Commit | Line | Data |
---|---|---|
0f9dd46c JB |
1 | /* |
2 | * Copyright (C) 2008 Red Hat. All rights reserved. | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or | |
5 | * modify it under the terms of the GNU General Public | |
6 | * License v2 as published by the Free Software Foundation. | |
7 | * | |
8 | * This program is distributed in the hope that it will be useful, | |
9 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
11 | * General Public License for more details. | |
12 | * | |
13 | * You should have received a copy of the GNU General Public | |
14 | * License along with this program; if not, write to the | |
15 | * Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
16 | * Boston, MA 021110-1307, USA. | |
17 | */ | |
18 | ||
96303081 | 19 | #include <linux/pagemap.h> |
0f9dd46c | 20 | #include <linux/sched.h> |
5a0e3ad6 | 21 | #include <linux/slab.h> |
96303081 | 22 | #include <linux/math64.h> |
6ab60601 | 23 | #include <linux/ratelimit.h> |
0f9dd46c | 24 | #include "ctree.h" |
fa9c0d79 CM |
25 | #include "free-space-cache.h" |
26 | #include "transaction.h" | |
0af3d00b | 27 | #include "disk-io.h" |
43be2146 | 28 | #include "extent_io.h" |
581bb050 | 29 | #include "inode-map.h" |
fa9c0d79 | 30 | |
96303081 JB |
31 | #define BITS_PER_BITMAP (PAGE_CACHE_SIZE * 8) |
32 | #define MAX_CACHE_BYTES_PER_GIG (32 * 1024) | |
0f9dd46c | 33 | |
34d52cb6 | 34 | static int link_free_space(struct btrfs_free_space_ctl *ctl, |
0cb59c99 | 35 | struct btrfs_free_space *info); |
cd023e7b JB |
36 | static void unlink_free_space(struct btrfs_free_space_ctl *ctl, |
37 | struct btrfs_free_space *info); | |
0cb59c99 | 38 | |
0414efae LZ |
39 | static struct inode *__lookup_free_space_inode(struct btrfs_root *root, |
40 | struct btrfs_path *path, | |
41 | u64 offset) | |
0af3d00b JB |
42 | { |
43 | struct btrfs_key key; | |
44 | struct btrfs_key location; | |
45 | struct btrfs_disk_key disk_key; | |
46 | struct btrfs_free_space_header *header; | |
47 | struct extent_buffer *leaf; | |
48 | struct inode *inode = NULL; | |
49 | int ret; | |
50 | ||
0af3d00b | 51 | key.objectid = BTRFS_FREE_SPACE_OBJECTID; |
0414efae | 52 | key.offset = offset; |
0af3d00b JB |
53 | key.type = 0; |
54 | ||
55 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
56 | if (ret < 0) | |
57 | return ERR_PTR(ret); | |
58 | if (ret > 0) { | |
b3b4aa74 | 59 | btrfs_release_path(path); |
0af3d00b JB |
60 | return ERR_PTR(-ENOENT); |
61 | } | |
62 | ||
63 | leaf = path->nodes[0]; | |
64 | header = btrfs_item_ptr(leaf, path->slots[0], | |
65 | struct btrfs_free_space_header); | |
66 | btrfs_free_space_key(leaf, header, &disk_key); | |
67 | btrfs_disk_key_to_cpu(&location, &disk_key); | |
b3b4aa74 | 68 | btrfs_release_path(path); |
0af3d00b JB |
69 | |
70 | inode = btrfs_iget(root->fs_info->sb, &location, root, NULL); | |
71 | if (!inode) | |
72 | return ERR_PTR(-ENOENT); | |
73 | if (IS_ERR(inode)) | |
74 | return inode; | |
75 | if (is_bad_inode(inode)) { | |
76 | iput(inode); | |
77 | return ERR_PTR(-ENOENT); | |
78 | } | |
79 | ||
528c0327 AV |
80 | mapping_set_gfp_mask(inode->i_mapping, |
81 | mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS); | |
adae52b9 | 82 | |
0414efae LZ |
83 | return inode; |
84 | } | |
85 | ||
86 | struct inode *lookup_free_space_inode(struct btrfs_root *root, | |
87 | struct btrfs_block_group_cache | |
88 | *block_group, struct btrfs_path *path) | |
89 | { | |
90 | struct inode *inode = NULL; | |
5b0e95bf | 91 | u32 flags = BTRFS_INODE_NODATASUM | BTRFS_INODE_NODATACOW; |
0414efae LZ |
92 | |
93 | spin_lock(&block_group->lock); | |
94 | if (block_group->inode) | |
95 | inode = igrab(block_group->inode); | |
96 | spin_unlock(&block_group->lock); | |
97 | if (inode) | |
98 | return inode; | |
99 | ||
100 | inode = __lookup_free_space_inode(root, path, | |
101 | block_group->key.objectid); | |
102 | if (IS_ERR(inode)) | |
103 | return inode; | |
104 | ||
0af3d00b | 105 | spin_lock(&block_group->lock); |
5b0e95bf | 106 | if (!((BTRFS_I(inode)->flags & flags) == flags)) { |
c2cf52eb SK |
107 | btrfs_info(root->fs_info, |
108 | "Old style space inode found, converting."); | |
5b0e95bf JB |
109 | BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM | |
110 | BTRFS_INODE_NODATACOW; | |
2f356126 JB |
111 | block_group->disk_cache_state = BTRFS_DC_CLEAR; |
112 | } | |
113 | ||
300e4f8a | 114 | if (!block_group->iref) { |
0af3d00b JB |
115 | block_group->inode = igrab(inode); |
116 | block_group->iref = 1; | |
117 | } | |
118 | spin_unlock(&block_group->lock); | |
119 | ||
120 | return inode; | |
121 | } | |
122 | ||
48a3b636 ES |
123 | static int __create_free_space_inode(struct btrfs_root *root, |
124 | struct btrfs_trans_handle *trans, | |
125 | struct btrfs_path *path, | |
126 | u64 ino, u64 offset) | |
0af3d00b JB |
127 | { |
128 | struct btrfs_key key; | |
129 | struct btrfs_disk_key disk_key; | |
130 | struct btrfs_free_space_header *header; | |
131 | struct btrfs_inode_item *inode_item; | |
132 | struct extent_buffer *leaf; | |
5b0e95bf | 133 | u64 flags = BTRFS_INODE_NOCOMPRESS | BTRFS_INODE_PREALLOC; |
0af3d00b JB |
134 | int ret; |
135 | ||
0414efae | 136 | ret = btrfs_insert_empty_inode(trans, root, path, ino); |
0af3d00b JB |
137 | if (ret) |
138 | return ret; | |
139 | ||
5b0e95bf JB |
140 | /* We inline crc's for the free disk space cache */ |
141 | if (ino != BTRFS_FREE_INO_OBJECTID) | |
142 | flags |= BTRFS_INODE_NODATASUM | BTRFS_INODE_NODATACOW; | |
143 | ||
0af3d00b JB |
144 | leaf = path->nodes[0]; |
145 | inode_item = btrfs_item_ptr(leaf, path->slots[0], | |
146 | struct btrfs_inode_item); | |
147 | btrfs_item_key(leaf, &disk_key, path->slots[0]); | |
148 | memset_extent_buffer(leaf, 0, (unsigned long)inode_item, | |
149 | sizeof(*inode_item)); | |
150 | btrfs_set_inode_generation(leaf, inode_item, trans->transid); | |
151 | btrfs_set_inode_size(leaf, inode_item, 0); | |
152 | btrfs_set_inode_nbytes(leaf, inode_item, 0); | |
153 | btrfs_set_inode_uid(leaf, inode_item, 0); | |
154 | btrfs_set_inode_gid(leaf, inode_item, 0); | |
155 | btrfs_set_inode_mode(leaf, inode_item, S_IFREG | 0600); | |
5b0e95bf | 156 | btrfs_set_inode_flags(leaf, inode_item, flags); |
0af3d00b JB |
157 | btrfs_set_inode_nlink(leaf, inode_item, 1); |
158 | btrfs_set_inode_transid(leaf, inode_item, trans->transid); | |
0414efae | 159 | btrfs_set_inode_block_group(leaf, inode_item, offset); |
0af3d00b | 160 | btrfs_mark_buffer_dirty(leaf); |
b3b4aa74 | 161 | btrfs_release_path(path); |
0af3d00b JB |
162 | |
163 | key.objectid = BTRFS_FREE_SPACE_OBJECTID; | |
0414efae | 164 | key.offset = offset; |
0af3d00b JB |
165 | key.type = 0; |
166 | ||
167 | ret = btrfs_insert_empty_item(trans, root, path, &key, | |
168 | sizeof(struct btrfs_free_space_header)); | |
169 | if (ret < 0) { | |
b3b4aa74 | 170 | btrfs_release_path(path); |
0af3d00b JB |
171 | return ret; |
172 | } | |
173 | leaf = path->nodes[0]; | |
174 | header = btrfs_item_ptr(leaf, path->slots[0], | |
175 | struct btrfs_free_space_header); | |
176 | memset_extent_buffer(leaf, 0, (unsigned long)header, sizeof(*header)); | |
177 | btrfs_set_free_space_key(leaf, header, &disk_key); | |
178 | btrfs_mark_buffer_dirty(leaf); | |
b3b4aa74 | 179 | btrfs_release_path(path); |
0af3d00b JB |
180 | |
181 | return 0; | |
182 | } | |
183 | ||
0414efae LZ |
184 | int create_free_space_inode(struct btrfs_root *root, |
185 | struct btrfs_trans_handle *trans, | |
186 | struct btrfs_block_group_cache *block_group, | |
187 | struct btrfs_path *path) | |
188 | { | |
189 | int ret; | |
190 | u64 ino; | |
191 | ||
192 | ret = btrfs_find_free_objectid(root, &ino); | |
193 | if (ret < 0) | |
194 | return ret; | |
195 | ||
196 | return __create_free_space_inode(root, trans, path, ino, | |
197 | block_group->key.objectid); | |
198 | } | |
199 | ||
7b61cd92 MX |
200 | int btrfs_check_trunc_cache_free_space(struct btrfs_root *root, |
201 | struct btrfs_block_rsv *rsv) | |
0af3d00b | 202 | { |
c8174313 | 203 | u64 needed_bytes; |
7b61cd92 | 204 | int ret; |
c8174313 JB |
205 | |
206 | /* 1 for slack space, 1 for updating the inode */ | |
207 | needed_bytes = btrfs_calc_trunc_metadata_size(root, 1) + | |
208 | btrfs_calc_trans_metadata_size(root, 1); | |
209 | ||
7b61cd92 MX |
210 | spin_lock(&rsv->lock); |
211 | if (rsv->reserved < needed_bytes) | |
212 | ret = -ENOSPC; | |
213 | else | |
214 | ret = 0; | |
215 | spin_unlock(&rsv->lock); | |
4b286cd1 | 216 | return ret; |
7b61cd92 MX |
217 | } |
218 | ||
219 | int btrfs_truncate_free_space_cache(struct btrfs_root *root, | |
220 | struct btrfs_trans_handle *trans, | |
7b61cd92 MX |
221 | struct inode *inode) |
222 | { | |
7b61cd92 | 223 | int ret = 0; |
0af3d00b | 224 | |
0af3d00b | 225 | btrfs_i_size_write(inode, 0); |
7caef267 | 226 | truncate_pagecache(inode, 0); |
0af3d00b JB |
227 | |
228 | /* | |
229 | * We don't need an orphan item because truncating the free space cache | |
230 | * will never be split across transactions. | |
231 | */ | |
232 | ret = btrfs_truncate_inode_items(trans, root, inode, | |
233 | 0, BTRFS_EXTENT_DATA_KEY); | |
234 | if (ret) { | |
79787eaa | 235 | btrfs_abort_transaction(trans, root, ret); |
0af3d00b JB |
236 | return ret; |
237 | } | |
238 | ||
82d5902d | 239 | ret = btrfs_update_inode(trans, root, inode); |
79787eaa JM |
240 | if (ret) |
241 | btrfs_abort_transaction(trans, root, ret); | |
c8174313 | 242 | |
82d5902d | 243 | return ret; |
0af3d00b JB |
244 | } |
245 | ||
9d66e233 JB |
246 | static int readahead_cache(struct inode *inode) |
247 | { | |
248 | struct file_ra_state *ra; | |
249 | unsigned long last_index; | |
250 | ||
251 | ra = kzalloc(sizeof(*ra), GFP_NOFS); | |
252 | if (!ra) | |
253 | return -ENOMEM; | |
254 | ||
255 | file_ra_state_init(ra, inode->i_mapping); | |
256 | last_index = (i_size_read(inode) - 1) >> PAGE_CACHE_SHIFT; | |
257 | ||
258 | page_cache_sync_readahead(inode->i_mapping, ra, NULL, 0, last_index); | |
259 | ||
260 | kfree(ra); | |
261 | ||
262 | return 0; | |
263 | } | |
264 | ||
a67509c3 JB |
265 | struct io_ctl { |
266 | void *cur, *orig; | |
267 | struct page *page; | |
268 | struct page **pages; | |
269 | struct btrfs_root *root; | |
270 | unsigned long size; | |
271 | int index; | |
272 | int num_pages; | |
5b0e95bf | 273 | unsigned check_crcs:1; |
a67509c3 JB |
274 | }; |
275 | ||
276 | static int io_ctl_init(struct io_ctl *io_ctl, struct inode *inode, | |
277 | struct btrfs_root *root) | |
278 | { | |
279 | memset(io_ctl, 0, sizeof(struct io_ctl)); | |
280 | io_ctl->num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> | |
281 | PAGE_CACHE_SHIFT; | |
282 | io_ctl->pages = kzalloc(sizeof(struct page *) * io_ctl->num_pages, | |
283 | GFP_NOFS); | |
284 | if (!io_ctl->pages) | |
285 | return -ENOMEM; | |
286 | io_ctl->root = root; | |
5b0e95bf JB |
287 | if (btrfs_ino(inode) != BTRFS_FREE_INO_OBJECTID) |
288 | io_ctl->check_crcs = 1; | |
a67509c3 JB |
289 | return 0; |
290 | } | |
291 | ||
292 | static void io_ctl_free(struct io_ctl *io_ctl) | |
293 | { | |
294 | kfree(io_ctl->pages); | |
295 | } | |
296 | ||
297 | static void io_ctl_unmap_page(struct io_ctl *io_ctl) | |
298 | { | |
299 | if (io_ctl->cur) { | |
300 | kunmap(io_ctl->page); | |
301 | io_ctl->cur = NULL; | |
302 | io_ctl->orig = NULL; | |
303 | } | |
304 | } | |
305 | ||
306 | static void io_ctl_map_page(struct io_ctl *io_ctl, int clear) | |
307 | { | |
b12d6869 | 308 | ASSERT(io_ctl->index < io_ctl->num_pages); |
a67509c3 JB |
309 | io_ctl->page = io_ctl->pages[io_ctl->index++]; |
310 | io_ctl->cur = kmap(io_ctl->page); | |
311 | io_ctl->orig = io_ctl->cur; | |
312 | io_ctl->size = PAGE_CACHE_SIZE; | |
313 | if (clear) | |
314 | memset(io_ctl->cur, 0, PAGE_CACHE_SIZE); | |
315 | } | |
316 | ||
317 | static void io_ctl_drop_pages(struct io_ctl *io_ctl) | |
318 | { | |
319 | int i; | |
320 | ||
321 | io_ctl_unmap_page(io_ctl); | |
322 | ||
323 | for (i = 0; i < io_ctl->num_pages; i++) { | |
a1ee5a45 LZ |
324 | if (io_ctl->pages[i]) { |
325 | ClearPageChecked(io_ctl->pages[i]); | |
326 | unlock_page(io_ctl->pages[i]); | |
327 | page_cache_release(io_ctl->pages[i]); | |
328 | } | |
a67509c3 JB |
329 | } |
330 | } | |
331 | ||
332 | static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct inode *inode, | |
333 | int uptodate) | |
334 | { | |
335 | struct page *page; | |
336 | gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); | |
337 | int i; | |
338 | ||
339 | for (i = 0; i < io_ctl->num_pages; i++) { | |
340 | page = find_or_create_page(inode->i_mapping, i, mask); | |
341 | if (!page) { | |
342 | io_ctl_drop_pages(io_ctl); | |
343 | return -ENOMEM; | |
344 | } | |
345 | io_ctl->pages[i] = page; | |
346 | if (uptodate && !PageUptodate(page)) { | |
347 | btrfs_readpage(NULL, page); | |
348 | lock_page(page); | |
349 | if (!PageUptodate(page)) { | |
350 | printk(KERN_ERR "btrfs: error reading free " | |
351 | "space cache\n"); | |
352 | io_ctl_drop_pages(io_ctl); | |
353 | return -EIO; | |
354 | } | |
355 | } | |
356 | } | |
357 | ||
f7d61dcd JB |
358 | for (i = 0; i < io_ctl->num_pages; i++) { |
359 | clear_page_dirty_for_io(io_ctl->pages[i]); | |
360 | set_page_extent_mapped(io_ctl->pages[i]); | |
361 | } | |
362 | ||
a67509c3 JB |
363 | return 0; |
364 | } | |
365 | ||
366 | static void io_ctl_set_generation(struct io_ctl *io_ctl, u64 generation) | |
367 | { | |
528c0327 | 368 | __le64 *val; |
a67509c3 JB |
369 | |
370 | io_ctl_map_page(io_ctl, 1); | |
371 | ||
372 | /* | |
5b0e95bf JB |
373 | * Skip the csum areas. If we don't check crcs then we just have a |
374 | * 64bit chunk at the front of the first page. | |
a67509c3 | 375 | */ |
5b0e95bf JB |
376 | if (io_ctl->check_crcs) { |
377 | io_ctl->cur += (sizeof(u32) * io_ctl->num_pages); | |
378 | io_ctl->size -= sizeof(u64) + (sizeof(u32) * io_ctl->num_pages); | |
379 | } else { | |
380 | io_ctl->cur += sizeof(u64); | |
381 | io_ctl->size -= sizeof(u64) * 2; | |
382 | } | |
a67509c3 JB |
383 | |
384 | val = io_ctl->cur; | |
385 | *val = cpu_to_le64(generation); | |
386 | io_ctl->cur += sizeof(u64); | |
a67509c3 JB |
387 | } |
388 | ||
389 | static int io_ctl_check_generation(struct io_ctl *io_ctl, u64 generation) | |
390 | { | |
528c0327 | 391 | __le64 *gen; |
a67509c3 | 392 | |
5b0e95bf JB |
393 | /* |
394 | * Skip the crc area. If we don't check crcs then we just have a 64bit | |
395 | * chunk at the front of the first page. | |
396 | */ | |
397 | if (io_ctl->check_crcs) { | |
398 | io_ctl->cur += sizeof(u32) * io_ctl->num_pages; | |
399 | io_ctl->size -= sizeof(u64) + | |
400 | (sizeof(u32) * io_ctl->num_pages); | |
401 | } else { | |
402 | io_ctl->cur += sizeof(u64); | |
403 | io_ctl->size -= sizeof(u64) * 2; | |
404 | } | |
a67509c3 | 405 | |
a67509c3 JB |
406 | gen = io_ctl->cur; |
407 | if (le64_to_cpu(*gen) != generation) { | |
408 | printk_ratelimited(KERN_ERR "btrfs: space cache generation " | |
409 | "(%Lu) does not match inode (%Lu)\n", *gen, | |
410 | generation); | |
411 | io_ctl_unmap_page(io_ctl); | |
412 | return -EIO; | |
413 | } | |
414 | io_ctl->cur += sizeof(u64); | |
5b0e95bf JB |
415 | return 0; |
416 | } | |
417 | ||
418 | static void io_ctl_set_crc(struct io_ctl *io_ctl, int index) | |
419 | { | |
420 | u32 *tmp; | |
421 | u32 crc = ~(u32)0; | |
422 | unsigned offset = 0; | |
423 | ||
424 | if (!io_ctl->check_crcs) { | |
425 | io_ctl_unmap_page(io_ctl); | |
426 | return; | |
427 | } | |
428 | ||
429 | if (index == 0) | |
cb54f257 | 430 | offset = sizeof(u32) * io_ctl->num_pages; |
5b0e95bf | 431 | |
b0496686 | 432 | crc = btrfs_csum_data(io_ctl->orig + offset, crc, |
5b0e95bf JB |
433 | PAGE_CACHE_SIZE - offset); |
434 | btrfs_csum_final(crc, (char *)&crc); | |
435 | io_ctl_unmap_page(io_ctl); | |
436 | tmp = kmap(io_ctl->pages[0]); | |
437 | tmp += index; | |
438 | *tmp = crc; | |
439 | kunmap(io_ctl->pages[0]); | |
440 | } | |
441 | ||
442 | static int io_ctl_check_crc(struct io_ctl *io_ctl, int index) | |
443 | { | |
444 | u32 *tmp, val; | |
445 | u32 crc = ~(u32)0; | |
446 | unsigned offset = 0; | |
447 | ||
448 | if (!io_ctl->check_crcs) { | |
449 | io_ctl_map_page(io_ctl, 0); | |
450 | return 0; | |
451 | } | |
452 | ||
453 | if (index == 0) | |
454 | offset = sizeof(u32) * io_ctl->num_pages; | |
455 | ||
456 | tmp = kmap(io_ctl->pages[0]); | |
457 | tmp += index; | |
458 | val = *tmp; | |
459 | kunmap(io_ctl->pages[0]); | |
460 | ||
461 | io_ctl_map_page(io_ctl, 0); | |
b0496686 | 462 | crc = btrfs_csum_data(io_ctl->orig + offset, crc, |
5b0e95bf JB |
463 | PAGE_CACHE_SIZE - offset); |
464 | btrfs_csum_final(crc, (char *)&crc); | |
465 | if (val != crc) { | |
466 | printk_ratelimited(KERN_ERR "btrfs: csum mismatch on free " | |
467 | "space cache\n"); | |
468 | io_ctl_unmap_page(io_ctl); | |
469 | return -EIO; | |
470 | } | |
471 | ||
a67509c3 JB |
472 | return 0; |
473 | } | |
474 | ||
475 | static int io_ctl_add_entry(struct io_ctl *io_ctl, u64 offset, u64 bytes, | |
476 | void *bitmap) | |
477 | { | |
478 | struct btrfs_free_space_entry *entry; | |
479 | ||
480 | if (!io_ctl->cur) | |
481 | return -ENOSPC; | |
482 | ||
483 | entry = io_ctl->cur; | |
484 | entry->offset = cpu_to_le64(offset); | |
485 | entry->bytes = cpu_to_le64(bytes); | |
486 | entry->type = (bitmap) ? BTRFS_FREE_SPACE_BITMAP : | |
487 | BTRFS_FREE_SPACE_EXTENT; | |
488 | io_ctl->cur += sizeof(struct btrfs_free_space_entry); | |
489 | io_ctl->size -= sizeof(struct btrfs_free_space_entry); | |
490 | ||
491 | if (io_ctl->size >= sizeof(struct btrfs_free_space_entry)) | |
492 | return 0; | |
493 | ||
5b0e95bf | 494 | io_ctl_set_crc(io_ctl, io_ctl->index - 1); |
a67509c3 JB |
495 | |
496 | /* No more pages to map */ | |
497 | if (io_ctl->index >= io_ctl->num_pages) | |
498 | return 0; | |
499 | ||
500 | /* map the next page */ | |
501 | io_ctl_map_page(io_ctl, 1); | |
502 | return 0; | |
503 | } | |
504 | ||
505 | static int io_ctl_add_bitmap(struct io_ctl *io_ctl, void *bitmap) | |
506 | { | |
507 | if (!io_ctl->cur) | |
508 | return -ENOSPC; | |
509 | ||
510 | /* | |
511 | * If we aren't at the start of the current page, unmap this one and | |
512 | * map the next one if there is any left. | |
513 | */ | |
514 | if (io_ctl->cur != io_ctl->orig) { | |
5b0e95bf | 515 | io_ctl_set_crc(io_ctl, io_ctl->index - 1); |
a67509c3 JB |
516 | if (io_ctl->index >= io_ctl->num_pages) |
517 | return -ENOSPC; | |
518 | io_ctl_map_page(io_ctl, 0); | |
519 | } | |
520 | ||
521 | memcpy(io_ctl->cur, bitmap, PAGE_CACHE_SIZE); | |
5b0e95bf | 522 | io_ctl_set_crc(io_ctl, io_ctl->index - 1); |
a67509c3 JB |
523 | if (io_ctl->index < io_ctl->num_pages) |
524 | io_ctl_map_page(io_ctl, 0); | |
525 | return 0; | |
526 | } | |
527 | ||
528 | static void io_ctl_zero_remaining_pages(struct io_ctl *io_ctl) | |
529 | { | |
5b0e95bf JB |
530 | /* |
531 | * If we're not on the boundary we know we've modified the page and we | |
532 | * need to crc the page. | |
533 | */ | |
534 | if (io_ctl->cur != io_ctl->orig) | |
535 | io_ctl_set_crc(io_ctl, io_ctl->index - 1); | |
536 | else | |
537 | io_ctl_unmap_page(io_ctl); | |
a67509c3 JB |
538 | |
539 | while (io_ctl->index < io_ctl->num_pages) { | |
540 | io_ctl_map_page(io_ctl, 1); | |
5b0e95bf | 541 | io_ctl_set_crc(io_ctl, io_ctl->index - 1); |
a67509c3 JB |
542 | } |
543 | } | |
544 | ||
5b0e95bf JB |
545 | static int io_ctl_read_entry(struct io_ctl *io_ctl, |
546 | struct btrfs_free_space *entry, u8 *type) | |
a67509c3 JB |
547 | { |
548 | struct btrfs_free_space_entry *e; | |
2f120c05 JB |
549 | int ret; |
550 | ||
551 | if (!io_ctl->cur) { | |
552 | ret = io_ctl_check_crc(io_ctl, io_ctl->index); | |
553 | if (ret) | |
554 | return ret; | |
555 | } | |
a67509c3 JB |
556 | |
557 | e = io_ctl->cur; | |
558 | entry->offset = le64_to_cpu(e->offset); | |
559 | entry->bytes = le64_to_cpu(e->bytes); | |
5b0e95bf | 560 | *type = e->type; |
a67509c3 JB |
561 | io_ctl->cur += sizeof(struct btrfs_free_space_entry); |
562 | io_ctl->size -= sizeof(struct btrfs_free_space_entry); | |
563 | ||
564 | if (io_ctl->size >= sizeof(struct btrfs_free_space_entry)) | |
5b0e95bf | 565 | return 0; |
a67509c3 JB |
566 | |
567 | io_ctl_unmap_page(io_ctl); | |
568 | ||
2f120c05 | 569 | return 0; |
a67509c3 JB |
570 | } |
571 | ||
5b0e95bf JB |
572 | static int io_ctl_read_bitmap(struct io_ctl *io_ctl, |
573 | struct btrfs_free_space *entry) | |
a67509c3 | 574 | { |
5b0e95bf JB |
575 | int ret; |
576 | ||
5b0e95bf JB |
577 | ret = io_ctl_check_crc(io_ctl, io_ctl->index); |
578 | if (ret) | |
579 | return ret; | |
580 | ||
a67509c3 JB |
581 | memcpy(entry->bitmap, io_ctl->cur, PAGE_CACHE_SIZE); |
582 | io_ctl_unmap_page(io_ctl); | |
5b0e95bf JB |
583 | |
584 | return 0; | |
a67509c3 JB |
585 | } |
586 | ||
cd023e7b JB |
587 | /* |
588 | * Since we attach pinned extents after the fact we can have contiguous sections | |
589 | * of free space that are split up in entries. This poses a problem with the | |
590 | * tree logging stuff since it could have allocated across what appears to be 2 | |
591 | * entries since we would have merged the entries when adding the pinned extents | |
592 | * back to the free space cache. So run through the space cache that we just | |
593 | * loaded and merge contiguous entries. This will make the log replay stuff not | |
594 | * blow up and it will make for nicer allocator behavior. | |
595 | */ | |
596 | static void merge_space_tree(struct btrfs_free_space_ctl *ctl) | |
597 | { | |
598 | struct btrfs_free_space *e, *prev = NULL; | |
599 | struct rb_node *n; | |
600 | ||
601 | again: | |
602 | spin_lock(&ctl->tree_lock); | |
603 | for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) { | |
604 | e = rb_entry(n, struct btrfs_free_space, offset_index); | |
605 | if (!prev) | |
606 | goto next; | |
607 | if (e->bitmap || prev->bitmap) | |
608 | goto next; | |
609 | if (prev->offset + prev->bytes == e->offset) { | |
610 | unlink_free_space(ctl, prev); | |
611 | unlink_free_space(ctl, e); | |
612 | prev->bytes += e->bytes; | |
613 | kmem_cache_free(btrfs_free_space_cachep, e); | |
614 | link_free_space(ctl, prev); | |
615 | prev = NULL; | |
616 | spin_unlock(&ctl->tree_lock); | |
617 | goto again; | |
618 | } | |
619 | next: | |
620 | prev = e; | |
621 | } | |
622 | spin_unlock(&ctl->tree_lock); | |
623 | } | |
624 | ||
48a3b636 ES |
625 | static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, |
626 | struct btrfs_free_space_ctl *ctl, | |
627 | struct btrfs_path *path, u64 offset) | |
9d66e233 | 628 | { |
9d66e233 JB |
629 | struct btrfs_free_space_header *header; |
630 | struct extent_buffer *leaf; | |
a67509c3 | 631 | struct io_ctl io_ctl; |
9d66e233 | 632 | struct btrfs_key key; |
a67509c3 | 633 | struct btrfs_free_space *e, *n; |
9d66e233 JB |
634 | struct list_head bitmaps; |
635 | u64 num_entries; | |
636 | u64 num_bitmaps; | |
637 | u64 generation; | |
a67509c3 | 638 | u8 type; |
f6a39829 | 639 | int ret = 0; |
9d66e233 JB |
640 | |
641 | INIT_LIST_HEAD(&bitmaps); | |
642 | ||
9d66e233 | 643 | /* Nothing in the space cache, goodbye */ |
0414efae | 644 | if (!i_size_read(inode)) |
a67509c3 | 645 | return 0; |
9d66e233 JB |
646 | |
647 | key.objectid = BTRFS_FREE_SPACE_OBJECTID; | |
0414efae | 648 | key.offset = offset; |
9d66e233 JB |
649 | key.type = 0; |
650 | ||
651 | ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); | |
0414efae | 652 | if (ret < 0) |
a67509c3 | 653 | return 0; |
0414efae | 654 | else if (ret > 0) { |
945d8962 | 655 | btrfs_release_path(path); |
a67509c3 | 656 | return 0; |
9d66e233 JB |
657 | } |
658 | ||
0414efae LZ |
659 | ret = -1; |
660 | ||
9d66e233 JB |
661 | leaf = path->nodes[0]; |
662 | header = btrfs_item_ptr(leaf, path->slots[0], | |
663 | struct btrfs_free_space_header); | |
664 | num_entries = btrfs_free_space_entries(leaf, header); | |
665 | num_bitmaps = btrfs_free_space_bitmaps(leaf, header); | |
666 | generation = btrfs_free_space_generation(leaf, header); | |
945d8962 | 667 | btrfs_release_path(path); |
9d66e233 JB |
668 | |
669 | if (BTRFS_I(inode)->generation != generation) { | |
c2cf52eb SK |
670 | btrfs_err(root->fs_info, |
671 | "free space inode generation (%llu) " | |
672 | "did not match free space cache generation (%llu)", | |
c1c9ff7c | 673 | BTRFS_I(inode)->generation, generation); |
a67509c3 | 674 | return 0; |
9d66e233 JB |
675 | } |
676 | ||
677 | if (!num_entries) | |
a67509c3 | 678 | return 0; |
9d66e233 | 679 | |
706efc66 LZ |
680 | ret = io_ctl_init(&io_ctl, inode, root); |
681 | if (ret) | |
682 | return ret; | |
683 | ||
9d66e233 | 684 | ret = readahead_cache(inode); |
0414efae | 685 | if (ret) |
9d66e233 | 686 | goto out; |
9d66e233 | 687 | |
a67509c3 JB |
688 | ret = io_ctl_prepare_pages(&io_ctl, inode, 1); |
689 | if (ret) | |
690 | goto out; | |
9d66e233 | 691 | |
5b0e95bf JB |
692 | ret = io_ctl_check_crc(&io_ctl, 0); |
693 | if (ret) | |
694 | goto free_cache; | |
695 | ||
a67509c3 JB |
696 | ret = io_ctl_check_generation(&io_ctl, generation); |
697 | if (ret) | |
698 | goto free_cache; | |
9d66e233 | 699 | |
a67509c3 JB |
700 | while (num_entries) { |
701 | e = kmem_cache_zalloc(btrfs_free_space_cachep, | |
702 | GFP_NOFS); | |
703 | if (!e) | |
9d66e233 | 704 | goto free_cache; |
9d66e233 | 705 | |
5b0e95bf JB |
706 | ret = io_ctl_read_entry(&io_ctl, e, &type); |
707 | if (ret) { | |
708 | kmem_cache_free(btrfs_free_space_cachep, e); | |
709 | goto free_cache; | |
710 | } | |
711 | ||
a67509c3 JB |
712 | if (!e->bytes) { |
713 | kmem_cache_free(btrfs_free_space_cachep, e); | |
714 | goto free_cache; | |
9d66e233 | 715 | } |
a67509c3 JB |
716 | |
717 | if (type == BTRFS_FREE_SPACE_EXTENT) { | |
718 | spin_lock(&ctl->tree_lock); | |
719 | ret = link_free_space(ctl, e); | |
720 | spin_unlock(&ctl->tree_lock); | |
721 | if (ret) { | |
c2cf52eb SK |
722 | btrfs_err(root->fs_info, |
723 | "Duplicate entries in free space cache, dumping"); | |
a67509c3 | 724 | kmem_cache_free(btrfs_free_space_cachep, e); |
9d66e233 JB |
725 | goto free_cache; |
726 | } | |
a67509c3 | 727 | } else { |
b12d6869 | 728 | ASSERT(num_bitmaps); |
a67509c3 JB |
729 | num_bitmaps--; |
730 | e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); | |
731 | if (!e->bitmap) { | |
732 | kmem_cache_free( | |
733 | btrfs_free_space_cachep, e); | |
9d66e233 JB |
734 | goto free_cache; |
735 | } | |
a67509c3 JB |
736 | spin_lock(&ctl->tree_lock); |
737 | ret = link_free_space(ctl, e); | |
738 | ctl->total_bitmaps++; | |
739 | ctl->op->recalc_thresholds(ctl); | |
740 | spin_unlock(&ctl->tree_lock); | |
741 | if (ret) { | |
c2cf52eb SK |
742 | btrfs_err(root->fs_info, |
743 | "Duplicate entries in free space cache, dumping"); | |
dc89e982 | 744 | kmem_cache_free(btrfs_free_space_cachep, e); |
9d66e233 JB |
745 | goto free_cache; |
746 | } | |
a67509c3 | 747 | list_add_tail(&e->list, &bitmaps); |
9d66e233 JB |
748 | } |
749 | ||
a67509c3 JB |
750 | num_entries--; |
751 | } | |
9d66e233 | 752 | |
2f120c05 JB |
753 | io_ctl_unmap_page(&io_ctl); |
754 | ||
a67509c3 JB |
755 | /* |
756 | * We add the bitmaps at the end of the entries in order that | |
757 | * the bitmap entries are added to the cache. | |
758 | */ | |
759 | list_for_each_entry_safe(e, n, &bitmaps, list) { | |
9d66e233 | 760 | list_del_init(&e->list); |
5b0e95bf JB |
761 | ret = io_ctl_read_bitmap(&io_ctl, e); |
762 | if (ret) | |
763 | goto free_cache; | |
9d66e233 JB |
764 | } |
765 | ||
a67509c3 | 766 | io_ctl_drop_pages(&io_ctl); |
cd023e7b | 767 | merge_space_tree(ctl); |
9d66e233 JB |
768 | ret = 1; |
769 | out: | |
a67509c3 | 770 | io_ctl_free(&io_ctl); |
9d66e233 | 771 | return ret; |
9d66e233 | 772 | free_cache: |
a67509c3 | 773 | io_ctl_drop_pages(&io_ctl); |
0414efae | 774 | __btrfs_remove_free_space_cache(ctl); |
9d66e233 JB |
775 | goto out; |
776 | } | |
777 | ||
0414efae LZ |
778 | int load_free_space_cache(struct btrfs_fs_info *fs_info, |
779 | struct btrfs_block_group_cache *block_group) | |
0cb59c99 | 780 | { |
34d52cb6 | 781 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
0414efae LZ |
782 | struct btrfs_root *root = fs_info->tree_root; |
783 | struct inode *inode; | |
784 | struct btrfs_path *path; | |
5b0e95bf | 785 | int ret = 0; |
0414efae LZ |
786 | bool matched; |
787 | u64 used = btrfs_block_group_used(&block_group->item); | |
788 | ||
0414efae LZ |
789 | /* |
790 | * If this block group has been marked to be cleared for one reason or | |
791 | * another then we can't trust the on disk cache, so just return. | |
792 | */ | |
9d66e233 | 793 | spin_lock(&block_group->lock); |
0414efae LZ |
794 | if (block_group->disk_cache_state != BTRFS_DC_WRITTEN) { |
795 | spin_unlock(&block_group->lock); | |
796 | return 0; | |
797 | } | |
9d66e233 | 798 | spin_unlock(&block_group->lock); |
0414efae LZ |
799 | |
800 | path = btrfs_alloc_path(); | |
801 | if (!path) | |
802 | return 0; | |
d53ba474 JB |
803 | path->search_commit_root = 1; |
804 | path->skip_locking = 1; | |
0414efae LZ |
805 | |
806 | inode = lookup_free_space_inode(root, block_group, path); | |
807 | if (IS_ERR(inode)) { | |
808 | btrfs_free_path(path); | |
809 | return 0; | |
810 | } | |
811 | ||
5b0e95bf JB |
812 | /* We may have converted the inode and made the cache invalid. */ |
813 | spin_lock(&block_group->lock); | |
814 | if (block_group->disk_cache_state != BTRFS_DC_WRITTEN) { | |
815 | spin_unlock(&block_group->lock); | |
a7e221e9 | 816 | btrfs_free_path(path); |
5b0e95bf JB |
817 | goto out; |
818 | } | |
819 | spin_unlock(&block_group->lock); | |
820 | ||
0414efae LZ |
821 | ret = __load_free_space_cache(fs_info->tree_root, inode, ctl, |
822 | path, block_group->key.objectid); | |
823 | btrfs_free_path(path); | |
824 | if (ret <= 0) | |
825 | goto out; | |
826 | ||
827 | spin_lock(&ctl->tree_lock); | |
828 | matched = (ctl->free_space == (block_group->key.offset - used - | |
829 | block_group->bytes_super)); | |
830 | spin_unlock(&ctl->tree_lock); | |
831 | ||
832 | if (!matched) { | |
833 | __btrfs_remove_free_space_cache(ctl); | |
c2cf52eb SK |
834 | btrfs_err(fs_info, "block group %llu has wrong amount of free space", |
835 | block_group->key.objectid); | |
0414efae LZ |
836 | ret = -1; |
837 | } | |
838 | out: | |
839 | if (ret < 0) { | |
840 | /* This cache is bogus, make sure it gets cleared */ | |
841 | spin_lock(&block_group->lock); | |
842 | block_group->disk_cache_state = BTRFS_DC_CLEAR; | |
843 | spin_unlock(&block_group->lock); | |
82d5902d | 844 | ret = 0; |
0414efae | 845 | |
c2cf52eb SK |
846 | btrfs_err(fs_info, "failed to load free space cache for block group %llu", |
847 | block_group->key.objectid); | |
0414efae LZ |
848 | } |
849 | ||
850 | iput(inode); | |
851 | return ret; | |
9d66e233 JB |
852 | } |
853 | ||
c09544e0 JB |
854 | /** |
855 | * __btrfs_write_out_cache - write out cached info to an inode | |
856 | * @root - the root the inode belongs to | |
857 | * @ctl - the free space cache we are going to write out | |
858 | * @block_group - the block_group for this cache if it belongs to a block_group | |
859 | * @trans - the trans handle | |
860 | * @path - the path to use | |
861 | * @offset - the offset for the key we'll insert | |
862 | * | |
863 | * This function writes out a free space cache struct to disk for quick recovery | |
864 | * on mount. This will return 0 if it was successfull in writing the cache out, | |
865 | * and -1 if it was not. | |
866 | */ | |
48a3b636 ES |
867 | static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, |
868 | struct btrfs_free_space_ctl *ctl, | |
869 | struct btrfs_block_group_cache *block_group, | |
870 | struct btrfs_trans_handle *trans, | |
871 | struct btrfs_path *path, u64 offset) | |
0cb59c99 JB |
872 | { |
873 | struct btrfs_free_space_header *header; | |
874 | struct extent_buffer *leaf; | |
0cb59c99 JB |
875 | struct rb_node *node; |
876 | struct list_head *pos, *n; | |
0cb59c99 | 877 | struct extent_state *cached_state = NULL; |
43be2146 JB |
878 | struct btrfs_free_cluster *cluster = NULL; |
879 | struct extent_io_tree *unpin = NULL; | |
a67509c3 | 880 | struct io_ctl io_ctl; |
0cb59c99 JB |
881 | struct list_head bitmap_list; |
882 | struct btrfs_key key; | |
db804f23 | 883 | u64 start, extent_start, extent_end, len; |
0cb59c99 JB |
884 | int entries = 0; |
885 | int bitmaps = 0; | |
c09544e0 JB |
886 | int ret; |
887 | int err = -1; | |
0cb59c99 | 888 | |
0cb59c99 JB |
889 | INIT_LIST_HEAD(&bitmap_list); |
890 | ||
0414efae LZ |
891 | if (!i_size_read(inode)) |
892 | return -1; | |
2b20982e | 893 | |
706efc66 LZ |
894 | ret = io_ctl_init(&io_ctl, inode, root); |
895 | if (ret) | |
896 | return -1; | |
be1a12a0 | 897 | |
43be2146 | 898 | /* Get the cluster for this block_group if it exists */ |
0414efae | 899 | if (block_group && !list_empty(&block_group->cluster_list)) |
43be2146 JB |
900 | cluster = list_entry(block_group->cluster_list.next, |
901 | struct btrfs_free_cluster, | |
902 | block_group_list); | |
903 | ||
a67509c3 JB |
904 | /* Lock all pages first so we can lock the extent safely. */ |
905 | io_ctl_prepare_pages(&io_ctl, inode, 0); | |
0cb59c99 | 906 | |
0cb59c99 | 907 | lock_extent_bits(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, |
d0082371 | 908 | 0, &cached_state); |
0cb59c99 | 909 | |
f75b130e JB |
910 | node = rb_first(&ctl->free_space_offset); |
911 | if (!node && cluster) { | |
912 | node = rb_first(&cluster->root); | |
913 | cluster = NULL; | |
914 | } | |
915 | ||
5b0e95bf JB |
916 | /* Make sure we can fit our crcs into the first page */ |
917 | if (io_ctl.check_crcs && | |
73e1e61f | 918 | (io_ctl.num_pages * sizeof(u32)) >= PAGE_CACHE_SIZE) |
5b0e95bf | 919 | goto out_nospc; |
5b0e95bf | 920 | |
a67509c3 | 921 | io_ctl_set_generation(&io_ctl, trans->transid); |
43be2146 | 922 | |
a67509c3 JB |
923 | /* Write out the extent entries */ |
924 | while (node) { | |
925 | struct btrfs_free_space *e; | |
0cb59c99 | 926 | |
a67509c3 JB |
927 | e = rb_entry(node, struct btrfs_free_space, offset_index); |
928 | entries++; | |
0cb59c99 | 929 | |
a67509c3 JB |
930 | ret = io_ctl_add_entry(&io_ctl, e->offset, e->bytes, |
931 | e->bitmap); | |
932 | if (ret) | |
933 | goto out_nospc; | |
2f356126 | 934 | |
a67509c3 JB |
935 | if (e->bitmap) { |
936 | list_add_tail(&e->list, &bitmap_list); | |
937 | bitmaps++; | |
2f356126 | 938 | } |
a67509c3 JB |
939 | node = rb_next(node); |
940 | if (!node && cluster) { | |
941 | node = rb_first(&cluster->root); | |
942 | cluster = NULL; | |
43be2146 | 943 | } |
a67509c3 | 944 | } |
43be2146 | 945 | |
a67509c3 JB |
946 | /* |
947 | * We want to add any pinned extents to our free space cache | |
948 | * so we don't leak the space | |
949 | */ | |
db804f23 LZ |
950 | |
951 | /* | |
952 | * We shouldn't have switched the pinned extents yet so this is the | |
953 | * right one | |
954 | */ | |
955 | unpin = root->fs_info->pinned_extents; | |
956 | ||
957 | if (block_group) | |
958 | start = block_group->key.objectid; | |
959 | ||
a67509c3 JB |
960 | while (block_group && (start < block_group->key.objectid + |
961 | block_group->key.offset)) { | |
db804f23 LZ |
962 | ret = find_first_extent_bit(unpin, start, |
963 | &extent_start, &extent_end, | |
e6138876 | 964 | EXTENT_DIRTY, NULL); |
a67509c3 JB |
965 | if (ret) { |
966 | ret = 0; | |
967 | break; | |
0cb59c99 | 968 | } |
0cb59c99 | 969 | |
a67509c3 | 970 | /* This pinned extent is out of our range */ |
db804f23 | 971 | if (extent_start >= block_group->key.objectid + |
a67509c3 JB |
972 | block_group->key.offset) |
973 | break; | |
2f356126 | 974 | |
db804f23 LZ |
975 | extent_start = max(extent_start, start); |
976 | extent_end = min(block_group->key.objectid + | |
977 | block_group->key.offset, extent_end + 1); | |
978 | len = extent_end - extent_start; | |
0cb59c99 | 979 | |
a67509c3 | 980 | entries++; |
db804f23 | 981 | ret = io_ctl_add_entry(&io_ctl, extent_start, len, NULL); |
a67509c3 JB |
982 | if (ret) |
983 | goto out_nospc; | |
0cb59c99 | 984 | |
db804f23 | 985 | start = extent_end; |
a67509c3 | 986 | } |
0cb59c99 JB |
987 | |
988 | /* Write out the bitmaps */ | |
989 | list_for_each_safe(pos, n, &bitmap_list) { | |
0cb59c99 JB |
990 | struct btrfs_free_space *entry = |
991 | list_entry(pos, struct btrfs_free_space, list); | |
992 | ||
a67509c3 JB |
993 | ret = io_ctl_add_bitmap(&io_ctl, entry->bitmap); |
994 | if (ret) | |
995 | goto out_nospc; | |
0cb59c99 | 996 | list_del_init(&entry->list); |
be1a12a0 JB |
997 | } |
998 | ||
0cb59c99 | 999 | /* Zero out the rest of the pages just to make sure */ |
a67509c3 | 1000 | io_ctl_zero_remaining_pages(&io_ctl); |
0cb59c99 | 1001 | |
a67509c3 JB |
1002 | ret = btrfs_dirty_pages(root, inode, io_ctl.pages, io_ctl.num_pages, |
1003 | 0, i_size_read(inode), &cached_state); | |
1004 | io_ctl_drop_pages(&io_ctl); | |
0cb59c99 JB |
1005 | unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, |
1006 | i_size_read(inode) - 1, &cached_state, GFP_NOFS); | |
1007 | ||
c09544e0 | 1008 | if (ret) |
2f356126 | 1009 | goto out; |
be1a12a0 | 1010 | |
be1a12a0 | 1011 | |
5fd02043 | 1012 | btrfs_wait_ordered_range(inode, 0, (u64)-1); |
0cb59c99 JB |
1013 | |
1014 | key.objectid = BTRFS_FREE_SPACE_OBJECTID; | |
0414efae | 1015 | key.offset = offset; |
0cb59c99 JB |
1016 | key.type = 0; |
1017 | ||
a9b5fcdd | 1018 | ret = btrfs_search_slot(trans, root, &key, path, 0, 1); |
0cb59c99 | 1019 | if (ret < 0) { |
a67509c3 | 1020 | clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, inode->i_size - 1, |
5b0e95bf JB |
1021 | EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, NULL, |
1022 | GFP_NOFS); | |
2f356126 | 1023 | goto out; |
0cb59c99 JB |
1024 | } |
1025 | leaf = path->nodes[0]; | |
1026 | if (ret > 0) { | |
1027 | struct btrfs_key found_key; | |
b12d6869 | 1028 | ASSERT(path->slots[0]); |
0cb59c99 JB |
1029 | path->slots[0]--; |
1030 | btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); | |
1031 | if (found_key.objectid != BTRFS_FREE_SPACE_OBJECTID || | |
0414efae | 1032 | found_key.offset != offset) { |
a67509c3 JB |
1033 | clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, |
1034 | inode->i_size - 1, | |
5b0e95bf JB |
1035 | EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, |
1036 | NULL, GFP_NOFS); | |
b3b4aa74 | 1037 | btrfs_release_path(path); |
2f356126 | 1038 | goto out; |
0cb59c99 JB |
1039 | } |
1040 | } | |
549b4fdb JB |
1041 | |
1042 | BTRFS_I(inode)->generation = trans->transid; | |
0cb59c99 JB |
1043 | header = btrfs_item_ptr(leaf, path->slots[0], |
1044 | struct btrfs_free_space_header); | |
1045 | btrfs_set_free_space_entries(leaf, header, entries); | |
1046 | btrfs_set_free_space_bitmaps(leaf, header, bitmaps); | |
1047 | btrfs_set_free_space_generation(leaf, header, trans->transid); | |
1048 | btrfs_mark_buffer_dirty(leaf); | |
b3b4aa74 | 1049 | btrfs_release_path(path); |
0cb59c99 | 1050 | |
c09544e0 | 1051 | err = 0; |
2f356126 | 1052 | out: |
a67509c3 | 1053 | io_ctl_free(&io_ctl); |
c09544e0 | 1054 | if (err) { |
a67509c3 | 1055 | invalidate_inode_pages2(inode->i_mapping); |
0cb59c99 JB |
1056 | BTRFS_I(inode)->generation = 0; |
1057 | } | |
0cb59c99 | 1058 | btrfs_update_inode(trans, root, inode); |
c09544e0 | 1059 | return err; |
a67509c3 JB |
1060 | |
1061 | out_nospc: | |
1062 | list_for_each_safe(pos, n, &bitmap_list) { | |
1063 | struct btrfs_free_space *entry = | |
1064 | list_entry(pos, struct btrfs_free_space, list); | |
1065 | list_del_init(&entry->list); | |
1066 | } | |
1067 | io_ctl_drop_pages(&io_ctl); | |
1068 | unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, | |
1069 | i_size_read(inode) - 1, &cached_state, GFP_NOFS); | |
1070 | goto out; | |
0414efae LZ |
1071 | } |
1072 | ||
1073 | int btrfs_write_out_cache(struct btrfs_root *root, | |
1074 | struct btrfs_trans_handle *trans, | |
1075 | struct btrfs_block_group_cache *block_group, | |
1076 | struct btrfs_path *path) | |
1077 | { | |
1078 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; | |
1079 | struct inode *inode; | |
1080 | int ret = 0; | |
1081 | ||
1082 | root = root->fs_info->tree_root; | |
1083 | ||
1084 | spin_lock(&block_group->lock); | |
1085 | if (block_group->disk_cache_state < BTRFS_DC_SETUP) { | |
1086 | spin_unlock(&block_group->lock); | |
1087 | return 0; | |
1088 | } | |
1089 | spin_unlock(&block_group->lock); | |
1090 | ||
1091 | inode = lookup_free_space_inode(root, block_group, path); | |
1092 | if (IS_ERR(inode)) | |
1093 | return 0; | |
1094 | ||
1095 | ret = __btrfs_write_out_cache(root, inode, ctl, block_group, trans, | |
1096 | path, block_group->key.objectid); | |
c09544e0 | 1097 | if (ret) { |
0414efae LZ |
1098 | spin_lock(&block_group->lock); |
1099 | block_group->disk_cache_state = BTRFS_DC_ERROR; | |
1100 | spin_unlock(&block_group->lock); | |
82d5902d | 1101 | ret = 0; |
c09544e0 | 1102 | #ifdef DEBUG |
c2cf52eb SK |
1103 | btrfs_err(root->fs_info, |
1104 | "failed to write free space cache for block group %llu", | |
1105 | block_group->key.objectid); | |
c09544e0 | 1106 | #endif |
0414efae LZ |
1107 | } |
1108 | ||
0cb59c99 JB |
1109 | iput(inode); |
1110 | return ret; | |
1111 | } | |
1112 | ||
34d52cb6 | 1113 | static inline unsigned long offset_to_bit(u64 bitmap_start, u32 unit, |
96303081 | 1114 | u64 offset) |
0f9dd46c | 1115 | { |
b12d6869 | 1116 | ASSERT(offset >= bitmap_start); |
96303081 | 1117 | offset -= bitmap_start; |
34d52cb6 | 1118 | return (unsigned long)(div_u64(offset, unit)); |
96303081 | 1119 | } |
0f9dd46c | 1120 | |
34d52cb6 | 1121 | static inline unsigned long bytes_to_bits(u64 bytes, u32 unit) |
96303081 | 1122 | { |
34d52cb6 | 1123 | return (unsigned long)(div_u64(bytes, unit)); |
96303081 | 1124 | } |
0f9dd46c | 1125 | |
34d52cb6 | 1126 | static inline u64 offset_to_bitmap(struct btrfs_free_space_ctl *ctl, |
96303081 JB |
1127 | u64 offset) |
1128 | { | |
1129 | u64 bitmap_start; | |
1130 | u64 bytes_per_bitmap; | |
0f9dd46c | 1131 | |
34d52cb6 LZ |
1132 | bytes_per_bitmap = BITS_PER_BITMAP * ctl->unit; |
1133 | bitmap_start = offset - ctl->start; | |
96303081 JB |
1134 | bitmap_start = div64_u64(bitmap_start, bytes_per_bitmap); |
1135 | bitmap_start *= bytes_per_bitmap; | |
34d52cb6 | 1136 | bitmap_start += ctl->start; |
0f9dd46c | 1137 | |
96303081 | 1138 | return bitmap_start; |
0f9dd46c JB |
1139 | } |
1140 | ||
96303081 JB |
1141 | static int tree_insert_offset(struct rb_root *root, u64 offset, |
1142 | struct rb_node *node, int bitmap) | |
0f9dd46c JB |
1143 | { |
1144 | struct rb_node **p = &root->rb_node; | |
1145 | struct rb_node *parent = NULL; | |
1146 | struct btrfs_free_space *info; | |
1147 | ||
1148 | while (*p) { | |
1149 | parent = *p; | |
96303081 | 1150 | info = rb_entry(parent, struct btrfs_free_space, offset_index); |
0f9dd46c | 1151 | |
96303081 | 1152 | if (offset < info->offset) { |
0f9dd46c | 1153 | p = &(*p)->rb_left; |
96303081 | 1154 | } else if (offset > info->offset) { |
0f9dd46c | 1155 | p = &(*p)->rb_right; |
96303081 JB |
1156 | } else { |
1157 | /* | |
1158 | * we could have a bitmap entry and an extent entry | |
1159 | * share the same offset. If this is the case, we want | |
1160 | * the extent entry to always be found first if we do a | |
1161 | * linear search through the tree, since we want to have | |
1162 | * the quickest allocation time, and allocating from an | |
1163 | * extent is faster than allocating from a bitmap. So | |
1164 | * if we're inserting a bitmap and we find an entry at | |
1165 | * this offset, we want to go right, or after this entry | |
1166 | * logically. If we are inserting an extent and we've | |
1167 | * found a bitmap, we want to go left, or before | |
1168 | * logically. | |
1169 | */ | |
1170 | if (bitmap) { | |
207dde82 JB |
1171 | if (info->bitmap) { |
1172 | WARN_ON_ONCE(1); | |
1173 | return -EEXIST; | |
1174 | } | |
96303081 JB |
1175 | p = &(*p)->rb_right; |
1176 | } else { | |
207dde82 JB |
1177 | if (!info->bitmap) { |
1178 | WARN_ON_ONCE(1); | |
1179 | return -EEXIST; | |
1180 | } | |
96303081 JB |
1181 | p = &(*p)->rb_left; |
1182 | } | |
1183 | } | |
0f9dd46c JB |
1184 | } |
1185 | ||
1186 | rb_link_node(node, parent, p); | |
1187 | rb_insert_color(node, root); | |
1188 | ||
1189 | return 0; | |
1190 | } | |
1191 | ||
1192 | /* | |
70cb0743 JB |
1193 | * searches the tree for the given offset. |
1194 | * | |
96303081 JB |
1195 | * fuzzy - If this is set, then we are trying to make an allocation, and we just |
1196 | * want a section that has at least bytes size and comes at or after the given | |
1197 | * offset. | |
0f9dd46c | 1198 | */ |
96303081 | 1199 | static struct btrfs_free_space * |
34d52cb6 | 1200 | tree_search_offset(struct btrfs_free_space_ctl *ctl, |
96303081 | 1201 | u64 offset, int bitmap_only, int fuzzy) |
0f9dd46c | 1202 | { |
34d52cb6 | 1203 | struct rb_node *n = ctl->free_space_offset.rb_node; |
96303081 JB |
1204 | struct btrfs_free_space *entry, *prev = NULL; |
1205 | ||
1206 | /* find entry that is closest to the 'offset' */ | |
1207 | while (1) { | |
1208 | if (!n) { | |
1209 | entry = NULL; | |
1210 | break; | |
1211 | } | |
0f9dd46c | 1212 | |
0f9dd46c | 1213 | entry = rb_entry(n, struct btrfs_free_space, offset_index); |
96303081 | 1214 | prev = entry; |
0f9dd46c | 1215 | |
96303081 | 1216 | if (offset < entry->offset) |
0f9dd46c | 1217 | n = n->rb_left; |
96303081 | 1218 | else if (offset > entry->offset) |
0f9dd46c | 1219 | n = n->rb_right; |
96303081 | 1220 | else |
0f9dd46c | 1221 | break; |
0f9dd46c JB |
1222 | } |
1223 | ||
96303081 JB |
1224 | if (bitmap_only) { |
1225 | if (!entry) | |
1226 | return NULL; | |
1227 | if (entry->bitmap) | |
1228 | return entry; | |
0f9dd46c | 1229 | |
96303081 JB |
1230 | /* |
1231 | * bitmap entry and extent entry may share same offset, | |
1232 | * in that case, bitmap entry comes after extent entry. | |
1233 | */ | |
1234 | n = rb_next(n); | |
1235 | if (!n) | |
1236 | return NULL; | |
1237 | entry = rb_entry(n, struct btrfs_free_space, offset_index); | |
1238 | if (entry->offset != offset) | |
1239 | return NULL; | |
0f9dd46c | 1240 | |
96303081 JB |
1241 | WARN_ON(!entry->bitmap); |
1242 | return entry; | |
1243 | } else if (entry) { | |
1244 | if (entry->bitmap) { | |
0f9dd46c | 1245 | /* |
96303081 JB |
1246 | * if previous extent entry covers the offset, |
1247 | * we should return it instead of the bitmap entry | |
0f9dd46c | 1248 | */ |
de6c4115 MX |
1249 | n = rb_prev(&entry->offset_index); |
1250 | if (n) { | |
96303081 JB |
1251 | prev = rb_entry(n, struct btrfs_free_space, |
1252 | offset_index); | |
de6c4115 MX |
1253 | if (!prev->bitmap && |
1254 | prev->offset + prev->bytes > offset) | |
1255 | entry = prev; | |
0f9dd46c | 1256 | } |
96303081 JB |
1257 | } |
1258 | return entry; | |
1259 | } | |
1260 | ||
1261 | if (!prev) | |
1262 | return NULL; | |
1263 | ||
1264 | /* find last entry before the 'offset' */ | |
1265 | entry = prev; | |
1266 | if (entry->offset > offset) { | |
1267 | n = rb_prev(&entry->offset_index); | |
1268 | if (n) { | |
1269 | entry = rb_entry(n, struct btrfs_free_space, | |
1270 | offset_index); | |
b12d6869 | 1271 | ASSERT(entry->offset <= offset); |
0f9dd46c | 1272 | } else { |
96303081 JB |
1273 | if (fuzzy) |
1274 | return entry; | |
1275 | else | |
1276 | return NULL; | |
0f9dd46c JB |
1277 | } |
1278 | } | |
1279 | ||
96303081 | 1280 | if (entry->bitmap) { |
de6c4115 MX |
1281 | n = rb_prev(&entry->offset_index); |
1282 | if (n) { | |
96303081 JB |
1283 | prev = rb_entry(n, struct btrfs_free_space, |
1284 | offset_index); | |
de6c4115 MX |
1285 | if (!prev->bitmap && |
1286 | prev->offset + prev->bytes > offset) | |
1287 | return prev; | |
96303081 | 1288 | } |
34d52cb6 | 1289 | if (entry->offset + BITS_PER_BITMAP * ctl->unit > offset) |
96303081 JB |
1290 | return entry; |
1291 | } else if (entry->offset + entry->bytes > offset) | |
1292 | return entry; | |
1293 | ||
1294 | if (!fuzzy) | |
1295 | return NULL; | |
1296 | ||
1297 | while (1) { | |
1298 | if (entry->bitmap) { | |
1299 | if (entry->offset + BITS_PER_BITMAP * | |
34d52cb6 | 1300 | ctl->unit > offset) |
96303081 JB |
1301 | break; |
1302 | } else { | |
1303 | if (entry->offset + entry->bytes > offset) | |
1304 | break; | |
1305 | } | |
1306 | ||
1307 | n = rb_next(&entry->offset_index); | |
1308 | if (!n) | |
1309 | return NULL; | |
1310 | entry = rb_entry(n, struct btrfs_free_space, offset_index); | |
1311 | } | |
1312 | return entry; | |
0f9dd46c JB |
1313 | } |
1314 | ||
f333adb5 | 1315 | static inline void |
34d52cb6 | 1316 | __unlink_free_space(struct btrfs_free_space_ctl *ctl, |
f333adb5 | 1317 | struct btrfs_free_space *info) |
0f9dd46c | 1318 | { |
34d52cb6 LZ |
1319 | rb_erase(&info->offset_index, &ctl->free_space_offset); |
1320 | ctl->free_extents--; | |
f333adb5 LZ |
1321 | } |
1322 | ||
34d52cb6 | 1323 | static void unlink_free_space(struct btrfs_free_space_ctl *ctl, |
f333adb5 LZ |
1324 | struct btrfs_free_space *info) |
1325 | { | |
34d52cb6 LZ |
1326 | __unlink_free_space(ctl, info); |
1327 | ctl->free_space -= info->bytes; | |
0f9dd46c JB |
1328 | } |
1329 | ||
34d52cb6 | 1330 | static int link_free_space(struct btrfs_free_space_ctl *ctl, |
0f9dd46c JB |
1331 | struct btrfs_free_space *info) |
1332 | { | |
1333 | int ret = 0; | |
1334 | ||
b12d6869 | 1335 | ASSERT(info->bytes || info->bitmap); |
34d52cb6 | 1336 | ret = tree_insert_offset(&ctl->free_space_offset, info->offset, |
96303081 | 1337 | &info->offset_index, (info->bitmap != NULL)); |
0f9dd46c JB |
1338 | if (ret) |
1339 | return ret; | |
1340 | ||
34d52cb6 LZ |
1341 | ctl->free_space += info->bytes; |
1342 | ctl->free_extents++; | |
96303081 JB |
1343 | return ret; |
1344 | } | |
1345 | ||
34d52cb6 | 1346 | static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl) |
96303081 | 1347 | { |
34d52cb6 | 1348 | struct btrfs_block_group_cache *block_group = ctl->private; |
25891f79 JB |
1349 | u64 max_bytes; |
1350 | u64 bitmap_bytes; | |
1351 | u64 extent_bytes; | |
8eb2d829 | 1352 | u64 size = block_group->key.offset; |
96009762 | 1353 | u64 bytes_per_bg = BITS_PER_BITMAP * ctl->unit; |
34d52cb6 LZ |
1354 | int max_bitmaps = div64_u64(size + bytes_per_bg - 1, bytes_per_bg); |
1355 | ||
dde5740f JB |
1356 | max_bitmaps = max(max_bitmaps, 1); |
1357 | ||
b12d6869 | 1358 | ASSERT(ctl->total_bitmaps <= max_bitmaps); |
96303081 JB |
1359 | |
1360 | /* | |
1361 | * The goal is to keep the total amount of memory used per 1gb of space | |
1362 | * at or below 32k, so we need to adjust how much memory we allow to be | |
1363 | * used by extent based free space tracking | |
1364 | */ | |
8eb2d829 LZ |
1365 | if (size < 1024 * 1024 * 1024) |
1366 | max_bytes = MAX_CACHE_BYTES_PER_GIG; | |
1367 | else | |
1368 | max_bytes = MAX_CACHE_BYTES_PER_GIG * | |
1369 | div64_u64(size, 1024 * 1024 * 1024); | |
96303081 | 1370 | |
25891f79 JB |
1371 | /* |
1372 | * we want to account for 1 more bitmap than what we have so we can make | |
1373 | * sure we don't go over our overall goal of MAX_CACHE_BYTES_PER_GIG as | |
1374 | * we add more bitmaps. | |
1375 | */ | |
34d52cb6 | 1376 | bitmap_bytes = (ctl->total_bitmaps + 1) * PAGE_CACHE_SIZE; |
96303081 | 1377 | |
25891f79 | 1378 | if (bitmap_bytes >= max_bytes) { |
34d52cb6 | 1379 | ctl->extents_thresh = 0; |
25891f79 JB |
1380 | return; |
1381 | } | |
96303081 | 1382 | |
25891f79 JB |
1383 | /* |
1384 | * we want the extent entry threshold to always be at most 1/2 the maxw | |
1385 | * bytes we can have, or whatever is less than that. | |
1386 | */ | |
1387 | extent_bytes = max_bytes - bitmap_bytes; | |
1388 | extent_bytes = min_t(u64, extent_bytes, div64_u64(max_bytes, 2)); | |
96303081 | 1389 | |
34d52cb6 | 1390 | ctl->extents_thresh = |
25891f79 | 1391 | div64_u64(extent_bytes, (sizeof(struct btrfs_free_space))); |
96303081 JB |
1392 | } |
1393 | ||
bb3ac5a4 MX |
1394 | static inline void __bitmap_clear_bits(struct btrfs_free_space_ctl *ctl, |
1395 | struct btrfs_free_space *info, | |
1396 | u64 offset, u64 bytes) | |
96303081 | 1397 | { |
f38b6e75 | 1398 | unsigned long start, count; |
96303081 | 1399 | |
34d52cb6 LZ |
1400 | start = offset_to_bit(info->offset, ctl->unit, offset); |
1401 | count = bytes_to_bits(bytes, ctl->unit); | |
b12d6869 | 1402 | ASSERT(start + count <= BITS_PER_BITMAP); |
96303081 | 1403 | |
f38b6e75 | 1404 | bitmap_clear(info->bitmap, start, count); |
96303081 JB |
1405 | |
1406 | info->bytes -= bytes; | |
bb3ac5a4 MX |
1407 | } |
1408 | ||
1409 | static void bitmap_clear_bits(struct btrfs_free_space_ctl *ctl, | |
1410 | struct btrfs_free_space *info, u64 offset, | |
1411 | u64 bytes) | |
1412 | { | |
1413 | __bitmap_clear_bits(ctl, info, offset, bytes); | |
34d52cb6 | 1414 | ctl->free_space -= bytes; |
96303081 JB |
1415 | } |
1416 | ||
34d52cb6 | 1417 | static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl, |
817d52f8 JB |
1418 | struct btrfs_free_space *info, u64 offset, |
1419 | u64 bytes) | |
96303081 | 1420 | { |
f38b6e75 | 1421 | unsigned long start, count; |
96303081 | 1422 | |
34d52cb6 LZ |
1423 | start = offset_to_bit(info->offset, ctl->unit, offset); |
1424 | count = bytes_to_bits(bytes, ctl->unit); | |
b12d6869 | 1425 | ASSERT(start + count <= BITS_PER_BITMAP); |
96303081 | 1426 | |
f38b6e75 | 1427 | bitmap_set(info->bitmap, start, count); |
96303081 JB |
1428 | |
1429 | info->bytes += bytes; | |
34d52cb6 | 1430 | ctl->free_space += bytes; |
96303081 JB |
1431 | } |
1432 | ||
a4820398 MX |
1433 | /* |
1434 | * If we can not find suitable extent, we will use bytes to record | |
1435 | * the size of the max extent. | |
1436 | */ | |
34d52cb6 | 1437 | static int search_bitmap(struct btrfs_free_space_ctl *ctl, |
96303081 JB |
1438 | struct btrfs_free_space *bitmap_info, u64 *offset, |
1439 | u64 *bytes) | |
1440 | { | |
1441 | unsigned long found_bits = 0; | |
a4820398 | 1442 | unsigned long max_bits = 0; |
96303081 JB |
1443 | unsigned long bits, i; |
1444 | unsigned long next_zero; | |
a4820398 | 1445 | unsigned long extent_bits; |
96303081 | 1446 | |
34d52cb6 | 1447 | i = offset_to_bit(bitmap_info->offset, ctl->unit, |
96303081 | 1448 | max_t(u64, *offset, bitmap_info->offset)); |
34d52cb6 | 1449 | bits = bytes_to_bits(*bytes, ctl->unit); |
96303081 | 1450 | |
ebb3dad4 | 1451 | for_each_set_bit_from(i, bitmap_info->bitmap, BITS_PER_BITMAP) { |
96303081 JB |
1452 | next_zero = find_next_zero_bit(bitmap_info->bitmap, |
1453 | BITS_PER_BITMAP, i); | |
a4820398 MX |
1454 | extent_bits = next_zero - i; |
1455 | if (extent_bits >= bits) { | |
1456 | found_bits = extent_bits; | |
96303081 | 1457 | break; |
a4820398 MX |
1458 | } else if (extent_bits > max_bits) { |
1459 | max_bits = extent_bits; | |
96303081 JB |
1460 | } |
1461 | i = next_zero; | |
1462 | } | |
1463 | ||
1464 | if (found_bits) { | |
34d52cb6 LZ |
1465 | *offset = (u64)(i * ctl->unit) + bitmap_info->offset; |
1466 | *bytes = (u64)(found_bits) * ctl->unit; | |
96303081 JB |
1467 | return 0; |
1468 | } | |
1469 | ||
a4820398 | 1470 | *bytes = (u64)(max_bits) * ctl->unit; |
96303081 JB |
1471 | return -1; |
1472 | } | |
1473 | ||
a4820398 | 1474 | /* Cache the size of the max extent in bytes */ |
34d52cb6 | 1475 | static struct btrfs_free_space * |
53b381b3 | 1476 | find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes, |
a4820398 | 1477 | unsigned long align, u64 *max_extent_size) |
96303081 JB |
1478 | { |
1479 | struct btrfs_free_space *entry; | |
1480 | struct rb_node *node; | |
53b381b3 DW |
1481 | u64 tmp; |
1482 | u64 align_off; | |
96303081 JB |
1483 | int ret; |
1484 | ||
34d52cb6 | 1485 | if (!ctl->free_space_offset.rb_node) |
a4820398 | 1486 | goto out; |
96303081 | 1487 | |
34d52cb6 | 1488 | entry = tree_search_offset(ctl, offset_to_bitmap(ctl, *offset), 0, 1); |
96303081 | 1489 | if (!entry) |
a4820398 | 1490 | goto out; |
96303081 JB |
1491 | |
1492 | for (node = &entry->offset_index; node; node = rb_next(node)) { | |
1493 | entry = rb_entry(node, struct btrfs_free_space, offset_index); | |
a4820398 MX |
1494 | if (entry->bytes < *bytes) { |
1495 | if (entry->bytes > *max_extent_size) | |
1496 | *max_extent_size = entry->bytes; | |
96303081 | 1497 | continue; |
a4820398 | 1498 | } |
96303081 | 1499 | |
53b381b3 DW |
1500 | /* make sure the space returned is big enough |
1501 | * to match our requested alignment | |
1502 | */ | |
1503 | if (*bytes >= align) { | |
a4820398 | 1504 | tmp = entry->offset - ctl->start + align - 1; |
53b381b3 DW |
1505 | do_div(tmp, align); |
1506 | tmp = tmp * align + ctl->start; | |
1507 | align_off = tmp - entry->offset; | |
1508 | } else { | |
1509 | align_off = 0; | |
1510 | tmp = entry->offset; | |
1511 | } | |
1512 | ||
a4820398 MX |
1513 | if (entry->bytes < *bytes + align_off) { |
1514 | if (entry->bytes > *max_extent_size) | |
1515 | *max_extent_size = entry->bytes; | |
53b381b3 | 1516 | continue; |
a4820398 | 1517 | } |
53b381b3 | 1518 | |
96303081 | 1519 | if (entry->bitmap) { |
a4820398 MX |
1520 | u64 size = *bytes; |
1521 | ||
1522 | ret = search_bitmap(ctl, entry, &tmp, &size); | |
53b381b3 DW |
1523 | if (!ret) { |
1524 | *offset = tmp; | |
a4820398 | 1525 | *bytes = size; |
96303081 | 1526 | return entry; |
a4820398 MX |
1527 | } else if (size > *max_extent_size) { |
1528 | *max_extent_size = size; | |
53b381b3 | 1529 | } |
96303081 JB |
1530 | continue; |
1531 | } | |
1532 | ||
53b381b3 DW |
1533 | *offset = tmp; |
1534 | *bytes = entry->bytes - align_off; | |
96303081 JB |
1535 | return entry; |
1536 | } | |
a4820398 | 1537 | out: |
96303081 JB |
1538 | return NULL; |
1539 | } | |
1540 | ||
34d52cb6 | 1541 | static void add_new_bitmap(struct btrfs_free_space_ctl *ctl, |
96303081 JB |
1542 | struct btrfs_free_space *info, u64 offset) |
1543 | { | |
34d52cb6 | 1544 | info->offset = offset_to_bitmap(ctl, offset); |
f019f426 | 1545 | info->bytes = 0; |
f2d0f676 | 1546 | INIT_LIST_HEAD(&info->list); |
34d52cb6 LZ |
1547 | link_free_space(ctl, info); |
1548 | ctl->total_bitmaps++; | |
96303081 | 1549 | |
34d52cb6 | 1550 | ctl->op->recalc_thresholds(ctl); |
96303081 JB |
1551 | } |
1552 | ||
34d52cb6 | 1553 | static void free_bitmap(struct btrfs_free_space_ctl *ctl, |
edf6e2d1 LZ |
1554 | struct btrfs_free_space *bitmap_info) |
1555 | { | |
34d52cb6 | 1556 | unlink_free_space(ctl, bitmap_info); |
edf6e2d1 | 1557 | kfree(bitmap_info->bitmap); |
dc89e982 | 1558 | kmem_cache_free(btrfs_free_space_cachep, bitmap_info); |
34d52cb6 LZ |
1559 | ctl->total_bitmaps--; |
1560 | ctl->op->recalc_thresholds(ctl); | |
edf6e2d1 LZ |
1561 | } |
1562 | ||
34d52cb6 | 1563 | static noinline int remove_from_bitmap(struct btrfs_free_space_ctl *ctl, |
96303081 JB |
1564 | struct btrfs_free_space *bitmap_info, |
1565 | u64 *offset, u64 *bytes) | |
1566 | { | |
1567 | u64 end; | |
6606bb97 JB |
1568 | u64 search_start, search_bytes; |
1569 | int ret; | |
96303081 JB |
1570 | |
1571 | again: | |
34d52cb6 | 1572 | end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * ctl->unit) - 1; |
96303081 | 1573 | |
6606bb97 | 1574 | /* |
bdb7d303 JB |
1575 | * We need to search for bits in this bitmap. We could only cover some |
1576 | * of the extent in this bitmap thanks to how we add space, so we need | |
1577 | * to search for as much as it as we can and clear that amount, and then | |
1578 | * go searching for the next bit. | |
6606bb97 JB |
1579 | */ |
1580 | search_start = *offset; | |
bdb7d303 | 1581 | search_bytes = ctl->unit; |
13dbc089 | 1582 | search_bytes = min(search_bytes, end - search_start + 1); |
34d52cb6 | 1583 | ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes); |
b50c6e25 JB |
1584 | if (ret < 0 || search_start != *offset) |
1585 | return -EINVAL; | |
6606bb97 | 1586 | |
bdb7d303 JB |
1587 | /* We may have found more bits than what we need */ |
1588 | search_bytes = min(search_bytes, *bytes); | |
1589 | ||
1590 | /* Cannot clear past the end of the bitmap */ | |
1591 | search_bytes = min(search_bytes, end - search_start + 1); | |
1592 | ||
1593 | bitmap_clear_bits(ctl, bitmap_info, search_start, search_bytes); | |
1594 | *offset += search_bytes; | |
1595 | *bytes -= search_bytes; | |
96303081 JB |
1596 | |
1597 | if (*bytes) { | |
6606bb97 | 1598 | struct rb_node *next = rb_next(&bitmap_info->offset_index); |
edf6e2d1 | 1599 | if (!bitmap_info->bytes) |
34d52cb6 | 1600 | free_bitmap(ctl, bitmap_info); |
96303081 | 1601 | |
6606bb97 JB |
1602 | /* |
1603 | * no entry after this bitmap, but we still have bytes to | |
1604 | * remove, so something has gone wrong. | |
1605 | */ | |
1606 | if (!next) | |
96303081 JB |
1607 | return -EINVAL; |
1608 | ||
6606bb97 JB |
1609 | bitmap_info = rb_entry(next, struct btrfs_free_space, |
1610 | offset_index); | |
1611 | ||
1612 | /* | |
1613 | * if the next entry isn't a bitmap we need to return to let the | |
1614 | * extent stuff do its work. | |
1615 | */ | |
96303081 JB |
1616 | if (!bitmap_info->bitmap) |
1617 | return -EAGAIN; | |
1618 | ||
6606bb97 JB |
1619 | /* |
1620 | * Ok the next item is a bitmap, but it may not actually hold | |
1621 | * the information for the rest of this free space stuff, so | |
1622 | * look for it, and if we don't find it return so we can try | |
1623 | * everything over again. | |
1624 | */ | |
1625 | search_start = *offset; | |
bdb7d303 | 1626 | search_bytes = ctl->unit; |
34d52cb6 | 1627 | ret = search_bitmap(ctl, bitmap_info, &search_start, |
6606bb97 JB |
1628 | &search_bytes); |
1629 | if (ret < 0 || search_start != *offset) | |
1630 | return -EAGAIN; | |
1631 | ||
96303081 | 1632 | goto again; |
edf6e2d1 | 1633 | } else if (!bitmap_info->bytes) |
34d52cb6 | 1634 | free_bitmap(ctl, bitmap_info); |
96303081 JB |
1635 | |
1636 | return 0; | |
1637 | } | |
1638 | ||
2cdc342c JB |
1639 | static u64 add_bytes_to_bitmap(struct btrfs_free_space_ctl *ctl, |
1640 | struct btrfs_free_space *info, u64 offset, | |
1641 | u64 bytes) | |
1642 | { | |
1643 | u64 bytes_to_set = 0; | |
1644 | u64 end; | |
1645 | ||
1646 | end = info->offset + (u64)(BITS_PER_BITMAP * ctl->unit); | |
1647 | ||
1648 | bytes_to_set = min(end - offset, bytes); | |
1649 | ||
1650 | bitmap_set_bits(ctl, info, offset, bytes_to_set); | |
1651 | ||
1652 | return bytes_to_set; | |
1653 | ||
1654 | } | |
1655 | ||
34d52cb6 LZ |
1656 | static bool use_bitmap(struct btrfs_free_space_ctl *ctl, |
1657 | struct btrfs_free_space *info) | |
96303081 | 1658 | { |
34d52cb6 | 1659 | struct btrfs_block_group_cache *block_group = ctl->private; |
96303081 JB |
1660 | |
1661 | /* | |
1662 | * If we are below the extents threshold then we can add this as an | |
1663 | * extent, and don't have to deal with the bitmap | |
1664 | */ | |
34d52cb6 | 1665 | if (ctl->free_extents < ctl->extents_thresh) { |
32cb0840 JB |
1666 | /* |
1667 | * If this block group has some small extents we don't want to | |
1668 | * use up all of our free slots in the cache with them, we want | |
1669 | * to reserve them to larger extents, however if we have plent | |
1670 | * of cache left then go ahead an dadd them, no sense in adding | |
1671 | * the overhead of a bitmap if we don't have to. | |
1672 | */ | |
1673 | if (info->bytes <= block_group->sectorsize * 4) { | |
34d52cb6 LZ |
1674 | if (ctl->free_extents * 2 <= ctl->extents_thresh) |
1675 | return false; | |
32cb0840 | 1676 | } else { |
34d52cb6 | 1677 | return false; |
32cb0840 JB |
1678 | } |
1679 | } | |
96303081 JB |
1680 | |
1681 | /* | |
dde5740f JB |
1682 | * The original block groups from mkfs can be really small, like 8 |
1683 | * megabytes, so don't bother with a bitmap for those entries. However | |
1684 | * some block groups can be smaller than what a bitmap would cover but | |
1685 | * are still large enough that they could overflow the 32k memory limit, | |
1686 | * so allow those block groups to still be allowed to have a bitmap | |
1687 | * entry. | |
96303081 | 1688 | */ |
dde5740f | 1689 | if (((BITS_PER_BITMAP * ctl->unit) >> 1) > block_group->key.offset) |
34d52cb6 LZ |
1690 | return false; |
1691 | ||
1692 | return true; | |
1693 | } | |
1694 | ||
2cdc342c JB |
1695 | static struct btrfs_free_space_op free_space_op = { |
1696 | .recalc_thresholds = recalculate_thresholds, | |
1697 | .use_bitmap = use_bitmap, | |
1698 | }; | |
1699 | ||
34d52cb6 LZ |
1700 | static int insert_into_bitmap(struct btrfs_free_space_ctl *ctl, |
1701 | struct btrfs_free_space *info) | |
1702 | { | |
1703 | struct btrfs_free_space *bitmap_info; | |
2cdc342c | 1704 | struct btrfs_block_group_cache *block_group = NULL; |
34d52cb6 | 1705 | int added = 0; |
2cdc342c | 1706 | u64 bytes, offset, bytes_added; |
34d52cb6 | 1707 | int ret; |
96303081 JB |
1708 | |
1709 | bytes = info->bytes; | |
1710 | offset = info->offset; | |
1711 | ||
34d52cb6 LZ |
1712 | if (!ctl->op->use_bitmap(ctl, info)) |
1713 | return 0; | |
1714 | ||
2cdc342c JB |
1715 | if (ctl->op == &free_space_op) |
1716 | block_group = ctl->private; | |
38e87880 | 1717 | again: |
2cdc342c JB |
1718 | /* |
1719 | * Since we link bitmaps right into the cluster we need to see if we | |
1720 | * have a cluster here, and if so and it has our bitmap we need to add | |
1721 | * the free space to that bitmap. | |
1722 | */ | |
1723 | if (block_group && !list_empty(&block_group->cluster_list)) { | |
1724 | struct btrfs_free_cluster *cluster; | |
1725 | struct rb_node *node; | |
1726 | struct btrfs_free_space *entry; | |
1727 | ||
1728 | cluster = list_entry(block_group->cluster_list.next, | |
1729 | struct btrfs_free_cluster, | |
1730 | block_group_list); | |
1731 | spin_lock(&cluster->lock); | |
1732 | node = rb_first(&cluster->root); | |
1733 | if (!node) { | |
1734 | spin_unlock(&cluster->lock); | |
38e87880 | 1735 | goto no_cluster_bitmap; |
2cdc342c JB |
1736 | } |
1737 | ||
1738 | entry = rb_entry(node, struct btrfs_free_space, offset_index); | |
1739 | if (!entry->bitmap) { | |
1740 | spin_unlock(&cluster->lock); | |
38e87880 | 1741 | goto no_cluster_bitmap; |
2cdc342c JB |
1742 | } |
1743 | ||
1744 | if (entry->offset == offset_to_bitmap(ctl, offset)) { | |
1745 | bytes_added = add_bytes_to_bitmap(ctl, entry, | |
1746 | offset, bytes); | |
1747 | bytes -= bytes_added; | |
1748 | offset += bytes_added; | |
1749 | } | |
1750 | spin_unlock(&cluster->lock); | |
1751 | if (!bytes) { | |
1752 | ret = 1; | |
1753 | goto out; | |
1754 | } | |
1755 | } | |
38e87880 CM |
1756 | |
1757 | no_cluster_bitmap: | |
34d52cb6 | 1758 | bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), |
96303081 JB |
1759 | 1, 0); |
1760 | if (!bitmap_info) { | |
b12d6869 | 1761 | ASSERT(added == 0); |
96303081 JB |
1762 | goto new_bitmap; |
1763 | } | |
1764 | ||
2cdc342c JB |
1765 | bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes); |
1766 | bytes -= bytes_added; | |
1767 | offset += bytes_added; | |
1768 | added = 0; | |
96303081 JB |
1769 | |
1770 | if (!bytes) { | |
1771 | ret = 1; | |
1772 | goto out; | |
1773 | } else | |
1774 | goto again; | |
1775 | ||
1776 | new_bitmap: | |
1777 | if (info && info->bitmap) { | |
34d52cb6 | 1778 | add_new_bitmap(ctl, info, offset); |
96303081 JB |
1779 | added = 1; |
1780 | info = NULL; | |
1781 | goto again; | |
1782 | } else { | |
34d52cb6 | 1783 | spin_unlock(&ctl->tree_lock); |
96303081 JB |
1784 | |
1785 | /* no pre-allocated info, allocate a new one */ | |
1786 | if (!info) { | |
dc89e982 JB |
1787 | info = kmem_cache_zalloc(btrfs_free_space_cachep, |
1788 | GFP_NOFS); | |
96303081 | 1789 | if (!info) { |
34d52cb6 | 1790 | spin_lock(&ctl->tree_lock); |
96303081 JB |
1791 | ret = -ENOMEM; |
1792 | goto out; | |
1793 | } | |
1794 | } | |
1795 | ||
1796 | /* allocate the bitmap */ | |
1797 | info->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); | |
34d52cb6 | 1798 | spin_lock(&ctl->tree_lock); |
96303081 JB |
1799 | if (!info->bitmap) { |
1800 | ret = -ENOMEM; | |
1801 | goto out; | |
1802 | } | |
1803 | goto again; | |
1804 | } | |
1805 | ||
1806 | out: | |
1807 | if (info) { | |
1808 | if (info->bitmap) | |
1809 | kfree(info->bitmap); | |
dc89e982 | 1810 | kmem_cache_free(btrfs_free_space_cachep, info); |
96303081 | 1811 | } |
0f9dd46c JB |
1812 | |
1813 | return ret; | |
1814 | } | |
1815 | ||
945d8962 | 1816 | static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, |
f333adb5 | 1817 | struct btrfs_free_space *info, bool update_stat) |
0f9dd46c | 1818 | { |
120d66ee LZ |
1819 | struct btrfs_free_space *left_info; |
1820 | struct btrfs_free_space *right_info; | |
1821 | bool merged = false; | |
1822 | u64 offset = info->offset; | |
1823 | u64 bytes = info->bytes; | |
6226cb0a | 1824 | |
0f9dd46c JB |
1825 | /* |
1826 | * first we want to see if there is free space adjacent to the range we | |
1827 | * are adding, if there is remove that struct and add a new one to | |
1828 | * cover the entire range | |
1829 | */ | |
34d52cb6 | 1830 | right_info = tree_search_offset(ctl, offset + bytes, 0, 0); |
96303081 JB |
1831 | if (right_info && rb_prev(&right_info->offset_index)) |
1832 | left_info = rb_entry(rb_prev(&right_info->offset_index), | |
1833 | struct btrfs_free_space, offset_index); | |
1834 | else | |
34d52cb6 | 1835 | left_info = tree_search_offset(ctl, offset - 1, 0, 0); |
0f9dd46c | 1836 | |
96303081 | 1837 | if (right_info && !right_info->bitmap) { |
f333adb5 | 1838 | if (update_stat) |
34d52cb6 | 1839 | unlink_free_space(ctl, right_info); |
f333adb5 | 1840 | else |
34d52cb6 | 1841 | __unlink_free_space(ctl, right_info); |
6226cb0a | 1842 | info->bytes += right_info->bytes; |
dc89e982 | 1843 | kmem_cache_free(btrfs_free_space_cachep, right_info); |
120d66ee | 1844 | merged = true; |
0f9dd46c JB |
1845 | } |
1846 | ||
96303081 JB |
1847 | if (left_info && !left_info->bitmap && |
1848 | left_info->offset + left_info->bytes == offset) { | |
f333adb5 | 1849 | if (update_stat) |
34d52cb6 | 1850 | unlink_free_space(ctl, left_info); |
f333adb5 | 1851 | else |
34d52cb6 | 1852 | __unlink_free_space(ctl, left_info); |
6226cb0a JB |
1853 | info->offset = left_info->offset; |
1854 | info->bytes += left_info->bytes; | |
dc89e982 | 1855 | kmem_cache_free(btrfs_free_space_cachep, left_info); |
120d66ee | 1856 | merged = true; |
0f9dd46c JB |
1857 | } |
1858 | ||
120d66ee LZ |
1859 | return merged; |
1860 | } | |
1861 | ||
581bb050 LZ |
1862 | int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl, |
1863 | u64 offset, u64 bytes) | |
120d66ee LZ |
1864 | { |
1865 | struct btrfs_free_space *info; | |
1866 | int ret = 0; | |
1867 | ||
dc89e982 | 1868 | info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS); |
120d66ee LZ |
1869 | if (!info) |
1870 | return -ENOMEM; | |
1871 | ||
1872 | info->offset = offset; | |
1873 | info->bytes = bytes; | |
1874 | ||
34d52cb6 | 1875 | spin_lock(&ctl->tree_lock); |
120d66ee | 1876 | |
34d52cb6 | 1877 | if (try_merge_free_space(ctl, info, true)) |
120d66ee LZ |
1878 | goto link; |
1879 | ||
1880 | /* | |
1881 | * There was no extent directly to the left or right of this new | |
1882 | * extent then we know we're going to have to allocate a new extent, so | |
1883 | * before we do that see if we need to drop this into a bitmap | |
1884 | */ | |
34d52cb6 | 1885 | ret = insert_into_bitmap(ctl, info); |
120d66ee LZ |
1886 | if (ret < 0) { |
1887 | goto out; | |
1888 | } else if (ret) { | |
1889 | ret = 0; | |
1890 | goto out; | |
1891 | } | |
1892 | link: | |
34d52cb6 | 1893 | ret = link_free_space(ctl, info); |
0f9dd46c | 1894 | if (ret) |
dc89e982 | 1895 | kmem_cache_free(btrfs_free_space_cachep, info); |
96303081 | 1896 | out: |
34d52cb6 | 1897 | spin_unlock(&ctl->tree_lock); |
6226cb0a | 1898 | |
0f9dd46c | 1899 | if (ret) { |
96303081 | 1900 | printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret); |
b12d6869 | 1901 | ASSERT(ret != -EEXIST); |
0f9dd46c JB |
1902 | } |
1903 | ||
0f9dd46c JB |
1904 | return ret; |
1905 | } | |
1906 | ||
6226cb0a JB |
1907 | int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, |
1908 | u64 offset, u64 bytes) | |
0f9dd46c | 1909 | { |
34d52cb6 | 1910 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
0f9dd46c | 1911 | struct btrfs_free_space *info; |
b0175117 JB |
1912 | int ret; |
1913 | bool re_search = false; | |
0f9dd46c | 1914 | |
34d52cb6 | 1915 | spin_lock(&ctl->tree_lock); |
6226cb0a | 1916 | |
96303081 | 1917 | again: |
b0175117 | 1918 | ret = 0; |
bdb7d303 JB |
1919 | if (!bytes) |
1920 | goto out_lock; | |
1921 | ||
34d52cb6 | 1922 | info = tree_search_offset(ctl, offset, 0, 0); |
96303081 | 1923 | if (!info) { |
6606bb97 JB |
1924 | /* |
1925 | * oops didn't find an extent that matched the space we wanted | |
1926 | * to remove, look for a bitmap instead | |
1927 | */ | |
34d52cb6 | 1928 | info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), |
6606bb97 JB |
1929 | 1, 0); |
1930 | if (!info) { | |
b0175117 JB |
1931 | /* |
1932 | * If we found a partial bit of our free space in a | |
1933 | * bitmap but then couldn't find the other part this may | |
1934 | * be a problem, so WARN about it. | |
24a70313 | 1935 | */ |
b0175117 | 1936 | WARN_ON(re_search); |
6606bb97 JB |
1937 | goto out_lock; |
1938 | } | |
96303081 JB |
1939 | } |
1940 | ||
b0175117 | 1941 | re_search = false; |
bdb7d303 | 1942 | if (!info->bitmap) { |
34d52cb6 | 1943 | unlink_free_space(ctl, info); |
bdb7d303 JB |
1944 | if (offset == info->offset) { |
1945 | u64 to_free = min(bytes, info->bytes); | |
1946 | ||
1947 | info->bytes -= to_free; | |
1948 | info->offset += to_free; | |
1949 | if (info->bytes) { | |
1950 | ret = link_free_space(ctl, info); | |
1951 | WARN_ON(ret); | |
1952 | } else { | |
1953 | kmem_cache_free(btrfs_free_space_cachep, info); | |
1954 | } | |
0f9dd46c | 1955 | |
bdb7d303 JB |
1956 | offset += to_free; |
1957 | bytes -= to_free; | |
1958 | goto again; | |
1959 | } else { | |
1960 | u64 old_end = info->bytes + info->offset; | |
9b49c9b9 | 1961 | |
bdb7d303 | 1962 | info->bytes = offset - info->offset; |
34d52cb6 | 1963 | ret = link_free_space(ctl, info); |
96303081 JB |
1964 | WARN_ON(ret); |
1965 | if (ret) | |
1966 | goto out_lock; | |
96303081 | 1967 | |
bdb7d303 JB |
1968 | /* Not enough bytes in this entry to satisfy us */ |
1969 | if (old_end < offset + bytes) { | |
1970 | bytes -= old_end - offset; | |
1971 | offset = old_end; | |
1972 | goto again; | |
1973 | } else if (old_end == offset + bytes) { | |
1974 | /* all done */ | |
1975 | goto out_lock; | |
1976 | } | |
1977 | spin_unlock(&ctl->tree_lock); | |
1978 | ||
1979 | ret = btrfs_add_free_space(block_group, offset + bytes, | |
1980 | old_end - (offset + bytes)); | |
1981 | WARN_ON(ret); | |
1982 | goto out; | |
1983 | } | |
0f9dd46c | 1984 | } |
96303081 | 1985 | |
34d52cb6 | 1986 | ret = remove_from_bitmap(ctl, info, &offset, &bytes); |
b0175117 JB |
1987 | if (ret == -EAGAIN) { |
1988 | re_search = true; | |
96303081 | 1989 | goto again; |
b0175117 | 1990 | } |
96303081 | 1991 | out_lock: |
34d52cb6 | 1992 | spin_unlock(&ctl->tree_lock); |
0f9dd46c | 1993 | out: |
25179201 JB |
1994 | return ret; |
1995 | } | |
1996 | ||
0f9dd46c JB |
1997 | void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group, |
1998 | u64 bytes) | |
1999 | { | |
34d52cb6 | 2000 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
0f9dd46c JB |
2001 | struct btrfs_free_space *info; |
2002 | struct rb_node *n; | |
2003 | int count = 0; | |
2004 | ||
34d52cb6 | 2005 | for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) { |
0f9dd46c | 2006 | info = rb_entry(n, struct btrfs_free_space, offset_index); |
f6175efa | 2007 | if (info->bytes >= bytes && !block_group->ro) |
0f9dd46c | 2008 | count++; |
96303081 | 2009 | printk(KERN_CRIT "entry offset %llu, bytes %llu, bitmap %s\n", |
c1c9ff7c | 2010 | info->offset, info->bytes, |
96303081 | 2011 | (info->bitmap) ? "yes" : "no"); |
0f9dd46c | 2012 | } |
96303081 JB |
2013 | printk(KERN_INFO "block group has cluster?: %s\n", |
2014 | list_empty(&block_group->cluster_list) ? "no" : "yes"); | |
0f9dd46c JB |
2015 | printk(KERN_INFO "%d blocks of free space at or bigger than bytes is" |
2016 | "\n", count); | |
2017 | } | |
2018 | ||
34d52cb6 | 2019 | void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group) |
0f9dd46c | 2020 | { |
34d52cb6 | 2021 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
0f9dd46c | 2022 | |
34d52cb6 LZ |
2023 | spin_lock_init(&ctl->tree_lock); |
2024 | ctl->unit = block_group->sectorsize; | |
2025 | ctl->start = block_group->key.objectid; | |
2026 | ctl->private = block_group; | |
2027 | ctl->op = &free_space_op; | |
0f9dd46c | 2028 | |
34d52cb6 LZ |
2029 | /* |
2030 | * we only want to have 32k of ram per block group for keeping | |
2031 | * track of free space, and if we pass 1/2 of that we want to | |
2032 | * start converting things over to using bitmaps | |
2033 | */ | |
2034 | ctl->extents_thresh = ((1024 * 32) / 2) / | |
2035 | sizeof(struct btrfs_free_space); | |
0f9dd46c JB |
2036 | } |
2037 | ||
fa9c0d79 CM |
2038 | /* |
2039 | * for a given cluster, put all of its extents back into the free | |
2040 | * space cache. If the block group passed doesn't match the block group | |
2041 | * pointed to by the cluster, someone else raced in and freed the | |
2042 | * cluster already. In that case, we just return without changing anything | |
2043 | */ | |
2044 | static int | |
2045 | __btrfs_return_cluster_to_free_space( | |
2046 | struct btrfs_block_group_cache *block_group, | |
2047 | struct btrfs_free_cluster *cluster) | |
2048 | { | |
34d52cb6 | 2049 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
fa9c0d79 CM |
2050 | struct btrfs_free_space *entry; |
2051 | struct rb_node *node; | |
2052 | ||
2053 | spin_lock(&cluster->lock); | |
2054 | if (cluster->block_group != block_group) | |
2055 | goto out; | |
2056 | ||
96303081 | 2057 | cluster->block_group = NULL; |
fa9c0d79 | 2058 | cluster->window_start = 0; |
96303081 | 2059 | list_del_init(&cluster->block_group_list); |
96303081 | 2060 | |
fa9c0d79 | 2061 | node = rb_first(&cluster->root); |
96303081 | 2062 | while (node) { |
4e69b598 JB |
2063 | bool bitmap; |
2064 | ||
fa9c0d79 CM |
2065 | entry = rb_entry(node, struct btrfs_free_space, offset_index); |
2066 | node = rb_next(&entry->offset_index); | |
2067 | rb_erase(&entry->offset_index, &cluster->root); | |
4e69b598 JB |
2068 | |
2069 | bitmap = (entry->bitmap != NULL); | |
2070 | if (!bitmap) | |
34d52cb6 LZ |
2071 | try_merge_free_space(ctl, entry, false); |
2072 | tree_insert_offset(&ctl->free_space_offset, | |
4e69b598 | 2073 | entry->offset, &entry->offset_index, bitmap); |
fa9c0d79 | 2074 | } |
6bef4d31 | 2075 | cluster->root = RB_ROOT; |
96303081 | 2076 | |
fa9c0d79 CM |
2077 | out: |
2078 | spin_unlock(&cluster->lock); | |
96303081 | 2079 | btrfs_put_block_group(block_group); |
fa9c0d79 CM |
2080 | return 0; |
2081 | } | |
2082 | ||
48a3b636 ES |
2083 | static void __btrfs_remove_free_space_cache_locked( |
2084 | struct btrfs_free_space_ctl *ctl) | |
0f9dd46c JB |
2085 | { |
2086 | struct btrfs_free_space *info; | |
2087 | struct rb_node *node; | |
581bb050 | 2088 | |
581bb050 LZ |
2089 | while ((node = rb_last(&ctl->free_space_offset)) != NULL) { |
2090 | info = rb_entry(node, struct btrfs_free_space, offset_index); | |
9b90f513 JB |
2091 | if (!info->bitmap) { |
2092 | unlink_free_space(ctl, info); | |
2093 | kmem_cache_free(btrfs_free_space_cachep, info); | |
2094 | } else { | |
2095 | free_bitmap(ctl, info); | |
2096 | } | |
581bb050 LZ |
2097 | if (need_resched()) { |
2098 | spin_unlock(&ctl->tree_lock); | |
2099 | cond_resched(); | |
2100 | spin_lock(&ctl->tree_lock); | |
2101 | } | |
2102 | } | |
09655373 CM |
2103 | } |
2104 | ||
2105 | void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl) | |
2106 | { | |
2107 | spin_lock(&ctl->tree_lock); | |
2108 | __btrfs_remove_free_space_cache_locked(ctl); | |
581bb050 LZ |
2109 | spin_unlock(&ctl->tree_lock); |
2110 | } | |
2111 | ||
2112 | void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) | |
2113 | { | |
2114 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; | |
fa9c0d79 | 2115 | struct btrfs_free_cluster *cluster; |
96303081 | 2116 | struct list_head *head; |
0f9dd46c | 2117 | |
34d52cb6 | 2118 | spin_lock(&ctl->tree_lock); |
96303081 JB |
2119 | while ((head = block_group->cluster_list.next) != |
2120 | &block_group->cluster_list) { | |
2121 | cluster = list_entry(head, struct btrfs_free_cluster, | |
2122 | block_group_list); | |
fa9c0d79 CM |
2123 | |
2124 | WARN_ON(cluster->block_group != block_group); | |
2125 | __btrfs_return_cluster_to_free_space(block_group, cluster); | |
96303081 | 2126 | if (need_resched()) { |
34d52cb6 | 2127 | spin_unlock(&ctl->tree_lock); |
96303081 | 2128 | cond_resched(); |
34d52cb6 | 2129 | spin_lock(&ctl->tree_lock); |
96303081 | 2130 | } |
fa9c0d79 | 2131 | } |
09655373 | 2132 | __btrfs_remove_free_space_cache_locked(ctl); |
34d52cb6 | 2133 | spin_unlock(&ctl->tree_lock); |
fa9c0d79 | 2134 | |
0f9dd46c JB |
2135 | } |
2136 | ||
6226cb0a | 2137 | u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, |
a4820398 MX |
2138 | u64 offset, u64 bytes, u64 empty_size, |
2139 | u64 *max_extent_size) | |
0f9dd46c | 2140 | { |
34d52cb6 | 2141 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
6226cb0a | 2142 | struct btrfs_free_space *entry = NULL; |
96303081 | 2143 | u64 bytes_search = bytes + empty_size; |
6226cb0a | 2144 | u64 ret = 0; |
53b381b3 DW |
2145 | u64 align_gap = 0; |
2146 | u64 align_gap_len = 0; | |
0f9dd46c | 2147 | |
34d52cb6 | 2148 | spin_lock(&ctl->tree_lock); |
53b381b3 | 2149 | entry = find_free_space(ctl, &offset, &bytes_search, |
a4820398 | 2150 | block_group->full_stripe_len, max_extent_size); |
6226cb0a | 2151 | if (!entry) |
96303081 JB |
2152 | goto out; |
2153 | ||
2154 | ret = offset; | |
2155 | if (entry->bitmap) { | |
34d52cb6 | 2156 | bitmap_clear_bits(ctl, entry, offset, bytes); |
edf6e2d1 | 2157 | if (!entry->bytes) |
34d52cb6 | 2158 | free_bitmap(ctl, entry); |
96303081 | 2159 | } else { |
34d52cb6 | 2160 | unlink_free_space(ctl, entry); |
53b381b3 DW |
2161 | align_gap_len = offset - entry->offset; |
2162 | align_gap = entry->offset; | |
2163 | ||
2164 | entry->offset = offset + bytes; | |
2165 | WARN_ON(entry->bytes < bytes + align_gap_len); | |
2166 | ||
2167 | entry->bytes -= bytes + align_gap_len; | |
6226cb0a | 2168 | if (!entry->bytes) |
dc89e982 | 2169 | kmem_cache_free(btrfs_free_space_cachep, entry); |
6226cb0a | 2170 | else |
34d52cb6 | 2171 | link_free_space(ctl, entry); |
6226cb0a | 2172 | } |
96303081 | 2173 | out: |
34d52cb6 | 2174 | spin_unlock(&ctl->tree_lock); |
817d52f8 | 2175 | |
53b381b3 DW |
2176 | if (align_gap_len) |
2177 | __btrfs_add_free_space(ctl, align_gap, align_gap_len); | |
0f9dd46c JB |
2178 | return ret; |
2179 | } | |
fa9c0d79 CM |
2180 | |
2181 | /* | |
2182 | * given a cluster, put all of its extents back into the free space | |
2183 | * cache. If a block group is passed, this function will only free | |
2184 | * a cluster that belongs to the passed block group. | |
2185 | * | |
2186 | * Otherwise, it'll get a reference on the block group pointed to by the | |
2187 | * cluster and remove the cluster from it. | |
2188 | */ | |
2189 | int btrfs_return_cluster_to_free_space( | |
2190 | struct btrfs_block_group_cache *block_group, | |
2191 | struct btrfs_free_cluster *cluster) | |
2192 | { | |
34d52cb6 | 2193 | struct btrfs_free_space_ctl *ctl; |
fa9c0d79 CM |
2194 | int ret; |
2195 | ||
2196 | /* first, get a safe pointer to the block group */ | |
2197 | spin_lock(&cluster->lock); | |
2198 | if (!block_group) { | |
2199 | block_group = cluster->block_group; | |
2200 | if (!block_group) { | |
2201 | spin_unlock(&cluster->lock); | |
2202 | return 0; | |
2203 | } | |
2204 | } else if (cluster->block_group != block_group) { | |
2205 | /* someone else has already freed it don't redo their work */ | |
2206 | spin_unlock(&cluster->lock); | |
2207 | return 0; | |
2208 | } | |
2209 | atomic_inc(&block_group->count); | |
2210 | spin_unlock(&cluster->lock); | |
2211 | ||
34d52cb6 LZ |
2212 | ctl = block_group->free_space_ctl; |
2213 | ||
fa9c0d79 | 2214 | /* now return any extents the cluster had on it */ |
34d52cb6 | 2215 | spin_lock(&ctl->tree_lock); |
fa9c0d79 | 2216 | ret = __btrfs_return_cluster_to_free_space(block_group, cluster); |
34d52cb6 | 2217 | spin_unlock(&ctl->tree_lock); |
fa9c0d79 CM |
2218 | |
2219 | /* finally drop our ref */ | |
2220 | btrfs_put_block_group(block_group); | |
2221 | return ret; | |
2222 | } | |
2223 | ||
96303081 JB |
2224 | static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, |
2225 | struct btrfs_free_cluster *cluster, | |
4e69b598 | 2226 | struct btrfs_free_space *entry, |
a4820398 MX |
2227 | u64 bytes, u64 min_start, |
2228 | u64 *max_extent_size) | |
96303081 | 2229 | { |
34d52cb6 | 2230 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
96303081 JB |
2231 | int err; |
2232 | u64 search_start = cluster->window_start; | |
2233 | u64 search_bytes = bytes; | |
2234 | u64 ret = 0; | |
2235 | ||
96303081 JB |
2236 | search_start = min_start; |
2237 | search_bytes = bytes; | |
2238 | ||
34d52cb6 | 2239 | err = search_bitmap(ctl, entry, &search_start, &search_bytes); |
a4820398 MX |
2240 | if (err) { |
2241 | if (search_bytes > *max_extent_size) | |
2242 | *max_extent_size = search_bytes; | |
4e69b598 | 2243 | return 0; |
a4820398 | 2244 | } |
96303081 JB |
2245 | |
2246 | ret = search_start; | |
bb3ac5a4 | 2247 | __bitmap_clear_bits(ctl, entry, ret, bytes); |
96303081 JB |
2248 | |
2249 | return ret; | |
2250 | } | |
2251 | ||
fa9c0d79 CM |
2252 | /* |
2253 | * given a cluster, try to allocate 'bytes' from it, returns 0 | |
2254 | * if it couldn't find anything suitably large, or a logical disk offset | |
2255 | * if things worked out | |
2256 | */ | |
2257 | u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, | |
2258 | struct btrfs_free_cluster *cluster, u64 bytes, | |
a4820398 | 2259 | u64 min_start, u64 *max_extent_size) |
fa9c0d79 | 2260 | { |
34d52cb6 | 2261 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
fa9c0d79 CM |
2262 | struct btrfs_free_space *entry = NULL; |
2263 | struct rb_node *node; | |
2264 | u64 ret = 0; | |
2265 | ||
2266 | spin_lock(&cluster->lock); | |
2267 | if (bytes > cluster->max_size) | |
2268 | goto out; | |
2269 | ||
2270 | if (cluster->block_group != block_group) | |
2271 | goto out; | |
2272 | ||
2273 | node = rb_first(&cluster->root); | |
2274 | if (!node) | |
2275 | goto out; | |
2276 | ||
2277 | entry = rb_entry(node, struct btrfs_free_space, offset_index); | |
fa9c0d79 | 2278 | while(1) { |
a4820398 MX |
2279 | if (entry->bytes < bytes && entry->bytes > *max_extent_size) |
2280 | *max_extent_size = entry->bytes; | |
2281 | ||
4e69b598 JB |
2282 | if (entry->bytes < bytes || |
2283 | (!entry->bitmap && entry->offset < min_start)) { | |
fa9c0d79 CM |
2284 | node = rb_next(&entry->offset_index); |
2285 | if (!node) | |
2286 | break; | |
2287 | entry = rb_entry(node, struct btrfs_free_space, | |
2288 | offset_index); | |
2289 | continue; | |
2290 | } | |
fa9c0d79 | 2291 | |
4e69b598 JB |
2292 | if (entry->bitmap) { |
2293 | ret = btrfs_alloc_from_bitmap(block_group, | |
2294 | cluster, entry, bytes, | |
a4820398 MX |
2295 | cluster->window_start, |
2296 | max_extent_size); | |
4e69b598 | 2297 | if (ret == 0) { |
4e69b598 JB |
2298 | node = rb_next(&entry->offset_index); |
2299 | if (!node) | |
2300 | break; | |
2301 | entry = rb_entry(node, struct btrfs_free_space, | |
2302 | offset_index); | |
2303 | continue; | |
2304 | } | |
9b230628 | 2305 | cluster->window_start += bytes; |
4e69b598 | 2306 | } else { |
4e69b598 JB |
2307 | ret = entry->offset; |
2308 | ||
2309 | entry->offset += bytes; | |
2310 | entry->bytes -= bytes; | |
2311 | } | |
fa9c0d79 | 2312 | |
5e71b5d5 | 2313 | if (entry->bytes == 0) |
fa9c0d79 | 2314 | rb_erase(&entry->offset_index, &cluster->root); |
fa9c0d79 CM |
2315 | break; |
2316 | } | |
2317 | out: | |
2318 | spin_unlock(&cluster->lock); | |
96303081 | 2319 | |
5e71b5d5 LZ |
2320 | if (!ret) |
2321 | return 0; | |
2322 | ||
34d52cb6 | 2323 | spin_lock(&ctl->tree_lock); |
5e71b5d5 | 2324 | |
34d52cb6 | 2325 | ctl->free_space -= bytes; |
5e71b5d5 | 2326 | if (entry->bytes == 0) { |
34d52cb6 | 2327 | ctl->free_extents--; |
4e69b598 JB |
2328 | if (entry->bitmap) { |
2329 | kfree(entry->bitmap); | |
34d52cb6 LZ |
2330 | ctl->total_bitmaps--; |
2331 | ctl->op->recalc_thresholds(ctl); | |
4e69b598 | 2332 | } |
dc89e982 | 2333 | kmem_cache_free(btrfs_free_space_cachep, entry); |
5e71b5d5 LZ |
2334 | } |
2335 | ||
34d52cb6 | 2336 | spin_unlock(&ctl->tree_lock); |
5e71b5d5 | 2337 | |
fa9c0d79 CM |
2338 | return ret; |
2339 | } | |
2340 | ||
96303081 JB |
2341 | static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group, |
2342 | struct btrfs_free_space *entry, | |
2343 | struct btrfs_free_cluster *cluster, | |
1bb91902 AO |
2344 | u64 offset, u64 bytes, |
2345 | u64 cont1_bytes, u64 min_bytes) | |
96303081 | 2346 | { |
34d52cb6 | 2347 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
96303081 JB |
2348 | unsigned long next_zero; |
2349 | unsigned long i; | |
1bb91902 AO |
2350 | unsigned long want_bits; |
2351 | unsigned long min_bits; | |
96303081 JB |
2352 | unsigned long found_bits; |
2353 | unsigned long start = 0; | |
2354 | unsigned long total_found = 0; | |
4e69b598 | 2355 | int ret; |
96303081 | 2356 | |
96009762 | 2357 | i = offset_to_bit(entry->offset, ctl->unit, |
96303081 | 2358 | max_t(u64, offset, entry->offset)); |
96009762 WSH |
2359 | want_bits = bytes_to_bits(bytes, ctl->unit); |
2360 | min_bits = bytes_to_bits(min_bytes, ctl->unit); | |
96303081 JB |
2361 | |
2362 | again: | |
2363 | found_bits = 0; | |
ebb3dad4 | 2364 | for_each_set_bit_from(i, entry->bitmap, BITS_PER_BITMAP) { |
96303081 JB |
2365 | next_zero = find_next_zero_bit(entry->bitmap, |
2366 | BITS_PER_BITMAP, i); | |
1bb91902 | 2367 | if (next_zero - i >= min_bits) { |
96303081 JB |
2368 | found_bits = next_zero - i; |
2369 | break; | |
2370 | } | |
2371 | i = next_zero; | |
2372 | } | |
2373 | ||
2374 | if (!found_bits) | |
4e69b598 | 2375 | return -ENOSPC; |
96303081 | 2376 | |
1bb91902 | 2377 | if (!total_found) { |
96303081 | 2378 | start = i; |
b78d09bc | 2379 | cluster->max_size = 0; |
96303081 JB |
2380 | } |
2381 | ||
2382 | total_found += found_bits; | |
2383 | ||
96009762 WSH |
2384 | if (cluster->max_size < found_bits * ctl->unit) |
2385 | cluster->max_size = found_bits * ctl->unit; | |
96303081 | 2386 | |
1bb91902 AO |
2387 | if (total_found < want_bits || cluster->max_size < cont1_bytes) { |
2388 | i = next_zero + 1; | |
96303081 JB |
2389 | goto again; |
2390 | } | |
2391 | ||
96009762 | 2392 | cluster->window_start = start * ctl->unit + entry->offset; |
34d52cb6 | 2393 | rb_erase(&entry->offset_index, &ctl->free_space_offset); |
4e69b598 JB |
2394 | ret = tree_insert_offset(&cluster->root, entry->offset, |
2395 | &entry->offset_index, 1); | |
b12d6869 | 2396 | ASSERT(!ret); /* -EEXIST; Logic error */ |
96303081 | 2397 | |
3f7de037 | 2398 | trace_btrfs_setup_cluster(block_group, cluster, |
96009762 | 2399 | total_found * ctl->unit, 1); |
96303081 JB |
2400 | return 0; |
2401 | } | |
2402 | ||
4e69b598 JB |
2403 | /* |
2404 | * This searches the block group for just extents to fill the cluster with. | |
1bb91902 AO |
2405 | * Try to find a cluster with at least bytes total bytes, at least one |
2406 | * extent of cont1_bytes, and other clusters of at least min_bytes. | |
4e69b598 | 2407 | */ |
3de85bb9 JB |
2408 | static noinline int |
2409 | setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, | |
2410 | struct btrfs_free_cluster *cluster, | |
2411 | struct list_head *bitmaps, u64 offset, u64 bytes, | |
1bb91902 | 2412 | u64 cont1_bytes, u64 min_bytes) |
4e69b598 | 2413 | { |
34d52cb6 | 2414 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
4e69b598 JB |
2415 | struct btrfs_free_space *first = NULL; |
2416 | struct btrfs_free_space *entry = NULL; | |
4e69b598 JB |
2417 | struct btrfs_free_space *last; |
2418 | struct rb_node *node; | |
2419 | u64 window_start; | |
2420 | u64 window_free; | |
2421 | u64 max_extent; | |
3f7de037 | 2422 | u64 total_size = 0; |
4e69b598 | 2423 | |
34d52cb6 | 2424 | entry = tree_search_offset(ctl, offset, 0, 1); |
4e69b598 JB |
2425 | if (!entry) |
2426 | return -ENOSPC; | |
2427 | ||
2428 | /* | |
2429 | * We don't want bitmaps, so just move along until we find a normal | |
2430 | * extent entry. | |
2431 | */ | |
1bb91902 AO |
2432 | while (entry->bitmap || entry->bytes < min_bytes) { |
2433 | if (entry->bitmap && list_empty(&entry->list)) | |
86d4a77b | 2434 | list_add_tail(&entry->list, bitmaps); |
4e69b598 JB |
2435 | node = rb_next(&entry->offset_index); |
2436 | if (!node) | |
2437 | return -ENOSPC; | |
2438 | entry = rb_entry(node, struct btrfs_free_space, offset_index); | |
2439 | } | |
2440 | ||
2441 | window_start = entry->offset; | |
2442 | window_free = entry->bytes; | |
2443 | max_extent = entry->bytes; | |
2444 | first = entry; | |
2445 | last = entry; | |
4e69b598 | 2446 | |
1bb91902 AO |
2447 | for (node = rb_next(&entry->offset_index); node; |
2448 | node = rb_next(&entry->offset_index)) { | |
4e69b598 JB |
2449 | entry = rb_entry(node, struct btrfs_free_space, offset_index); |
2450 | ||
86d4a77b JB |
2451 | if (entry->bitmap) { |
2452 | if (list_empty(&entry->list)) | |
2453 | list_add_tail(&entry->list, bitmaps); | |
4e69b598 | 2454 | continue; |
86d4a77b JB |
2455 | } |
2456 | ||
1bb91902 AO |
2457 | if (entry->bytes < min_bytes) |
2458 | continue; | |
2459 | ||
2460 | last = entry; | |
2461 | window_free += entry->bytes; | |
2462 | if (entry->bytes > max_extent) | |
4e69b598 | 2463 | max_extent = entry->bytes; |
4e69b598 JB |
2464 | } |
2465 | ||
1bb91902 AO |
2466 | if (window_free < bytes || max_extent < cont1_bytes) |
2467 | return -ENOSPC; | |
2468 | ||
4e69b598 JB |
2469 | cluster->window_start = first->offset; |
2470 | ||
2471 | node = &first->offset_index; | |
2472 | ||
2473 | /* | |
2474 | * now we've found our entries, pull them out of the free space | |
2475 | * cache and put them into the cluster rbtree | |
2476 | */ | |
2477 | do { | |
2478 | int ret; | |
2479 | ||
2480 | entry = rb_entry(node, struct btrfs_free_space, offset_index); | |
2481 | node = rb_next(&entry->offset_index); | |
1bb91902 | 2482 | if (entry->bitmap || entry->bytes < min_bytes) |
4e69b598 JB |
2483 | continue; |
2484 | ||
34d52cb6 | 2485 | rb_erase(&entry->offset_index, &ctl->free_space_offset); |
4e69b598 JB |
2486 | ret = tree_insert_offset(&cluster->root, entry->offset, |
2487 | &entry->offset_index, 0); | |
3f7de037 | 2488 | total_size += entry->bytes; |
b12d6869 | 2489 | ASSERT(!ret); /* -EEXIST; Logic error */ |
4e69b598 JB |
2490 | } while (node && entry != last); |
2491 | ||
2492 | cluster->max_size = max_extent; | |
3f7de037 | 2493 | trace_btrfs_setup_cluster(block_group, cluster, total_size, 0); |
4e69b598 JB |
2494 | return 0; |
2495 | } | |
2496 | ||
2497 | /* | |
2498 | * This specifically looks for bitmaps that may work in the cluster, we assume | |
2499 | * that we have already failed to find extents that will work. | |
2500 | */ | |
3de85bb9 JB |
2501 | static noinline int |
2502 | setup_cluster_bitmap(struct btrfs_block_group_cache *block_group, | |
2503 | struct btrfs_free_cluster *cluster, | |
2504 | struct list_head *bitmaps, u64 offset, u64 bytes, | |
1bb91902 | 2505 | u64 cont1_bytes, u64 min_bytes) |
4e69b598 | 2506 | { |
34d52cb6 | 2507 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
4e69b598 | 2508 | struct btrfs_free_space *entry; |
4e69b598 | 2509 | int ret = -ENOSPC; |
0f0fbf1d | 2510 | u64 bitmap_offset = offset_to_bitmap(ctl, offset); |
4e69b598 | 2511 | |
34d52cb6 | 2512 | if (ctl->total_bitmaps == 0) |
4e69b598 JB |
2513 | return -ENOSPC; |
2514 | ||
0f0fbf1d LZ |
2515 | /* |
2516 | * The bitmap that covers offset won't be in the list unless offset | |
2517 | * is just its start offset. | |
2518 | */ | |
2519 | entry = list_first_entry(bitmaps, struct btrfs_free_space, list); | |
2520 | if (entry->offset != bitmap_offset) { | |
2521 | entry = tree_search_offset(ctl, bitmap_offset, 1, 0); | |
2522 | if (entry && list_empty(&entry->list)) | |
2523 | list_add(&entry->list, bitmaps); | |
2524 | } | |
2525 | ||
86d4a77b | 2526 | list_for_each_entry(entry, bitmaps, list) { |
357b9784 | 2527 | if (entry->bytes < bytes) |
86d4a77b JB |
2528 | continue; |
2529 | ret = btrfs_bitmap_cluster(block_group, entry, cluster, offset, | |
1bb91902 | 2530 | bytes, cont1_bytes, min_bytes); |
86d4a77b JB |
2531 | if (!ret) |
2532 | return 0; | |
2533 | } | |
2534 | ||
2535 | /* | |
52621cb6 LZ |
2536 | * The bitmaps list has all the bitmaps that record free space |
2537 | * starting after offset, so no more search is required. | |
86d4a77b | 2538 | */ |
52621cb6 | 2539 | return -ENOSPC; |
4e69b598 JB |
2540 | } |
2541 | ||
fa9c0d79 CM |
2542 | /* |
2543 | * here we try to find a cluster of blocks in a block group. The goal | |
1bb91902 | 2544 | * is to find at least bytes+empty_size. |
fa9c0d79 CM |
2545 | * We might not find them all in one contiguous area. |
2546 | * | |
2547 | * returns zero and sets up cluster if things worked out, otherwise | |
2548 | * it returns -enospc | |
2549 | */ | |
00361589 | 2550 | int btrfs_find_space_cluster(struct btrfs_root *root, |
fa9c0d79 CM |
2551 | struct btrfs_block_group_cache *block_group, |
2552 | struct btrfs_free_cluster *cluster, | |
2553 | u64 offset, u64 bytes, u64 empty_size) | |
2554 | { | |
34d52cb6 | 2555 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; |
86d4a77b | 2556 | struct btrfs_free_space *entry, *tmp; |
52621cb6 | 2557 | LIST_HEAD(bitmaps); |
fa9c0d79 | 2558 | u64 min_bytes; |
1bb91902 | 2559 | u64 cont1_bytes; |
fa9c0d79 CM |
2560 | int ret; |
2561 | ||
1bb91902 AO |
2562 | /* |
2563 | * Choose the minimum extent size we'll require for this | |
2564 | * cluster. For SSD_SPREAD, don't allow any fragmentation. | |
2565 | * For metadata, allow allocates with smaller extents. For | |
2566 | * data, keep it dense. | |
2567 | */ | |
451d7585 | 2568 | if (btrfs_test_opt(root, SSD_SPREAD)) { |
1bb91902 | 2569 | cont1_bytes = min_bytes = bytes + empty_size; |
451d7585 | 2570 | } else if (block_group->flags & BTRFS_BLOCK_GROUP_METADATA) { |
1bb91902 AO |
2571 | cont1_bytes = bytes; |
2572 | min_bytes = block_group->sectorsize; | |
2573 | } else { | |
2574 | cont1_bytes = max(bytes, (bytes + empty_size) >> 2); | |
2575 | min_bytes = block_group->sectorsize; | |
2576 | } | |
fa9c0d79 | 2577 | |
34d52cb6 | 2578 | spin_lock(&ctl->tree_lock); |
7d0d2e8e JB |
2579 | |
2580 | /* | |
2581 | * If we know we don't have enough space to make a cluster don't even | |
2582 | * bother doing all the work to try and find one. | |
2583 | */ | |
1bb91902 | 2584 | if (ctl->free_space < bytes) { |
34d52cb6 | 2585 | spin_unlock(&ctl->tree_lock); |
7d0d2e8e JB |
2586 | return -ENOSPC; |
2587 | } | |
2588 | ||
fa9c0d79 CM |
2589 | spin_lock(&cluster->lock); |
2590 | ||
2591 | /* someone already found a cluster, hooray */ | |
2592 | if (cluster->block_group) { | |
2593 | ret = 0; | |
2594 | goto out; | |
2595 | } | |
fa9c0d79 | 2596 | |
3f7de037 JB |
2597 | trace_btrfs_find_cluster(block_group, offset, bytes, empty_size, |
2598 | min_bytes); | |
2599 | ||
2600 | INIT_LIST_HEAD(&bitmaps); | |
86d4a77b | 2601 | ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset, |
1bb91902 AO |
2602 | bytes + empty_size, |
2603 | cont1_bytes, min_bytes); | |
4e69b598 | 2604 | if (ret) |
86d4a77b | 2605 | ret = setup_cluster_bitmap(block_group, cluster, &bitmaps, |
1bb91902 AO |
2606 | offset, bytes + empty_size, |
2607 | cont1_bytes, min_bytes); | |
86d4a77b JB |
2608 | |
2609 | /* Clear our temporary list */ | |
2610 | list_for_each_entry_safe(entry, tmp, &bitmaps, list) | |
2611 | list_del_init(&entry->list); | |
fa9c0d79 | 2612 | |
4e69b598 JB |
2613 | if (!ret) { |
2614 | atomic_inc(&block_group->count); | |
2615 | list_add_tail(&cluster->block_group_list, | |
2616 | &block_group->cluster_list); | |
2617 | cluster->block_group = block_group; | |
3f7de037 JB |
2618 | } else { |
2619 | trace_btrfs_failed_cluster_setup(block_group); | |
fa9c0d79 | 2620 | } |
fa9c0d79 CM |
2621 | out: |
2622 | spin_unlock(&cluster->lock); | |
34d52cb6 | 2623 | spin_unlock(&ctl->tree_lock); |
fa9c0d79 CM |
2624 | |
2625 | return ret; | |
2626 | } | |
2627 | ||
2628 | /* | |
2629 | * simple code to zero out a cluster | |
2630 | */ | |
2631 | void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster) | |
2632 | { | |
2633 | spin_lock_init(&cluster->lock); | |
2634 | spin_lock_init(&cluster->refill_lock); | |
6bef4d31 | 2635 | cluster->root = RB_ROOT; |
fa9c0d79 CM |
2636 | cluster->max_size = 0; |
2637 | INIT_LIST_HEAD(&cluster->block_group_list); | |
2638 | cluster->block_group = NULL; | |
2639 | } | |
2640 | ||
7fe1e641 LZ |
2641 | static int do_trimming(struct btrfs_block_group_cache *block_group, |
2642 | u64 *total_trimmed, u64 start, u64 bytes, | |
2643 | u64 reserved_start, u64 reserved_bytes) | |
f7039b1d | 2644 | { |
7fe1e641 | 2645 | struct btrfs_space_info *space_info = block_group->space_info; |
f7039b1d | 2646 | struct btrfs_fs_info *fs_info = block_group->fs_info; |
7fe1e641 LZ |
2647 | int ret; |
2648 | int update = 0; | |
2649 | u64 trimmed = 0; | |
f7039b1d | 2650 | |
7fe1e641 LZ |
2651 | spin_lock(&space_info->lock); |
2652 | spin_lock(&block_group->lock); | |
2653 | if (!block_group->ro) { | |
2654 | block_group->reserved += reserved_bytes; | |
2655 | space_info->bytes_reserved += reserved_bytes; | |
2656 | update = 1; | |
2657 | } | |
2658 | spin_unlock(&block_group->lock); | |
2659 | spin_unlock(&space_info->lock); | |
2660 | ||
2661 | ret = btrfs_error_discard_extent(fs_info->extent_root, | |
2662 | start, bytes, &trimmed); | |
2663 | if (!ret) | |
2664 | *total_trimmed += trimmed; | |
2665 | ||
2666 | btrfs_add_free_space(block_group, reserved_start, reserved_bytes); | |
2667 | ||
2668 | if (update) { | |
2669 | spin_lock(&space_info->lock); | |
2670 | spin_lock(&block_group->lock); | |
2671 | if (block_group->ro) | |
2672 | space_info->bytes_readonly += reserved_bytes; | |
2673 | block_group->reserved -= reserved_bytes; | |
2674 | space_info->bytes_reserved -= reserved_bytes; | |
2675 | spin_unlock(&space_info->lock); | |
2676 | spin_unlock(&block_group->lock); | |
2677 | } | |
2678 | ||
2679 | return ret; | |
2680 | } | |
2681 | ||
2682 | static int trim_no_bitmap(struct btrfs_block_group_cache *block_group, | |
2683 | u64 *total_trimmed, u64 start, u64 end, u64 minlen) | |
2684 | { | |
2685 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; | |
2686 | struct btrfs_free_space *entry; | |
2687 | struct rb_node *node; | |
2688 | int ret = 0; | |
2689 | u64 extent_start; | |
2690 | u64 extent_bytes; | |
2691 | u64 bytes; | |
f7039b1d LD |
2692 | |
2693 | while (start < end) { | |
34d52cb6 | 2694 | spin_lock(&ctl->tree_lock); |
f7039b1d | 2695 | |
34d52cb6 LZ |
2696 | if (ctl->free_space < minlen) { |
2697 | spin_unlock(&ctl->tree_lock); | |
f7039b1d LD |
2698 | break; |
2699 | } | |
2700 | ||
34d52cb6 | 2701 | entry = tree_search_offset(ctl, start, 0, 1); |
7fe1e641 | 2702 | if (!entry) { |
34d52cb6 | 2703 | spin_unlock(&ctl->tree_lock); |
f7039b1d LD |
2704 | break; |
2705 | } | |
2706 | ||
7fe1e641 LZ |
2707 | /* skip bitmaps */ |
2708 | while (entry->bitmap) { | |
2709 | node = rb_next(&entry->offset_index); | |
2710 | if (!node) { | |
34d52cb6 | 2711 | spin_unlock(&ctl->tree_lock); |
7fe1e641 | 2712 | goto out; |
f7039b1d | 2713 | } |
7fe1e641 LZ |
2714 | entry = rb_entry(node, struct btrfs_free_space, |
2715 | offset_index); | |
f7039b1d LD |
2716 | } |
2717 | ||
7fe1e641 LZ |
2718 | if (entry->offset >= end) { |
2719 | spin_unlock(&ctl->tree_lock); | |
2720 | break; | |
f7039b1d LD |
2721 | } |
2722 | ||
7fe1e641 LZ |
2723 | extent_start = entry->offset; |
2724 | extent_bytes = entry->bytes; | |
2725 | start = max(start, extent_start); | |
2726 | bytes = min(extent_start + extent_bytes, end) - start; | |
2727 | if (bytes < minlen) { | |
2728 | spin_unlock(&ctl->tree_lock); | |
2729 | goto next; | |
f7039b1d LD |
2730 | } |
2731 | ||
7fe1e641 LZ |
2732 | unlink_free_space(ctl, entry); |
2733 | kmem_cache_free(btrfs_free_space_cachep, entry); | |
2734 | ||
34d52cb6 | 2735 | spin_unlock(&ctl->tree_lock); |
f7039b1d | 2736 | |
7fe1e641 LZ |
2737 | ret = do_trimming(block_group, total_trimmed, start, bytes, |
2738 | extent_start, extent_bytes); | |
2739 | if (ret) | |
2740 | break; | |
2741 | next: | |
2742 | start += bytes; | |
f7039b1d | 2743 | |
7fe1e641 LZ |
2744 | if (fatal_signal_pending(current)) { |
2745 | ret = -ERESTARTSYS; | |
2746 | break; | |
2747 | } | |
2748 | ||
2749 | cond_resched(); | |
2750 | } | |
2751 | out: | |
2752 | return ret; | |
2753 | } | |
2754 | ||
2755 | static int trim_bitmaps(struct btrfs_block_group_cache *block_group, | |
2756 | u64 *total_trimmed, u64 start, u64 end, u64 minlen) | |
2757 | { | |
2758 | struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; | |
2759 | struct btrfs_free_space *entry; | |
2760 | int ret = 0; | |
2761 | int ret2; | |
2762 | u64 bytes; | |
2763 | u64 offset = offset_to_bitmap(ctl, start); | |
2764 | ||
2765 | while (offset < end) { | |
2766 | bool next_bitmap = false; | |
2767 | ||
2768 | spin_lock(&ctl->tree_lock); | |
2769 | ||
2770 | if (ctl->free_space < minlen) { | |
2771 | spin_unlock(&ctl->tree_lock); | |
2772 | break; | |
2773 | } | |
2774 | ||
2775 | entry = tree_search_offset(ctl, offset, 1, 0); | |
2776 | if (!entry) { | |
2777 | spin_unlock(&ctl->tree_lock); | |
2778 | next_bitmap = true; | |
2779 | goto next; | |
2780 | } | |
2781 | ||
2782 | bytes = minlen; | |
2783 | ret2 = search_bitmap(ctl, entry, &start, &bytes); | |
2784 | if (ret2 || start >= end) { | |
2785 | spin_unlock(&ctl->tree_lock); | |
2786 | next_bitmap = true; | |
2787 | goto next; | |
2788 | } | |
2789 | ||
2790 | bytes = min(bytes, end - start); | |
2791 | if (bytes < minlen) { | |
2792 | spin_unlock(&ctl->tree_lock); | |
2793 | goto next; | |
2794 | } | |
2795 | ||
2796 | bitmap_clear_bits(ctl, entry, start, bytes); | |
2797 | if (entry->bytes == 0) | |
2798 | free_bitmap(ctl, entry); | |
2799 | ||
2800 | spin_unlock(&ctl->tree_lock); | |
2801 | ||
2802 | ret = do_trimming(block_group, total_trimmed, start, bytes, | |
2803 | start, bytes); | |
2804 | if (ret) | |
2805 | break; | |
2806 | next: | |
2807 | if (next_bitmap) { | |
2808 | offset += BITS_PER_BITMAP * ctl->unit; | |
2809 | } else { | |
2810 | start += bytes; | |
2811 | if (start >= offset + BITS_PER_BITMAP * ctl->unit) | |
2812 | offset += BITS_PER_BITMAP * ctl->unit; | |
f7039b1d | 2813 | } |
f7039b1d LD |
2814 | |
2815 | if (fatal_signal_pending(current)) { | |
2816 | ret = -ERESTARTSYS; | |
2817 | break; | |
2818 | } | |
2819 | ||
2820 | cond_resched(); | |
2821 | } | |
2822 | ||
2823 | return ret; | |
2824 | } | |
581bb050 | 2825 | |
7fe1e641 LZ |
2826 | int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, |
2827 | u64 *trimmed, u64 start, u64 end, u64 minlen) | |
2828 | { | |
2829 | int ret; | |
2830 | ||
2831 | *trimmed = 0; | |
2832 | ||
2833 | ret = trim_no_bitmap(block_group, trimmed, start, end, minlen); | |
2834 | if (ret) | |
2835 | return ret; | |
2836 | ||
2837 | ret = trim_bitmaps(block_group, trimmed, start, end, minlen); | |
2838 | ||
2839 | return ret; | |
2840 | } | |
2841 | ||
581bb050 LZ |
2842 | /* |
2843 | * Find the left-most item in the cache tree, and then return the | |
2844 | * smallest inode number in the item. | |
2845 | * | |
2846 | * Note: the returned inode number may not be the smallest one in | |
2847 | * the tree, if the left-most item is a bitmap. | |
2848 | */ | |
2849 | u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root) | |
2850 | { | |
2851 | struct btrfs_free_space_ctl *ctl = fs_root->free_ino_ctl; | |
2852 | struct btrfs_free_space *entry = NULL; | |
2853 | u64 ino = 0; | |
2854 | ||
2855 | spin_lock(&ctl->tree_lock); | |
2856 | ||
2857 | if (RB_EMPTY_ROOT(&ctl->free_space_offset)) | |
2858 | goto out; | |
2859 | ||
2860 | entry = rb_entry(rb_first(&ctl->free_space_offset), | |
2861 | struct btrfs_free_space, offset_index); | |
2862 | ||
2863 | if (!entry->bitmap) { | |
2864 | ino = entry->offset; | |
2865 | ||
2866 | unlink_free_space(ctl, entry); | |
2867 | entry->offset++; | |
2868 | entry->bytes--; | |
2869 | if (!entry->bytes) | |
2870 | kmem_cache_free(btrfs_free_space_cachep, entry); | |
2871 | else | |
2872 | link_free_space(ctl, entry); | |
2873 | } else { | |
2874 | u64 offset = 0; | |
2875 | u64 count = 1; | |
2876 | int ret; | |
2877 | ||
2878 | ret = search_bitmap(ctl, entry, &offset, &count); | |
79787eaa | 2879 | /* Logic error; Should be empty if it can't find anything */ |
b12d6869 | 2880 | ASSERT(!ret); |
581bb050 LZ |
2881 | |
2882 | ino = offset; | |
2883 | bitmap_clear_bits(ctl, entry, offset, 1); | |
2884 | if (entry->bytes == 0) | |
2885 | free_bitmap(ctl, entry); | |
2886 | } | |
2887 | out: | |
2888 | spin_unlock(&ctl->tree_lock); | |
2889 | ||
2890 | return ino; | |
2891 | } | |
82d5902d LZ |
2892 | |
2893 | struct inode *lookup_free_ino_inode(struct btrfs_root *root, | |
2894 | struct btrfs_path *path) | |
2895 | { | |
2896 | struct inode *inode = NULL; | |
2897 | ||
2898 | spin_lock(&root->cache_lock); | |
2899 | if (root->cache_inode) | |
2900 | inode = igrab(root->cache_inode); | |
2901 | spin_unlock(&root->cache_lock); | |
2902 | if (inode) | |
2903 | return inode; | |
2904 | ||
2905 | inode = __lookup_free_space_inode(root, path, 0); | |
2906 | if (IS_ERR(inode)) | |
2907 | return inode; | |
2908 | ||
2909 | spin_lock(&root->cache_lock); | |
7841cb28 | 2910 | if (!btrfs_fs_closing(root->fs_info)) |
82d5902d LZ |
2911 | root->cache_inode = igrab(inode); |
2912 | spin_unlock(&root->cache_lock); | |
2913 | ||
2914 | return inode; | |
2915 | } | |
2916 | ||
2917 | int create_free_ino_inode(struct btrfs_root *root, | |
2918 | struct btrfs_trans_handle *trans, | |
2919 | struct btrfs_path *path) | |
2920 | { | |
2921 | return __create_free_space_inode(root, trans, path, | |
2922 | BTRFS_FREE_INO_OBJECTID, 0); | |
2923 | } | |
2924 | ||
2925 | int load_free_ino_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root) | |
2926 | { | |
2927 | struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; | |
2928 | struct btrfs_path *path; | |
2929 | struct inode *inode; | |
2930 | int ret = 0; | |
2931 | u64 root_gen = btrfs_root_generation(&root->root_item); | |
2932 | ||
4b9465cb CM |
2933 | if (!btrfs_test_opt(root, INODE_MAP_CACHE)) |
2934 | return 0; | |
2935 | ||
82d5902d LZ |
2936 | /* |
2937 | * If we're unmounting then just return, since this does a search on the | |
2938 | * normal root and not the commit root and we could deadlock. | |
2939 | */ | |
7841cb28 | 2940 | if (btrfs_fs_closing(fs_info)) |
82d5902d LZ |
2941 | return 0; |
2942 | ||
2943 | path = btrfs_alloc_path(); | |
2944 | if (!path) | |
2945 | return 0; | |
2946 | ||
2947 | inode = lookup_free_ino_inode(root, path); | |
2948 | if (IS_ERR(inode)) | |
2949 | goto out; | |
2950 | ||
2951 | if (root_gen != BTRFS_I(inode)->generation) | |
2952 | goto out_put; | |
2953 | ||
2954 | ret = __load_free_space_cache(root, inode, ctl, path, 0); | |
2955 | ||
2956 | if (ret < 0) | |
c2cf52eb SK |
2957 | btrfs_err(fs_info, |
2958 | "failed to load free ino cache for root %llu", | |
2959 | root->root_key.objectid); | |
82d5902d LZ |
2960 | out_put: |
2961 | iput(inode); | |
2962 | out: | |
2963 | btrfs_free_path(path); | |
2964 | return ret; | |
2965 | } | |
2966 | ||
2967 | int btrfs_write_out_ino_cache(struct btrfs_root *root, | |
2968 | struct btrfs_trans_handle *trans, | |
53645a91 FDBM |
2969 | struct btrfs_path *path, |
2970 | struct inode *inode) | |
82d5902d LZ |
2971 | { |
2972 | struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; | |
82d5902d LZ |
2973 | int ret; |
2974 | ||
4b9465cb CM |
2975 | if (!btrfs_test_opt(root, INODE_MAP_CACHE)) |
2976 | return 0; | |
2977 | ||
82d5902d | 2978 | ret = __btrfs_write_out_cache(root, inode, ctl, NULL, trans, path, 0); |
c09544e0 JB |
2979 | if (ret) { |
2980 | btrfs_delalloc_release_metadata(inode, inode->i_size); | |
2981 | #ifdef DEBUG | |
c2cf52eb SK |
2982 | btrfs_err(root->fs_info, |
2983 | "failed to write free ino cache for root %llu", | |
2984 | root->root_key.objectid); | |
c09544e0 JB |
2985 | #endif |
2986 | } | |
82d5902d | 2987 | |
82d5902d LZ |
2988 | return ret; |
2989 | } | |
74255aa0 JB |
2990 | |
2991 | #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS | |
dc11dd5d JB |
2992 | /* |
2993 | * Use this if you need to make a bitmap or extent entry specifically, it | |
2994 | * doesn't do any of the merging that add_free_space does, this acts a lot like | |
2995 | * how the free space cache loading stuff works, so you can get really weird | |
2996 | * configurations. | |
2997 | */ | |
2998 | int test_add_free_space_entry(struct btrfs_block_group_cache *cache, | |
2999 | u64 offset, u64 bytes, bool bitmap) | |
74255aa0 | 3000 | { |
dc11dd5d JB |
3001 | struct btrfs_free_space_ctl *ctl = cache->free_space_ctl; |
3002 | struct btrfs_free_space *info = NULL, *bitmap_info; | |
3003 | void *map = NULL; | |
3004 | u64 bytes_added; | |
3005 | int ret; | |
74255aa0 | 3006 | |
dc11dd5d JB |
3007 | again: |
3008 | if (!info) { | |
3009 | info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS); | |
3010 | if (!info) | |
3011 | return -ENOMEM; | |
74255aa0 JB |
3012 | } |
3013 | ||
dc11dd5d JB |
3014 | if (!bitmap) { |
3015 | spin_lock(&ctl->tree_lock); | |
3016 | info->offset = offset; | |
3017 | info->bytes = bytes; | |
3018 | ret = link_free_space(ctl, info); | |
3019 | spin_unlock(&ctl->tree_lock); | |
3020 | if (ret) | |
3021 | kmem_cache_free(btrfs_free_space_cachep, info); | |
3022 | return ret; | |
3023 | } | |
3024 | ||
3025 | if (!map) { | |
3026 | map = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); | |
3027 | if (!map) { | |
3028 | kmem_cache_free(btrfs_free_space_cachep, info); | |
3029 | return -ENOMEM; | |
3030 | } | |
3031 | } | |
3032 | ||
3033 | spin_lock(&ctl->tree_lock); | |
3034 | bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), | |
3035 | 1, 0); | |
3036 | if (!bitmap_info) { | |
3037 | info->bitmap = map; | |
3038 | map = NULL; | |
3039 | add_new_bitmap(ctl, info, offset); | |
3040 | bitmap_info = info; | |
3041 | } | |
74255aa0 | 3042 | |
dc11dd5d JB |
3043 | bytes_added = add_bytes_to_bitmap(ctl, bitmap_info, offset, bytes); |
3044 | bytes -= bytes_added; | |
3045 | offset += bytes_added; | |
3046 | spin_unlock(&ctl->tree_lock); | |
74255aa0 | 3047 | |
dc11dd5d JB |
3048 | if (bytes) |
3049 | goto again; | |
74255aa0 | 3050 | |
dc11dd5d JB |
3051 | if (map) |
3052 | kfree(map); | |
3053 | return 0; | |
74255aa0 JB |
3054 | } |
3055 | ||
3056 | /* | |
3057 | * Checks to see if the given range is in the free space cache. This is really | |
3058 | * just used to check the absence of space, so if there is free space in the | |
3059 | * range at all we will return 1. | |
3060 | */ | |
dc11dd5d JB |
3061 | int test_check_exists(struct btrfs_block_group_cache *cache, |
3062 | u64 offset, u64 bytes) | |
74255aa0 JB |
3063 | { |
3064 | struct btrfs_free_space_ctl *ctl = cache->free_space_ctl; | |
3065 | struct btrfs_free_space *info; | |
3066 | int ret = 0; | |
3067 | ||
3068 | spin_lock(&ctl->tree_lock); | |
3069 | info = tree_search_offset(ctl, offset, 0, 0); | |
3070 | if (!info) { | |
3071 | info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), | |
3072 | 1, 0); | |
3073 | if (!info) | |
3074 | goto out; | |
3075 | } | |
3076 | ||
3077 | have_info: | |
3078 | if (info->bitmap) { | |
3079 | u64 bit_off, bit_bytes; | |
3080 | struct rb_node *n; | |
3081 | struct btrfs_free_space *tmp; | |
3082 | ||
3083 | bit_off = offset; | |
3084 | bit_bytes = ctl->unit; | |
3085 | ret = search_bitmap(ctl, info, &bit_off, &bit_bytes); | |
3086 | if (!ret) { | |
3087 | if (bit_off == offset) { | |
3088 | ret = 1; | |
3089 | goto out; | |
3090 | } else if (bit_off > offset && | |
3091 | offset + bytes > bit_off) { | |
3092 | ret = 1; | |
3093 | goto out; | |
3094 | } | |
3095 | } | |
3096 | ||
3097 | n = rb_prev(&info->offset_index); | |
3098 | while (n) { | |
3099 | tmp = rb_entry(n, struct btrfs_free_space, | |
3100 | offset_index); | |
3101 | if (tmp->offset + tmp->bytes < offset) | |
3102 | break; | |
3103 | if (offset + bytes < tmp->offset) { | |
3104 | n = rb_prev(&info->offset_index); | |
3105 | continue; | |
3106 | } | |
3107 | info = tmp; | |
3108 | goto have_info; | |
3109 | } | |
3110 | ||
3111 | n = rb_next(&info->offset_index); | |
3112 | while (n) { | |
3113 | tmp = rb_entry(n, struct btrfs_free_space, | |
3114 | offset_index); | |
3115 | if (offset + bytes < tmp->offset) | |
3116 | break; | |
3117 | if (tmp->offset + tmp->bytes < offset) { | |
3118 | n = rb_next(&info->offset_index); | |
3119 | continue; | |
3120 | } | |
3121 | info = tmp; | |
3122 | goto have_info; | |
3123 | } | |
3124 | ||
3125 | goto out; | |
3126 | } | |
3127 | ||
3128 | if (info->offset == offset) { | |
3129 | ret = 1; | |
3130 | goto out; | |
3131 | } | |
3132 | ||
3133 | if (offset > info->offset && offset < info->offset + info->bytes) | |
3134 | ret = 1; | |
3135 | out: | |
3136 | spin_unlock(&ctl->tree_lock); | |
3137 | return ret; | |
3138 | } | |
dc11dd5d | 3139 | #endif /* CONFIG_BTRFS_FS_RUN_SANITY_TESTS */ |