]>
Commit | Line | Data |
---|---|---|
a52d9a80 CM |
1 | #include <linux/bitops.h> |
2 | #include <linux/slab.h> | |
3 | #include <linux/bio.h> | |
4 | #include <linux/mm.h> | |
5 | #include <linux/gfp.h> | |
6 | #include <linux/pagemap.h> | |
7 | #include <linux/page-flags.h> | |
8 | #include <linux/module.h> | |
9 | #include <linux/spinlock.h> | |
10 | #include <linux/blkdev.h> | |
4dc11904 | 11 | #include <linux/swap.h> |
a52d9a80 CM |
12 | #include "extent_map.h" |
13 | ||
86479a04 CM |
14 | /* temporary define until extent_map moves out of btrfs */ |
15 | struct kmem_cache *btrfs_cache_create(const char *name, size_t size, | |
16 | unsigned long extra_flags, | |
17 | void (*ctor)(void *, struct kmem_cache *, | |
18 | unsigned long)); | |
19 | ||
a52d9a80 CM |
20 | static struct kmem_cache *extent_map_cache; |
21 | static struct kmem_cache *extent_state_cache; | |
6d36dcd4 | 22 | static struct kmem_cache *extent_buffer_cache; |
f510cfec | 23 | |
f510cfec CM |
24 | static LIST_HEAD(buffers); |
25 | static LIST_HEAD(states); | |
26 | ||
f510cfec | 27 | static spinlock_t state_lock = SPIN_LOCK_UNLOCKED; |
4dc11904 | 28 | #define BUFFER_LRU_MAX 64 |
a52d9a80 CM |
29 | |
30 | struct tree_entry { | |
31 | u64 start; | |
32 | u64 end; | |
33 | int in_tree; | |
34 | struct rb_node rb_node; | |
35 | }; | |
36 | ||
a52d9a80 CM |
37 | void __init extent_map_init(void) |
38 | { | |
86479a04 | 39 | extent_map_cache = btrfs_cache_create("extent_map", |
6d36dcd4 | 40 | sizeof(struct extent_map), 0, |
a52d9a80 | 41 | NULL); |
86479a04 | 42 | extent_state_cache = btrfs_cache_create("extent_state", |
6d36dcd4 | 43 | sizeof(struct extent_state), 0, |
a52d9a80 | 44 | NULL); |
6d36dcd4 CM |
45 | extent_buffer_cache = btrfs_cache_create("extent_buffers", |
46 | sizeof(struct extent_buffer), 0, | |
47 | NULL); | |
a52d9a80 CM |
48 | } |
49 | ||
50 | void __exit extent_map_exit(void) | |
51 | { | |
f510cfec | 52 | struct extent_state *state; |
6d36dcd4 | 53 | |
f510cfec CM |
54 | while (!list_empty(&states)) { |
55 | state = list_entry(states.next, struct extent_state, list); | |
56 | printk("state leak: start %Lu end %Lu state %lu in tree %d refs %d\n", state->start, state->end, state->state, state->in_tree, atomic_read(&state->refs)); | |
57 | list_del(&state->list); | |
58 | kmem_cache_free(extent_state_cache, state); | |
59 | ||
60 | } | |
f510cfec | 61 | |
a52d9a80 CM |
62 | if (extent_map_cache) |
63 | kmem_cache_destroy(extent_map_cache); | |
64 | if (extent_state_cache) | |
65 | kmem_cache_destroy(extent_state_cache); | |
6d36dcd4 CM |
66 | if (extent_buffer_cache) |
67 | kmem_cache_destroy(extent_buffer_cache); | |
a52d9a80 CM |
68 | } |
69 | ||
70 | void extent_map_tree_init(struct extent_map_tree *tree, | |
71 | struct address_space *mapping, gfp_t mask) | |
72 | { | |
73 | tree->map.rb_node = NULL; | |
74 | tree->state.rb_node = NULL; | |
07157aac | 75 | tree->ops = NULL; |
a52d9a80 | 76 | rwlock_init(&tree->lock); |
4dc11904 | 77 | spin_lock_init(&tree->lru_lock); |
a52d9a80 | 78 | tree->mapping = mapping; |
4dc11904 CM |
79 | INIT_LIST_HEAD(&tree->buffer_lru); |
80 | tree->lru_size = 0; | |
a52d9a80 CM |
81 | } |
82 | EXPORT_SYMBOL(extent_map_tree_init); | |
83 | ||
19c00ddc | 84 | void extent_map_tree_empty_lru(struct extent_map_tree *tree) |
4dc11904 CM |
85 | { |
86 | struct extent_buffer *eb; | |
87 | while(!list_empty(&tree->buffer_lru)) { | |
88 | eb = list_entry(tree->buffer_lru.next, struct extent_buffer, | |
89 | lru); | |
90 | list_del(&eb->lru); | |
91 | free_extent_buffer(eb); | |
92 | } | |
93 | } | |
19c00ddc | 94 | EXPORT_SYMBOL(extent_map_tree_empty_lru); |
4dc11904 | 95 | |
a52d9a80 CM |
96 | struct extent_map *alloc_extent_map(gfp_t mask) |
97 | { | |
98 | struct extent_map *em; | |
99 | em = kmem_cache_alloc(extent_map_cache, mask); | |
100 | if (!em || IS_ERR(em)) | |
101 | return em; | |
102 | em->in_tree = 0; | |
103 | atomic_set(&em->refs, 1); | |
104 | return em; | |
105 | } | |
106 | EXPORT_SYMBOL(alloc_extent_map); | |
107 | ||
108 | void free_extent_map(struct extent_map *em) | |
109 | { | |
2bf5a725 CM |
110 | if (!em) |
111 | return; | |
a52d9a80 CM |
112 | if (atomic_dec_and_test(&em->refs)) { |
113 | WARN_ON(em->in_tree); | |
114 | kmem_cache_free(extent_map_cache, em); | |
115 | } | |
116 | } | |
117 | EXPORT_SYMBOL(free_extent_map); | |
118 | ||
119 | ||
120 | struct extent_state *alloc_extent_state(gfp_t mask) | |
121 | { | |
122 | struct extent_state *state; | |
f510cfec CM |
123 | unsigned long flags; |
124 | ||
a52d9a80 CM |
125 | state = kmem_cache_alloc(extent_state_cache, mask); |
126 | if (!state || IS_ERR(state)) | |
127 | return state; | |
128 | state->state = 0; | |
129 | state->in_tree = 0; | |
07157aac | 130 | state->private = 0; |
f510cfec CM |
131 | |
132 | spin_lock_irqsave(&state_lock, flags); | |
133 | list_add(&state->list, &states); | |
134 | spin_unlock_irqrestore(&state_lock, flags); | |
135 | ||
a52d9a80 CM |
136 | atomic_set(&state->refs, 1); |
137 | init_waitqueue_head(&state->wq); | |
a52d9a80 CM |
138 | return state; |
139 | } | |
140 | EXPORT_SYMBOL(alloc_extent_state); | |
141 | ||
142 | void free_extent_state(struct extent_state *state) | |
143 | { | |
f510cfec | 144 | unsigned long flags; |
2bf5a725 CM |
145 | if (!state) |
146 | return; | |
a52d9a80 CM |
147 | if (atomic_dec_and_test(&state->refs)) { |
148 | WARN_ON(state->in_tree); | |
f510cfec CM |
149 | spin_lock_irqsave(&state_lock, flags); |
150 | list_del(&state->list); | |
151 | spin_unlock_irqrestore(&state_lock, flags); | |
a52d9a80 CM |
152 | kmem_cache_free(extent_state_cache, state); |
153 | } | |
154 | } | |
155 | EXPORT_SYMBOL(free_extent_state); | |
156 | ||
157 | static struct rb_node *tree_insert(struct rb_root *root, u64 offset, | |
158 | struct rb_node *node) | |
159 | { | |
160 | struct rb_node ** p = &root->rb_node; | |
161 | struct rb_node * parent = NULL; | |
162 | struct tree_entry *entry; | |
163 | ||
164 | while(*p) { | |
165 | parent = *p; | |
166 | entry = rb_entry(parent, struct tree_entry, rb_node); | |
167 | ||
168 | if (offset < entry->start) | |
169 | p = &(*p)->rb_left; | |
170 | else if (offset > entry->end) | |
171 | p = &(*p)->rb_right; | |
172 | else | |
173 | return parent; | |
174 | } | |
175 | ||
176 | entry = rb_entry(node, struct tree_entry, rb_node); | |
177 | entry->in_tree = 1; | |
178 | rb_link_node(node, parent, p); | |
179 | rb_insert_color(node, root); | |
180 | return NULL; | |
181 | } | |
182 | ||
183 | static struct rb_node *__tree_search(struct rb_root *root, u64 offset, | |
184 | struct rb_node **prev_ret) | |
185 | { | |
186 | struct rb_node * n = root->rb_node; | |
187 | struct rb_node *prev = NULL; | |
188 | struct tree_entry *entry; | |
189 | struct tree_entry *prev_entry = NULL; | |
190 | ||
191 | while(n) { | |
192 | entry = rb_entry(n, struct tree_entry, rb_node); | |
193 | prev = n; | |
194 | prev_entry = entry; | |
195 | ||
196 | if (offset < entry->start) | |
197 | n = n->rb_left; | |
198 | else if (offset > entry->end) | |
199 | n = n->rb_right; | |
200 | else | |
201 | return n; | |
202 | } | |
203 | if (!prev_ret) | |
204 | return NULL; | |
205 | while(prev && offset > prev_entry->end) { | |
206 | prev = rb_next(prev); | |
207 | prev_entry = rb_entry(prev, struct tree_entry, rb_node); | |
208 | } | |
209 | *prev_ret = prev; | |
210 | return NULL; | |
211 | } | |
212 | ||
213 | static inline struct rb_node *tree_search(struct rb_root *root, u64 offset) | |
214 | { | |
215 | struct rb_node *prev; | |
216 | struct rb_node *ret; | |
217 | ret = __tree_search(root, offset, &prev); | |
218 | if (!ret) | |
219 | return prev; | |
220 | return ret; | |
221 | } | |
222 | ||
223 | static int tree_delete(struct rb_root *root, u64 offset) | |
224 | { | |
225 | struct rb_node *node; | |
226 | struct tree_entry *entry; | |
227 | ||
228 | node = __tree_search(root, offset, NULL); | |
229 | if (!node) | |
230 | return -ENOENT; | |
231 | entry = rb_entry(node, struct tree_entry, rb_node); | |
232 | entry->in_tree = 0; | |
233 | rb_erase(node, root); | |
234 | return 0; | |
235 | } | |
236 | ||
237 | /* | |
238 | * add_extent_mapping tries a simple backward merge with existing | |
239 | * mappings. The extent_map struct passed in will be inserted into | |
240 | * the tree directly (no copies made, just a reference taken). | |
241 | */ | |
242 | int add_extent_mapping(struct extent_map_tree *tree, | |
243 | struct extent_map *em) | |
244 | { | |
245 | int ret = 0; | |
246 | struct extent_map *prev = NULL; | |
247 | struct rb_node *rb; | |
248 | ||
249 | write_lock_irq(&tree->lock); | |
250 | rb = tree_insert(&tree->map, em->end, &em->rb_node); | |
251 | if (rb) { | |
252 | prev = rb_entry(rb, struct extent_map, rb_node); | |
253 | printk("found extent map %Lu %Lu on insert of %Lu %Lu\n", prev->start, prev->end, em->start, em->end); | |
254 | ret = -EEXIST; | |
255 | goto out; | |
256 | } | |
257 | atomic_inc(&em->refs); | |
258 | if (em->start != 0) { | |
259 | rb = rb_prev(&em->rb_node); | |
260 | if (rb) | |
261 | prev = rb_entry(rb, struct extent_map, rb_node); | |
262 | if (prev && prev->end + 1 == em->start && | |
5f39d397 CM |
263 | ((em->block_start == EXTENT_MAP_HOLE && |
264 | prev->block_start == EXTENT_MAP_HOLE) || | |
a52d9a80 CM |
265 | (em->block_start == prev->block_end + 1))) { |
266 | em->start = prev->start; | |
267 | em->block_start = prev->block_start; | |
268 | rb_erase(&prev->rb_node, &tree->map); | |
269 | prev->in_tree = 0; | |
270 | free_extent_map(prev); | |
271 | } | |
272 | } | |
273 | out: | |
274 | write_unlock_irq(&tree->lock); | |
275 | return ret; | |
276 | } | |
277 | EXPORT_SYMBOL(add_extent_mapping); | |
278 | ||
279 | /* | |
280 | * lookup_extent_mapping returns the first extent_map struct in the | |
281 | * tree that intersects the [start, end] (inclusive) range. There may | |
282 | * be additional objects in the tree that intersect, so check the object | |
283 | * returned carefully to make sure you don't need additional lookups. | |
284 | */ | |
285 | struct extent_map *lookup_extent_mapping(struct extent_map_tree *tree, | |
286 | u64 start, u64 end) | |
287 | { | |
288 | struct extent_map *em; | |
289 | struct rb_node *rb_node; | |
290 | ||
291 | read_lock_irq(&tree->lock); | |
292 | rb_node = tree_search(&tree->map, start); | |
293 | if (!rb_node) { | |
294 | em = NULL; | |
295 | goto out; | |
296 | } | |
297 | if (IS_ERR(rb_node)) { | |
298 | em = ERR_PTR(PTR_ERR(rb_node)); | |
299 | goto out; | |
300 | } | |
301 | em = rb_entry(rb_node, struct extent_map, rb_node); | |
302 | if (em->end < start || em->start > end) { | |
303 | em = NULL; | |
304 | goto out; | |
305 | } | |
306 | atomic_inc(&em->refs); | |
307 | out: | |
308 | read_unlock_irq(&tree->lock); | |
309 | return em; | |
310 | } | |
311 | EXPORT_SYMBOL(lookup_extent_mapping); | |
312 | ||
313 | /* | |
314 | * removes an extent_map struct from the tree. No reference counts are | |
315 | * dropped, and no checks are done to see if the range is in use | |
316 | */ | |
317 | int remove_extent_mapping(struct extent_map_tree *tree, struct extent_map *em) | |
318 | { | |
319 | int ret; | |
320 | ||
321 | write_lock_irq(&tree->lock); | |
322 | ret = tree_delete(&tree->map, em->end); | |
323 | write_unlock_irq(&tree->lock); | |
324 | return ret; | |
325 | } | |
326 | EXPORT_SYMBOL(remove_extent_mapping); | |
327 | ||
328 | /* | |
329 | * utility function to look for merge candidates inside a given range. | |
330 | * Any extents with matching state are merged together into a single | |
331 | * extent in the tree. Extents with EXTENT_IO in their state field | |
332 | * are not merged because the end_io handlers need to be able to do | |
333 | * operations on them without sleeping (or doing allocations/splits). | |
334 | * | |
335 | * This should be called with the tree lock held. | |
336 | */ | |
337 | static int merge_state(struct extent_map_tree *tree, | |
338 | struct extent_state *state) | |
339 | { | |
340 | struct extent_state *other; | |
341 | struct rb_node *other_node; | |
342 | ||
343 | if (state->state & EXTENT_IOBITS) | |
344 | return 0; | |
345 | ||
346 | other_node = rb_prev(&state->rb_node); | |
347 | if (other_node) { | |
348 | other = rb_entry(other_node, struct extent_state, rb_node); | |
349 | if (other->end == state->start - 1 && | |
350 | other->state == state->state) { | |
351 | state->start = other->start; | |
352 | other->in_tree = 0; | |
353 | rb_erase(&other->rb_node, &tree->state); | |
354 | free_extent_state(other); | |
355 | } | |
356 | } | |
357 | other_node = rb_next(&state->rb_node); | |
358 | if (other_node) { | |
359 | other = rb_entry(other_node, struct extent_state, rb_node); | |
360 | if (other->start == state->end + 1 && | |
361 | other->state == state->state) { | |
362 | other->start = state->start; | |
363 | state->in_tree = 0; | |
364 | rb_erase(&state->rb_node, &tree->state); | |
365 | free_extent_state(state); | |
366 | } | |
367 | } | |
368 | return 0; | |
369 | } | |
370 | ||
371 | /* | |
372 | * insert an extent_state struct into the tree. 'bits' are set on the | |
373 | * struct before it is inserted. | |
374 | * | |
375 | * This may return -EEXIST if the extent is already there, in which case the | |
376 | * state struct is freed. | |
377 | * | |
378 | * The tree lock is not taken internally. This is a utility function and | |
379 | * probably isn't what you want to call (see set/clear_extent_bit). | |
380 | */ | |
381 | static int insert_state(struct extent_map_tree *tree, | |
382 | struct extent_state *state, u64 start, u64 end, | |
383 | int bits) | |
384 | { | |
385 | struct rb_node *node; | |
386 | ||
387 | if (end < start) { | |
388 | printk("end < start %Lu %Lu\n", end, start); | |
389 | WARN_ON(1); | |
390 | } | |
391 | state->state |= bits; | |
392 | state->start = start; | |
393 | state->end = end; | |
a52d9a80 CM |
394 | node = tree_insert(&tree->state, end, &state->rb_node); |
395 | if (node) { | |
396 | struct extent_state *found; | |
397 | found = rb_entry(node, struct extent_state, rb_node); | |
b888db2b | 398 | printk("found node %Lu %Lu on insert of %Lu %Lu\n", found->start, found->end, start, end); |
a52d9a80 CM |
399 | free_extent_state(state); |
400 | return -EEXIST; | |
401 | } | |
402 | merge_state(tree, state); | |
403 | return 0; | |
404 | } | |
405 | ||
406 | /* | |
407 | * split a given extent state struct in two, inserting the preallocated | |
408 | * struct 'prealloc' as the newly created second half. 'split' indicates an | |
409 | * offset inside 'orig' where it should be split. | |
410 | * | |
411 | * Before calling, | |
412 | * the tree has 'orig' at [orig->start, orig->end]. After calling, there | |
413 | * are two extent state structs in the tree: | |
414 | * prealloc: [orig->start, split - 1] | |
415 | * orig: [ split, orig->end ] | |
416 | * | |
417 | * The tree locks are not taken by this function. They need to be held | |
418 | * by the caller. | |
419 | */ | |
420 | static int split_state(struct extent_map_tree *tree, struct extent_state *orig, | |
421 | struct extent_state *prealloc, u64 split) | |
422 | { | |
423 | struct rb_node *node; | |
424 | prealloc->start = orig->start; | |
425 | prealloc->end = split - 1; | |
426 | prealloc->state = orig->state; | |
427 | orig->start = split; | |
f510cfec | 428 | |
a52d9a80 CM |
429 | node = tree_insert(&tree->state, prealloc->end, &prealloc->rb_node); |
430 | if (node) { | |
431 | struct extent_state *found; | |
432 | found = rb_entry(node, struct extent_state, rb_node); | |
b888db2b | 433 | printk("found node %Lu %Lu on insert of %Lu %Lu\n", found->start, found->end, prealloc->start, prealloc->end); |
a52d9a80 CM |
434 | free_extent_state(prealloc); |
435 | return -EEXIST; | |
436 | } | |
437 | return 0; | |
438 | } | |
439 | ||
440 | /* | |
441 | * utility function to clear some bits in an extent state struct. | |
442 | * it will optionally wake up any one waiting on this state (wake == 1), or | |
443 | * forcibly remove the state from the tree (delete == 1). | |
444 | * | |
445 | * If no bits are set on the state struct after clearing things, the | |
446 | * struct is freed and removed from the tree | |
447 | */ | |
448 | static int clear_state_bit(struct extent_map_tree *tree, | |
449 | struct extent_state *state, int bits, int wake, | |
450 | int delete) | |
451 | { | |
452 | int ret = state->state & bits; | |
453 | state->state &= ~bits; | |
454 | if (wake) | |
455 | wake_up(&state->wq); | |
456 | if (delete || state->state == 0) { | |
457 | if (state->in_tree) { | |
458 | rb_erase(&state->rb_node, &tree->state); | |
459 | state->in_tree = 0; | |
460 | free_extent_state(state); | |
461 | } else { | |
462 | WARN_ON(1); | |
463 | } | |
464 | } else { | |
465 | merge_state(tree, state); | |
466 | } | |
467 | return ret; | |
468 | } | |
469 | ||
470 | /* | |
471 | * clear some bits on a range in the tree. This may require splitting | |
472 | * or inserting elements in the tree, so the gfp mask is used to | |
473 | * indicate which allocations or sleeping are allowed. | |
474 | * | |
475 | * pass 'wake' == 1 to kick any sleepers, and 'delete' == 1 to remove | |
476 | * the given range from the tree regardless of state (ie for truncate). | |
477 | * | |
478 | * the range [start, end] is inclusive. | |
479 | * | |
480 | * This takes the tree lock, and returns < 0 on error, > 0 if any of the | |
481 | * bits were already set, or zero if none of the bits were already set. | |
482 | */ | |
483 | int clear_extent_bit(struct extent_map_tree *tree, u64 start, u64 end, | |
484 | int bits, int wake, int delete, gfp_t mask) | |
485 | { | |
486 | struct extent_state *state; | |
487 | struct extent_state *prealloc = NULL; | |
488 | struct rb_node *node; | |
90f1c19a | 489 | unsigned long flags; |
a52d9a80 CM |
490 | int err; |
491 | int set = 0; | |
492 | ||
493 | again: | |
494 | if (!prealloc && (mask & __GFP_WAIT)) { | |
495 | prealloc = alloc_extent_state(mask); | |
496 | if (!prealloc) | |
497 | return -ENOMEM; | |
498 | } | |
499 | ||
90f1c19a | 500 | write_lock_irqsave(&tree->lock, flags); |
a52d9a80 CM |
501 | /* |
502 | * this search will find the extents that end after | |
503 | * our range starts | |
504 | */ | |
505 | node = tree_search(&tree->state, start); | |
506 | if (!node) | |
507 | goto out; | |
508 | state = rb_entry(node, struct extent_state, rb_node); | |
509 | if (state->start > end) | |
510 | goto out; | |
511 | WARN_ON(state->end < start); | |
512 | ||
513 | /* | |
514 | * | ---- desired range ---- | | |
515 | * | state | or | |
516 | * | ------------- state -------------- | | |
517 | * | |
518 | * We need to split the extent we found, and may flip | |
519 | * bits on second half. | |
520 | * | |
521 | * If the extent we found extends past our range, we | |
522 | * just split and search again. It'll get split again | |
523 | * the next time though. | |
524 | * | |
525 | * If the extent we found is inside our range, we clear | |
526 | * the desired bit on it. | |
527 | */ | |
528 | ||
529 | if (state->start < start) { | |
530 | err = split_state(tree, state, prealloc, start); | |
531 | BUG_ON(err == -EEXIST); | |
532 | prealloc = NULL; | |
533 | if (err) | |
534 | goto out; | |
535 | if (state->end <= end) { | |
536 | start = state->end + 1; | |
537 | set |= clear_state_bit(tree, state, bits, | |
538 | wake, delete); | |
539 | } else { | |
540 | start = state->start; | |
541 | } | |
542 | goto search_again; | |
543 | } | |
544 | /* | |
545 | * | ---- desired range ---- | | |
546 | * | state | | |
547 | * We need to split the extent, and clear the bit | |
548 | * on the first half | |
549 | */ | |
550 | if (state->start <= end && state->end > end) { | |
551 | err = split_state(tree, state, prealloc, end + 1); | |
552 | BUG_ON(err == -EEXIST); | |
553 | ||
554 | if (wake) | |
555 | wake_up(&state->wq); | |
556 | set |= clear_state_bit(tree, prealloc, bits, | |
557 | wake, delete); | |
558 | prealloc = NULL; | |
559 | goto out; | |
560 | } | |
561 | ||
562 | start = state->end + 1; | |
563 | set |= clear_state_bit(tree, state, bits, wake, delete); | |
564 | goto search_again; | |
565 | ||
566 | out: | |
90f1c19a | 567 | write_unlock_irqrestore(&tree->lock, flags); |
a52d9a80 CM |
568 | if (prealloc) |
569 | free_extent_state(prealloc); | |
570 | ||
571 | return set; | |
572 | ||
573 | search_again: | |
96b5179d | 574 | if (start > end) |
a52d9a80 | 575 | goto out; |
90f1c19a | 576 | write_unlock_irqrestore(&tree->lock, flags); |
a52d9a80 CM |
577 | if (mask & __GFP_WAIT) |
578 | cond_resched(); | |
579 | goto again; | |
580 | } | |
581 | EXPORT_SYMBOL(clear_extent_bit); | |
582 | ||
583 | static int wait_on_state(struct extent_map_tree *tree, | |
584 | struct extent_state *state) | |
585 | { | |
586 | DEFINE_WAIT(wait); | |
587 | prepare_to_wait(&state->wq, &wait, TASK_UNINTERRUPTIBLE); | |
588 | read_unlock_irq(&tree->lock); | |
589 | schedule(); | |
590 | read_lock_irq(&tree->lock); | |
591 | finish_wait(&state->wq, &wait); | |
592 | return 0; | |
593 | } | |
594 | ||
595 | /* | |
596 | * waits for one or more bits to clear on a range in the state tree. | |
597 | * The range [start, end] is inclusive. | |
598 | * The tree lock is taken by this function | |
599 | */ | |
600 | int wait_extent_bit(struct extent_map_tree *tree, u64 start, u64 end, int bits) | |
601 | { | |
602 | struct extent_state *state; | |
603 | struct rb_node *node; | |
604 | ||
605 | read_lock_irq(&tree->lock); | |
606 | again: | |
607 | while (1) { | |
608 | /* | |
609 | * this search will find all the extents that end after | |
610 | * our range starts | |
611 | */ | |
612 | node = tree_search(&tree->state, start); | |
613 | if (!node) | |
614 | break; | |
615 | ||
616 | state = rb_entry(node, struct extent_state, rb_node); | |
617 | ||
618 | if (state->start > end) | |
619 | goto out; | |
620 | ||
621 | if (state->state & bits) { | |
622 | start = state->start; | |
623 | atomic_inc(&state->refs); | |
624 | wait_on_state(tree, state); | |
625 | free_extent_state(state); | |
626 | goto again; | |
627 | } | |
628 | start = state->end + 1; | |
629 | ||
630 | if (start > end) | |
631 | break; | |
632 | ||
633 | if (need_resched()) { | |
634 | read_unlock_irq(&tree->lock); | |
635 | cond_resched(); | |
636 | read_lock_irq(&tree->lock); | |
637 | } | |
638 | } | |
639 | out: | |
640 | read_unlock_irq(&tree->lock); | |
641 | return 0; | |
642 | } | |
643 | EXPORT_SYMBOL(wait_extent_bit); | |
644 | ||
645 | /* | |
646 | * set some bits on a range in the tree. This may require allocations | |
647 | * or sleeping, so the gfp mask is used to indicate what is allowed. | |
648 | * | |
649 | * If 'exclusive' == 1, this will fail with -EEXIST if some part of the | |
650 | * range already has the desired bits set. The start of the existing | |
651 | * range is returned in failed_start in this case. | |
652 | * | |
653 | * [start, end] is inclusive | |
654 | * This takes the tree lock. | |
655 | */ | |
656 | int set_extent_bit(struct extent_map_tree *tree, u64 start, u64 end, int bits, | |
657 | int exclusive, u64 *failed_start, gfp_t mask) | |
658 | { | |
659 | struct extent_state *state; | |
660 | struct extent_state *prealloc = NULL; | |
661 | struct rb_node *node; | |
90f1c19a | 662 | unsigned long flags; |
a52d9a80 CM |
663 | int err = 0; |
664 | int set; | |
665 | u64 last_start; | |
666 | u64 last_end; | |
667 | again: | |
668 | if (!prealloc && (mask & __GFP_WAIT)) { | |
669 | prealloc = alloc_extent_state(mask); | |
670 | if (!prealloc) | |
671 | return -ENOMEM; | |
672 | } | |
673 | ||
90f1c19a | 674 | write_lock_irqsave(&tree->lock, flags); |
a52d9a80 CM |
675 | /* |
676 | * this search will find all the extents that end after | |
677 | * our range starts. | |
678 | */ | |
679 | node = tree_search(&tree->state, start); | |
680 | if (!node) { | |
681 | err = insert_state(tree, prealloc, start, end, bits); | |
682 | prealloc = NULL; | |
683 | BUG_ON(err == -EEXIST); | |
684 | goto out; | |
685 | } | |
686 | ||
687 | state = rb_entry(node, struct extent_state, rb_node); | |
688 | last_start = state->start; | |
689 | last_end = state->end; | |
690 | ||
691 | /* | |
692 | * | ---- desired range ---- | | |
693 | * | state | | |
694 | * | |
695 | * Just lock what we found and keep going | |
696 | */ | |
697 | if (state->start == start && state->end <= end) { | |
698 | set = state->state & bits; | |
699 | if (set && exclusive) { | |
700 | *failed_start = state->start; | |
701 | err = -EEXIST; | |
702 | goto out; | |
703 | } | |
704 | state->state |= bits; | |
705 | start = state->end + 1; | |
706 | merge_state(tree, state); | |
707 | goto search_again; | |
708 | } | |
709 | ||
710 | /* | |
711 | * | ---- desired range ---- | | |
712 | * | state | | |
713 | * or | |
714 | * | ------------- state -------------- | | |
715 | * | |
716 | * We need to split the extent we found, and may flip bits on | |
717 | * second half. | |
718 | * | |
719 | * If the extent we found extends past our | |
720 | * range, we just split and search again. It'll get split | |
721 | * again the next time though. | |
722 | * | |
723 | * If the extent we found is inside our range, we set the | |
724 | * desired bit on it. | |
725 | */ | |
726 | if (state->start < start) { | |
727 | set = state->state & bits; | |
728 | if (exclusive && set) { | |
729 | *failed_start = start; | |
730 | err = -EEXIST; | |
731 | goto out; | |
732 | } | |
733 | err = split_state(tree, state, prealloc, start); | |
734 | BUG_ON(err == -EEXIST); | |
735 | prealloc = NULL; | |
736 | if (err) | |
737 | goto out; | |
738 | if (state->end <= end) { | |
739 | state->state |= bits; | |
740 | start = state->end + 1; | |
741 | merge_state(tree, state); | |
742 | } else { | |
743 | start = state->start; | |
744 | } | |
745 | goto search_again; | |
746 | } | |
a52d9a80 CM |
747 | /* |
748 | * | ---- desired range ---- | | |
749 | * | state | or | state | | |
750 | * | |
751 | * There's a hole, we need to insert something in it and | |
752 | * ignore the extent we found. | |
753 | */ | |
754 | if (state->start > start) { | |
755 | u64 this_end; | |
756 | if (end < last_start) | |
757 | this_end = end; | |
758 | else | |
759 | this_end = last_start -1; | |
760 | err = insert_state(tree, prealloc, start, this_end, | |
761 | bits); | |
762 | prealloc = NULL; | |
763 | BUG_ON(err == -EEXIST); | |
764 | if (err) | |
765 | goto out; | |
766 | start = this_end + 1; | |
767 | goto search_again; | |
768 | } | |
a8c450b2 CM |
769 | /* |
770 | * | ---- desired range ---- | | |
771 | * | state | | |
772 | * We need to split the extent, and set the bit | |
773 | * on the first half | |
774 | */ | |
775 | if (state->start <= end && state->end > end) { | |
776 | set = state->state & bits; | |
777 | if (exclusive && set) { | |
778 | *failed_start = start; | |
779 | err = -EEXIST; | |
780 | goto out; | |
781 | } | |
782 | err = split_state(tree, state, prealloc, end + 1); | |
783 | BUG_ON(err == -EEXIST); | |
784 | ||
785 | prealloc->state |= bits; | |
786 | merge_state(tree, prealloc); | |
787 | prealloc = NULL; | |
788 | goto out; | |
789 | } | |
790 | ||
a52d9a80 CM |
791 | goto search_again; |
792 | ||
793 | out: | |
90f1c19a | 794 | write_unlock_irqrestore(&tree->lock, flags); |
a52d9a80 CM |
795 | if (prealloc) |
796 | free_extent_state(prealloc); | |
797 | ||
798 | return err; | |
799 | ||
800 | search_again: | |
801 | if (start > end) | |
802 | goto out; | |
90f1c19a | 803 | write_unlock_irqrestore(&tree->lock, flags); |
a52d9a80 CM |
804 | if (mask & __GFP_WAIT) |
805 | cond_resched(); | |
806 | goto again; | |
807 | } | |
808 | EXPORT_SYMBOL(set_extent_bit); | |
809 | ||
810 | /* wrappers around set/clear extent bit */ | |
811 | int set_extent_dirty(struct extent_map_tree *tree, u64 start, u64 end, | |
812 | gfp_t mask) | |
813 | { | |
814 | return set_extent_bit(tree, start, end, EXTENT_DIRTY, 0, NULL, | |
815 | mask); | |
816 | } | |
817 | EXPORT_SYMBOL(set_extent_dirty); | |
818 | ||
96b5179d CM |
819 | int set_extent_bits(struct extent_map_tree *tree, u64 start, u64 end, |
820 | int bits, gfp_t mask) | |
821 | { | |
822 | return set_extent_bit(tree, start, end, bits, 0, NULL, | |
823 | mask); | |
824 | } | |
825 | EXPORT_SYMBOL(set_extent_bits); | |
826 | ||
827 | int clear_extent_bits(struct extent_map_tree *tree, u64 start, u64 end, | |
828 | int bits, gfp_t mask) | |
829 | { | |
830 | return clear_extent_bit(tree, start, end, bits, 0, 0, mask); | |
831 | } | |
832 | EXPORT_SYMBOL(clear_extent_bits); | |
833 | ||
b888db2b CM |
834 | int set_extent_delalloc(struct extent_map_tree *tree, u64 start, u64 end, |
835 | gfp_t mask) | |
836 | { | |
837 | return set_extent_bit(tree, start, end, | |
838 | EXTENT_DELALLOC | EXTENT_DIRTY, 0, NULL, | |
839 | mask); | |
840 | } | |
841 | EXPORT_SYMBOL(set_extent_delalloc); | |
842 | ||
a52d9a80 CM |
843 | int clear_extent_dirty(struct extent_map_tree *tree, u64 start, u64 end, |
844 | gfp_t mask) | |
845 | { | |
b888db2b CM |
846 | return clear_extent_bit(tree, start, end, |
847 | EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, mask); | |
a52d9a80 CM |
848 | } |
849 | EXPORT_SYMBOL(clear_extent_dirty); | |
850 | ||
851 | int set_extent_new(struct extent_map_tree *tree, u64 start, u64 end, | |
852 | gfp_t mask) | |
853 | { | |
854 | return set_extent_bit(tree, start, end, EXTENT_NEW, 0, NULL, | |
855 | mask); | |
856 | } | |
857 | EXPORT_SYMBOL(set_extent_new); | |
858 | ||
859 | int clear_extent_new(struct extent_map_tree *tree, u64 start, u64 end, | |
860 | gfp_t mask) | |
861 | { | |
862 | return clear_extent_bit(tree, start, end, EXTENT_NEW, 0, 0, mask); | |
863 | } | |
864 | EXPORT_SYMBOL(clear_extent_new); | |
865 | ||
866 | int set_extent_uptodate(struct extent_map_tree *tree, u64 start, u64 end, | |
867 | gfp_t mask) | |
868 | { | |
869 | return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, NULL, | |
870 | mask); | |
871 | } | |
872 | EXPORT_SYMBOL(set_extent_uptodate); | |
873 | ||
874 | int clear_extent_uptodate(struct extent_map_tree *tree, u64 start, u64 end, | |
875 | gfp_t mask) | |
876 | { | |
877 | return clear_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, 0, mask); | |
878 | } | |
879 | EXPORT_SYMBOL(clear_extent_uptodate); | |
880 | ||
881 | int set_extent_writeback(struct extent_map_tree *tree, u64 start, u64 end, | |
882 | gfp_t mask) | |
883 | { | |
884 | return set_extent_bit(tree, start, end, EXTENT_WRITEBACK, | |
885 | 0, NULL, mask); | |
886 | } | |
887 | EXPORT_SYMBOL(set_extent_writeback); | |
888 | ||
889 | int clear_extent_writeback(struct extent_map_tree *tree, u64 start, u64 end, | |
890 | gfp_t mask) | |
891 | { | |
892 | return clear_extent_bit(tree, start, end, EXTENT_WRITEBACK, 1, 0, mask); | |
893 | } | |
894 | EXPORT_SYMBOL(clear_extent_writeback); | |
895 | ||
896 | int wait_on_extent_writeback(struct extent_map_tree *tree, u64 start, u64 end) | |
897 | { | |
898 | return wait_extent_bit(tree, start, end, EXTENT_WRITEBACK); | |
899 | } | |
900 | EXPORT_SYMBOL(wait_on_extent_writeback); | |
901 | ||
902 | /* | |
903 | * locks a range in ascending order, waiting for any locked regions | |
904 | * it hits on the way. [start,end] are inclusive, and this will sleep. | |
905 | */ | |
906 | int lock_extent(struct extent_map_tree *tree, u64 start, u64 end, gfp_t mask) | |
907 | { | |
908 | int err; | |
909 | u64 failed_start; | |
910 | while (1) { | |
911 | err = set_extent_bit(tree, start, end, EXTENT_LOCKED, 1, | |
912 | &failed_start, mask); | |
913 | if (err == -EEXIST && (mask & __GFP_WAIT)) { | |
914 | wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED); | |
915 | start = failed_start; | |
916 | } else { | |
917 | break; | |
918 | } | |
919 | WARN_ON(start > end); | |
920 | } | |
921 | return err; | |
922 | } | |
923 | EXPORT_SYMBOL(lock_extent); | |
924 | ||
925 | int unlock_extent(struct extent_map_tree *tree, u64 start, u64 end, | |
926 | gfp_t mask) | |
927 | { | |
928 | return clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, mask); | |
929 | } | |
930 | EXPORT_SYMBOL(unlock_extent); | |
931 | ||
932 | /* | |
933 | * helper function to set pages and extents in the tree dirty | |
934 | */ | |
935 | int set_range_dirty(struct extent_map_tree *tree, u64 start, u64 end) | |
936 | { | |
937 | unsigned long index = start >> PAGE_CACHE_SHIFT; | |
938 | unsigned long end_index = end >> PAGE_CACHE_SHIFT; | |
939 | struct page *page; | |
940 | ||
941 | while (index <= end_index) { | |
942 | page = find_get_page(tree->mapping, index); | |
943 | BUG_ON(!page); | |
944 | __set_page_dirty_nobuffers(page); | |
945 | page_cache_release(page); | |
946 | index++; | |
947 | } | |
948 | set_extent_dirty(tree, start, end, GFP_NOFS); | |
949 | return 0; | |
950 | } | |
951 | EXPORT_SYMBOL(set_range_dirty); | |
952 | ||
953 | /* | |
954 | * helper function to set both pages and extents in the tree writeback | |
955 | */ | |
956 | int set_range_writeback(struct extent_map_tree *tree, u64 start, u64 end) | |
957 | { | |
958 | unsigned long index = start >> PAGE_CACHE_SHIFT; | |
959 | unsigned long end_index = end >> PAGE_CACHE_SHIFT; | |
960 | struct page *page; | |
961 | ||
962 | while (index <= end_index) { | |
963 | page = find_get_page(tree->mapping, index); | |
964 | BUG_ON(!page); | |
965 | set_page_writeback(page); | |
966 | page_cache_release(page); | |
967 | index++; | |
968 | } | |
969 | set_extent_writeback(tree, start, end, GFP_NOFS); | |
970 | return 0; | |
971 | } | |
972 | EXPORT_SYMBOL(set_range_writeback); | |
973 | ||
5f39d397 CM |
974 | int find_first_extent_bit(struct extent_map_tree *tree, u64 start, |
975 | u64 *start_ret, u64 *end_ret, int bits) | |
976 | { | |
977 | struct rb_node *node; | |
978 | struct extent_state *state; | |
979 | int ret = 1; | |
980 | ||
e19caa5f | 981 | read_lock_irq(&tree->lock); |
5f39d397 CM |
982 | /* |
983 | * this search will find all the extents that end after | |
984 | * our range starts. | |
985 | */ | |
986 | node = tree_search(&tree->state, start); | |
987 | if (!node || IS_ERR(node)) { | |
988 | goto out; | |
989 | } | |
990 | ||
991 | while(1) { | |
992 | state = rb_entry(node, struct extent_state, rb_node); | |
e19caa5f | 993 | if (state->end >= start && (state->state & bits)) { |
5f39d397 CM |
994 | *start_ret = state->start; |
995 | *end_ret = state->end; | |
996 | ret = 0; | |
f510cfec | 997 | break; |
5f39d397 CM |
998 | } |
999 | node = rb_next(node); | |
1000 | if (!node) | |
1001 | break; | |
1002 | } | |
1003 | out: | |
e19caa5f | 1004 | read_unlock_irq(&tree->lock); |
5f39d397 CM |
1005 | return ret; |
1006 | } | |
1007 | EXPORT_SYMBOL(find_first_extent_bit); | |
1008 | ||
b888db2b CM |
1009 | u64 find_lock_delalloc_range(struct extent_map_tree *tree, |
1010 | u64 start, u64 lock_start, u64 *end, u64 max_bytes) | |
1011 | { | |
1012 | struct rb_node *node; | |
1013 | struct extent_state *state; | |
1014 | u64 cur_start = start; | |
1015 | u64 found = 0; | |
1016 | u64 total_bytes = 0; | |
1017 | ||
1018 | write_lock_irq(&tree->lock); | |
1019 | /* | |
1020 | * this search will find all the extents that end after | |
1021 | * our range starts. | |
1022 | */ | |
1023 | search_again: | |
1024 | node = tree_search(&tree->state, cur_start); | |
1025 | if (!node || IS_ERR(node)) { | |
1026 | goto out; | |
1027 | } | |
1028 | ||
1029 | while(1) { | |
1030 | state = rb_entry(node, struct extent_state, rb_node); | |
1031 | if (state->start != cur_start) { | |
1032 | goto out; | |
1033 | } | |
1034 | if (!(state->state & EXTENT_DELALLOC)) { | |
1035 | goto out; | |
1036 | } | |
1037 | if (state->start >= lock_start) { | |
1038 | if (state->state & EXTENT_LOCKED) { | |
1039 | DEFINE_WAIT(wait); | |
1040 | atomic_inc(&state->refs); | |
1041 | write_unlock_irq(&tree->lock); | |
1042 | schedule(); | |
1043 | write_lock_irq(&tree->lock); | |
1044 | finish_wait(&state->wq, &wait); | |
1045 | free_extent_state(state); | |
1046 | goto search_again; | |
1047 | } | |
1048 | state->state |= EXTENT_LOCKED; | |
1049 | } | |
1050 | found++; | |
1051 | *end = state->end; | |
1052 | cur_start = state->end + 1; | |
1053 | node = rb_next(node); | |
1054 | if (!node) | |
1055 | break; | |
1056 | total_bytes = state->end - state->start + 1; | |
1057 | if (total_bytes >= max_bytes) | |
1058 | break; | |
1059 | } | |
1060 | out: | |
1061 | write_unlock_irq(&tree->lock); | |
1062 | return found; | |
1063 | } | |
1064 | ||
a52d9a80 CM |
1065 | /* |
1066 | * helper function to lock both pages and extents in the tree. | |
1067 | * pages must be locked first. | |
1068 | */ | |
1069 | int lock_range(struct extent_map_tree *tree, u64 start, u64 end) | |
1070 | { | |
1071 | unsigned long index = start >> PAGE_CACHE_SHIFT; | |
1072 | unsigned long end_index = end >> PAGE_CACHE_SHIFT; | |
1073 | struct page *page; | |
1074 | int err; | |
1075 | ||
1076 | while (index <= end_index) { | |
1077 | page = grab_cache_page(tree->mapping, index); | |
1078 | if (!page) { | |
1079 | err = -ENOMEM; | |
1080 | goto failed; | |
1081 | } | |
1082 | if (IS_ERR(page)) { | |
1083 | err = PTR_ERR(page); | |
1084 | goto failed; | |
1085 | } | |
1086 | index++; | |
1087 | } | |
1088 | lock_extent(tree, start, end, GFP_NOFS); | |
1089 | return 0; | |
1090 | ||
1091 | failed: | |
1092 | /* | |
1093 | * we failed above in getting the page at 'index', so we undo here | |
1094 | * up to but not including the page at 'index' | |
1095 | */ | |
1096 | end_index = index; | |
1097 | index = start >> PAGE_CACHE_SHIFT; | |
1098 | while (index < end_index) { | |
1099 | page = find_get_page(tree->mapping, index); | |
1100 | unlock_page(page); | |
1101 | page_cache_release(page); | |
1102 | index++; | |
1103 | } | |
1104 | return err; | |
1105 | } | |
1106 | EXPORT_SYMBOL(lock_range); | |
1107 | ||
1108 | /* | |
1109 | * helper function to unlock both pages and extents in the tree. | |
1110 | */ | |
1111 | int unlock_range(struct extent_map_tree *tree, u64 start, u64 end) | |
1112 | { | |
1113 | unsigned long index = start >> PAGE_CACHE_SHIFT; | |
1114 | unsigned long end_index = end >> PAGE_CACHE_SHIFT; | |
1115 | struct page *page; | |
1116 | ||
1117 | while (index <= end_index) { | |
1118 | page = find_get_page(tree->mapping, index); | |
1119 | unlock_page(page); | |
1120 | page_cache_release(page); | |
1121 | index++; | |
1122 | } | |
1123 | unlock_extent(tree, start, end, GFP_NOFS); | |
1124 | return 0; | |
1125 | } | |
1126 | EXPORT_SYMBOL(unlock_range); | |
1127 | ||
07157aac CM |
1128 | int set_state_private(struct extent_map_tree *tree, u64 start, u64 private) |
1129 | { | |
1130 | struct rb_node *node; | |
1131 | struct extent_state *state; | |
1132 | int ret = 0; | |
1133 | ||
1134 | write_lock_irq(&tree->lock); | |
1135 | /* | |
1136 | * this search will find all the extents that end after | |
1137 | * our range starts. | |
1138 | */ | |
1139 | node = tree_search(&tree->state, start); | |
1140 | if (!node || IS_ERR(node)) { | |
1141 | ret = -ENOENT; | |
1142 | goto out; | |
1143 | } | |
1144 | state = rb_entry(node, struct extent_state, rb_node); | |
1145 | if (state->start != start) { | |
1146 | ret = -ENOENT; | |
1147 | goto out; | |
1148 | } | |
1149 | state->private = private; | |
1150 | out: | |
1151 | write_unlock_irq(&tree->lock); | |
1152 | return ret; | |
07157aac CM |
1153 | } |
1154 | ||
1155 | int get_state_private(struct extent_map_tree *tree, u64 start, u64 *private) | |
1156 | { | |
1157 | struct rb_node *node; | |
1158 | struct extent_state *state; | |
1159 | int ret = 0; | |
1160 | ||
1161 | read_lock_irq(&tree->lock); | |
1162 | /* | |
1163 | * this search will find all the extents that end after | |
1164 | * our range starts. | |
1165 | */ | |
1166 | node = tree_search(&tree->state, start); | |
1167 | if (!node || IS_ERR(node)) { | |
1168 | ret = -ENOENT; | |
1169 | goto out; | |
1170 | } | |
1171 | state = rb_entry(node, struct extent_state, rb_node); | |
1172 | if (state->start != start) { | |
1173 | ret = -ENOENT; | |
1174 | goto out; | |
1175 | } | |
1176 | *private = state->private; | |
1177 | out: | |
1178 | read_unlock_irq(&tree->lock); | |
1179 | return ret; | |
1180 | } | |
1181 | ||
a52d9a80 CM |
1182 | /* |
1183 | * searches a range in the state tree for a given mask. | |
1184 | * If 'filled' == 1, this returns 1 only if ever extent in the tree | |
1185 | * has the bits set. Otherwise, 1 is returned if any bit in the | |
1186 | * range is found set. | |
1187 | */ | |
1a5bc167 CM |
1188 | int test_range_bit(struct extent_map_tree *tree, u64 start, u64 end, |
1189 | int bits, int filled) | |
a52d9a80 CM |
1190 | { |
1191 | struct extent_state *state = NULL; | |
1192 | struct rb_node *node; | |
1193 | int bitset = 0; | |
1194 | ||
1195 | read_lock_irq(&tree->lock); | |
1196 | node = tree_search(&tree->state, start); | |
1197 | while (node && start <= end) { | |
1198 | state = rb_entry(node, struct extent_state, rb_node); | |
1199 | if (state->start > end) | |
1200 | break; | |
1201 | ||
1202 | if (filled && state->start > start) { | |
1203 | bitset = 0; | |
1204 | break; | |
1205 | } | |
1206 | if (state->state & bits) { | |
1207 | bitset = 1; | |
1208 | if (!filled) | |
1209 | break; | |
1210 | } else if (filled) { | |
1211 | bitset = 0; | |
1212 | break; | |
1213 | } | |
1214 | start = state->end + 1; | |
1215 | if (start > end) | |
1216 | break; | |
1217 | node = rb_next(node); | |
1218 | } | |
1219 | read_unlock_irq(&tree->lock); | |
1220 | return bitset; | |
1221 | } | |
1a5bc167 | 1222 | EXPORT_SYMBOL(test_range_bit); |
a52d9a80 CM |
1223 | |
1224 | /* | |
1225 | * helper function to set a given page up to date if all the | |
1226 | * extents in the tree for that page are up to date | |
1227 | */ | |
1228 | static int check_page_uptodate(struct extent_map_tree *tree, | |
1229 | struct page *page) | |
1230 | { | |
1231 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1232 | u64 end = start + PAGE_CACHE_SIZE - 1; | |
1233 | if (test_range_bit(tree, start, end, EXTENT_UPTODATE, 1)) | |
1234 | SetPageUptodate(page); | |
1235 | return 0; | |
1236 | } | |
1237 | ||
1238 | /* | |
1239 | * helper function to unlock a page if all the extents in the tree | |
1240 | * for that page are unlocked | |
1241 | */ | |
1242 | static int check_page_locked(struct extent_map_tree *tree, | |
1243 | struct page *page) | |
1244 | { | |
1245 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1246 | u64 end = start + PAGE_CACHE_SIZE - 1; | |
1247 | if (!test_range_bit(tree, start, end, EXTENT_LOCKED, 0)) | |
1248 | unlock_page(page); | |
1249 | return 0; | |
1250 | } | |
1251 | ||
1252 | /* | |
1253 | * helper function to end page writeback if all the extents | |
1254 | * in the tree for that page are done with writeback | |
1255 | */ | |
1256 | static int check_page_writeback(struct extent_map_tree *tree, | |
1257 | struct page *page) | |
1258 | { | |
1259 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1260 | u64 end = start + PAGE_CACHE_SIZE - 1; | |
1261 | if (!test_range_bit(tree, start, end, EXTENT_WRITEBACK, 0)) | |
1262 | end_page_writeback(page); | |
1263 | return 0; | |
1264 | } | |
1265 | ||
1266 | /* lots and lots of room for performance fixes in the end_bio funcs */ | |
1267 | ||
1268 | /* | |
1269 | * after a writepage IO is done, we need to: | |
1270 | * clear the uptodate bits on error | |
1271 | * clear the writeback bits in the extent tree for this IO | |
1272 | * end_page_writeback if the page has no more pending IO | |
1273 | * | |
1274 | * Scheduling is not allowed, so the extent state tree is expected | |
1275 | * to have one and only one object corresponding to this IO. | |
1276 | */ | |
1277 | static int end_bio_extent_writepage(struct bio *bio, | |
1278 | unsigned int bytes_done, int err) | |
1279 | { | |
1280 | const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); | |
1281 | struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1; | |
1282 | struct extent_map_tree *tree = bio->bi_private; | |
1283 | u64 start; | |
1284 | u64 end; | |
1285 | int whole_page; | |
1286 | ||
1287 | if (bio->bi_size) | |
1288 | return 1; | |
1289 | ||
1290 | do { | |
1291 | struct page *page = bvec->bv_page; | |
1292 | start = (page->index << PAGE_CACHE_SHIFT) + bvec->bv_offset; | |
1293 | end = start + bvec->bv_len - 1; | |
1294 | ||
1295 | if (bvec->bv_offset == 0 && bvec->bv_len == PAGE_CACHE_SIZE) | |
1296 | whole_page = 1; | |
1297 | else | |
1298 | whole_page = 0; | |
1299 | ||
1300 | if (--bvec >= bio->bi_io_vec) | |
1301 | prefetchw(&bvec->bv_page->flags); | |
1302 | ||
1303 | if (!uptodate) { | |
1304 | clear_extent_uptodate(tree, start, end, GFP_ATOMIC); | |
1305 | ClearPageUptodate(page); | |
1306 | SetPageError(page); | |
1307 | } | |
1308 | clear_extent_writeback(tree, start, end, GFP_ATOMIC); | |
1309 | ||
1310 | if (whole_page) | |
1311 | end_page_writeback(page); | |
1312 | else | |
1313 | check_page_writeback(tree, page); | |
0e2752a7 CH |
1314 | if (tree->ops && tree->ops->writepage_end_io_hook) |
1315 | tree->ops->writepage_end_io_hook(page, start, end); | |
a52d9a80 CM |
1316 | } while (bvec >= bio->bi_io_vec); |
1317 | ||
1318 | bio_put(bio); | |
1319 | return 0; | |
1320 | } | |
1321 | ||
1322 | /* | |
1323 | * after a readpage IO is done, we need to: | |
1324 | * clear the uptodate bits on error | |
1325 | * set the uptodate bits if things worked | |
1326 | * set the page up to date if all extents in the tree are uptodate | |
1327 | * clear the lock bit in the extent tree | |
1328 | * unlock the page if there are no other extents locked for it | |
1329 | * | |
1330 | * Scheduling is not allowed, so the extent state tree is expected | |
1331 | * to have one and only one object corresponding to this IO. | |
1332 | */ | |
1333 | static int end_bio_extent_readpage(struct bio *bio, | |
1334 | unsigned int bytes_done, int err) | |
1335 | { | |
07157aac | 1336 | int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); |
a52d9a80 CM |
1337 | struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1; |
1338 | struct extent_map_tree *tree = bio->bi_private; | |
1339 | u64 start; | |
1340 | u64 end; | |
1341 | int whole_page; | |
07157aac | 1342 | int ret; |
a52d9a80 CM |
1343 | |
1344 | if (bio->bi_size) | |
1345 | return 1; | |
1346 | ||
1347 | do { | |
1348 | struct page *page = bvec->bv_page; | |
1349 | start = (page->index << PAGE_CACHE_SHIFT) + bvec->bv_offset; | |
1350 | end = start + bvec->bv_len - 1; | |
1351 | ||
1352 | if (bvec->bv_offset == 0 && bvec->bv_len == PAGE_CACHE_SIZE) | |
1353 | whole_page = 1; | |
1354 | else | |
1355 | whole_page = 0; | |
1356 | ||
1357 | if (--bvec >= bio->bi_io_vec) | |
1358 | prefetchw(&bvec->bv_page->flags); | |
1359 | ||
07157aac CM |
1360 | if (uptodate && tree->ops && tree->ops->readpage_end_io_hook) { |
1361 | ret = tree->ops->readpage_end_io_hook(page, start, end); | |
1362 | if (ret) | |
1363 | uptodate = 0; | |
1364 | } | |
a52d9a80 CM |
1365 | if (uptodate) { |
1366 | set_extent_uptodate(tree, start, end, GFP_ATOMIC); | |
1367 | if (whole_page) | |
1368 | SetPageUptodate(page); | |
1369 | else | |
1370 | check_page_uptodate(tree, page); | |
1371 | } else { | |
1372 | ClearPageUptodate(page); | |
1373 | SetPageError(page); | |
1374 | } | |
1375 | ||
1376 | unlock_extent(tree, start, end, GFP_ATOMIC); | |
1377 | ||
1378 | if (whole_page) | |
1379 | unlock_page(page); | |
1380 | else | |
1381 | check_page_locked(tree, page); | |
1382 | } while (bvec >= bio->bi_io_vec); | |
1383 | ||
1384 | bio_put(bio); | |
1385 | return 0; | |
1386 | } | |
1387 | ||
1388 | /* | |
1389 | * IO done from prepare_write is pretty simple, we just unlock | |
1390 | * the structs in the extent tree when done, and set the uptodate bits | |
1391 | * as appropriate. | |
1392 | */ | |
1393 | static int end_bio_extent_preparewrite(struct bio *bio, | |
1394 | unsigned int bytes_done, int err) | |
1395 | { | |
1396 | const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); | |
1397 | struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1; | |
1398 | struct extent_map_tree *tree = bio->bi_private; | |
1399 | u64 start; | |
1400 | u64 end; | |
1401 | ||
1402 | if (bio->bi_size) | |
1403 | return 1; | |
1404 | ||
1405 | do { | |
1406 | struct page *page = bvec->bv_page; | |
1407 | start = (page->index << PAGE_CACHE_SHIFT) + bvec->bv_offset; | |
1408 | end = start + bvec->bv_len - 1; | |
1409 | ||
1410 | if (--bvec >= bio->bi_io_vec) | |
1411 | prefetchw(&bvec->bv_page->flags); | |
1412 | ||
1413 | if (uptodate) { | |
1414 | set_extent_uptodate(tree, start, end, GFP_ATOMIC); | |
1415 | } else { | |
1416 | ClearPageUptodate(page); | |
1417 | SetPageError(page); | |
1418 | } | |
1419 | ||
1420 | unlock_extent(tree, start, end, GFP_ATOMIC); | |
1421 | ||
1422 | } while (bvec >= bio->bi_io_vec); | |
1423 | ||
1424 | bio_put(bio); | |
1425 | return 0; | |
1426 | } | |
1427 | ||
1428 | static int submit_extent_page(int rw, struct extent_map_tree *tree, | |
1429 | struct page *page, sector_t sector, | |
1430 | size_t size, unsigned long offset, | |
1431 | struct block_device *bdev, | |
1432 | bio_end_io_t end_io_func) | |
1433 | { | |
1434 | struct bio *bio; | |
1435 | int ret = 0; | |
1436 | ||
1437 | bio = bio_alloc(GFP_NOIO, 1); | |
1438 | ||
1439 | bio->bi_sector = sector; | |
1440 | bio->bi_bdev = bdev; | |
1441 | bio->bi_io_vec[0].bv_page = page; | |
1442 | bio->bi_io_vec[0].bv_len = size; | |
1443 | bio->bi_io_vec[0].bv_offset = offset; | |
1444 | ||
1445 | bio->bi_vcnt = 1; | |
1446 | bio->bi_idx = 0; | |
1447 | bio->bi_size = size; | |
1448 | ||
1449 | bio->bi_end_io = end_io_func; | |
1450 | bio->bi_private = tree; | |
1451 | ||
1452 | bio_get(bio); | |
1453 | submit_bio(rw, bio); | |
1454 | ||
1455 | if (bio_flagged(bio, BIO_EOPNOTSUPP)) | |
1456 | ret = -EOPNOTSUPP; | |
1457 | ||
1458 | bio_put(bio); | |
1459 | return ret; | |
1460 | } | |
1461 | ||
b3cfa35a CH |
1462 | void set_page_extent_mapped(struct page *page) |
1463 | { | |
1464 | if (!PagePrivate(page)) { | |
1465 | SetPagePrivate(page); | |
1466 | WARN_ON(!page->mapping->a_ops->invalidatepage); | |
19c00ddc | 1467 | set_page_private(page, EXTENT_PAGE_PRIVATE); |
b3cfa35a CH |
1468 | page_cache_get(page); |
1469 | } | |
1470 | } | |
1471 | ||
a52d9a80 CM |
1472 | /* |
1473 | * basic readpage implementation. Locked extent state structs are inserted | |
1474 | * into the tree that are removed when the IO is done (by the end_io | |
1475 | * handlers) | |
1476 | */ | |
1477 | int extent_read_full_page(struct extent_map_tree *tree, struct page *page, | |
1478 | get_extent_t *get_extent) | |
1479 | { | |
1480 | struct inode *inode = page->mapping->host; | |
1481 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1482 | u64 page_end = start + PAGE_CACHE_SIZE - 1; | |
1483 | u64 end; | |
1484 | u64 cur = start; | |
1485 | u64 extent_offset; | |
1486 | u64 last_byte = i_size_read(inode); | |
1487 | u64 block_start; | |
1488 | u64 cur_end; | |
1489 | sector_t sector; | |
1490 | struct extent_map *em; | |
1491 | struct block_device *bdev; | |
1492 | int ret; | |
1493 | int nr = 0; | |
1494 | size_t page_offset = 0; | |
1495 | size_t iosize; | |
1496 | size_t blocksize = inode->i_sb->s_blocksize; | |
1497 | ||
b3cfa35a | 1498 | set_page_extent_mapped(page); |
a52d9a80 CM |
1499 | |
1500 | end = page_end; | |
1501 | lock_extent(tree, start, end, GFP_NOFS); | |
1502 | ||
1503 | while (cur <= end) { | |
1504 | if (cur >= last_byte) { | |
1505 | iosize = PAGE_CACHE_SIZE - page_offset; | |
1506 | zero_user_page(page, page_offset, iosize, KM_USER0); | |
1507 | set_extent_uptodate(tree, cur, cur + iosize - 1, | |
1508 | GFP_NOFS); | |
1509 | unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); | |
1510 | break; | |
1511 | } | |
1512 | em = get_extent(inode, page, page_offset, cur, end, 0); | |
1513 | if (IS_ERR(em) || !em) { | |
1514 | SetPageError(page); | |
1515 | unlock_extent(tree, cur, end, GFP_NOFS); | |
1516 | break; | |
1517 | } | |
1518 | ||
1519 | extent_offset = cur - em->start; | |
1520 | BUG_ON(em->end < cur); | |
1521 | BUG_ON(end < cur); | |
1522 | ||
1523 | iosize = min(em->end - cur, end - cur) + 1; | |
1524 | cur_end = min(em->end, end); | |
1525 | iosize = (iosize + blocksize - 1) & ~((u64)blocksize - 1); | |
1526 | sector = (em->block_start + extent_offset) >> 9; | |
1527 | bdev = em->bdev; | |
1528 | block_start = em->block_start; | |
1529 | free_extent_map(em); | |
1530 | em = NULL; | |
1531 | ||
1532 | /* we've found a hole, just zero and go on */ | |
5f39d397 | 1533 | if (block_start == EXTENT_MAP_HOLE) { |
a52d9a80 CM |
1534 | zero_user_page(page, page_offset, iosize, KM_USER0); |
1535 | set_extent_uptodate(tree, cur, cur + iosize - 1, | |
1536 | GFP_NOFS); | |
1537 | unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); | |
1538 | cur = cur + iosize; | |
1539 | page_offset += iosize; | |
1540 | continue; | |
1541 | } | |
1542 | /* the get_extent function already copied into the page */ | |
1543 | if (test_range_bit(tree, cur, cur_end, EXTENT_UPTODATE, 1)) { | |
1544 | unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); | |
1545 | cur = cur + iosize; | |
1546 | page_offset += iosize; | |
1547 | continue; | |
1548 | } | |
1549 | ||
07157aac CM |
1550 | ret = 0; |
1551 | if (tree->ops && tree->ops->readpage_io_hook) { | |
1552 | ret = tree->ops->readpage_io_hook(page, cur, | |
1553 | cur + iosize - 1); | |
1554 | } | |
1555 | if (!ret) { | |
1556 | ret = submit_extent_page(READ, tree, page, | |
1557 | sector, iosize, page_offset, | |
1558 | bdev, end_bio_extent_readpage); | |
1559 | } | |
a52d9a80 CM |
1560 | if (ret) |
1561 | SetPageError(page); | |
1562 | cur = cur + iosize; | |
1563 | page_offset += iosize; | |
1564 | nr++; | |
1565 | } | |
1566 | if (!nr) { | |
1567 | if (!PageError(page)) | |
1568 | SetPageUptodate(page); | |
1569 | unlock_page(page); | |
1570 | } | |
1571 | return 0; | |
1572 | } | |
1573 | EXPORT_SYMBOL(extent_read_full_page); | |
1574 | ||
1575 | /* | |
1576 | * the writepage semantics are similar to regular writepage. extent | |
1577 | * records are inserted to lock ranges in the tree, and as dirty areas | |
1578 | * are found, they are marked writeback. Then the lock bits are removed | |
1579 | * and the end_io handler clears the writeback ranges | |
1580 | */ | |
1581 | int extent_write_full_page(struct extent_map_tree *tree, struct page *page, | |
1582 | get_extent_t *get_extent, | |
1583 | struct writeback_control *wbc) | |
1584 | { | |
1585 | struct inode *inode = page->mapping->host; | |
1586 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1587 | u64 page_end = start + PAGE_CACHE_SIZE - 1; | |
1588 | u64 end; | |
1589 | u64 cur = start; | |
1590 | u64 extent_offset; | |
1591 | u64 last_byte = i_size_read(inode); | |
1592 | u64 block_start; | |
1593 | sector_t sector; | |
1594 | struct extent_map *em; | |
1595 | struct block_device *bdev; | |
1596 | int ret; | |
1597 | int nr = 0; | |
1598 | size_t page_offset = 0; | |
1599 | size_t iosize; | |
1600 | size_t blocksize; | |
1601 | loff_t i_size = i_size_read(inode); | |
1602 | unsigned long end_index = i_size >> PAGE_CACHE_SHIFT; | |
b888db2b CM |
1603 | u64 nr_delalloc; |
1604 | u64 delalloc_end; | |
a52d9a80 | 1605 | |
b888db2b | 1606 | WARN_ON(!PageLocked(page)); |
a52d9a80 CM |
1607 | if (page->index > end_index) { |
1608 | clear_extent_dirty(tree, start, page_end, GFP_NOFS); | |
1609 | unlock_page(page); | |
1610 | return 0; | |
1611 | } | |
1612 | ||
1613 | if (page->index == end_index) { | |
1614 | size_t offset = i_size & (PAGE_CACHE_SIZE - 1); | |
1615 | zero_user_page(page, offset, | |
1616 | PAGE_CACHE_SIZE - offset, KM_USER0); | |
1617 | } | |
1618 | ||
b3cfa35a | 1619 | set_page_extent_mapped(page); |
a52d9a80 | 1620 | |
a52d9a80 | 1621 | lock_extent(tree, start, page_end, GFP_NOFS); |
b888db2b CM |
1622 | nr_delalloc = find_lock_delalloc_range(tree, start, page_end + 1, |
1623 | &delalloc_end, | |
1624 | 128 * 1024 * 1024); | |
1625 | if (nr_delalloc) { | |
07157aac | 1626 | tree->ops->fill_delalloc(inode, start, delalloc_end); |
b888db2b CM |
1627 | if (delalloc_end >= page_end + 1) { |
1628 | clear_extent_bit(tree, page_end + 1, delalloc_end, | |
1629 | EXTENT_LOCKED | EXTENT_DELALLOC, | |
1630 | 1, 0, GFP_NOFS); | |
1631 | } | |
1632 | clear_extent_bit(tree, start, page_end, EXTENT_DELALLOC, | |
1633 | 0, 0, GFP_NOFS); | |
1634 | if (test_range_bit(tree, start, page_end, EXTENT_DELALLOC, 0)) { | |
1635 | printk("found delalloc bits after clear extent_bit\n"); | |
1636 | } | |
1637 | } else if (test_range_bit(tree, start, page_end, EXTENT_DELALLOC, 0)) { | |
1638 | printk("found delalloc bits after find_delalloc_range returns 0\n"); | |
1639 | } | |
1640 | ||
1641 | end = page_end; | |
1642 | if (test_range_bit(tree, start, page_end, EXTENT_DELALLOC, 0)) { | |
1643 | printk("found delalloc bits after lock_extent\n"); | |
1644 | } | |
a52d9a80 CM |
1645 | |
1646 | if (last_byte <= start) { | |
1647 | clear_extent_dirty(tree, start, page_end, GFP_NOFS); | |
1648 | goto done; | |
1649 | } | |
1650 | ||
1651 | set_extent_uptodate(tree, start, page_end, GFP_NOFS); | |
1652 | blocksize = inode->i_sb->s_blocksize; | |
1653 | ||
1654 | while (cur <= end) { | |
1655 | if (cur >= last_byte) { | |
1656 | clear_extent_dirty(tree, cur, page_end, GFP_NOFS); | |
1657 | break; | |
1658 | } | |
b888db2b | 1659 | em = get_extent(inode, page, page_offset, cur, end, 0); |
a52d9a80 CM |
1660 | if (IS_ERR(em) || !em) { |
1661 | SetPageError(page); | |
1662 | break; | |
1663 | } | |
1664 | ||
1665 | extent_offset = cur - em->start; | |
1666 | BUG_ON(em->end < cur); | |
1667 | BUG_ON(end < cur); | |
1668 | iosize = min(em->end - cur, end - cur) + 1; | |
1669 | iosize = (iosize + blocksize - 1) & ~((u64)blocksize - 1); | |
1670 | sector = (em->block_start + extent_offset) >> 9; | |
1671 | bdev = em->bdev; | |
1672 | block_start = em->block_start; | |
1673 | free_extent_map(em); | |
1674 | em = NULL; | |
1675 | ||
5f39d397 CM |
1676 | if (block_start == EXTENT_MAP_HOLE || |
1677 | block_start == EXTENT_MAP_INLINE) { | |
a52d9a80 CM |
1678 | clear_extent_dirty(tree, cur, |
1679 | cur + iosize - 1, GFP_NOFS); | |
1680 | cur = cur + iosize; | |
1681 | page_offset += iosize; | |
1682 | continue; | |
1683 | } | |
1684 | ||
1685 | /* leave this out until we have a page_mkwrite call */ | |
1686 | if (0 && !test_range_bit(tree, cur, cur + iosize - 1, | |
1687 | EXTENT_DIRTY, 0)) { | |
1688 | cur = cur + iosize; | |
1689 | page_offset += iosize; | |
1690 | continue; | |
1691 | } | |
1692 | clear_extent_dirty(tree, cur, cur + iosize - 1, GFP_NOFS); | |
b06355f0 CH |
1693 | if (tree->ops && tree->ops->writepage_io_hook) { |
1694 | ret = tree->ops->writepage_io_hook(page, cur, | |
1695 | cur + iosize - 1); | |
1696 | } else { | |
1697 | ret = 0; | |
1698 | } | |
a52d9a80 CM |
1699 | if (ret) |
1700 | SetPageError(page); | |
07157aac CM |
1701 | else { |
1702 | set_range_writeback(tree, cur, cur + iosize - 1); | |
1703 | ret = submit_extent_page(WRITE, tree, page, sector, | |
1704 | iosize, page_offset, bdev, | |
1705 | end_bio_extent_writepage); | |
1706 | if (ret) | |
1707 | SetPageError(page); | |
1708 | } | |
a52d9a80 CM |
1709 | cur = cur + iosize; |
1710 | page_offset += iosize; | |
1711 | nr++; | |
1712 | } | |
1713 | done: | |
a52d9a80 CM |
1714 | unlock_extent(tree, start, page_end, GFP_NOFS); |
1715 | unlock_page(page); | |
1716 | return 0; | |
1717 | } | |
1718 | EXPORT_SYMBOL(extent_write_full_page); | |
1719 | ||
1720 | /* | |
1721 | * basic invalidatepage code, this waits on any locked or writeback | |
1722 | * ranges corresponding to the page, and then deletes any extent state | |
1723 | * records from the tree | |
1724 | */ | |
1725 | int extent_invalidatepage(struct extent_map_tree *tree, | |
1726 | struct page *page, unsigned long offset) | |
1727 | { | |
1728 | u64 start = (page->index << PAGE_CACHE_SHIFT); | |
1729 | u64 end = start + PAGE_CACHE_SIZE - 1; | |
1730 | size_t blocksize = page->mapping->host->i_sb->s_blocksize; | |
1731 | ||
1732 | start += (offset + blocksize -1) & ~(blocksize - 1); | |
1733 | if (start > end) | |
1734 | return 0; | |
1735 | ||
1736 | lock_extent(tree, start, end, GFP_NOFS); | |
1737 | wait_on_extent_writeback(tree, start, end); | |
2bf5a725 CM |
1738 | clear_extent_bit(tree, start, end, |
1739 | EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC, | |
a52d9a80 CM |
1740 | 1, 1, GFP_NOFS); |
1741 | return 0; | |
1742 | } | |
1743 | EXPORT_SYMBOL(extent_invalidatepage); | |
1744 | ||
1745 | /* | |
1746 | * simple commit_write call, set_range_dirty is used to mark both | |
1747 | * the pages and the extent records as dirty | |
1748 | */ | |
1749 | int extent_commit_write(struct extent_map_tree *tree, | |
1750 | struct inode *inode, struct page *page, | |
1751 | unsigned from, unsigned to) | |
1752 | { | |
1753 | loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to; | |
1754 | ||
b3cfa35a | 1755 | set_page_extent_mapped(page); |
a52d9a80 CM |
1756 | set_page_dirty(page); |
1757 | ||
1758 | if (pos > inode->i_size) { | |
1759 | i_size_write(inode, pos); | |
1760 | mark_inode_dirty(inode); | |
1761 | } | |
1762 | return 0; | |
1763 | } | |
1764 | EXPORT_SYMBOL(extent_commit_write); | |
1765 | ||
1766 | int extent_prepare_write(struct extent_map_tree *tree, | |
1767 | struct inode *inode, struct page *page, | |
1768 | unsigned from, unsigned to, get_extent_t *get_extent) | |
1769 | { | |
1770 | u64 page_start = page->index << PAGE_CACHE_SHIFT; | |
1771 | u64 page_end = page_start + PAGE_CACHE_SIZE - 1; | |
1772 | u64 block_start; | |
1773 | u64 orig_block_start; | |
1774 | u64 block_end; | |
1775 | u64 cur_end; | |
1776 | struct extent_map *em; | |
1777 | unsigned blocksize = 1 << inode->i_blkbits; | |
1778 | size_t page_offset = 0; | |
1779 | size_t block_off_start; | |
1780 | size_t block_off_end; | |
1781 | int err = 0; | |
1782 | int iocount = 0; | |
1783 | int ret = 0; | |
1784 | int isnew; | |
1785 | ||
b3cfa35a CH |
1786 | set_page_extent_mapped(page); |
1787 | ||
a52d9a80 CM |
1788 | block_start = (page_start + from) & ~((u64)blocksize - 1); |
1789 | block_end = (page_start + to - 1) | (blocksize - 1); | |
1790 | orig_block_start = block_start; | |
1791 | ||
1792 | lock_extent(tree, page_start, page_end, GFP_NOFS); | |
1793 | while(block_start <= block_end) { | |
1794 | em = get_extent(inode, page, page_offset, block_start, | |
1795 | block_end, 1); | |
1796 | if (IS_ERR(em) || !em) { | |
1797 | goto err; | |
1798 | } | |
1799 | cur_end = min(block_end, em->end); | |
1800 | block_off_start = block_start & (PAGE_CACHE_SIZE - 1); | |
1801 | block_off_end = block_off_start + blocksize; | |
1802 | isnew = clear_extent_new(tree, block_start, cur_end, GFP_NOFS); | |
1803 | ||
1804 | if (!PageUptodate(page) && isnew && | |
1805 | (block_off_end > to || block_off_start < from)) { | |
1806 | void *kaddr; | |
1807 | ||
1808 | kaddr = kmap_atomic(page, KM_USER0); | |
1809 | if (block_off_end > to) | |
1810 | memset(kaddr + to, 0, block_off_end - to); | |
1811 | if (block_off_start < from) | |
1812 | memset(kaddr + block_off_start, 0, | |
1813 | from - block_off_start); | |
1814 | flush_dcache_page(page); | |
1815 | kunmap_atomic(kaddr, KM_USER0); | |
1816 | } | |
1817 | if (!isnew && !PageUptodate(page) && | |
1818 | (block_off_end > to || block_off_start < from) && | |
1819 | !test_range_bit(tree, block_start, cur_end, | |
1820 | EXTENT_UPTODATE, 1)) { | |
1821 | u64 sector; | |
1822 | u64 extent_offset = block_start - em->start; | |
1823 | size_t iosize; | |
1824 | sector = (em->block_start + extent_offset) >> 9; | |
1825 | iosize = (cur_end - block_start + blocksize - 1) & | |
1826 | ~((u64)blocksize - 1); | |
1827 | /* | |
1828 | * we've already got the extent locked, but we | |
1829 | * need to split the state such that our end_bio | |
1830 | * handler can clear the lock. | |
1831 | */ | |
1832 | set_extent_bit(tree, block_start, | |
1833 | block_start + iosize - 1, | |
1834 | EXTENT_LOCKED, 0, NULL, GFP_NOFS); | |
1835 | ret = submit_extent_page(READ, tree, page, | |
1836 | sector, iosize, page_offset, em->bdev, | |
1837 | end_bio_extent_preparewrite); | |
1838 | iocount++; | |
1839 | block_start = block_start + iosize; | |
1840 | } else { | |
1841 | set_extent_uptodate(tree, block_start, cur_end, | |
1842 | GFP_NOFS); | |
1843 | unlock_extent(tree, block_start, cur_end, GFP_NOFS); | |
1844 | block_start = cur_end + 1; | |
1845 | } | |
1846 | page_offset = block_start & (PAGE_CACHE_SIZE - 1); | |
1847 | free_extent_map(em); | |
1848 | } | |
1849 | if (iocount) { | |
1850 | wait_extent_bit(tree, orig_block_start, | |
1851 | block_end, EXTENT_LOCKED); | |
1852 | } | |
1853 | check_page_uptodate(tree, page); | |
1854 | err: | |
1855 | /* FIXME, zero out newly allocated blocks on error */ | |
1856 | return err; | |
1857 | } | |
1858 | EXPORT_SYMBOL(extent_prepare_write); | |
1859 | ||
1860 | /* | |
1861 | * a helper for releasepage. As long as there are no locked extents | |
1862 | * in the range corresponding to the page, both state records and extent | |
1863 | * map records are removed | |
1864 | */ | |
1865 | int try_release_extent_mapping(struct extent_map_tree *tree, struct page *page) | |
1866 | { | |
1867 | struct extent_map *em; | |
1868 | u64 start = page->index << PAGE_CACHE_SHIFT; | |
1869 | u64 end = start + PAGE_CACHE_SIZE - 1; | |
1870 | u64 orig_start = start; | |
b888db2b | 1871 | int ret = 1; |
a52d9a80 CM |
1872 | |
1873 | while (start <= end) { | |
1874 | em = lookup_extent_mapping(tree, start, end); | |
1875 | if (!em || IS_ERR(em)) | |
1876 | break; | |
b888db2b CM |
1877 | if (!test_range_bit(tree, em->start, em->end, |
1878 | EXTENT_LOCKED, 0)) { | |
1879 | remove_extent_mapping(tree, em); | |
1880 | /* once for the rb tree */ | |
a52d9a80 | 1881 | free_extent_map(em); |
a52d9a80 | 1882 | } |
a52d9a80 | 1883 | start = em->end + 1; |
a52d9a80 CM |
1884 | /* once for us */ |
1885 | free_extent_map(em); | |
1886 | } | |
b888db2b CM |
1887 | if (test_range_bit(tree, orig_start, end, EXTENT_LOCKED, 0)) |
1888 | ret = 0; | |
1889 | else | |
1890 | clear_extent_bit(tree, orig_start, end, EXTENT_UPTODATE, | |
1891 | 1, 1, GFP_NOFS); | |
1892 | return ret; | |
a52d9a80 CM |
1893 | } |
1894 | EXPORT_SYMBOL(try_release_extent_mapping); | |
1895 | ||
d396c6f5 CH |
1896 | sector_t extent_bmap(struct address_space *mapping, sector_t iblock, |
1897 | get_extent_t *get_extent) | |
1898 | { | |
1899 | struct inode *inode = mapping->host; | |
1900 | u64 start = iblock << inode->i_blkbits; | |
1901 | u64 end = start + (1 << inode->i_blkbits) - 1; | |
1902 | struct extent_map *em; | |
1903 | ||
1904 | em = get_extent(inode, NULL, 0, start, end, 0); | |
1905 | if (!em || IS_ERR(em)) | |
1906 | return 0; | |
1907 | ||
d396c6f5 | 1908 | if (em->block_start == EXTENT_MAP_INLINE || |
5f39d397 CM |
1909 | em->block_start == EXTENT_MAP_HOLE) |
1910 | return 0; | |
d396c6f5 CH |
1911 | |
1912 | return (em->block_start + start - em->start) >> inode->i_blkbits; | |
1913 | } | |
5f39d397 | 1914 | |
4dc11904 | 1915 | static int add_lru(struct extent_map_tree *tree, struct extent_buffer *eb) |
6d36dcd4 | 1916 | { |
4dc11904 CM |
1917 | if (list_empty(&eb->lru)) { |
1918 | extent_buffer_get(eb); | |
1919 | list_add(&eb->lru, &tree->buffer_lru); | |
1920 | tree->lru_size++; | |
1921 | if (tree->lru_size >= BUFFER_LRU_MAX) { | |
1922 | struct extent_buffer *rm; | |
1923 | rm = list_entry(tree->buffer_lru.prev, | |
1924 | struct extent_buffer, lru); | |
1925 | tree->lru_size--; | |
1926 | list_del(&rm->lru); | |
1927 | free_extent_buffer(rm); | |
1928 | } | |
1929 | } else | |
1930 | list_move(&eb->lru, &tree->buffer_lru); | |
1931 | return 0; | |
1932 | } | |
1933 | static struct extent_buffer *find_lru(struct extent_map_tree *tree, | |
1934 | u64 start, unsigned long len) | |
1935 | { | |
1936 | struct list_head *lru = &tree->buffer_lru; | |
1937 | struct list_head *cur = lru->next; | |
1938 | struct extent_buffer *eb; | |
f510cfec | 1939 | |
4dc11904 CM |
1940 | if (list_empty(lru)) |
1941 | return NULL; | |
f510cfec | 1942 | |
4dc11904 CM |
1943 | do { |
1944 | eb = list_entry(cur, struct extent_buffer, lru); | |
1945 | if (eb->start == start && eb->len == len) { | |
1946 | extent_buffer_get(eb); | |
1947 | return eb; | |
1948 | } | |
1949 | cur = cur->next; | |
1950 | } while (cur != lru); | |
1951 | return NULL; | |
6d36dcd4 CM |
1952 | } |
1953 | ||
4dc11904 | 1954 | static inline unsigned long num_extent_pages(u64 start, u64 len) |
6d36dcd4 | 1955 | { |
4dc11904 CM |
1956 | return ((start + len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT) - |
1957 | (start >> PAGE_CACHE_SHIFT); | |
6d36dcd4 CM |
1958 | } |
1959 | ||
4dc11904 CM |
1960 | static inline struct page *extent_buffer_page(struct extent_buffer *eb, |
1961 | unsigned long i) | |
6d36dcd4 CM |
1962 | { |
1963 | struct page *p; | |
db94535d | 1964 | |
4dc11904 | 1965 | if (i == 0) |
810191ff | 1966 | return eb->first_page; |
6d36dcd4 | 1967 | i += eb->start >> PAGE_CACHE_SHIFT; |
810191ff | 1968 | p = find_get_page(eb->first_page->mapping, i); |
6d36dcd4 CM |
1969 | page_cache_release(p); |
1970 | return p; | |
1971 | } | |
1972 | ||
4dc11904 CM |
1973 | static struct extent_buffer *__alloc_extent_buffer(struct extent_map_tree *tree, |
1974 | u64 start, | |
1975 | unsigned long len, | |
1976 | gfp_t mask) | |
db94535d | 1977 | { |
4dc11904 CM |
1978 | struct extent_buffer *eb = NULL; |
1979 | ||
1980 | spin_lock(&tree->lru_lock); | |
1981 | eb = find_lru(tree, start, len); | |
19c00ddc | 1982 | if (eb) { |
4dc11904 | 1983 | goto lru_add; |
19c00ddc | 1984 | } |
4dc11904 CM |
1985 | spin_unlock(&tree->lru_lock); |
1986 | ||
1987 | if (eb) { | |
1988 | memset(eb, 0, sizeof(*eb)); | |
1989 | } else { | |
1990 | eb = kmem_cache_zalloc(extent_buffer_cache, mask); | |
1991 | } | |
1992 | INIT_LIST_HEAD(&eb->lru); | |
1993 | eb->start = start; | |
1994 | eb->len = len; | |
1995 | atomic_set(&eb->refs, 1); | |
1996 | ||
1997 | spin_lock(&tree->lru_lock); | |
1998 | lru_add: | |
1999 | add_lru(tree, eb); | |
2000 | spin_unlock(&tree->lru_lock); | |
2001 | return eb; | |
2002 | } | |
2003 | ||
2004 | static void __free_extent_buffer(struct extent_buffer *eb) | |
2005 | { | |
2006 | kmem_cache_free(extent_buffer_cache, eb); | |
db94535d | 2007 | } |
4dc11904 | 2008 | |
5f39d397 CM |
2009 | struct extent_buffer *alloc_extent_buffer(struct extent_map_tree *tree, |
2010 | u64 start, unsigned long len, | |
19c00ddc | 2011 | struct page *page0, |
5f39d397 CM |
2012 | gfp_t mask) |
2013 | { | |
db94535d | 2014 | unsigned long num_pages = num_extent_pages(start, len); |
5f39d397 CM |
2015 | unsigned long i; |
2016 | unsigned long index = start >> PAGE_CACHE_SHIFT; | |
2017 | struct extent_buffer *eb; | |
2018 | struct page *p; | |
2019 | struct address_space *mapping = tree->mapping; | |
2020 | int uptodate = 0; | |
2021 | ||
4dc11904 | 2022 | eb = __alloc_extent_buffer(tree, start, len, mask); |
5f39d397 CM |
2023 | if (!eb || IS_ERR(eb)) |
2024 | return NULL; | |
2025 | ||
4dc11904 CM |
2026 | if (eb->flags & EXTENT_BUFFER_FILLED) |
2027 | return eb; | |
5f39d397 | 2028 | |
19c00ddc CM |
2029 | if (page0) { |
2030 | eb->first_page = page0; | |
2031 | i = 1; | |
2032 | index++; | |
2033 | page_cache_get(page0); | |
ff79f819 | 2034 | mark_page_accessed(page0); |
19c00ddc CM |
2035 | set_page_extent_mapped(page0); |
2036 | set_page_private(page0, EXTENT_PAGE_PRIVATE_FIRST_PAGE | | |
2037 | len << 2); | |
2038 | } else { | |
2039 | i = 0; | |
2040 | } | |
2041 | for (; i < num_pages; i++, index++) { | |
5f39d397 | 2042 | p = find_or_create_page(mapping, index, mask | __GFP_HIGHMEM); |
6d36dcd4 | 2043 | if (!p) { |
db94535d | 2044 | WARN_ON(1); |
6d36dcd4 CM |
2045 | /* make sure the free only frees the pages we've |
2046 | * grabbed a reference on | |
2047 | */ | |
2048 | eb->len = i << PAGE_CACHE_SHIFT; | |
2049 | eb->start &= ~((u64)PAGE_CACHE_SIZE - 1); | |
5f39d397 | 2050 | goto fail; |
6d36dcd4 | 2051 | } |
f510cfec | 2052 | set_page_extent_mapped(p); |
ff79f819 | 2053 | mark_page_accessed(p); |
19c00ddc | 2054 | if (i == 0) { |
810191ff | 2055 | eb->first_page = p; |
19c00ddc CM |
2056 | set_page_private(p, EXTENT_PAGE_PRIVATE_FIRST_PAGE | |
2057 | len << 2); | |
2058 | } else { | |
2059 | set_page_private(p, EXTENT_PAGE_PRIVATE); | |
2060 | } | |
5f39d397 CM |
2061 | if (!PageUptodate(p)) |
2062 | uptodate = 0; | |
2063 | unlock_page(p); | |
2064 | } | |
2065 | if (uptodate) | |
2066 | eb->flags |= EXTENT_UPTODATE; | |
4dc11904 | 2067 | eb->flags |= EXTENT_BUFFER_FILLED; |
5f39d397 CM |
2068 | return eb; |
2069 | fail: | |
2070 | free_extent_buffer(eb); | |
2071 | return NULL; | |
2072 | } | |
2073 | EXPORT_SYMBOL(alloc_extent_buffer); | |
2074 | ||
2075 | struct extent_buffer *find_extent_buffer(struct extent_map_tree *tree, | |
2076 | u64 start, unsigned long len, | |
2077 | gfp_t mask) | |
2078 | { | |
db94535d | 2079 | unsigned long num_pages = num_extent_pages(start, len); |
19c00ddc | 2080 | unsigned long i; unsigned long index = start >> PAGE_CACHE_SHIFT; |
5f39d397 CM |
2081 | struct extent_buffer *eb; |
2082 | struct page *p; | |
2083 | struct address_space *mapping = tree->mapping; | |
14048ed0 | 2084 | int uptodate = 1; |
5f39d397 | 2085 | |
4dc11904 | 2086 | eb = __alloc_extent_buffer(tree, start, len, mask); |
5f39d397 CM |
2087 | if (!eb || IS_ERR(eb)) |
2088 | return NULL; | |
2089 | ||
4dc11904 CM |
2090 | if (eb->flags & EXTENT_BUFFER_FILLED) |
2091 | return eb; | |
5f39d397 CM |
2092 | |
2093 | for (i = 0; i < num_pages; i++, index++) { | |
14048ed0 | 2094 | p = find_lock_page(mapping, index); |
6d36dcd4 CM |
2095 | if (!p) { |
2096 | /* make sure the free only frees the pages we've | |
2097 | * grabbed a reference on | |
2098 | */ | |
2099 | eb->len = i << PAGE_CACHE_SHIFT; | |
2100 | eb->start &= ~((u64)PAGE_CACHE_SIZE - 1); | |
5f39d397 | 2101 | goto fail; |
6d36dcd4 | 2102 | } |
f510cfec | 2103 | set_page_extent_mapped(p); |
ff79f819 | 2104 | mark_page_accessed(p); |
19c00ddc CM |
2105 | |
2106 | if (i == 0) { | |
810191ff | 2107 | eb->first_page = p; |
19c00ddc CM |
2108 | set_page_private(p, EXTENT_PAGE_PRIVATE_FIRST_PAGE | |
2109 | len << 2); | |
2110 | } else { | |
2111 | set_page_private(p, EXTENT_PAGE_PRIVATE); | |
2112 | } | |
2113 | ||
14048ed0 CM |
2114 | if (!PageUptodate(p)) |
2115 | uptodate = 0; | |
2116 | unlock_page(p); | |
5f39d397 | 2117 | } |
14048ed0 CM |
2118 | if (uptodate) |
2119 | eb->flags |= EXTENT_UPTODATE; | |
4dc11904 | 2120 | eb->flags |= EXTENT_BUFFER_FILLED; |
5f39d397 CM |
2121 | return eb; |
2122 | fail: | |
2123 | free_extent_buffer(eb); | |
2124 | return NULL; | |
2125 | } | |
2126 | EXPORT_SYMBOL(find_extent_buffer); | |
2127 | ||
2128 | void free_extent_buffer(struct extent_buffer *eb) | |
2129 | { | |
2130 | unsigned long i; | |
2131 | unsigned long num_pages; | |
2132 | ||
2133 | if (!eb) | |
2134 | return; | |
2135 | ||
2136 | if (!atomic_dec_and_test(&eb->refs)) | |
2137 | return; | |
2138 | ||
db94535d | 2139 | num_pages = num_extent_pages(eb->start, eb->len); |
5f39d397 | 2140 | |
09e71a32 | 2141 | for (i = 0; i < num_pages; i++) { |
6d36dcd4 | 2142 | page_cache_release(extent_buffer_page(eb, i)); |
5f39d397 | 2143 | } |
6d36dcd4 | 2144 | __free_extent_buffer(eb); |
5f39d397 CM |
2145 | } |
2146 | EXPORT_SYMBOL(free_extent_buffer); | |
2147 | ||
2148 | int clear_extent_buffer_dirty(struct extent_map_tree *tree, | |
2149 | struct extent_buffer *eb) | |
2150 | { | |
2151 | int set; | |
2152 | unsigned long i; | |
2153 | unsigned long num_pages; | |
2154 | struct page *page; | |
2155 | ||
2156 | u64 start = eb->start; | |
2157 | u64 end = start + eb->len - 1; | |
2158 | ||
2159 | set = clear_extent_dirty(tree, start, end, GFP_NOFS); | |
db94535d | 2160 | num_pages = num_extent_pages(eb->start, eb->len); |
5f39d397 CM |
2161 | |
2162 | for (i = 0; i < num_pages; i++) { | |
6d36dcd4 | 2163 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2164 | lock_page(page); |
2165 | /* | |
2166 | * if we're on the last page or the first page and the | |
2167 | * block isn't aligned on a page boundary, do extra checks | |
2168 | * to make sure we don't clean page that is partially dirty | |
2169 | */ | |
2170 | if ((i == 0 && (eb->start & (PAGE_CACHE_SIZE - 1))) || | |
2171 | ((i == num_pages - 1) && | |
2172 | ((eb->start + eb->len - 1) & (PAGE_CACHE_SIZE - 1)))) { | |
2173 | start = page->index << PAGE_CACHE_SHIFT; | |
2174 | end = start + PAGE_CACHE_SIZE - 1; | |
2175 | if (test_range_bit(tree, start, end, | |
2176 | EXTENT_DIRTY, 0)) { | |
2177 | unlock_page(page); | |
2178 | continue; | |
2179 | } | |
2180 | } | |
2181 | clear_page_dirty_for_io(page); | |
2182 | unlock_page(page); | |
2183 | } | |
2184 | return 0; | |
2185 | } | |
2186 | EXPORT_SYMBOL(clear_extent_buffer_dirty); | |
2187 | ||
2188 | int wait_on_extent_buffer_writeback(struct extent_map_tree *tree, | |
2189 | struct extent_buffer *eb) | |
2190 | { | |
2191 | return wait_on_extent_writeback(tree, eb->start, | |
2192 | eb->start + eb->len - 1); | |
2193 | } | |
2194 | EXPORT_SYMBOL(wait_on_extent_buffer_writeback); | |
2195 | ||
2196 | int set_extent_buffer_dirty(struct extent_map_tree *tree, | |
2197 | struct extent_buffer *eb) | |
2198 | { | |
810191ff CM |
2199 | unsigned long i; |
2200 | unsigned long num_pages; | |
2201 | ||
2202 | num_pages = num_extent_pages(eb->start, eb->len); | |
2203 | for (i = 0; i < num_pages; i++) { | |
19c00ddc CM |
2204 | struct page *page = extent_buffer_page(eb, i); |
2205 | /* writepage may need to do something special for the | |
2206 | * first page, we have to make sure page->private is | |
2207 | * properly set. releasepage may drop page->private | |
2208 | * on us if the page isn't already dirty. | |
2209 | */ | |
2210 | if (i == 0) { | |
2211 | lock_page(page); | |
2212 | set_page_private(page, | |
2213 | EXTENT_PAGE_PRIVATE_FIRST_PAGE | | |
2214 | eb->len << 2); | |
2215 | } | |
810191ff | 2216 | __set_page_dirty_nobuffers(extent_buffer_page(eb, i)); |
19c00ddc CM |
2217 | if (i == 0) |
2218 | unlock_page(page); | |
810191ff CM |
2219 | } |
2220 | return set_extent_dirty(tree, eb->start, | |
2221 | eb->start + eb->len - 1, GFP_NOFS); | |
5f39d397 CM |
2222 | } |
2223 | EXPORT_SYMBOL(set_extent_buffer_dirty); | |
2224 | ||
2225 | int set_extent_buffer_uptodate(struct extent_map_tree *tree, | |
2226 | struct extent_buffer *eb) | |
2227 | { | |
2228 | unsigned long i; | |
2229 | struct page *page; | |
2230 | unsigned long num_pages; | |
2231 | ||
db94535d | 2232 | num_pages = num_extent_pages(eb->start, eb->len); |
5f39d397 CM |
2233 | |
2234 | set_extent_uptodate(tree, eb->start, eb->start + eb->len - 1, | |
2235 | GFP_NOFS); | |
2236 | for (i = 0; i < num_pages; i++) { | |
6d36dcd4 | 2237 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2238 | if ((i == 0 && (eb->start & (PAGE_CACHE_SIZE - 1))) || |
2239 | ((i == num_pages - 1) && | |
2240 | ((eb->start + eb->len - 1) & (PAGE_CACHE_SIZE - 1)))) { | |
2241 | check_page_uptodate(tree, page); | |
2242 | continue; | |
2243 | } | |
2244 | SetPageUptodate(page); | |
2245 | } | |
2246 | return 0; | |
2247 | } | |
2248 | EXPORT_SYMBOL(set_extent_buffer_uptodate); | |
2249 | ||
2250 | int extent_buffer_uptodate(struct extent_map_tree *tree, | |
2251 | struct extent_buffer *eb) | |
2252 | { | |
2253 | if (eb->flags & EXTENT_UPTODATE) | |
2254 | return 1; | |
2255 | return test_range_bit(tree, eb->start, eb->start + eb->len - 1, | |
2256 | EXTENT_UPTODATE, 1); | |
2257 | } | |
2258 | EXPORT_SYMBOL(extent_buffer_uptodate); | |
2259 | ||
2260 | int read_extent_buffer_pages(struct extent_map_tree *tree, | |
19c00ddc CM |
2261 | struct extent_buffer *eb, |
2262 | u64 start, | |
2263 | int wait) | |
5f39d397 CM |
2264 | { |
2265 | unsigned long i; | |
19c00ddc | 2266 | unsigned long start_i; |
5f39d397 CM |
2267 | struct page *page; |
2268 | int err; | |
2269 | int ret = 0; | |
2270 | unsigned long num_pages; | |
2271 | ||
2272 | if (eb->flags & EXTENT_UPTODATE) | |
2273 | return 0; | |
2274 | ||
14048ed0 | 2275 | if (0 && test_range_bit(tree, eb->start, eb->start + eb->len - 1, |
5f39d397 CM |
2276 | EXTENT_UPTODATE, 1)) { |
2277 | return 0; | |
2278 | } | |
19c00ddc CM |
2279 | if (start) { |
2280 | WARN_ON(start < eb->start); | |
2281 | start_i = (start >> PAGE_CACHE_SHIFT) - | |
2282 | (eb->start >> PAGE_CACHE_SHIFT); | |
2283 | } else { | |
2284 | start_i = 0; | |
2285 | } | |
5f39d397 | 2286 | |
db94535d | 2287 | num_pages = num_extent_pages(eb->start, eb->len); |
19c00ddc | 2288 | for (i = start_i; i < num_pages; i++) { |
6d36dcd4 | 2289 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2290 | if (PageUptodate(page)) { |
2291 | continue; | |
2292 | } | |
2293 | if (!wait) { | |
2294 | if (TestSetPageLocked(page)) { | |
2295 | continue; | |
2296 | } | |
2297 | } else { | |
2298 | lock_page(page); | |
2299 | } | |
2300 | if (!PageUptodate(page)) { | |
2301 | err = page->mapping->a_ops->readpage(NULL, page); | |
2302 | if (err) { | |
2303 | ret = err; | |
2304 | } | |
2305 | } else { | |
2306 | unlock_page(page); | |
2307 | } | |
2308 | } | |
2309 | ||
2310 | if (ret || !wait) { | |
2311 | return ret; | |
2312 | } | |
2313 | ||
19c00ddc | 2314 | for (i = start_i; i < num_pages; i++) { |
6d36dcd4 | 2315 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2316 | wait_on_page_locked(page); |
2317 | if (!PageUptodate(page)) { | |
2318 | ret = -EIO; | |
2319 | } | |
2320 | } | |
4dc11904 CM |
2321 | if (!ret) |
2322 | eb->flags |= EXTENT_UPTODATE; | |
5f39d397 CM |
2323 | return ret; |
2324 | } | |
2325 | EXPORT_SYMBOL(read_extent_buffer_pages); | |
2326 | ||
2327 | void read_extent_buffer(struct extent_buffer *eb, void *dstv, | |
2328 | unsigned long start, | |
2329 | unsigned long len) | |
2330 | { | |
2331 | size_t cur; | |
2332 | size_t offset; | |
2333 | struct page *page; | |
2334 | char *kaddr; | |
2335 | char *dst = (char *)dstv; | |
2336 | size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2337 | unsigned long i = (start_offset + start) >> PAGE_CACHE_SHIFT; | |
14048ed0 | 2338 | unsigned long num_pages = num_extent_pages(eb->start, eb->len); |
5f39d397 CM |
2339 | |
2340 | WARN_ON(start > eb->len); | |
2341 | WARN_ON(start + len > eb->start + eb->len); | |
2342 | ||
5f39d397 CM |
2343 | offset = start & ((unsigned long)PAGE_CACHE_SIZE - 1); |
2344 | if (i == 0) | |
2345 | offset += start_offset; | |
2346 | ||
2347 | while(len > 0) { | |
6d36dcd4 | 2348 | page = extent_buffer_page(eb, i); |
14048ed0 CM |
2349 | if (!PageUptodate(page)) { |
2350 | printk("page %lu not up to date i %lu, total %lu, len %lu\n", page->index, i, num_pages, eb->len); | |
2351 | WARN_ON(1); | |
2352 | } | |
5f39d397 CM |
2353 | WARN_ON(!PageUptodate(page)); |
2354 | ||
2355 | cur = min(len, (PAGE_CACHE_SIZE - offset)); | |
ae5252bd | 2356 | kaddr = kmap_atomic(page, KM_USER0); |
5f39d397 | 2357 | memcpy(dst, kaddr + offset, cur); |
ae5252bd | 2358 | kunmap_atomic(kaddr, KM_USER0); |
5f39d397 CM |
2359 | |
2360 | dst += cur; | |
2361 | len -= cur; | |
2362 | offset = 0; | |
2363 | i++; | |
5f39d397 CM |
2364 | } |
2365 | } | |
2366 | EXPORT_SYMBOL(read_extent_buffer); | |
2367 | ||
19c00ddc | 2368 | int map_private_extent_buffer(struct extent_buffer *eb, unsigned long start, |
db94535d CM |
2369 | unsigned long min_len, char **token, char **map, |
2370 | unsigned long *map_start, | |
2371 | unsigned long *map_len, int km) | |
5f39d397 | 2372 | { |
479965d6 | 2373 | size_t offset = start & (PAGE_CACHE_SIZE - 1); |
5f39d397 | 2374 | char *kaddr; |
db94535d | 2375 | struct page *p; |
5f39d397 CM |
2376 | size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1); |
2377 | unsigned long i = (start_offset + start) >> PAGE_CACHE_SHIFT; | |
479965d6 | 2378 | unsigned long end_i = (start_offset + start + min_len) >> |
810191ff | 2379 | PAGE_CACHE_SHIFT; |
479965d6 CM |
2380 | |
2381 | if (i != end_i) | |
2382 | return -EINVAL; | |
5f39d397 | 2383 | |
5f39d397 CM |
2384 | if (i == 0) { |
2385 | offset = start_offset; | |
2386 | *map_start = 0; | |
2387 | } else { | |
db94535d | 2388 | offset = 0; |
479965d6 | 2389 | *map_start = (i << PAGE_CACHE_SHIFT) - start_offset; |
5f39d397 | 2390 | } |
19c00ddc CM |
2391 | if (start + min_len >= eb->len) { |
2392 | printk("bad mapping eb start %Lu len %lu, wanted %lu %lu\n", eb->start, eb->len, start, min_len); | |
2393 | WARN_ON(1); | |
2394 | } | |
5f39d397 | 2395 | |
db94535d CM |
2396 | p = extent_buffer_page(eb, i); |
2397 | WARN_ON(!PageUptodate(p)); | |
2398 | kaddr = kmap_atomic(p, km); | |
5f39d397 CM |
2399 | *token = kaddr; |
2400 | *map = kaddr + offset; | |
2401 | *map_len = PAGE_CACHE_SIZE - offset; | |
2402 | return 0; | |
2403 | } | |
19c00ddc | 2404 | EXPORT_SYMBOL(map_private_extent_buffer); |
db94535d CM |
2405 | |
2406 | int map_extent_buffer(struct extent_buffer *eb, unsigned long start, | |
2407 | unsigned long min_len, | |
2408 | char **token, char **map, | |
2409 | unsigned long *map_start, | |
2410 | unsigned long *map_len, int km) | |
2411 | { | |
2412 | int err; | |
2413 | int save = 0; | |
2414 | if (eb->map_token) { | |
db94535d CM |
2415 | unmap_extent_buffer(eb, eb->map_token, km); |
2416 | eb->map_token = NULL; | |
2417 | save = 1; | |
2418 | } | |
19c00ddc CM |
2419 | err = map_private_extent_buffer(eb, start, min_len, token, map, |
2420 | map_start, map_len, km); | |
db94535d CM |
2421 | if (!err && save) { |
2422 | eb->map_token = *token; | |
2423 | eb->kaddr = *map; | |
2424 | eb->map_start = *map_start; | |
2425 | eb->map_len = *map_len; | |
2426 | } | |
2427 | return err; | |
2428 | } | |
5f39d397 CM |
2429 | EXPORT_SYMBOL(map_extent_buffer); |
2430 | ||
2431 | void unmap_extent_buffer(struct extent_buffer *eb, char *token, int km) | |
2432 | { | |
ae5252bd | 2433 | kunmap_atomic(token, km); |
5f39d397 CM |
2434 | } |
2435 | EXPORT_SYMBOL(unmap_extent_buffer); | |
2436 | ||
2437 | int memcmp_extent_buffer(struct extent_buffer *eb, const void *ptrv, | |
2438 | unsigned long start, | |
2439 | unsigned long len) | |
2440 | { | |
2441 | size_t cur; | |
2442 | size_t offset; | |
2443 | struct page *page; | |
2444 | char *kaddr; | |
2445 | char *ptr = (char *)ptrv; | |
2446 | size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2447 | unsigned long i = (start_offset + start) >> PAGE_CACHE_SHIFT; | |
2448 | int ret = 0; | |
2449 | ||
2450 | WARN_ON(start > eb->len); | |
2451 | WARN_ON(start + len > eb->start + eb->len); | |
2452 | ||
5f39d397 CM |
2453 | offset = start & ((unsigned long)PAGE_CACHE_SIZE - 1); |
2454 | if (i == 0) | |
2455 | offset += start_offset; | |
2456 | ||
2457 | while(len > 0) { | |
6d36dcd4 | 2458 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2459 | WARN_ON(!PageUptodate(page)); |
2460 | ||
2461 | cur = min(len, (PAGE_CACHE_SIZE - offset)); | |
2462 | ||
ae5252bd | 2463 | kaddr = kmap_atomic(page, KM_USER0); |
5f39d397 | 2464 | ret = memcmp(ptr, kaddr + offset, cur); |
ae5252bd | 2465 | kunmap_atomic(kaddr, KM_USER0); |
5f39d397 CM |
2466 | if (ret) |
2467 | break; | |
2468 | ||
2469 | ptr += cur; | |
2470 | len -= cur; | |
2471 | offset = 0; | |
2472 | i++; | |
5f39d397 CM |
2473 | } |
2474 | return ret; | |
2475 | } | |
2476 | EXPORT_SYMBOL(memcmp_extent_buffer); | |
2477 | ||
2478 | void write_extent_buffer(struct extent_buffer *eb, const void *srcv, | |
2479 | unsigned long start, unsigned long len) | |
2480 | { | |
2481 | size_t cur; | |
2482 | size_t offset; | |
2483 | struct page *page; | |
2484 | char *kaddr; | |
2485 | char *src = (char *)srcv; | |
2486 | size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2487 | unsigned long i = (start_offset + start) >> PAGE_CACHE_SHIFT; | |
2488 | ||
2489 | WARN_ON(start > eb->len); | |
2490 | WARN_ON(start + len > eb->start + eb->len); | |
2491 | ||
5f39d397 CM |
2492 | offset = start & ((unsigned long)PAGE_CACHE_SIZE - 1); |
2493 | if (i == 0) | |
2494 | offset += start_offset; | |
2495 | ||
2496 | while(len > 0) { | |
6d36dcd4 | 2497 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2498 | WARN_ON(!PageUptodate(page)); |
2499 | ||
2500 | cur = min(len, PAGE_CACHE_SIZE - offset); | |
ae5252bd | 2501 | kaddr = kmap_atomic(page, KM_USER0); |
5f39d397 | 2502 | memcpy(kaddr + offset, src, cur); |
ae5252bd | 2503 | kunmap_atomic(kaddr, KM_USER0); |
5f39d397 CM |
2504 | |
2505 | src += cur; | |
2506 | len -= cur; | |
2507 | offset = 0; | |
2508 | i++; | |
5f39d397 CM |
2509 | } |
2510 | } | |
2511 | EXPORT_SYMBOL(write_extent_buffer); | |
2512 | ||
2513 | void memset_extent_buffer(struct extent_buffer *eb, char c, | |
2514 | unsigned long start, unsigned long len) | |
2515 | { | |
2516 | size_t cur; | |
2517 | size_t offset; | |
2518 | struct page *page; | |
2519 | char *kaddr; | |
2520 | size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2521 | unsigned long i = (start_offset + start) >> PAGE_CACHE_SHIFT; | |
2522 | ||
2523 | WARN_ON(start > eb->len); | |
2524 | WARN_ON(start + len > eb->start + eb->len); | |
2525 | ||
5f39d397 CM |
2526 | offset = start & ((unsigned long)PAGE_CACHE_SIZE - 1); |
2527 | if (i == 0) | |
2528 | offset += start_offset; | |
2529 | ||
2530 | while(len > 0) { | |
6d36dcd4 | 2531 | page = extent_buffer_page(eb, i); |
5f39d397 CM |
2532 | WARN_ON(!PageUptodate(page)); |
2533 | ||
2534 | cur = min(len, PAGE_CACHE_SIZE - offset); | |
ae5252bd | 2535 | kaddr = kmap_atomic(page, KM_USER0); |
5f39d397 | 2536 | memset(kaddr + offset, c, cur); |
ae5252bd | 2537 | kunmap_atomic(kaddr, KM_USER0); |
5f39d397 CM |
2538 | |
2539 | len -= cur; | |
2540 | offset = 0; | |
2541 | i++; | |
5f39d397 CM |
2542 | } |
2543 | } | |
2544 | EXPORT_SYMBOL(memset_extent_buffer); | |
2545 | ||
2546 | void copy_extent_buffer(struct extent_buffer *dst, struct extent_buffer *src, | |
2547 | unsigned long dst_offset, unsigned long src_offset, | |
2548 | unsigned long len) | |
2549 | { | |
2550 | u64 dst_len = dst->len; | |
2551 | size_t cur; | |
2552 | size_t offset; | |
2553 | struct page *page; | |
2554 | char *kaddr; | |
2555 | size_t start_offset = dst->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2556 | unsigned long i = (start_offset + dst_offset) >> PAGE_CACHE_SHIFT; | |
2557 | ||
2558 | WARN_ON(src->len != dst_len); | |
2559 | ||
2560 | offset = dst_offset & ((unsigned long)PAGE_CACHE_SIZE - 1); | |
2561 | if (i == 0) | |
2562 | offset += start_offset; | |
2563 | ||
2564 | while(len > 0) { | |
6d36dcd4 | 2565 | page = extent_buffer_page(dst, i); |
5f39d397 CM |
2566 | WARN_ON(!PageUptodate(page)); |
2567 | ||
2568 | cur = min(len, (unsigned long)(PAGE_CACHE_SIZE - offset)); | |
2569 | ||
ae5252bd | 2570 | kaddr = kmap_atomic(page, KM_USER1); |
5f39d397 | 2571 | read_extent_buffer(src, kaddr + offset, src_offset, cur); |
ae5252bd | 2572 | kunmap_atomic(kaddr, KM_USER1); |
5f39d397 CM |
2573 | |
2574 | src_offset += cur; | |
2575 | len -= cur; | |
2576 | offset = 0; | |
2577 | i++; | |
2578 | } | |
2579 | } | |
2580 | EXPORT_SYMBOL(copy_extent_buffer); | |
2581 | ||
2582 | static void move_pages(struct page *dst_page, struct page *src_page, | |
2583 | unsigned long dst_off, unsigned long src_off, | |
2584 | unsigned long len) | |
2585 | { | |
ae5252bd | 2586 | char *dst_kaddr = kmap_atomic(dst_page, KM_USER0); |
5f39d397 CM |
2587 | if (dst_page == src_page) { |
2588 | memmove(dst_kaddr + dst_off, dst_kaddr + src_off, len); | |
2589 | } else { | |
ae5252bd | 2590 | char *src_kaddr = kmap_atomic(src_page, KM_USER1); |
5f39d397 CM |
2591 | char *p = dst_kaddr + dst_off + len; |
2592 | char *s = src_kaddr + src_off + len; | |
2593 | ||
2594 | while (len--) | |
2595 | *--p = *--s; | |
2596 | ||
ae5252bd | 2597 | kunmap_atomic(src_kaddr, KM_USER1); |
5f39d397 | 2598 | } |
ae5252bd | 2599 | kunmap_atomic(dst_kaddr, KM_USER0); |
5f39d397 CM |
2600 | } |
2601 | ||
2602 | static void copy_pages(struct page *dst_page, struct page *src_page, | |
2603 | unsigned long dst_off, unsigned long src_off, | |
2604 | unsigned long len) | |
2605 | { | |
ae5252bd | 2606 | char *dst_kaddr = kmap_atomic(dst_page, KM_USER0); |
5f39d397 CM |
2607 | char *src_kaddr; |
2608 | ||
2609 | if (dst_page != src_page) | |
ae5252bd | 2610 | src_kaddr = kmap_atomic(src_page, KM_USER1); |
5f39d397 CM |
2611 | else |
2612 | src_kaddr = dst_kaddr; | |
2613 | ||
2614 | memcpy(dst_kaddr + dst_off, src_kaddr + src_off, len); | |
5f39d397 CM |
2615 | kunmap_atomic(dst_kaddr, KM_USER0); |
2616 | if (dst_page != src_page) | |
2617 | kunmap_atomic(src_kaddr, KM_USER1); | |
5f39d397 CM |
2618 | } |
2619 | ||
2620 | void memcpy_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset, | |
2621 | unsigned long src_offset, unsigned long len) | |
2622 | { | |
2623 | size_t cur; | |
2624 | size_t dst_off_in_page; | |
2625 | size_t src_off_in_page; | |
2626 | size_t start_offset = dst->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2627 | unsigned long dst_i; | |
2628 | unsigned long src_i; | |
2629 | ||
2630 | if (src_offset + len > dst->len) { | |
2631 | printk("memmove bogus src_offset %lu move len %lu len %lu\n", | |
2632 | src_offset, len, dst->len); | |
2633 | BUG_ON(1); | |
2634 | } | |
2635 | if (dst_offset + len > dst->len) { | |
2636 | printk("memmove bogus dst_offset %lu move len %lu len %lu\n", | |
2637 | dst_offset, len, dst->len); | |
2638 | BUG_ON(1); | |
2639 | } | |
2640 | ||
2641 | while(len > 0) { | |
2642 | dst_off_in_page = dst_offset & | |
2643 | ((unsigned long)PAGE_CACHE_SIZE - 1); | |
2644 | src_off_in_page = src_offset & | |
2645 | ((unsigned long)PAGE_CACHE_SIZE - 1); | |
2646 | ||
2647 | dst_i = (start_offset + dst_offset) >> PAGE_CACHE_SHIFT; | |
2648 | src_i = (start_offset + src_offset) >> PAGE_CACHE_SHIFT; | |
2649 | ||
2650 | if (src_i == 0) | |
2651 | src_off_in_page += start_offset; | |
2652 | if (dst_i == 0) | |
2653 | dst_off_in_page += start_offset; | |
2654 | ||
2655 | cur = min(len, (unsigned long)(PAGE_CACHE_SIZE - | |
2656 | src_off_in_page)); | |
ae2f5411 JA |
2657 | cur = min_t(unsigned long, cur, |
2658 | (unsigned long)(PAGE_CACHE_SIZE - dst_off_in_page)); | |
5f39d397 | 2659 | |
6d36dcd4 CM |
2660 | copy_pages(extent_buffer_page(dst, dst_i), |
2661 | extent_buffer_page(dst, src_i), | |
5f39d397 CM |
2662 | dst_off_in_page, src_off_in_page, cur); |
2663 | ||
2664 | src_offset += cur; | |
2665 | dst_offset += cur; | |
2666 | len -= cur; | |
2667 | } | |
2668 | } | |
2669 | EXPORT_SYMBOL(memcpy_extent_buffer); | |
2670 | ||
2671 | void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset, | |
2672 | unsigned long src_offset, unsigned long len) | |
2673 | { | |
2674 | size_t cur; | |
2675 | size_t dst_off_in_page; | |
2676 | size_t src_off_in_page; | |
2677 | unsigned long dst_end = dst_offset + len - 1; | |
2678 | unsigned long src_end = src_offset + len - 1; | |
2679 | size_t start_offset = dst->start & ((u64)PAGE_CACHE_SIZE - 1); | |
2680 | unsigned long dst_i; | |
2681 | unsigned long src_i; | |
2682 | ||
2683 | if (src_offset + len > dst->len) { | |
2684 | printk("memmove bogus src_offset %lu move len %lu len %lu\n", | |
2685 | src_offset, len, dst->len); | |
2686 | BUG_ON(1); | |
2687 | } | |
2688 | if (dst_offset + len > dst->len) { | |
2689 | printk("memmove bogus dst_offset %lu move len %lu len %lu\n", | |
2690 | dst_offset, len, dst->len); | |
2691 | BUG_ON(1); | |
2692 | } | |
2693 | if (dst_offset < src_offset) { | |
2694 | memcpy_extent_buffer(dst, dst_offset, src_offset, len); | |
2695 | return; | |
2696 | } | |
2697 | while(len > 0) { | |
2698 | dst_i = (start_offset + dst_end) >> PAGE_CACHE_SHIFT; | |
2699 | src_i = (start_offset + src_end) >> PAGE_CACHE_SHIFT; | |
2700 | ||
2701 | dst_off_in_page = dst_end & | |
2702 | ((unsigned long)PAGE_CACHE_SIZE - 1); | |
2703 | src_off_in_page = src_end & | |
2704 | ((unsigned long)PAGE_CACHE_SIZE - 1); | |
5f39d397 CM |
2705 | if (src_i == 0) |
2706 | src_off_in_page += start_offset; | |
2707 | if (dst_i == 0) | |
2708 | dst_off_in_page += start_offset; | |
2709 | ||
ae2f5411 | 2710 | cur = min_t(unsigned long, len, src_off_in_page + 1); |
5f39d397 | 2711 | cur = min(cur, dst_off_in_page + 1); |
6d36dcd4 CM |
2712 | move_pages(extent_buffer_page(dst, dst_i), |
2713 | extent_buffer_page(dst, src_i), | |
5f39d397 CM |
2714 | dst_off_in_page - cur + 1, |
2715 | src_off_in_page - cur + 1, cur); | |
2716 | ||
db94535d CM |
2717 | dst_end -= cur; |
2718 | src_end -= cur; | |
5f39d397 CM |
2719 | len -= cur; |
2720 | } | |
2721 | } | |
2722 | EXPORT_SYMBOL(memmove_extent_buffer); |