inode(file); | |
4729 | handle_t *handle = NULL; | |
4730 | unsigned int max_blocks; | |
4731 | loff_t new_size = 0; | |
4732 | int ret = 0; | |
4733 | int flags; | |
4734 | int partial; | |
4735 | loff_t start, end; | |
4736 | ext4_lblk_t lblk; | |
4737 | struct address_space *mapping = inode->i_mapping; | |
4738 | unsigned int blkbits = inode->i_blkbits; | |
4739 | ||
4740 | trace_ext4_zero_range(inode, offset, len, mode); | |
4741 | ||
6c5e73d3 | 4742 | if (!S_ISREG(inode->i_mode)) |
4743 | return -EINVAL; | |
4744 | ||
e1ee60fd NJ |
4745 | /* Call ext4_force_commit to flush all data in case of data=journal. */ |
4746 | if (ext4_should_journal_data(inode)) { | |
4747 | ret = ext4_force_commit(inode->i_sb); | |
4748 | if (ret) | |
4749 | return ret; | |
4750 | } | |
4751 | ||
b8a86845 LC |
4752 | /* |
4753 | * Write out all dirty pages to avoid race conditions | |
4754 | * Then release them. | |
4755 | */ | |
4756 | if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { | |
4757 | ret = filemap_write_and_wait_range(mapping, offset, | |
4758 | offset + len - 1); | |
4759 | if (ret) | |
4760 | return ret; | |
4761 | } | |
4762 | ||
4763 | /* | |
4764 | * Round up offset. This is not fallocate, we neet to zero out | |
4765 | * blocks, so convert interior block aligned part of the range to | |
4766 | * unwritten and possibly manually zero out unaligned parts of the | |
4767 | * range. | |
4768 | */ | |
4769 | start = round_up(offset, 1 << blkbits); | |
4770 | end = round_down((offset + len), 1 << blkbits); | |
4771 | ||
4772 | if (start < offset || end > offset + len) | |
4773 | return -EINVAL; | |
4774 | partial = (offset + len) & ((1 << blkbits) - 1); | |
4775 | ||
4776 | lblk = start >> blkbits; | |
4777 | max_blocks = (end >> blkbits); | |
4778 | if (max_blocks < lblk) | |
4779 | max_blocks = 0; | |
4780 | else | |
4781 | max_blocks -= lblk; | |
4782 | ||
556615dc | 4783 | flags = EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT | |
b8a86845 LC |
4784 | EXT4_GET_BLOCKS_CONVERT_UNWRITTEN; |
4785 | if (mode & FALLOC_FL_KEEP_SIZE) | |
4786 | flags |= EXT4_GET_BLOCKS_KEEP_SIZE; | |
4787 | ||
4788 | mutex_lock(&inode->i_mutex); | |
4789 | ||
4790 | /* | |
4791 | * Indirect files do not support unwritten extnets | |
4792 | */ | |
4793 | if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { | |
4794 | ret = -EOPNOTSUPP; | |
4795 | goto out_mutex; | |
4796 | } | |
4797 | ||
4798 | if (!(mode & FALLOC_FL_KEEP_SIZE) && | |
4799 | offset + len > i_size_read(inode)) { | |
4800 | new_size = offset + len; | |
4801 | ret = inode_newsize_ok(inode, new_size); | |
4802 | if (ret) | |
4803 | goto out_mutex; | |
4804 | /* | |
4805 | * If we have a partial block after EOF we have to allocate | |
4806 | * the entire block. | |
4807 | */ | |
4808 | if (partial) | |
4809 | max_blocks += 1; | |
4810 | } | |
4811 | ||
4812 | if (max_blocks > 0) { | |
4813 | ||
4814 | /* Now release the pages and zero block aligned part of pages*/ | |
4815 | truncate_pagecache_range(inode, start, end - 1); | |
4816 | ||
4817 | /* Wait all existing dio workers, newcomers will block on i_mutex */ | |
4818 | ext4_inode_block_unlocked_dio(inode); | |
4819 | inode_dio_wait(inode); | |
4820 | ||
4821 | /* | |
4822 | * Remove entire range from the extent status tree. | |
4823 | */ | |
4824 | ret = ext4_es_remove_extent(inode, lblk, max_blocks); | |
4825 | if (ret) | |
4826 | goto out_dio; | |
4827 | ||
4828 | ret = ext4_alloc_file_blocks(file, lblk, max_blocks, flags, | |
4829 | mode); | |
4830 | if (ret) | |
4831 | goto out_dio; | |
4832 | } | |
4833 | ||
4834 | handle = ext4_journal_start(inode, EXT4_HT_MISC, 4); | |
4835 | if (IS_ERR(handle)) { | |
4836 | ret = PTR_ERR(handle); | |
4837 | ext4_std_error(inode->i_sb, ret); | |
4838 | goto out_dio; | |
4839 | } | |
4840 | ||
4841 | inode->i_mtime = inode->i_ctime = ext4_current_time(inode); | |
4842 | ||
e5b30416 | 4843 | if (new_size) { |
b8a86845 LC |
4844 | if (new_size > i_size_read(inode)) |
4845 | i_size_write(inode, new_size); | |
4846 | if (new_size > EXT4_I(inode)->i_disksize) | |
4847 | ext4_update_i_disksize(inode, new_size); | |
e5b30416 | 4848 | } else { |
b8a86845 LC |
4849 | /* |
4850 | * Mark that we allocate beyond EOF so the subsequent truncate | |
4851 | * can proceed even if the new size is the same as i_size. | |
4852 | */ | |
4853 | if ((offset + len) > i_size_read(inode)) | |
4854 | ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS); | |
4855 | } | |
4856 | ||
4857 | ext4_mark_inode_dirty(handle, inode); | |
4858 | ||
4859 | /* Zero out partial block at the edges of the range */ | |
4860 | ret = ext4_zero_partial_blocks(handle, inode, offset, len); | |
4861 | ||
4862 | if (file->f_flags & O_SYNC) | |
4863 | ext4_handle_sync(handle); | |
4864 | ||
4865 | ext4_journal_stop(handle); | |
4866 | out_dio: | |
4867 | ext4_inode_resume_unlocked_dio(inode); | |
4868 | out_mutex: | |
4869 | mutex_unlock(&inode->i_mutex); | |
4870 | return ret; | |
4871 | } | |
4872 | ||
a2df2a63 | 4873 | /* |
2fe17c10 | 4874 | * preallocate space for a file. This implements ext4's fallocate file |
a2df2a63 AA |
4875 | * operation, which gets called from sys_fallocate system call. |
4876 | * For block-mapped files, posix_fallocate should fall back to the method | |
4877 | * of writing zeroes to the required new blocks (the same behavior which is | |
4878 | * expected for file systems which do not support fallocate() system call). | |
4879 | */ | |
2fe17c10 | 4880 | long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) |
a2df2a63 | 4881 | { |
496ad9aa | 4882 | struct inode *inode = file_inode(file); |
a2df2a63 | 4883 | handle_t *handle; |
f282ac19 | 4884 | loff_t new_size = 0; |
498e5f24 | 4885 | unsigned int max_blocks; |
a2df2a63 | 4886 | int ret = 0; |
a4e5d88b | 4887 | int flags; |
0e8b6879 | 4888 | ext4_lblk_t lblk; |
f282ac19 | 4889 | struct timespec tv; |
0e8b6879 | 4890 | unsigned int blkbits = inode->i_blkbits; |
a2df2a63 | 4891 | |
a4bb6b64 | 4892 | /* Return error if mode is not supported */ |
9eb79482 | 4893 | if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | |
b8a86845 | 4894 | FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE)) |
a4bb6b64 AH |
4895 | return -EOPNOTSUPP; |
4896 | ||
4897 | if (mode & FALLOC_FL_PUNCH_HOLE) | |
aeb2817a | 4898 | return ext4_punch_hole(inode, offset, len); |
a4bb6b64 | 4899 | |
0c8d414f TM |
4900 | ret = ext4_convert_inline_data(inode); |
4901 | if (ret) | |
4902 | return ret; | |
4903 | ||
8bad6fc8 ZL |
4904 | /* |
4905 | * currently supporting (pre)allocate mode for extent-based | |
4906 | * files _only_ | |
4907 | */ | |
4908 | if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) | |
4909 | return -EOPNOTSUPP; | |
4910 | ||
40c406c7 TT |
4911 | if (mode & FALLOC_FL_COLLAPSE_RANGE) |
4912 | return ext4_collapse_range(inode, offset, len); | |
4913 | ||
b8a86845 LC |
4914 | if (mode & FALLOC_FL_ZERO_RANGE) |
4915 | return ext4_zero_range(file, offset, len, mode); | |
4916 | ||
0562e0ba | 4917 | trace_ext4_fallocate_enter(inode, offset, len, mode); |
0e8b6879 | 4918 | lblk = offset >> blkbits; |
fd28784a AK |
4919 | /* |
4920 | * We can't just convert len to max_blocks because | |
4921 | * If blocksize = 4096 offset = 3072 and len = 2048 | |
4922 | */ | |
a2df2a63 | 4923 | max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) |
0e8b6879 LC |
4924 | - lblk; |
4925 | ||
556615dc | 4926 | flags = EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT; |
0e8b6879 LC |
4927 | if (mode & FALLOC_FL_KEEP_SIZE) |
4928 | flags |= EXT4_GET_BLOCKS_KEEP_SIZE; | |
4929 | ||
55bd725a | 4930 | mutex_lock(&inode->i_mutex); |
f282ac19 LC |
4931 | |
4932 | if (!(mode & FALLOC_FL_KEEP_SIZE) && | |
4933 | offset + len > i_size_read(inode)) { | |
4934 | new_size = offset + len; | |
4935 | ret = inode_newsize_ok(inode, new_size); | |
4936 | if (ret) | |
4937 | goto out; | |
6d19c42b | 4938 | } |
f282ac19 | 4939 | |
0e8b6879 LC |
4940 | ret = ext4_alloc_file_blocks(file, lblk, max_blocks, flags, mode); |
4941 | if (ret) | |
4942 | goto out; | |
f282ac19 LC |
4943 | |
4944 | handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); | |
4945 | if (IS_ERR(handle)) | |
4946 | goto out; | |
4947 | ||
4948 | tv = inode->i_ctime = ext4_current_time(inode); | |
4949 | ||
e5b30416 | 4950 | if (new_size) { |
f282ac19 LC |
4951 | if (new_size > i_size_read(inode)) { |
4952 | i_size_write(inode, new_size); | |
4953 | inode->i_mtime = tv; | |
4954 | } | |
4955 | if (new_size > EXT4_I(inode)->i_disksize) | |
4956 | ext4_update_i_disksize(inode, new_size); | |
e5b30416 | 4957 | } else { |
f282ac19 LC |
4958 | /* |
4959 | * Mark that we allocate beyond EOF so the subsequent truncate | |
4960 | * can proceed even if the new size is the same as i_size. | |
4961 | */ | |
4962 | if ((offset + len) > i_size_read(inode)) | |
4963 | ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS); | |
4964 | } | |
4965 | ext4_mark_inode_dirty(handle, inode); | |
4966 | if (file->f_flags & O_SYNC) | |
4967 | ext4_handle_sync(handle); | |
4968 | ||
4969 | ext4_journal_stop(handle); | |
4970 | out: | |
55bd725a | 4971 | mutex_unlock(&inode->i_mutex); |
0e8b6879 LC |
4972 | trace_ext4_fallocate_exit(inode, offset, max_blocks, ret); |
4973 | return ret; | |
a2df2a63 | 4974 | } |
6873fa0d | 4975 | |
0031462b MC |
4976 | /* |
4977 | * This function convert a range of blocks to written extents | |
4978 | * The caller of this function will pass the start offset and the size. | |
4979 | * all unwritten extents within this range will be converted to | |
4980 | * written extents. | |
4981 | * | |
4982 | * This function is called from the direct IO end io call back | |
4983 | * function, to convert the fallocated extents after IO is completed. | |
109f5565 | 4984 | * Returns 0 on success. |
0031462b | 4985 | */ |
6b523df4 JK |
4986 | int ext4_convert_unwritten_extents(handle_t *handle, struct inode *inode, |
4987 | loff_t offset, ssize_t len) | |
0031462b | 4988 | { |
0031462b MC |
4989 | unsigned int max_blocks; |
4990 | int ret = 0; | |
4991 | int ret2 = 0; | |
2ed88685 | 4992 | struct ext4_map_blocks map; |
0031462b MC |
4993 | unsigned int credits, blkbits = inode->i_blkbits; |
4994 | ||
2ed88685 | 4995 | map.m_lblk = offset >> blkbits; |
0031462b MC |
4996 | /* |
4997 | * We can't just convert len to max_blocks because | |
4998 | * If blocksize = 4096 offset = 3072 and len = 2048 | |
4999 | */ | |
2ed88685 TT |
5000 | max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) - |
5001 | map.m_lblk); | |
0031462b | 5002 | /* |
6b523df4 JK |
5003 | * This is somewhat ugly but the idea is clear: When transaction is |
5004 | * reserved, everything goes into it. Otherwise we rather start several | |
5005 | * smaller transactions for conversion of each extent separately. | |
0031462b | 5006 | */ |
6b523df4 JK |
5007 | if (handle) { |
5008 | handle = ext4_journal_start_reserved(handle, | |
5009 | EXT4_HT_EXT_CONVERT); | |
5010 | if (IS_ERR(handle)) | |
5011 | return PTR_ERR(handle); | |
5012 | credits = 0; | |
5013 | } else { | |
5014 | /* | |
5015 | * credits to insert 1 extent into extent tree | |
5016 | */ | |
5017 | credits = ext4_chunk_trans_blocks(inode, max_blocks); | |
5018 | } | |
0031462b | 5019 | while (ret >= 0 && ret < max_blocks) { |
2ed88685 TT |
5020 | map.m_lblk += ret; |
5021 | map.m_len = (max_blocks -= ret); | |
6b523df4 JK |
5022 | if (credits) { |
5023 | handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, | |
5024 | credits); | |
5025 | if (IS_ERR(handle)) { | |
5026 | ret = PTR_ERR(handle); | |
5027 | break; | |
5028 | } | |
0031462b | 5029 | } |
2ed88685 | 5030 | ret = ext4_map_blocks(handle, inode, &map, |
c7064ef1 | 5031 | EXT4_GET_BLOCKS_IO_CONVERT_EXT); |
b06acd38 LC |
5032 | if (ret <= 0) |
5033 | ext4_warning(inode->i_sb, | |
5034 | "inode #%lu: block %u: len %u: " | |
5035 | "ext4_ext_map_blocks returned %d", | |
5036 | inode->i_ino, map.m_lblk, | |
5037 | map.m_len, ret); | |
0031462b | 5038 | ext4_mark_inode_dirty(handle, inode); |
6b523df4 JK |
5039 | if (credits) |
5040 | ret2 = ext4_journal_stop(handle); | |
5041 | if (ret <= 0 || ret2) | |
0031462b MC |
5042 | break; |
5043 | } | |
6b523df4 JK |
5044 | if (!credits) |
5045 | ret2 = ext4_journal_stop(handle); | |
0031462b MC |
5046 | return ret > 0 ? ret2 : ret; |
5047 | } | |
6d9c85eb | 5048 | |
6873fa0d | 5049 | /* |
69eb33dc ZL |
5050 | * If newes is not existing extent (newes->ec_pblk equals zero) find |
5051 | * delayed extent at start of newes and update newes accordingly and | |
91dd8c11 LC |
5052 | * return start of the next delayed extent. |
5053 | * | |
69eb33dc | 5054 | * If newes is existing extent (newes->ec_pblk is not equal zero) |
91dd8c11 | 5055 | * return start of next delayed extent or EXT_MAX_BLOCKS if no delayed |
69eb33dc | 5056 | * extent found. Leave newes unmodified. |
6873fa0d | 5057 | */ |
91dd8c11 | 5058 | static int ext4_find_delayed_extent(struct inode *inode, |
69eb33dc | 5059 | struct extent_status *newes) |
6873fa0d | 5060 | { |
b3aff3e3 | 5061 | struct extent_status es; |
be401363 | 5062 | ext4_lblk_t block, next_del; |
6873fa0d | 5063 | |
69eb33dc | 5064 | if (newes->es_pblk == 0) { |
e30b5dca YZ |
5065 | ext4_es_find_delayed_extent_range(inode, newes->es_lblk, |
5066 | newes->es_lblk + newes->es_len - 1, &es); | |
5067 | ||
6d9c85eb | 5068 | /* |
69eb33dc | 5069 | * No extent in extent-tree contains block @newes->es_pblk, |
6d9c85eb | 5070 | * then the block may stay in 1)a hole or 2)delayed-extent. |
6d9c85eb | 5071 | */ |
06b0c886 | 5072 | if (es.es_len == 0) |
b3aff3e3 | 5073 | /* A hole found. */ |
91dd8c11 | 5074 | return 0; |
b3aff3e3 | 5075 | |
69eb33dc | 5076 | if (es.es_lblk > newes->es_lblk) { |
b3aff3e3 | 5077 | /* A hole found. */ |
69eb33dc ZL |
5078 | newes->es_len = min(es.es_lblk - newes->es_lblk, |
5079 | newes->es_len); | |
91dd8c11 | 5080 | return 0; |
6873fa0d | 5081 | } |
6d9c85eb | 5082 | |
69eb33dc | 5083 | newes->es_len = es.es_lblk + es.es_len - newes->es_lblk; |
6873fa0d ES |
5084 | } |
5085 | ||
69eb33dc | 5086 | block = newes->es_lblk + newes->es_len; |
e30b5dca | 5087 | ext4_es_find_delayed_extent_range(inode, block, EXT_MAX_BLOCKS, &es); |
be401363 ZL |
5088 | if (es.es_len == 0) |
5089 | next_del = EXT_MAX_BLOCKS; | |
5090 | else | |
5091 | next_del = es.es_lblk; | |
5092 | ||
91dd8c11 | 5093 | return next_del; |
6873fa0d | 5094 | } |
6873fa0d ES |
5095 | /* fiemap flags we can handle specified here */ |
5096 | #define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) | |
5097 | ||
3a06d778 AK |
5098 | static int ext4_xattr_fiemap(struct inode *inode, |
5099 | struct fiemap_extent_info *fieinfo) | |
6873fa0d ES |
5100 | { |
5101 | __u64 physical = 0; | |
5102 | __u64 length; | |
5103 | __u32 flags = FIEMAP_EXTENT_LAST; | |
5104 | int blockbits = inode->i_sb->s_blocksize_bits; | |
5105 | int error = 0; | |
5106 | ||
5107 | /* in-inode? */ | |
19f5fb7a | 5108 | if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) { |
6873fa0d ES |
5109 | struct ext4_iloc iloc; |
5110 | int offset; /* offset of xattr in inode */ | |
5111 | ||
5112 | error = ext4_get_inode_loc(inode, &iloc); | |
5113 | if (error) | |
5114 | return error; | |
a60697f4 | 5115 | physical = (__u64)iloc.bh->b_blocknr << blockbits; |
6873fa0d ES |
5116 | offset = EXT4_GOOD_OLD_INODE_SIZE + |
5117 | EXT4_I(inode)->i_extra_isize; | |
5118 | physical += offset; | |
5119 | length = EXT4_SB(inode->i_sb)->s_inode_size - offset; | |
5120 | flags |= FIEMAP_EXTENT_DATA_INLINE; | |
fd2dd9fb | 5121 | brelse(iloc.bh); |
6873fa0d | 5122 | } else { /* external block */ |
a60697f4 | 5123 | physical = (__u64)EXT4_I(inode)->i_file_acl << blockbits; |
6873fa0d ES |
5124 | length = inode->i_sb->s_blocksize; |
5125 | } | |
5126 | ||
5127 | if (physical) | |
5128 | error = fiemap_fill_next_extent(fieinfo, 0, physical, | |
5129 | length, flags); | |
5130 | return (error < 0 ? error : 0); | |
5131 | } | |
5132 | ||
5133 | int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, | |
5134 | __u64 start, __u64 len) | |
5135 | { | |
5136 | ext4_lblk_t start_blk; | |
6873fa0d ES |
5137 | int error = 0; |
5138 | ||
94191985 TM |
5139 | if (ext4_has_inline_data(inode)) { |
5140 | int has_inline = 1; | |
5141 | ||
5142 | error = ext4_inline_data_fiemap(inode, fieinfo, &has_inline); | |
5143 | ||
5144 | if (has_inline) | |
5145 | return error; | |
5146 | } | |
5147 | ||
7869a4a6 TT |
5148 | if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) { |
5149 | error = ext4_ext_precache(inode); | |
5150 | if (error) | |
5151 | return error; | |
5152 | } | |
5153 | ||
6873fa0d | 5154 | /* fallback to generic here if not in extents fmt */ |
12e9b892 | 5155 | if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) |
6873fa0d ES |
5156 | return generic_block_fiemap(inode, fieinfo, start, len, |
5157 | ext4_get_block); | |
5158 | ||
5159 | if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS)) | |
5160 | return -EBADR; | |
5161 | ||
5162 | if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) { | |
5163 | error = ext4_xattr_fiemap(inode, fieinfo); | |
5164 | } else { | |
aca92ff6 LM |
5165 | ext4_lblk_t len_blks; |
5166 | __u64 last_blk; | |
5167 | ||
6873fa0d | 5168 | start_blk = start >> inode->i_sb->s_blocksize_bits; |
aca92ff6 | 5169 | last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits; |
f17722f9 LC |
5170 | if (last_blk >= EXT_MAX_BLOCKS) |
5171 | last_blk = EXT_MAX_BLOCKS-1; | |
aca92ff6 | 5172 | len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1; |
6873fa0d ES |
5173 | |
5174 | /* | |
91dd8c11 LC |
5175 | * Walk the extent tree gathering extent information |
5176 | * and pushing extents back to the user. | |
6873fa0d | 5177 | */ |
91dd8c11 LC |
5178 | error = ext4_fill_fiemap_extents(inode, start_blk, |
5179 | len_blks, fieinfo); | |
6873fa0d | 5180 | } |
107a7bd3 | 5181 | ext4_es_lru_add(inode); |
6873fa0d ES |
5182 | return error; |
5183 | } | |
9eb79482 NJ |
5184 | |
5185 | /* | |
5186 | * ext4_access_path: | |
5187 | * Function to access the path buffer for marking it dirty. | |
5188 | * It also checks if there are sufficient credits left in the journal handle | |
5189 | * to update path. | |
5190 | */ | |
5191 | static int | |
5192 | ext4_access_path(handle_t *handle, struct inode *inode, | |
5193 | struct ext4_ext_path *path) | |
5194 | { | |
5195 | int credits, err; | |
5196 | ||
5197 | if (!ext4_handle_valid(handle)) | |
5198 | return 0; | |
5199 | ||
5200 | /* | |
5201 | * Check if need to extend journal credits | |
5202 | * 3 for leaf, sb, and inode plus 2 (bmap and group | |
5203 | * descriptor) for each block group; assume two block | |
5204 | * groups | |
5205 | */ | |
5206 | if (handle->h_buffer_credits < 7) { | |
5207 | credits = ext4_writepage_trans_blocks(inode); | |
5208 | err = ext4_ext_truncate_extend_restart(handle, inode, credits); | |
5209 | /* EAGAIN is success */ | |
5210 | if (err && err != -EAGAIN) | |
5211 | return err; | |
5212 | } | |
5213 | ||
5214 | err = ext4_ext_get_access(handle, inode, path); | |
5215 | return err; | |
5216 | } | |
5217 | ||
5218 | /* | |
5219 | * ext4_ext_shift_path_extents: | |
5220 | * Shift the extents of a path structure lying between path[depth].p_ext | |
5221 | * and EXT_LAST_EXTENT(path[depth].p_hdr) downwards, by subtracting shift | |
5222 | * from starting block for each extent. | |
5223 | */ | |
5224 | static int | |
5225 | ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, | |
5226 | struct inode *inode, handle_t *handle, | |
5227 | ext4_lblk_t *start) | |
5228 | { | |
5229 | int depth, err = 0; | |
5230 | struct ext4_extent *ex_start, *ex_last; | |
5231 | bool update = 0; | |
5232 | depth = path->p_depth; | |
5233 | ||
5234 | while (depth >= 0) { | |
5235 | if (depth == path->p_depth) { | |
5236 | ex_start = path[depth].p_ext; | |
5237 | if (!ex_start) | |
5238 | return -EIO; | |
5239 | ||
5240 | ex_last = EXT_LAST_EXTENT(path[depth].p_hdr); | |
5241 | if (!ex_last) | |
5242 | return -EIO; | |
5243 | ||
5244 | err = ext4_access_path(handle, inode, path + depth); | |
5245 | if (err) | |
5246 | goto out; | |
5247 | ||
5248 | if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr)) | |
5249 | update = 1; | |
5250 | ||
847c6c42 | 5251 | *start = le32_to_cpu(ex_last->ee_block) + |
9eb79482 NJ |
5252 | ext4_ext_get_actual_len(ex_last); |
5253 | ||
5254 | while (ex_start <= ex_last) { | |
847c6c42 | 5255 | le32_add_cpu(&ex_start->ee_block, -shift); |
6dd834ef LC |
5256 | /* Try to merge to the left. */ |
5257 | if ((ex_start > | |
5258 | EXT_FIRST_EXTENT(path[depth].p_hdr)) && | |
5259 | ext4_ext_try_to_merge_right(inode, | |
5260 | path, ex_start - 1)) | |
5261 | ex_last--; | |
5262 | else | |
5263 | ex_start++; | |
9eb79482 NJ |
5264 | } |
5265 | err = ext4_ext_dirty(handle, inode, path + depth); | |
5266 | if (err) | |
5267 | goto out; | |
5268 | ||
5269 | if (--depth < 0 || !update) | |
5270 | break; | |
5271 | } | |
5272 | ||
5273 | /* Update index too */ | |
5274 | err = ext4_access_path(handle, inode, path + depth); | |
5275 | if (err) | |
5276 | goto out; | |
5277 | ||
847c6c42 | 5278 | le32_add_cpu(&path[depth].p_idx->ei_block, -shift); |
9eb79482 NJ |
5279 | err = ext4_ext_dirty(handle, inode, path + depth); |
5280 | if (err) | |
5281 | goto out; | |
5282 | ||
5283 | /* we are done if current index is not a starting index */ | |
5284 | if (path[depth].p_idx != EXT_FIRST_INDEX(path[depth].p_hdr)) | |
5285 | break; | |
5286 | ||
5287 | depth--; | |
5288 | } | |
5289 | ||
5290 | out: | |
5291 | return err; | |
5292 | } | |
5293 | ||
5294 | /* | |
5295 | * ext4_ext_shift_extents: | |
5296 | * All the extents which lies in the range from start to the last allocated | |
5297 | * block for the file are shifted downwards by shift blocks. | |
5298 | * On success, 0 is returned, error otherwise. | |
5299 | */ | |
5300 | static int | |
5301 | ext4_ext_shift_extents(struct inode *inode, handle_t *handle, | |
5302 | ext4_lblk_t start, ext4_lblk_t shift) | |
5303 | { | |
5304 | struct ext4_ext_path *path; | |
5305 | int ret = 0, depth; | |
5306 | struct ext4_extent *extent; | |
5307 | ext4_lblk_t stop_block, current_block; | |
5308 | ext4_lblk_t ex_start, ex_end; | |
5309 | ||
5310 | /* Let path point to the last extent */ | |
5311 | path = ext4_ext_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, 0); | |
5312 | if (IS_ERR(path)) | |
5313 | return PTR_ERR(path); | |
5314 | ||
5315 | depth = path->p_depth; | |
5316 | extent = path[depth].p_ext; | |
5317 | if (!extent) { | |
5318 | ext4_ext_drop_refs(path); | |
5319 | kfree(path); | |
5320 | return ret; | |
5321 | } | |
5322 | ||
847c6c42 ZL |
5323 | stop_block = le32_to_cpu(extent->ee_block) + |
5324 | ext4_ext_get_actual_len(extent); | |
9eb79482 NJ |
5325 | ext4_ext_drop_refs(path); |
5326 | kfree(path); | |
5327 | ||
5328 | /* Nothing to shift, if hole is at the end of file */ | |
5329 | if (start >= stop_block) | |
5330 | return ret; | |
5331 | ||
5332 | /* | |
5333 | * Don't start shifting extents until we make sure the hole is big | |
5334 | * enough to accomodate the shift. | |
5335 | */ | |
5336 | path = ext4_ext_find_extent(inode, start - 1, NULL, 0); | |
8dc79ec4 DM |
5337 | if (IS_ERR(path)) |
5338 | return PTR_ERR(path); | |
9eb79482 NJ |
5339 | depth = path->p_depth; |
5340 | extent = path[depth].p_ext; | |
8dc79ec4 DM |
5341 | if (extent) { |
5342 | ex_start = le32_to_cpu(extent->ee_block); | |
5343 | ex_end = le32_to_cpu(extent->ee_block) + | |
847c6c42 | 5344 | ext4_ext_get_actual_len(extent); |
8dc79ec4 DM |
5345 | } else { |
5346 | ex_start = 0; | |
5347 | ex_end = 0; | |
5348 | } | |
9eb79482 NJ |
5349 | ext4_ext_drop_refs(path); |
5350 | kfree(path); | |
5351 | ||
5352 | if ((start == ex_start && shift > ex_start) || | |
5353 | (shift > start - ex_end)) | |
5354 | return -EINVAL; | |
5355 | ||
5356 | /* Its safe to start updating extents */ | |
5357 | while (start < stop_block) { | |
5358 | path = ext4_ext_find_extent(inode, start, NULL, 0); | |
5359 | if (IS_ERR(path)) | |
5360 | return PTR_ERR(path); | |
5361 | depth = path->p_depth; | |
5362 | extent = path[depth].p_ext; | |
a18ed359 DM |
5363 | if (!extent) { |
5364 | EXT4_ERROR_INODE(inode, "unexpected hole at %lu", | |
5365 | (unsigned long) start); | |
5366 | return -EIO; | |
5367 | } | |
5368 | ||
847c6c42 | 5369 | current_block = le32_to_cpu(extent->ee_block); |
9eb79482 NJ |
5370 | if (start > current_block) { |
5371 | /* Hole, move to the next extent */ | |
5372 | ret = mext_next_extent(inode, path, &extent); | |
5373 | if (ret != 0) { | |
5374 | ext4_ext_drop_refs(path); | |
5375 | kfree(path); | |
5376 | if (ret == 1) | |
5377 | ret = 0; | |
5378 | break; | |
5379 | } | |
5380 | } | |
5381 | ret = ext4_ext_shift_path_extents(path, shift, inode, | |
5382 | handle, &start); | |
5383 | ext4_ext_drop_refs(path); | |
5384 | kfree(path); | |
5385 | if (ret) | |
5386 | break; | |
5387 | } | |
5388 | ||
5389 | return ret; | |
5390 | } | |
5391 | ||
5392 | /* | |
5393 | * ext4_collapse_range: | |
5394 | * This implements the fallocate's collapse range functionality for ext4 | |
5395 | * Returns: 0 and non-zero on error. | |
5396 | */ | |
5397 | int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) | |
5398 | { | |
5399 | struct super_block *sb = inode->i_sb; | |
5400 | ext4_lblk_t punch_start, punch_stop; | |
5401 | handle_t *handle; | |
5402 | unsigned int credits; | |
a8680e0d | 5403 | loff_t new_size, ioffset; |
9eb79482 NJ |
5404 | int ret; |
5405 | ||
9eb79482 NJ |
5406 | /* Collapse range works only on fs block size aligned offsets. */ |
5407 | if (offset & (EXT4_BLOCK_SIZE(sb) - 1) || | |
5408 | len & (EXT4_BLOCK_SIZE(sb) - 1)) | |
5409 | return -EINVAL; | |
5410 | ||
5411 | if (!S_ISREG(inode->i_mode)) | |
86f1ca38 | 5412 | return -EINVAL; |
9eb79482 | 5413 | |
0a04b248 NJ |
5414 | if (EXT4_SB(inode->i_sb)->s_cluster_ratio > 1) |
5415 | return -EOPNOTSUPP; | |
5416 | ||
9eb79482 NJ |
5417 | trace_ext4_collapse_range(inode, offset, len); |
5418 | ||
5419 | punch_start = offset >> EXT4_BLOCK_SIZE_BITS(sb); | |
5420 | punch_stop = (offset + len) >> EXT4_BLOCK_SIZE_BITS(sb); | |
5421 | ||
1ce01c4a NJ |
5422 | /* Call ext4_force_commit to flush all data in case of data=journal. */ |
5423 | if (ext4_should_journal_data(inode)) { | |
5424 | ret = ext4_force_commit(inode->i_sb); | |
5425 | if (ret) | |
5426 | return ret; | |
5427 | } | |
5428 | ||
a8680e0d NJ |
5429 | /* |
5430 | * Need to round down offset to be aligned with page size boundary | |
5431 | * for page size > block size. | |
5432 | */ | |
5433 | ioffset = round_down(offset, PAGE_SIZE); | |
5434 | ||
9eb79482 | 5435 | /* Write out all dirty pages */ |
a8680e0d NJ |
5436 | ret = filemap_write_and_wait_range(inode->i_mapping, ioffset, |
5437 | LLONG_MAX); | |
9eb79482 NJ |
5438 | if (ret) |
5439 | return ret; | |
5440 | ||
5441 | /* Take mutex lock */ | |
5442 | mutex_lock(&inode->i_mutex); | |
5443 | ||
23fffa92 LC |
5444 | /* |
5445 | * There is no need to overlap collapse range with EOF, in which case | |
5446 | * it is effectively a truncate operation | |
5447 | */ | |
5448 | if (offset + len >= i_size_read(inode)) { | |
5449 | ret = -EINVAL; | |
5450 | goto out_mutex; | |
5451 | } | |
5452 | ||
9eb79482 NJ |
5453 | /* Currently just for extent based files */ |
5454 | if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { | |
5455 | ret = -EOPNOTSUPP; | |
5456 | goto out_mutex; | |
5457 | } | |
5458 | ||
a8680e0d | 5459 | truncate_pagecache(inode, ioffset); |
9eb79482 NJ |
5460 | |
5461 | /* Wait for existing dio to complete */ | |
5462 | ext4_inode_block_unlocked_dio(inode); | |
5463 | inode_dio_wait(inode); | |
5464 | ||
5465 | credits = ext4_writepage_trans_blocks(inode); | |
5466 | handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); | |
5467 | if (IS_ERR(handle)) { | |
5468 | ret = PTR_ERR(handle); | |
5469 | goto out_dio; | |
5470 | } | |
5471 | ||
5472 | down_write(&EXT4_I(inode)->i_data_sem); | |
5473 | ext4_discard_preallocations(inode); | |
5474 | ||
5475 | ret = ext4_es_remove_extent(inode, punch_start, | |
2c1d2328 | 5476 | EXT_MAX_BLOCKS - punch_start); |
9eb79482 NJ |
5477 | if (ret) { |
5478 | up_write(&EXT4_I(inode)->i_data_sem); | |
5479 | goto out_stop; | |
5480 | } | |
5481 | ||
5482 | ret = ext4_ext_remove_space(inode, punch_start, punch_stop - 1); | |
5483 | if (ret) { | |
5484 | up_write(&EXT4_I(inode)->i_data_sem); | |
5485 | goto out_stop; | |
5486 | } | |
ef24f6c2 | 5487 | ext4_discard_preallocations(inode); |
9eb79482 NJ |
5488 | |
5489 | ret = ext4_ext_shift_extents(inode, handle, punch_stop, | |
5490 | punch_stop - punch_start); | |
5491 | if (ret) { | |
5492 | up_write(&EXT4_I(inode)->i_data_sem); | |
5493 | goto out_stop; | |
5494 | } | |
5495 | ||
5496 | new_size = i_size_read(inode) - len; | |
9337d5d3 | 5497 | i_size_write(inode, new_size); |
9eb79482 NJ |
5498 | EXT4_I(inode)->i_disksize = new_size; |
5499 | ||
9eb79482 NJ |
5500 | up_write(&EXT4_I(inode)->i_data_sem); |
5501 | if (IS_SYNC(inode)) | |
5502 | ext4_handle_sync(handle); | |
5503 | inode->i_mtime = inode->i_ctime = ext4_current_time(inode); | |
5504 | ext4_mark_inode_dirty(handle, inode); | |
5505 | ||
5506 | out_stop: | |
5507 | ext4_journal_stop(handle); | |
5508 | out_dio: | |
5509 | ext4_inode_resume_unlocked_dio(inode); | |
5510 | out_mutex: | |
5511 | mutex_unlock(&inode->i_mutex); | |
5512 | return ret; | |
5513 | } |