]>
Commit | Line | Data |
---|---|---|
c1d7c514 | 1 | // SPDX-License-Identifier: GPL-2.0 |
a5ed9182 OS |
2 | /* |
3 | * Copyright (C) 2015 Facebook. All rights reserved. | |
a5ed9182 OS |
4 | */ |
5 | ||
6 | #include <linux/kernel.h> | |
25ff17e8 | 7 | #include <linux/sched/mm.h> |
a5ed9182 OS |
8 | #include "ctree.h" |
9 | #include "disk-io.h" | |
10 | #include "locking.h" | |
11 | #include "free-space-tree.h" | |
12 | #include "transaction.h" | |
13 | ||
14 | static int __add_block_group_free_space(struct btrfs_trans_handle *trans, | |
15 | struct btrfs_fs_info *fs_info, | |
16 | struct btrfs_block_group_cache *block_group, | |
17 | struct btrfs_path *path); | |
18 | ||
19 | void set_free_space_tree_thresholds(struct btrfs_block_group_cache *cache) | |
20 | { | |
21 | u32 bitmap_range; | |
22 | size_t bitmap_size; | |
23 | u64 num_bitmaps, total_bitmap_size; | |
24 | ||
25 | /* | |
26 | * We convert to bitmaps when the disk space required for using extents | |
27 | * exceeds that required for using bitmaps. | |
28 | */ | |
da17066c | 29 | bitmap_range = cache->fs_info->sectorsize * BTRFS_FREE_SPACE_BITMAP_BITS; |
a5ed9182 OS |
30 | num_bitmaps = div_u64(cache->key.offset + bitmap_range - 1, |
31 | bitmap_range); | |
32 | bitmap_size = sizeof(struct btrfs_item) + BTRFS_FREE_SPACE_BITMAP_SIZE; | |
33 | total_bitmap_size = num_bitmaps * bitmap_size; | |
34 | cache->bitmap_high_thresh = div_u64(total_bitmap_size, | |
35 | sizeof(struct btrfs_item)); | |
36 | ||
37 | /* | |
38 | * We allow for a small buffer between the high threshold and low | |
39 | * threshold to avoid thrashing back and forth between the two formats. | |
40 | */ | |
41 | if (cache->bitmap_high_thresh > 100) | |
42 | cache->bitmap_low_thresh = cache->bitmap_high_thresh - 100; | |
43 | else | |
44 | cache->bitmap_low_thresh = 0; | |
45 | } | |
46 | ||
47 | static int add_new_free_space_info(struct btrfs_trans_handle *trans, | |
48 | struct btrfs_fs_info *fs_info, | |
49 | struct btrfs_block_group_cache *block_group, | |
50 | struct btrfs_path *path) | |
51 | { | |
52 | struct btrfs_root *root = fs_info->free_space_root; | |
53 | struct btrfs_free_space_info *info; | |
54 | struct btrfs_key key; | |
55 | struct extent_buffer *leaf; | |
56 | int ret; | |
57 | ||
58 | key.objectid = block_group->key.objectid; | |
59 | key.type = BTRFS_FREE_SPACE_INFO_KEY; | |
60 | key.offset = block_group->key.offset; | |
61 | ||
62 | ret = btrfs_insert_empty_item(trans, root, path, &key, sizeof(*info)); | |
63 | if (ret) | |
64 | goto out; | |
65 | ||
66 | leaf = path->nodes[0]; | |
67 | info = btrfs_item_ptr(leaf, path->slots[0], | |
68 | struct btrfs_free_space_info); | |
69 | btrfs_set_free_space_extent_count(leaf, info, 0); | |
70 | btrfs_set_free_space_flags(leaf, info, 0); | |
71 | btrfs_mark_buffer_dirty(leaf); | |
72 | ||
73 | ret = 0; | |
74 | out: | |
75 | btrfs_release_path(path); | |
76 | return ret; | |
77 | } | |
78 | ||
79 | struct btrfs_free_space_info * | |
80 | search_free_space_info(struct btrfs_trans_handle *trans, | |
81 | struct btrfs_fs_info *fs_info, | |
82 | struct btrfs_block_group_cache *block_group, | |
83 | struct btrfs_path *path, int cow) | |
84 | { | |
85 | struct btrfs_root *root = fs_info->free_space_root; | |
86 | struct btrfs_key key; | |
87 | int ret; | |
88 | ||
89 | key.objectid = block_group->key.objectid; | |
90 | key.type = BTRFS_FREE_SPACE_INFO_KEY; | |
91 | key.offset = block_group->key.offset; | |
92 | ||
93 | ret = btrfs_search_slot(trans, root, &key, path, 0, cow); | |
94 | if (ret < 0) | |
95 | return ERR_PTR(ret); | |
96 | if (ret != 0) { | |
ab8d0fc4 | 97 | btrfs_warn(fs_info, "missing free space info for %llu", |
a5ed9182 OS |
98 | block_group->key.objectid); |
99 | ASSERT(0); | |
100 | return ERR_PTR(-ENOENT); | |
101 | } | |
102 | ||
103 | return btrfs_item_ptr(path->nodes[0], path->slots[0], | |
104 | struct btrfs_free_space_info); | |
105 | } | |
106 | ||
107 | /* | |
108 | * btrfs_search_slot() but we're looking for the greatest key less than the | |
109 | * passed key. | |
110 | */ | |
111 | static int btrfs_search_prev_slot(struct btrfs_trans_handle *trans, | |
112 | struct btrfs_root *root, | |
113 | struct btrfs_key *key, struct btrfs_path *p, | |
114 | int ins_len, int cow) | |
115 | { | |
116 | int ret; | |
117 | ||
118 | ret = btrfs_search_slot(trans, root, key, p, ins_len, cow); | |
119 | if (ret < 0) | |
120 | return ret; | |
121 | ||
122 | if (ret == 0) { | |
123 | ASSERT(0); | |
124 | return -EIO; | |
125 | } | |
126 | ||
127 | if (p->slots[0] == 0) { | |
128 | ASSERT(0); | |
129 | return -EIO; | |
130 | } | |
131 | p->slots[0]--; | |
132 | ||
133 | return 0; | |
134 | } | |
135 | ||
136 | static inline u32 free_space_bitmap_size(u64 size, u32 sectorsize) | |
137 | { | |
138 | return DIV_ROUND_UP((u32)div_u64(size, sectorsize), BITS_PER_BYTE); | |
139 | } | |
140 | ||
a565971f | 141 | static unsigned long *alloc_bitmap(u32 bitmap_size) |
a5ed9182 | 142 | { |
a565971f | 143 | unsigned long *ret; |
25ff17e8 | 144 | unsigned int nofs_flag; |
a565971f | 145 | u32 bitmap_rounded_size = round_up(bitmap_size, sizeof(unsigned long)); |
79b134a2 DS |
146 | |
147 | /* | |
25ff17e8 OS |
148 | * GFP_NOFS doesn't work with kvmalloc(), but we really can't recurse |
149 | * into the filesystem as the free space bitmap can be modified in the | |
150 | * critical section of a transaction commit. | |
151 | * | |
152 | * TODO: push the memalloc_nofs_{save,restore}() to the caller where we | |
153 | * know that recursion is unsafe. | |
79b134a2 | 154 | */ |
25ff17e8 | 155 | nofs_flag = memalloc_nofs_save(); |
a565971f | 156 | ret = kvzalloc(bitmap_rounded_size, GFP_KERNEL); |
25ff17e8 OS |
157 | memalloc_nofs_restore(nofs_flag); |
158 | return ret; | |
a5ed9182 OS |
159 | } |
160 | ||
a565971f | 161 | static void le_bitmap_set(unsigned long *map, unsigned int start, int len) |
6faa8f47 | 162 | { |
a565971f | 163 | u8 *p = ((u8 *)map) + BIT_BYTE(start); |
6faa8f47 HM |
164 | const unsigned int size = start + len; |
165 | int bits_to_set = BITS_PER_BYTE - (start % BITS_PER_BYTE); | |
166 | u8 mask_to_set = BITMAP_FIRST_BYTE_MASK(start); | |
167 | ||
168 | while (len - bits_to_set >= 0) { | |
169 | *p |= mask_to_set; | |
170 | len -= bits_to_set; | |
171 | bits_to_set = BITS_PER_BYTE; | |
172 | mask_to_set = ~0; | |
173 | p++; | |
174 | } | |
175 | if (len) { | |
176 | mask_to_set &= BITMAP_LAST_BYTE_MASK(size); | |
177 | *p |= mask_to_set; | |
178 | } | |
179 | } | |
180 | ||
a5ed9182 OS |
181 | int convert_free_space_to_bitmaps(struct btrfs_trans_handle *trans, |
182 | struct btrfs_fs_info *fs_info, | |
183 | struct btrfs_block_group_cache *block_group, | |
184 | struct btrfs_path *path) | |
185 | { | |
186 | struct btrfs_root *root = fs_info->free_space_root; | |
187 | struct btrfs_free_space_info *info; | |
188 | struct btrfs_key key, found_key; | |
189 | struct extent_buffer *leaf; | |
a565971f HM |
190 | unsigned long *bitmap; |
191 | char *bitmap_cursor; | |
a5ed9182 OS |
192 | u64 start, end; |
193 | u64 bitmap_range, i; | |
194 | u32 bitmap_size, flags, expected_extent_count; | |
195 | u32 extent_count = 0; | |
196 | int done = 0, nr; | |
197 | int ret; | |
198 | ||
199 | bitmap_size = free_space_bitmap_size(block_group->key.offset, | |
0b246afa | 200 | fs_info->sectorsize); |
a5ed9182 OS |
201 | bitmap = alloc_bitmap(bitmap_size); |
202 | if (!bitmap) { | |
203 | ret = -ENOMEM; | |
204 | goto out; | |
205 | } | |
206 | ||
207 | start = block_group->key.objectid; | |
208 | end = block_group->key.objectid + block_group->key.offset; | |
209 | ||
210 | key.objectid = end - 1; | |
211 | key.type = (u8)-1; | |
212 | key.offset = (u64)-1; | |
213 | ||
214 | while (!done) { | |
215 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
216 | if (ret) | |
217 | goto out; | |
218 | ||
219 | leaf = path->nodes[0]; | |
220 | nr = 0; | |
221 | path->slots[0]++; | |
222 | while (path->slots[0] > 0) { | |
223 | btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1); | |
224 | ||
225 | if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) { | |
226 | ASSERT(found_key.objectid == block_group->key.objectid); | |
227 | ASSERT(found_key.offset == block_group->key.offset); | |
228 | done = 1; | |
229 | break; | |
230 | } else if (found_key.type == BTRFS_FREE_SPACE_EXTENT_KEY) { | |
231 | u64 first, last; | |
232 | ||
233 | ASSERT(found_key.objectid >= start); | |
234 | ASSERT(found_key.objectid < end); | |
235 | ASSERT(found_key.objectid + found_key.offset <= end); | |
236 | ||
237 | first = div_u64(found_key.objectid - start, | |
0b246afa | 238 | fs_info->sectorsize); |
a5ed9182 | 239 | last = div_u64(found_key.objectid + found_key.offset - start, |
0b246afa | 240 | fs_info->sectorsize); |
2fe1d551 | 241 | le_bitmap_set(bitmap, first, last - first); |
a5ed9182 OS |
242 | |
243 | extent_count++; | |
244 | nr++; | |
245 | path->slots[0]--; | |
246 | } else { | |
247 | ASSERT(0); | |
248 | } | |
249 | } | |
250 | ||
251 | ret = btrfs_del_items(trans, root, path, path->slots[0], nr); | |
252 | if (ret) | |
253 | goto out; | |
254 | btrfs_release_path(path); | |
255 | } | |
256 | ||
257 | info = search_free_space_info(trans, fs_info, block_group, path, 1); | |
258 | if (IS_ERR(info)) { | |
259 | ret = PTR_ERR(info); | |
260 | goto out; | |
261 | } | |
262 | leaf = path->nodes[0]; | |
263 | flags = btrfs_free_space_flags(leaf, info); | |
264 | flags |= BTRFS_FREE_SPACE_USING_BITMAPS; | |
265 | btrfs_set_free_space_flags(leaf, info, flags); | |
266 | expected_extent_count = btrfs_free_space_extent_count(leaf, info); | |
267 | btrfs_mark_buffer_dirty(leaf); | |
268 | btrfs_release_path(path); | |
269 | ||
270 | if (extent_count != expected_extent_count) { | |
5d163e0e JM |
271 | btrfs_err(fs_info, |
272 | "incorrect extent count for %llu; counted %u, expected %u", | |
a5ed9182 OS |
273 | block_group->key.objectid, extent_count, |
274 | expected_extent_count); | |
275 | ASSERT(0); | |
276 | ret = -EIO; | |
277 | goto out; | |
278 | } | |
279 | ||
a565971f | 280 | bitmap_cursor = (char *)bitmap; |
0b246afa | 281 | bitmap_range = fs_info->sectorsize * BTRFS_FREE_SPACE_BITMAP_BITS; |
a5ed9182 OS |
282 | i = start; |
283 | while (i < end) { | |
284 | unsigned long ptr; | |
285 | u64 extent_size; | |
286 | u32 data_size; | |
287 | ||
288 | extent_size = min(end - i, bitmap_range); | |
289 | data_size = free_space_bitmap_size(extent_size, | |
0b246afa | 290 | fs_info->sectorsize); |
a5ed9182 OS |
291 | |
292 | key.objectid = i; | |
293 | key.type = BTRFS_FREE_SPACE_BITMAP_KEY; | |
294 | key.offset = extent_size; | |
295 | ||
296 | ret = btrfs_insert_empty_item(trans, root, path, &key, | |
297 | data_size); | |
298 | if (ret) | |
299 | goto out; | |
300 | ||
301 | leaf = path->nodes[0]; | |
302 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); | |
303 | write_extent_buffer(leaf, bitmap_cursor, ptr, | |
304 | data_size); | |
305 | btrfs_mark_buffer_dirty(leaf); | |
306 | btrfs_release_path(path); | |
307 | ||
308 | i += extent_size; | |
309 | bitmap_cursor += data_size; | |
310 | } | |
311 | ||
312 | ret = 0; | |
313 | out: | |
79b134a2 | 314 | kvfree(bitmap); |
a5ed9182 | 315 | if (ret) |
66642832 | 316 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
317 | return ret; |
318 | } | |
319 | ||
320 | int convert_free_space_to_extents(struct btrfs_trans_handle *trans, | |
321 | struct btrfs_fs_info *fs_info, | |
322 | struct btrfs_block_group_cache *block_group, | |
323 | struct btrfs_path *path) | |
324 | { | |
325 | struct btrfs_root *root = fs_info->free_space_root; | |
326 | struct btrfs_free_space_info *info; | |
327 | struct btrfs_key key, found_key; | |
328 | struct extent_buffer *leaf; | |
a565971f | 329 | unsigned long *bitmap; |
a5ed9182 | 330 | u64 start, end; |
a5ed9182 | 331 | u32 bitmap_size, flags, expected_extent_count; |
a565971f | 332 | unsigned long nrbits, start_bit, end_bit; |
a5ed9182 OS |
333 | u32 extent_count = 0; |
334 | int done = 0, nr; | |
335 | int ret; | |
336 | ||
337 | bitmap_size = free_space_bitmap_size(block_group->key.offset, | |
0b246afa | 338 | fs_info->sectorsize); |
a5ed9182 OS |
339 | bitmap = alloc_bitmap(bitmap_size); |
340 | if (!bitmap) { | |
341 | ret = -ENOMEM; | |
342 | goto out; | |
343 | } | |
344 | ||
345 | start = block_group->key.objectid; | |
346 | end = block_group->key.objectid + block_group->key.offset; | |
347 | ||
348 | key.objectid = end - 1; | |
349 | key.type = (u8)-1; | |
350 | key.offset = (u64)-1; | |
351 | ||
352 | while (!done) { | |
353 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
354 | if (ret) | |
355 | goto out; | |
356 | ||
357 | leaf = path->nodes[0]; | |
358 | nr = 0; | |
359 | path->slots[0]++; | |
360 | while (path->slots[0] > 0) { | |
361 | btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1); | |
362 | ||
363 | if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) { | |
364 | ASSERT(found_key.objectid == block_group->key.objectid); | |
365 | ASSERT(found_key.offset == block_group->key.offset); | |
366 | done = 1; | |
367 | break; | |
368 | } else if (found_key.type == BTRFS_FREE_SPACE_BITMAP_KEY) { | |
369 | unsigned long ptr; | |
a565971f | 370 | char *bitmap_cursor; |
a5ed9182 OS |
371 | u32 bitmap_pos, data_size; |
372 | ||
373 | ASSERT(found_key.objectid >= start); | |
374 | ASSERT(found_key.objectid < end); | |
375 | ASSERT(found_key.objectid + found_key.offset <= end); | |
376 | ||
377 | bitmap_pos = div_u64(found_key.objectid - start, | |
0b246afa | 378 | fs_info->sectorsize * |
a5ed9182 | 379 | BITS_PER_BYTE); |
a565971f | 380 | bitmap_cursor = ((char *)bitmap) + bitmap_pos; |
a5ed9182 | 381 | data_size = free_space_bitmap_size(found_key.offset, |
0b246afa | 382 | fs_info->sectorsize); |
a5ed9182 OS |
383 | |
384 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0] - 1); | |
385 | read_extent_buffer(leaf, bitmap_cursor, ptr, | |
386 | data_size); | |
387 | ||
388 | nr++; | |
389 | path->slots[0]--; | |
390 | } else { | |
391 | ASSERT(0); | |
392 | } | |
393 | } | |
394 | ||
395 | ret = btrfs_del_items(trans, root, path, path->slots[0], nr); | |
396 | if (ret) | |
397 | goto out; | |
398 | btrfs_release_path(path); | |
399 | } | |
400 | ||
401 | info = search_free_space_info(trans, fs_info, block_group, path, 1); | |
402 | if (IS_ERR(info)) { | |
403 | ret = PTR_ERR(info); | |
404 | goto out; | |
405 | } | |
406 | leaf = path->nodes[0]; | |
407 | flags = btrfs_free_space_flags(leaf, info); | |
408 | flags &= ~BTRFS_FREE_SPACE_USING_BITMAPS; | |
409 | btrfs_set_free_space_flags(leaf, info, flags); | |
410 | expected_extent_count = btrfs_free_space_extent_count(leaf, info); | |
411 | btrfs_mark_buffer_dirty(leaf); | |
412 | btrfs_release_path(path); | |
413 | ||
a565971f HM |
414 | nrbits = div_u64(block_group->key.offset, block_group->fs_info->sectorsize); |
415 | start_bit = find_next_bit_le(bitmap, nrbits, 0); | |
416 | ||
417 | while (start_bit < nrbits) { | |
418 | end_bit = find_next_zero_bit_le(bitmap, nrbits, start_bit); | |
419 | ASSERT(start_bit < end_bit); | |
420 | ||
421 | key.objectid = start + start_bit * block_group->fs_info->sectorsize; | |
a5ed9182 | 422 | key.type = BTRFS_FREE_SPACE_EXTENT_KEY; |
a565971f | 423 | key.offset = (end_bit - start_bit) * block_group->fs_info->sectorsize; |
a5ed9182 OS |
424 | |
425 | ret = btrfs_insert_empty_item(trans, root, path, &key, 0); | |
426 | if (ret) | |
427 | goto out; | |
428 | btrfs_release_path(path); | |
429 | ||
430 | extent_count++; | |
a565971f HM |
431 | |
432 | start_bit = find_next_bit_le(bitmap, nrbits, end_bit); | |
a5ed9182 OS |
433 | } |
434 | ||
435 | if (extent_count != expected_extent_count) { | |
5d163e0e JM |
436 | btrfs_err(fs_info, |
437 | "incorrect extent count for %llu; counted %u, expected %u", | |
a5ed9182 OS |
438 | block_group->key.objectid, extent_count, |
439 | expected_extent_count); | |
440 | ASSERT(0); | |
441 | ret = -EIO; | |
442 | goto out; | |
443 | } | |
444 | ||
445 | ret = 0; | |
446 | out: | |
79b134a2 | 447 | kvfree(bitmap); |
a5ed9182 | 448 | if (ret) |
66642832 | 449 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
450 | return ret; |
451 | } | |
452 | ||
453 | static int update_free_space_extent_count(struct btrfs_trans_handle *trans, | |
454 | struct btrfs_fs_info *fs_info, | |
455 | struct btrfs_block_group_cache *block_group, | |
456 | struct btrfs_path *path, | |
457 | int new_extents) | |
458 | { | |
459 | struct btrfs_free_space_info *info; | |
460 | u32 flags; | |
461 | u32 extent_count; | |
462 | int ret = 0; | |
463 | ||
464 | if (new_extents == 0) | |
465 | return 0; | |
466 | ||
467 | info = search_free_space_info(trans, fs_info, block_group, path, 1); | |
468 | if (IS_ERR(info)) { | |
469 | ret = PTR_ERR(info); | |
470 | goto out; | |
471 | } | |
472 | flags = btrfs_free_space_flags(path->nodes[0], info); | |
473 | extent_count = btrfs_free_space_extent_count(path->nodes[0], info); | |
474 | ||
475 | extent_count += new_extents; | |
476 | btrfs_set_free_space_extent_count(path->nodes[0], info, extent_count); | |
477 | btrfs_mark_buffer_dirty(path->nodes[0]); | |
478 | btrfs_release_path(path); | |
479 | ||
480 | if (!(flags & BTRFS_FREE_SPACE_USING_BITMAPS) && | |
481 | extent_count > block_group->bitmap_high_thresh) { | |
482 | ret = convert_free_space_to_bitmaps(trans, fs_info, block_group, | |
483 | path); | |
484 | } else if ((flags & BTRFS_FREE_SPACE_USING_BITMAPS) && | |
485 | extent_count < block_group->bitmap_low_thresh) { | |
486 | ret = convert_free_space_to_extents(trans, fs_info, block_group, | |
487 | path); | |
488 | } | |
489 | ||
490 | out: | |
491 | return ret; | |
492 | } | |
493 | ||
494 | int free_space_test_bit(struct btrfs_block_group_cache *block_group, | |
495 | struct btrfs_path *path, u64 offset) | |
496 | { | |
497 | struct extent_buffer *leaf; | |
498 | struct btrfs_key key; | |
499 | u64 found_start, found_end; | |
500 | unsigned long ptr, i; | |
501 | ||
502 | leaf = path->nodes[0]; | |
503 | btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); | |
504 | ASSERT(key.type == BTRFS_FREE_SPACE_BITMAP_KEY); | |
505 | ||
506 | found_start = key.objectid; | |
507 | found_end = key.objectid + key.offset; | |
508 | ASSERT(offset >= found_start && offset < found_end); | |
509 | ||
510 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); | |
da17066c JM |
511 | i = div_u64(offset - found_start, |
512 | block_group->fs_info->sectorsize); | |
a5ed9182 OS |
513 | return !!extent_buffer_test_bit(leaf, ptr, i); |
514 | } | |
515 | ||
516 | static void free_space_set_bits(struct btrfs_block_group_cache *block_group, | |
517 | struct btrfs_path *path, u64 *start, u64 *size, | |
518 | int bit) | |
519 | { | |
0b246afa | 520 | struct btrfs_fs_info *fs_info = block_group->fs_info; |
a5ed9182 OS |
521 | struct extent_buffer *leaf; |
522 | struct btrfs_key key; | |
523 | u64 end = *start + *size; | |
524 | u64 found_start, found_end; | |
525 | unsigned long ptr, first, last; | |
526 | ||
527 | leaf = path->nodes[0]; | |
528 | btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); | |
529 | ASSERT(key.type == BTRFS_FREE_SPACE_BITMAP_KEY); | |
530 | ||
531 | found_start = key.objectid; | |
532 | found_end = key.objectid + key.offset; | |
533 | ASSERT(*start >= found_start && *start < found_end); | |
534 | ASSERT(end > found_start); | |
535 | ||
536 | if (end > found_end) | |
537 | end = found_end; | |
538 | ||
539 | ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); | |
0b246afa JM |
540 | first = div_u64(*start - found_start, fs_info->sectorsize); |
541 | last = div_u64(end - found_start, fs_info->sectorsize); | |
a5ed9182 OS |
542 | if (bit) |
543 | extent_buffer_bitmap_set(leaf, ptr, first, last - first); | |
544 | else | |
545 | extent_buffer_bitmap_clear(leaf, ptr, first, last - first); | |
546 | btrfs_mark_buffer_dirty(leaf); | |
547 | ||
548 | *size -= end - *start; | |
549 | *start = end; | |
550 | } | |
551 | ||
552 | /* | |
553 | * We can't use btrfs_next_item() in modify_free_space_bitmap() because | |
554 | * btrfs_next_leaf() doesn't get the path for writing. We can forgo the fancy | |
555 | * tree walking in btrfs_next_leaf() anyways because we know exactly what we're | |
556 | * looking for. | |
557 | */ | |
558 | static int free_space_next_bitmap(struct btrfs_trans_handle *trans, | |
559 | struct btrfs_root *root, struct btrfs_path *p) | |
560 | { | |
561 | struct btrfs_key key; | |
562 | ||
563 | if (p->slots[0] + 1 < btrfs_header_nritems(p->nodes[0])) { | |
564 | p->slots[0]++; | |
565 | return 0; | |
566 | } | |
567 | ||
568 | btrfs_item_key_to_cpu(p->nodes[0], &key, p->slots[0]); | |
569 | btrfs_release_path(p); | |
570 | ||
571 | key.objectid += key.offset; | |
572 | key.type = (u8)-1; | |
573 | key.offset = (u64)-1; | |
574 | ||
575 | return btrfs_search_prev_slot(trans, root, &key, p, 0, 1); | |
576 | } | |
577 | ||
578 | /* | |
579 | * If remove is 1, then we are removing free space, thus clearing bits in the | |
580 | * bitmap. If remove is 0, then we are adding free space, thus setting bits in | |
581 | * the bitmap. | |
582 | */ | |
583 | static int modify_free_space_bitmap(struct btrfs_trans_handle *trans, | |
584 | struct btrfs_fs_info *fs_info, | |
585 | struct btrfs_block_group_cache *block_group, | |
586 | struct btrfs_path *path, | |
587 | u64 start, u64 size, int remove) | |
588 | { | |
589 | struct btrfs_root *root = fs_info->free_space_root; | |
590 | struct btrfs_key key; | |
591 | u64 end = start + size; | |
592 | u64 cur_start, cur_size; | |
593 | int prev_bit, next_bit; | |
594 | int new_extents; | |
595 | int ret; | |
596 | ||
597 | /* | |
598 | * Read the bit for the block immediately before the extent of space if | |
599 | * that block is within the block group. | |
600 | */ | |
601 | if (start > block_group->key.objectid) { | |
da17066c | 602 | u64 prev_block = start - block_group->fs_info->sectorsize; |
a5ed9182 OS |
603 | |
604 | key.objectid = prev_block; | |
605 | key.type = (u8)-1; | |
606 | key.offset = (u64)-1; | |
607 | ||
608 | ret = btrfs_search_prev_slot(trans, root, &key, path, 0, 1); | |
609 | if (ret) | |
610 | goto out; | |
611 | ||
612 | prev_bit = free_space_test_bit(block_group, path, prev_block); | |
613 | ||
614 | /* The previous block may have been in the previous bitmap. */ | |
615 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
616 | if (start >= key.objectid + key.offset) { | |
617 | ret = free_space_next_bitmap(trans, root, path); | |
618 | if (ret) | |
619 | goto out; | |
620 | } | |
621 | } else { | |
622 | key.objectid = start; | |
623 | key.type = (u8)-1; | |
624 | key.offset = (u64)-1; | |
625 | ||
626 | ret = btrfs_search_prev_slot(trans, root, &key, path, 0, 1); | |
627 | if (ret) | |
628 | goto out; | |
629 | ||
630 | prev_bit = -1; | |
631 | } | |
632 | ||
633 | /* | |
634 | * Iterate over all of the bitmaps overlapped by the extent of space, | |
635 | * clearing/setting bits as required. | |
636 | */ | |
637 | cur_start = start; | |
638 | cur_size = size; | |
639 | while (1) { | |
640 | free_space_set_bits(block_group, path, &cur_start, &cur_size, | |
641 | !remove); | |
642 | if (cur_size == 0) | |
643 | break; | |
644 | ret = free_space_next_bitmap(trans, root, path); | |
645 | if (ret) | |
646 | goto out; | |
647 | } | |
648 | ||
649 | /* | |
650 | * Read the bit for the block immediately after the extent of space if | |
651 | * that block is within the block group. | |
652 | */ | |
653 | if (end < block_group->key.objectid + block_group->key.offset) { | |
654 | /* The next block may be in the next bitmap. */ | |
655 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
656 | if (end >= key.objectid + key.offset) { | |
657 | ret = free_space_next_bitmap(trans, root, path); | |
658 | if (ret) | |
659 | goto out; | |
660 | } | |
661 | ||
662 | next_bit = free_space_test_bit(block_group, path, end); | |
663 | } else { | |
664 | next_bit = -1; | |
665 | } | |
666 | ||
667 | if (remove) { | |
668 | new_extents = -1; | |
669 | if (prev_bit == 1) { | |
670 | /* Leftover on the left. */ | |
671 | new_extents++; | |
672 | } | |
673 | if (next_bit == 1) { | |
674 | /* Leftover on the right. */ | |
675 | new_extents++; | |
676 | } | |
677 | } else { | |
678 | new_extents = 1; | |
679 | if (prev_bit == 1) { | |
680 | /* Merging with neighbor on the left. */ | |
681 | new_extents--; | |
682 | } | |
683 | if (next_bit == 1) { | |
684 | /* Merging with neighbor on the right. */ | |
685 | new_extents--; | |
686 | } | |
687 | } | |
688 | ||
689 | btrfs_release_path(path); | |
690 | ret = update_free_space_extent_count(trans, fs_info, block_group, path, | |
691 | new_extents); | |
692 | ||
693 | out: | |
694 | return ret; | |
695 | } | |
696 | ||
697 | static int remove_free_space_extent(struct btrfs_trans_handle *trans, | |
698 | struct btrfs_fs_info *fs_info, | |
699 | struct btrfs_block_group_cache *block_group, | |
700 | struct btrfs_path *path, | |
701 | u64 start, u64 size) | |
702 | { | |
703 | struct btrfs_root *root = fs_info->free_space_root; | |
704 | struct btrfs_key key; | |
705 | u64 found_start, found_end; | |
706 | u64 end = start + size; | |
707 | int new_extents = -1; | |
708 | int ret; | |
709 | ||
710 | key.objectid = start; | |
711 | key.type = (u8)-1; | |
712 | key.offset = (u64)-1; | |
713 | ||
714 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
715 | if (ret) | |
716 | goto out; | |
717 | ||
718 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
719 | ||
720 | ASSERT(key.type == BTRFS_FREE_SPACE_EXTENT_KEY); | |
721 | ||
722 | found_start = key.objectid; | |
723 | found_end = key.objectid + key.offset; | |
724 | ASSERT(start >= found_start && end <= found_end); | |
725 | ||
726 | /* | |
727 | * Okay, now that we've found the free space extent which contains the | |
728 | * free space that we are removing, there are four cases: | |
729 | * | |
730 | * 1. We're using the whole extent: delete the key we found and | |
731 | * decrement the free space extent count. | |
732 | * 2. We are using part of the extent starting at the beginning: delete | |
733 | * the key we found and insert a new key representing the leftover at | |
734 | * the end. There is no net change in the number of extents. | |
735 | * 3. We are using part of the extent ending at the end: delete the key | |
736 | * we found and insert a new key representing the leftover at the | |
737 | * beginning. There is no net change in the number of extents. | |
738 | * 4. We are using part of the extent in the middle: delete the key we | |
739 | * found and insert two new keys representing the leftovers on each | |
740 | * side. Where we used to have one extent, we now have two, so increment | |
741 | * the extent count. We may need to convert the block group to bitmaps | |
742 | * as a result. | |
743 | */ | |
744 | ||
745 | /* Delete the existing key (cases 1-4). */ | |
746 | ret = btrfs_del_item(trans, root, path); | |
747 | if (ret) | |
748 | goto out; | |
749 | ||
750 | /* Add a key for leftovers at the beginning (cases 3 and 4). */ | |
751 | if (start > found_start) { | |
752 | key.objectid = found_start; | |
753 | key.type = BTRFS_FREE_SPACE_EXTENT_KEY; | |
754 | key.offset = start - found_start; | |
755 | ||
756 | btrfs_release_path(path); | |
757 | ret = btrfs_insert_empty_item(trans, root, path, &key, 0); | |
758 | if (ret) | |
759 | goto out; | |
760 | new_extents++; | |
761 | } | |
762 | ||
763 | /* Add a key for leftovers at the end (cases 2 and 4). */ | |
764 | if (end < found_end) { | |
765 | key.objectid = end; | |
766 | key.type = BTRFS_FREE_SPACE_EXTENT_KEY; | |
767 | key.offset = found_end - end; | |
768 | ||
769 | btrfs_release_path(path); | |
770 | ret = btrfs_insert_empty_item(trans, root, path, &key, 0); | |
771 | if (ret) | |
772 | goto out; | |
773 | new_extents++; | |
774 | } | |
775 | ||
776 | btrfs_release_path(path); | |
777 | ret = update_free_space_extent_count(trans, fs_info, block_group, path, | |
778 | new_extents); | |
779 | ||
780 | out: | |
781 | return ret; | |
782 | } | |
783 | ||
784 | int __remove_from_free_space_tree(struct btrfs_trans_handle *trans, | |
785 | struct btrfs_fs_info *fs_info, | |
786 | struct btrfs_block_group_cache *block_group, | |
787 | struct btrfs_path *path, u64 start, u64 size) | |
788 | { | |
789 | struct btrfs_free_space_info *info; | |
790 | u32 flags; | |
791 | int ret; | |
792 | ||
793 | if (block_group->needs_free_space) { | |
794 | ret = __add_block_group_free_space(trans, fs_info, block_group, | |
795 | path); | |
796 | if (ret) | |
797 | return ret; | |
798 | } | |
799 | ||
800 | info = search_free_space_info(NULL, fs_info, block_group, path, 0); | |
801 | if (IS_ERR(info)) | |
802 | return PTR_ERR(info); | |
803 | flags = btrfs_free_space_flags(path->nodes[0], info); | |
804 | btrfs_release_path(path); | |
805 | ||
806 | if (flags & BTRFS_FREE_SPACE_USING_BITMAPS) { | |
807 | return modify_free_space_bitmap(trans, fs_info, block_group, | |
808 | path, start, size, 1); | |
809 | } else { | |
810 | return remove_free_space_extent(trans, fs_info, block_group, | |
811 | path, start, size); | |
812 | } | |
813 | } | |
814 | ||
815 | int remove_from_free_space_tree(struct btrfs_trans_handle *trans, | |
816 | struct btrfs_fs_info *fs_info, | |
817 | u64 start, u64 size) | |
818 | { | |
819 | struct btrfs_block_group_cache *block_group; | |
820 | struct btrfs_path *path; | |
821 | int ret; | |
822 | ||
823 | if (!btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) | |
824 | return 0; | |
825 | ||
826 | path = btrfs_alloc_path(); | |
827 | if (!path) { | |
828 | ret = -ENOMEM; | |
829 | goto out; | |
830 | } | |
831 | ||
832 | block_group = btrfs_lookup_block_group(fs_info, start); | |
833 | if (!block_group) { | |
834 | ASSERT(0); | |
835 | ret = -ENOENT; | |
836 | goto out; | |
837 | } | |
838 | ||
839 | mutex_lock(&block_group->free_space_lock); | |
840 | ret = __remove_from_free_space_tree(trans, fs_info, block_group, path, | |
841 | start, size); | |
842 | mutex_unlock(&block_group->free_space_lock); | |
843 | ||
844 | btrfs_put_block_group(block_group); | |
845 | out: | |
846 | btrfs_free_path(path); | |
847 | if (ret) | |
66642832 | 848 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
849 | return ret; |
850 | } | |
851 | ||
852 | static int add_free_space_extent(struct btrfs_trans_handle *trans, | |
853 | struct btrfs_fs_info *fs_info, | |
854 | struct btrfs_block_group_cache *block_group, | |
855 | struct btrfs_path *path, | |
856 | u64 start, u64 size) | |
857 | { | |
858 | struct btrfs_root *root = fs_info->free_space_root; | |
859 | struct btrfs_key key, new_key; | |
860 | u64 found_start, found_end; | |
861 | u64 end = start + size; | |
862 | int new_extents = 1; | |
863 | int ret; | |
864 | ||
865 | /* | |
866 | * We are adding a new extent of free space, but we need to merge | |
867 | * extents. There are four cases here: | |
868 | * | |
869 | * 1. The new extent does not have any immediate neighbors to merge | |
870 | * with: add the new key and increment the free space extent count. We | |
871 | * may need to convert the block group to bitmaps as a result. | |
872 | * 2. The new extent has an immediate neighbor before it: remove the | |
873 | * previous key and insert a new key combining both of them. There is no | |
874 | * net change in the number of extents. | |
875 | * 3. The new extent has an immediate neighbor after it: remove the next | |
876 | * key and insert a new key combining both of them. There is no net | |
877 | * change in the number of extents. | |
878 | * 4. The new extent has immediate neighbors on both sides: remove both | |
879 | * of the keys and insert a new key combining all of them. Where we used | |
880 | * to have two extents, we now have one, so decrement the extent count. | |
881 | */ | |
882 | ||
883 | new_key.objectid = start; | |
884 | new_key.type = BTRFS_FREE_SPACE_EXTENT_KEY; | |
885 | new_key.offset = size; | |
886 | ||
887 | /* Search for a neighbor on the left. */ | |
888 | if (start == block_group->key.objectid) | |
889 | goto right; | |
890 | key.objectid = start - 1; | |
891 | key.type = (u8)-1; | |
892 | key.offset = (u64)-1; | |
893 | ||
894 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
895 | if (ret) | |
896 | goto out; | |
897 | ||
898 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
899 | ||
900 | if (key.type != BTRFS_FREE_SPACE_EXTENT_KEY) { | |
901 | ASSERT(key.type == BTRFS_FREE_SPACE_INFO_KEY); | |
902 | btrfs_release_path(path); | |
903 | goto right; | |
904 | } | |
905 | ||
906 | found_start = key.objectid; | |
907 | found_end = key.objectid + key.offset; | |
908 | ASSERT(found_start >= block_group->key.objectid && | |
909 | found_end > block_group->key.objectid); | |
910 | ASSERT(found_start < start && found_end <= start); | |
911 | ||
912 | /* | |
913 | * Delete the neighbor on the left and absorb it into the new key (cases | |
914 | * 2 and 4). | |
915 | */ | |
916 | if (found_end == start) { | |
917 | ret = btrfs_del_item(trans, root, path); | |
918 | if (ret) | |
919 | goto out; | |
920 | new_key.objectid = found_start; | |
921 | new_key.offset += key.offset; | |
922 | new_extents--; | |
923 | } | |
924 | btrfs_release_path(path); | |
925 | ||
926 | right: | |
927 | /* Search for a neighbor on the right. */ | |
928 | if (end == block_group->key.objectid + block_group->key.offset) | |
929 | goto insert; | |
930 | key.objectid = end; | |
931 | key.type = (u8)-1; | |
932 | key.offset = (u64)-1; | |
933 | ||
934 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
935 | if (ret) | |
936 | goto out; | |
937 | ||
938 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
939 | ||
940 | if (key.type != BTRFS_FREE_SPACE_EXTENT_KEY) { | |
941 | ASSERT(key.type == BTRFS_FREE_SPACE_INFO_KEY); | |
942 | btrfs_release_path(path); | |
943 | goto insert; | |
944 | } | |
945 | ||
946 | found_start = key.objectid; | |
947 | found_end = key.objectid + key.offset; | |
948 | ASSERT(found_start >= block_group->key.objectid && | |
949 | found_end > block_group->key.objectid); | |
950 | ASSERT((found_start < start && found_end <= start) || | |
951 | (found_start >= end && found_end > end)); | |
952 | ||
953 | /* | |
954 | * Delete the neighbor on the right and absorb it into the new key | |
955 | * (cases 3 and 4). | |
956 | */ | |
957 | if (found_start == end) { | |
958 | ret = btrfs_del_item(trans, root, path); | |
959 | if (ret) | |
960 | goto out; | |
961 | new_key.offset += key.offset; | |
962 | new_extents--; | |
963 | } | |
964 | btrfs_release_path(path); | |
965 | ||
966 | insert: | |
967 | /* Insert the new key (cases 1-4). */ | |
968 | ret = btrfs_insert_empty_item(trans, root, path, &new_key, 0); | |
969 | if (ret) | |
970 | goto out; | |
971 | ||
972 | btrfs_release_path(path); | |
973 | ret = update_free_space_extent_count(trans, fs_info, block_group, path, | |
974 | new_extents); | |
975 | ||
976 | out: | |
977 | return ret; | |
978 | } | |
979 | ||
980 | int __add_to_free_space_tree(struct btrfs_trans_handle *trans, | |
981 | struct btrfs_fs_info *fs_info, | |
982 | struct btrfs_block_group_cache *block_group, | |
983 | struct btrfs_path *path, u64 start, u64 size) | |
984 | { | |
985 | struct btrfs_free_space_info *info; | |
986 | u32 flags; | |
987 | int ret; | |
988 | ||
989 | if (block_group->needs_free_space) { | |
990 | ret = __add_block_group_free_space(trans, fs_info, block_group, | |
991 | path); | |
992 | if (ret) | |
993 | return ret; | |
994 | } | |
995 | ||
996 | info = search_free_space_info(NULL, fs_info, block_group, path, 0); | |
997 | if (IS_ERR(info)) | |
998 | return PTR_ERR(info); | |
999 | flags = btrfs_free_space_flags(path->nodes[0], info); | |
1000 | btrfs_release_path(path); | |
1001 | ||
1002 | if (flags & BTRFS_FREE_SPACE_USING_BITMAPS) { | |
1003 | return modify_free_space_bitmap(trans, fs_info, block_group, | |
1004 | path, start, size, 0); | |
1005 | } else { | |
1006 | return add_free_space_extent(trans, fs_info, block_group, path, | |
1007 | start, size); | |
1008 | } | |
1009 | } | |
1010 | ||
1011 | int add_to_free_space_tree(struct btrfs_trans_handle *trans, | |
1012 | struct btrfs_fs_info *fs_info, | |
1013 | u64 start, u64 size) | |
1014 | { | |
1015 | struct btrfs_block_group_cache *block_group; | |
1016 | struct btrfs_path *path; | |
1017 | int ret; | |
1018 | ||
1019 | if (!btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) | |
1020 | return 0; | |
1021 | ||
1022 | path = btrfs_alloc_path(); | |
1023 | if (!path) { | |
1024 | ret = -ENOMEM; | |
1025 | goto out; | |
1026 | } | |
1027 | ||
1028 | block_group = btrfs_lookup_block_group(fs_info, start); | |
1029 | if (!block_group) { | |
1030 | ASSERT(0); | |
1031 | ret = -ENOENT; | |
1032 | goto out; | |
1033 | } | |
1034 | ||
1035 | mutex_lock(&block_group->free_space_lock); | |
1036 | ret = __add_to_free_space_tree(trans, fs_info, block_group, path, start, | |
1037 | size); | |
1038 | mutex_unlock(&block_group->free_space_lock); | |
1039 | ||
1040 | btrfs_put_block_group(block_group); | |
1041 | out: | |
1042 | btrfs_free_path(path); | |
1043 | if (ret) | |
66642832 | 1044 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
1045 | return ret; |
1046 | } | |
1047 | ||
1048 | /* | |
1049 | * Populate the free space tree by walking the extent tree. Operations on the | |
1050 | * extent tree that happen as a result of writes to the free space tree will go | |
1051 | * through the normal add/remove hooks. | |
1052 | */ | |
1053 | static int populate_free_space_tree(struct btrfs_trans_handle *trans, | |
1054 | struct btrfs_fs_info *fs_info, | |
1055 | struct btrfs_block_group_cache *block_group) | |
1056 | { | |
1057 | struct btrfs_root *extent_root = fs_info->extent_root; | |
1058 | struct btrfs_path *path, *path2; | |
1059 | struct btrfs_key key; | |
1060 | u64 start, end; | |
1061 | int ret; | |
1062 | ||
1063 | path = btrfs_alloc_path(); | |
1064 | if (!path) | |
1065 | return -ENOMEM; | |
019599ad | 1066 | path->reada = READA_FORWARD; |
a5ed9182 OS |
1067 | |
1068 | path2 = btrfs_alloc_path(); | |
1069 | if (!path2) { | |
1070 | btrfs_free_path(path); | |
1071 | return -ENOMEM; | |
1072 | } | |
1073 | ||
1074 | ret = add_new_free_space_info(trans, fs_info, block_group, path2); | |
1075 | if (ret) | |
1076 | goto out; | |
1077 | ||
511711af CM |
1078 | mutex_lock(&block_group->free_space_lock); |
1079 | ||
a5ed9182 OS |
1080 | /* |
1081 | * Iterate through all of the extent and metadata items in this block | |
1082 | * group, adding the free space between them and the free space at the | |
1083 | * end. Note that EXTENT_ITEM and METADATA_ITEM are less than | |
1084 | * BLOCK_GROUP_ITEM, so an extent may precede the block group that it's | |
1085 | * contained in. | |
1086 | */ | |
1087 | key.objectid = block_group->key.objectid; | |
1088 | key.type = BTRFS_EXTENT_ITEM_KEY; | |
1089 | key.offset = 0; | |
1090 | ||
1091 | ret = btrfs_search_slot_for_read(extent_root, &key, path, 1, 0); | |
1092 | if (ret < 0) | |
511711af | 1093 | goto out_locked; |
a5ed9182 OS |
1094 | ASSERT(ret == 0); |
1095 | ||
1096 | start = block_group->key.objectid; | |
1097 | end = block_group->key.objectid + block_group->key.offset; | |
1098 | while (1) { | |
1099 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
1100 | ||
1101 | if (key.type == BTRFS_EXTENT_ITEM_KEY || | |
1102 | key.type == BTRFS_METADATA_ITEM_KEY) { | |
1103 | if (key.objectid >= end) | |
1104 | break; | |
1105 | ||
1106 | if (start < key.objectid) { | |
1107 | ret = __add_to_free_space_tree(trans, fs_info, | |
1108 | block_group, | |
1109 | path2, start, | |
1110 | key.objectid - | |
1111 | start); | |
1112 | if (ret) | |
511711af | 1113 | goto out_locked; |
a5ed9182 OS |
1114 | } |
1115 | start = key.objectid; | |
1116 | if (key.type == BTRFS_METADATA_ITEM_KEY) | |
da17066c | 1117 | start += fs_info->nodesize; |
a5ed9182 OS |
1118 | else |
1119 | start += key.offset; | |
1120 | } else if (key.type == BTRFS_BLOCK_GROUP_ITEM_KEY) { | |
1121 | if (key.objectid != block_group->key.objectid) | |
1122 | break; | |
1123 | } | |
1124 | ||
1125 | ret = btrfs_next_item(extent_root, path); | |
1126 | if (ret < 0) | |
511711af | 1127 | goto out_locked; |
a5ed9182 OS |
1128 | if (ret) |
1129 | break; | |
1130 | } | |
1131 | if (start < end) { | |
1132 | ret = __add_to_free_space_tree(trans, fs_info, block_group, | |
1133 | path2, start, end - start); | |
1134 | if (ret) | |
511711af | 1135 | goto out_locked; |
a5ed9182 OS |
1136 | } |
1137 | ||
1138 | ret = 0; | |
511711af CM |
1139 | out_locked: |
1140 | mutex_unlock(&block_group->free_space_lock); | |
a5ed9182 OS |
1141 | out: |
1142 | btrfs_free_path(path2); | |
1143 | btrfs_free_path(path); | |
1144 | return ret; | |
1145 | } | |
1146 | ||
1147 | int btrfs_create_free_space_tree(struct btrfs_fs_info *fs_info) | |
1148 | { | |
1149 | struct btrfs_trans_handle *trans; | |
1150 | struct btrfs_root *tree_root = fs_info->tree_root; | |
1151 | struct btrfs_root *free_space_root; | |
1152 | struct btrfs_block_group_cache *block_group; | |
1153 | struct rb_node *node; | |
1154 | int ret; | |
1155 | ||
1156 | trans = btrfs_start_transaction(tree_root, 0); | |
1157 | if (IS_ERR(trans)) | |
1158 | return PTR_ERR(trans); | |
1159 | ||
afcdd129 | 1160 | set_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags); |
a5ed9182 OS |
1161 | free_space_root = btrfs_create_tree(trans, fs_info, |
1162 | BTRFS_FREE_SPACE_TREE_OBJECTID); | |
1163 | if (IS_ERR(free_space_root)) { | |
1164 | ret = PTR_ERR(free_space_root); | |
1165 | goto abort; | |
1166 | } | |
1167 | fs_info->free_space_root = free_space_root; | |
1168 | ||
1169 | node = rb_first(&fs_info->block_group_cache_tree); | |
1170 | while (node) { | |
1171 | block_group = rb_entry(node, struct btrfs_block_group_cache, | |
1172 | cache_node); | |
1173 | ret = populate_free_space_tree(trans, fs_info, block_group); | |
1174 | if (ret) | |
1175 | goto abort; | |
1176 | node = rb_next(node); | |
1177 | } | |
1178 | ||
1179 | btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE); | |
6675df31 | 1180 | btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID); |
afcdd129 | 1181 | clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags); |
a5ed9182 | 1182 | |
0bef7109 | 1183 | return btrfs_commit_transaction(trans); |
a5ed9182 OS |
1184 | |
1185 | abort: | |
afcdd129 | 1186 | clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags); |
66642832 | 1187 | btrfs_abort_transaction(trans, ret); |
3a45bb20 | 1188 | btrfs_end_transaction(trans); |
a5ed9182 OS |
1189 | return ret; |
1190 | } | |
1191 | ||
1192 | static int clear_free_space_tree(struct btrfs_trans_handle *trans, | |
1193 | struct btrfs_root *root) | |
1194 | { | |
1195 | struct btrfs_path *path; | |
1196 | struct btrfs_key key; | |
1197 | int nr; | |
1198 | int ret; | |
1199 | ||
1200 | path = btrfs_alloc_path(); | |
1201 | if (!path) | |
1202 | return -ENOMEM; | |
1203 | ||
1204 | path->leave_spinning = 1; | |
1205 | ||
1206 | key.objectid = 0; | |
1207 | key.type = 0; | |
1208 | key.offset = 0; | |
1209 | ||
1210 | while (1) { | |
1211 | ret = btrfs_search_slot(trans, root, &key, path, -1, 1); | |
1212 | if (ret < 0) | |
1213 | goto out; | |
1214 | ||
1215 | nr = btrfs_header_nritems(path->nodes[0]); | |
1216 | if (!nr) | |
1217 | break; | |
1218 | ||
1219 | path->slots[0] = 0; | |
1220 | ret = btrfs_del_items(trans, root, path, 0, nr); | |
1221 | if (ret) | |
1222 | goto out; | |
1223 | ||
1224 | btrfs_release_path(path); | |
1225 | } | |
1226 | ||
1227 | ret = 0; | |
1228 | out: | |
1229 | btrfs_free_path(path); | |
1230 | return ret; | |
1231 | } | |
1232 | ||
1233 | int btrfs_clear_free_space_tree(struct btrfs_fs_info *fs_info) | |
1234 | { | |
1235 | struct btrfs_trans_handle *trans; | |
1236 | struct btrfs_root *tree_root = fs_info->tree_root; | |
1237 | struct btrfs_root *free_space_root = fs_info->free_space_root; | |
1238 | int ret; | |
1239 | ||
1240 | trans = btrfs_start_transaction(tree_root, 0); | |
1241 | if (IS_ERR(trans)) | |
1242 | return PTR_ERR(trans); | |
1243 | ||
1244 | btrfs_clear_fs_compat_ro(fs_info, FREE_SPACE_TREE); | |
6675df31 | 1245 | btrfs_clear_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID); |
a5ed9182 OS |
1246 | fs_info->free_space_root = NULL; |
1247 | ||
1248 | ret = clear_free_space_tree(trans, free_space_root); | |
1249 | if (ret) | |
1250 | goto abort; | |
1251 | ||
1cd5447e | 1252 | ret = btrfs_del_root(trans, fs_info, &free_space_root->root_key); |
a5ed9182 OS |
1253 | if (ret) |
1254 | goto abort; | |
1255 | ||
1256 | list_del(&free_space_root->dirty_list); | |
1257 | ||
1258 | btrfs_tree_lock(free_space_root->node); | |
7c302b49 | 1259 | clean_tree_block(fs_info, free_space_root->node); |
a5ed9182 OS |
1260 | btrfs_tree_unlock(free_space_root->node); |
1261 | btrfs_free_tree_block(trans, free_space_root, free_space_root->node, | |
1262 | 0, 1); | |
1263 | ||
1264 | free_extent_buffer(free_space_root->node); | |
1265 | free_extent_buffer(free_space_root->commit_root); | |
1266 | kfree(free_space_root); | |
1267 | ||
0bef7109 | 1268 | return btrfs_commit_transaction(trans); |
a5ed9182 OS |
1269 | |
1270 | abort: | |
66642832 | 1271 | btrfs_abort_transaction(trans, ret); |
3a45bb20 | 1272 | btrfs_end_transaction(trans); |
a5ed9182 OS |
1273 | return ret; |
1274 | } | |
1275 | ||
1276 | static int __add_block_group_free_space(struct btrfs_trans_handle *trans, | |
1277 | struct btrfs_fs_info *fs_info, | |
1278 | struct btrfs_block_group_cache *block_group, | |
1279 | struct btrfs_path *path) | |
1280 | { | |
a5ed9182 OS |
1281 | int ret; |
1282 | ||
a5ed9182 OS |
1283 | block_group->needs_free_space = 0; |
1284 | ||
1285 | ret = add_new_free_space_info(trans, fs_info, block_group, path); | |
1286 | if (ret) | |
1287 | return ret; | |
1288 | ||
1289 | return __add_to_free_space_tree(trans, fs_info, block_group, path, | |
1290 | block_group->key.objectid, | |
1291 | block_group->key.offset); | |
1292 | } | |
1293 | ||
1294 | int add_block_group_free_space(struct btrfs_trans_handle *trans, | |
a5ed9182 OS |
1295 | struct btrfs_block_group_cache *block_group) |
1296 | { | |
e4e0711c | 1297 | struct btrfs_fs_info *fs_info = trans->fs_info; |
a5ed9182 OS |
1298 | struct btrfs_path *path = NULL; |
1299 | int ret = 0; | |
1300 | ||
1301 | if (!btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) | |
1302 | return 0; | |
1303 | ||
1304 | mutex_lock(&block_group->free_space_lock); | |
1305 | if (!block_group->needs_free_space) | |
1306 | goto out; | |
1307 | ||
1308 | path = btrfs_alloc_path(); | |
1309 | if (!path) { | |
1310 | ret = -ENOMEM; | |
1311 | goto out; | |
1312 | } | |
1313 | ||
1314 | ret = __add_block_group_free_space(trans, fs_info, block_group, path); | |
1315 | ||
1316 | out: | |
1317 | btrfs_free_path(path); | |
1318 | mutex_unlock(&block_group->free_space_lock); | |
1319 | if (ret) | |
66642832 | 1320 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
1321 | return ret; |
1322 | } | |
1323 | ||
1324 | int remove_block_group_free_space(struct btrfs_trans_handle *trans, | |
1325 | struct btrfs_fs_info *fs_info, | |
1326 | struct btrfs_block_group_cache *block_group) | |
1327 | { | |
1328 | struct btrfs_root *root = fs_info->free_space_root; | |
1329 | struct btrfs_path *path; | |
1330 | struct btrfs_key key, found_key; | |
1331 | struct extent_buffer *leaf; | |
1332 | u64 start, end; | |
1333 | int done = 0, nr; | |
1334 | int ret; | |
1335 | ||
1336 | if (!btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE)) | |
1337 | return 0; | |
1338 | ||
1339 | if (block_group->needs_free_space) { | |
1340 | /* We never added this block group to the free space tree. */ | |
1341 | return 0; | |
1342 | } | |
1343 | ||
1344 | path = btrfs_alloc_path(); | |
1345 | if (!path) { | |
1346 | ret = -ENOMEM; | |
1347 | goto out; | |
1348 | } | |
1349 | ||
1350 | start = block_group->key.objectid; | |
1351 | end = block_group->key.objectid + block_group->key.offset; | |
1352 | ||
1353 | key.objectid = end - 1; | |
1354 | key.type = (u8)-1; | |
1355 | key.offset = (u64)-1; | |
1356 | ||
1357 | while (!done) { | |
1358 | ret = btrfs_search_prev_slot(trans, root, &key, path, -1, 1); | |
1359 | if (ret) | |
1360 | goto out; | |
1361 | ||
1362 | leaf = path->nodes[0]; | |
1363 | nr = 0; | |
1364 | path->slots[0]++; | |
1365 | while (path->slots[0] > 0) { | |
1366 | btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1); | |
1367 | ||
1368 | if (found_key.type == BTRFS_FREE_SPACE_INFO_KEY) { | |
1369 | ASSERT(found_key.objectid == block_group->key.objectid); | |
1370 | ASSERT(found_key.offset == block_group->key.offset); | |
1371 | done = 1; | |
1372 | nr++; | |
1373 | path->slots[0]--; | |
1374 | break; | |
1375 | } else if (found_key.type == BTRFS_FREE_SPACE_EXTENT_KEY || | |
1376 | found_key.type == BTRFS_FREE_SPACE_BITMAP_KEY) { | |
1377 | ASSERT(found_key.objectid >= start); | |
1378 | ASSERT(found_key.objectid < end); | |
1379 | ASSERT(found_key.objectid + found_key.offset <= end); | |
1380 | nr++; | |
1381 | path->slots[0]--; | |
1382 | } else { | |
1383 | ASSERT(0); | |
1384 | } | |
1385 | } | |
1386 | ||
1387 | ret = btrfs_del_items(trans, root, path, path->slots[0], nr); | |
1388 | if (ret) | |
1389 | goto out; | |
1390 | btrfs_release_path(path); | |
1391 | } | |
1392 | ||
1393 | ret = 0; | |
1394 | out: | |
1395 | btrfs_free_path(path); | |
1396 | if (ret) | |
66642832 | 1397 | btrfs_abort_transaction(trans, ret); |
a5ed9182 OS |
1398 | return ret; |
1399 | } | |
1400 | ||
1401 | static int load_free_space_bitmaps(struct btrfs_caching_control *caching_ctl, | |
1402 | struct btrfs_path *path, | |
1403 | u32 expected_extent_count) | |
1404 | { | |
1405 | struct btrfs_block_group_cache *block_group; | |
1406 | struct btrfs_fs_info *fs_info; | |
1407 | struct btrfs_root *root; | |
1408 | struct btrfs_key key; | |
1409 | int prev_bit = 0, bit; | |
1410 | /* Initialize to silence GCC. */ | |
1411 | u64 extent_start = 0; | |
1412 | u64 end, offset; | |
1413 | u64 total_found = 0; | |
1414 | u32 extent_count = 0; | |
1415 | int ret; | |
1416 | ||
1417 | block_group = caching_ctl->block_group; | |
1418 | fs_info = block_group->fs_info; | |
1419 | root = fs_info->free_space_root; | |
1420 | ||
1421 | end = block_group->key.objectid + block_group->key.offset; | |
1422 | ||
1423 | while (1) { | |
1424 | ret = btrfs_next_item(root, path); | |
1425 | if (ret < 0) | |
1426 | goto out; | |
1427 | if (ret) | |
1428 | break; | |
1429 | ||
1430 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
1431 | ||
1432 | if (key.type == BTRFS_FREE_SPACE_INFO_KEY) | |
1433 | break; | |
1434 | ||
1435 | ASSERT(key.type == BTRFS_FREE_SPACE_BITMAP_KEY); | |
1436 | ASSERT(key.objectid < end && key.objectid + key.offset <= end); | |
1437 | ||
1438 | caching_ctl->progress = key.objectid; | |
1439 | ||
1440 | offset = key.objectid; | |
1441 | while (offset < key.objectid + key.offset) { | |
1442 | bit = free_space_test_bit(block_group, path, offset); | |
1443 | if (prev_bit == 0 && bit == 1) { | |
1444 | extent_start = offset; | |
1445 | } else if (prev_bit == 1 && bit == 0) { | |
1446 | total_found += add_new_free_space(block_group, | |
1447 | fs_info, | |
1448 | extent_start, | |
1449 | offset); | |
1450 | if (total_found > CACHING_CTL_WAKE_UP) { | |
1451 | total_found = 0; | |
1452 | wake_up(&caching_ctl->wait); | |
1453 | } | |
1454 | extent_count++; | |
1455 | } | |
1456 | prev_bit = bit; | |
0b246afa | 1457 | offset += fs_info->sectorsize; |
a5ed9182 OS |
1458 | } |
1459 | } | |
1460 | if (prev_bit == 1) { | |
1461 | total_found += add_new_free_space(block_group, fs_info, | |
1462 | extent_start, end); | |
1463 | extent_count++; | |
1464 | } | |
1465 | ||
1466 | if (extent_count != expected_extent_count) { | |
5d163e0e JM |
1467 | btrfs_err(fs_info, |
1468 | "incorrect extent count for %llu; counted %u, expected %u", | |
a5ed9182 OS |
1469 | block_group->key.objectid, extent_count, |
1470 | expected_extent_count); | |
1471 | ASSERT(0); | |
1472 | ret = -EIO; | |
1473 | goto out; | |
1474 | } | |
1475 | ||
1476 | caching_ctl->progress = (u64)-1; | |
1477 | ||
1478 | ret = 0; | |
1479 | out: | |
1480 | return ret; | |
1481 | } | |
1482 | ||
1483 | static int load_free_space_extents(struct btrfs_caching_control *caching_ctl, | |
1484 | struct btrfs_path *path, | |
1485 | u32 expected_extent_count) | |
1486 | { | |
1487 | struct btrfs_block_group_cache *block_group; | |
1488 | struct btrfs_fs_info *fs_info; | |
1489 | struct btrfs_root *root; | |
1490 | struct btrfs_key key; | |
1491 | u64 end; | |
1492 | u64 total_found = 0; | |
1493 | u32 extent_count = 0; | |
1494 | int ret; | |
1495 | ||
1496 | block_group = caching_ctl->block_group; | |
1497 | fs_info = block_group->fs_info; | |
1498 | root = fs_info->free_space_root; | |
1499 | ||
1500 | end = block_group->key.objectid + block_group->key.offset; | |
1501 | ||
1502 | while (1) { | |
1503 | ret = btrfs_next_item(root, path); | |
1504 | if (ret < 0) | |
1505 | goto out; | |
1506 | if (ret) | |
1507 | break; | |
1508 | ||
1509 | btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); | |
1510 | ||
1511 | if (key.type == BTRFS_FREE_SPACE_INFO_KEY) | |
1512 | break; | |
1513 | ||
1514 | ASSERT(key.type == BTRFS_FREE_SPACE_EXTENT_KEY); | |
1515 | ASSERT(key.objectid < end && key.objectid + key.offset <= end); | |
1516 | ||
1517 | caching_ctl->progress = key.objectid; | |
1518 | ||
1519 | total_found += add_new_free_space(block_group, fs_info, | |
1520 | key.objectid, | |
1521 | key.objectid + key.offset); | |
1522 | if (total_found > CACHING_CTL_WAKE_UP) { | |
1523 | total_found = 0; | |
1524 | wake_up(&caching_ctl->wait); | |
1525 | } | |
1526 | extent_count++; | |
1527 | } | |
1528 | ||
1529 | if (extent_count != expected_extent_count) { | |
5d163e0e JM |
1530 | btrfs_err(fs_info, |
1531 | "incorrect extent count for %llu; counted %u, expected %u", | |
a5ed9182 OS |
1532 | block_group->key.objectid, extent_count, |
1533 | expected_extent_count); | |
1534 | ASSERT(0); | |
1535 | ret = -EIO; | |
1536 | goto out; | |
1537 | } | |
1538 | ||
1539 | caching_ctl->progress = (u64)-1; | |
1540 | ||
1541 | ret = 0; | |
1542 | out: | |
1543 | return ret; | |
1544 | } | |
1545 | ||
1546 | int load_free_space_tree(struct btrfs_caching_control *caching_ctl) | |
1547 | { | |
1548 | struct btrfs_block_group_cache *block_group; | |
1549 | struct btrfs_fs_info *fs_info; | |
1550 | struct btrfs_free_space_info *info; | |
1551 | struct btrfs_path *path; | |
1552 | u32 extent_count, flags; | |
1553 | int ret; | |
1554 | ||
1555 | block_group = caching_ctl->block_group; | |
1556 | fs_info = block_group->fs_info; | |
1557 | ||
1558 | path = btrfs_alloc_path(); | |
1559 | if (!path) | |
1560 | return -ENOMEM; | |
1561 | ||
1562 | /* | |
1563 | * Just like caching_thread() doesn't want to deadlock on the extent | |
1564 | * tree, we don't want to deadlock on the free space tree. | |
1565 | */ | |
1566 | path->skip_locking = 1; | |
1567 | path->search_commit_root = 1; | |
7ce311d5 | 1568 | path->reada = READA_FORWARD; |
a5ed9182 OS |
1569 | |
1570 | info = search_free_space_info(NULL, fs_info, block_group, path, 0); | |
1571 | if (IS_ERR(info)) { | |
1572 | ret = PTR_ERR(info); | |
1573 | goto out; | |
1574 | } | |
1575 | extent_count = btrfs_free_space_extent_count(path->nodes[0], info); | |
1576 | flags = btrfs_free_space_flags(path->nodes[0], info); | |
1577 | ||
1578 | /* | |
1579 | * We left path pointing to the free space info item, so now | |
1580 | * load_free_space_foo can just iterate through the free space tree from | |
1581 | * there. | |
1582 | */ | |
1583 | if (flags & BTRFS_FREE_SPACE_USING_BITMAPS) | |
1584 | ret = load_free_space_bitmaps(caching_ctl, path, extent_count); | |
1585 | else | |
1586 | ret = load_free_space_extents(caching_ctl, path, extent_count); | |
1587 | ||
1588 | out: | |
1589 | btrfs_free_path(path); | |
1590 | return ret; | |
1591 | } |