Root/
1 | /* |
2 | * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com |
3 | * Written by Alex Tomas <alex@clusterfs.com> |
4 | * |
5 | * This program is free software; you can redistribute it and/or modify |
6 | * it under the terms of the GNU General Public License version 2 as |
7 | * published by the Free Software Foundation. |
8 | * |
9 | * This program is distributed in the hope that it will be useful, |
10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
12 | * GNU General Public License for more details. |
13 | * |
14 | * You should have received a copy of the GNU General Public Licens |
15 | * along with this program; if not, write to the Free Software |
16 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111- |
17 | */ |
18 | |
19 | |
20 | /* |
21 | * mballoc.c contains the multiblocks allocation routines |
22 | */ |
23 | |
24 | #include "mballoc.h" |
25 | #include <trace/events/ext4.h> |
26 | |
27 | /* |
28 | * MUSTDO: |
29 | * - test ext4_ext_search_left() and ext4_ext_search_right() |
30 | * - search for metadata in few groups |
31 | * |
32 | * TODO v4: |
33 | * - normalization should take into account whether file is still open |
34 | * - discard preallocations if no free space left (policy?) |
35 | * - don't normalize tails |
36 | * - quota |
37 | * - reservation for superuser |
38 | * |
39 | * TODO v3: |
40 | * - bitmap read-ahead (proposed by Oleg Drokin aka green) |
41 | * - track min/max extents in each group for better group selection |
42 | * - mb_mark_used() may allocate chunk right after splitting buddy |
43 | * - tree of groups sorted by number of free blocks |
44 | * - error handling |
45 | */ |
46 | |
47 | /* |
48 | * The allocation request involve request for multiple number of blocks |
49 | * near to the goal(block) value specified. |
50 | * |
51 | * During initialization phase of the allocator we decide to use the |
52 | * group preallocation or inode preallocation depending on the size of |
53 | * the file. The size of the file could be the resulting file size we |
54 | * would have after allocation, or the current file size, which ever |
55 | * is larger. If the size is less than sbi->s_mb_stream_request we |
56 | * select to use the group preallocation. The default value of |
57 | * s_mb_stream_request is 16 blocks. This can also be tuned via |
58 | * /sys/fs/ext4/<partition>/mb_stream_req. The value is represented in |
59 | * terms of number of blocks. |
60 | * |
61 | * The main motivation for having small file use group preallocation is to |
62 | * ensure that we have small files closer together on the disk. |
63 | * |
64 | * First stage the allocator looks at the inode prealloc list, |
65 | * ext4_inode_info->i_prealloc_list, which contains list of prealloc |
66 | * spaces for this particular inode. The inode prealloc space is |
67 | * represented as: |
68 | * |
69 | * pa_lstart -> the logical start block for this prealloc space |
70 | * pa_pstart -> the physical start block for this prealloc space |
71 | * pa_len -> lenght for this prealloc space |
72 | * pa_free -> free space available in this prealloc space |
73 | * |
74 | * The inode preallocation space is used looking at the _logical_ start |
75 | * block. If only the logical file block falls within the range of prealloc |
76 | * space we will consume the particular prealloc space. This make sure that |
77 | * that the we have contiguous physical blocks representing the file blocks |
78 | * |
79 | * The important thing to be noted in case of inode prealloc space is that |
80 | * we don't modify the values associated to inode prealloc space except |
81 | * pa_free. |
82 | * |
83 | * If we are not able to find blocks in the inode prealloc space and if we |
84 | * have the group allocation flag set then we look at the locality group |
85 | * prealloc space. These are per CPU prealloc list repreasented as |
86 | * |
87 | * ext4_sb_info.s_locality_groups[smp_processor_id()] |
88 | * |
89 | * The reason for having a per cpu locality group is to reduce the contention |
90 | * between CPUs. It is possible to get scheduled at this point. |
91 | * |
92 | * The locality group prealloc space is used looking at whether we have |
93 | * enough free space (pa_free) withing the prealloc space. |
94 | * |
95 | * If we can't allocate blocks via inode prealloc or/and locality group |
96 | * prealloc then we look at the buddy cache. The buddy cache is represented |
97 | * by ext4_sb_info.s_buddy_cache (struct inode) whose file offset gets |
98 | * mapped to the buddy and bitmap information regarding different |
99 | * groups. The buddy information is attached to buddy cache inode so that |
100 | * we can access them through the page cache. The information regarding |
101 | * each group is loaded via ext4_mb_load_buddy. The information involve |
102 | * block bitmap and buddy information. The information are stored in the |
103 | * inode as: |
104 | * |
105 | * { page } |
106 | * [ group 0 bitmap][ group 0 buddy] [group 1][ group 1]... |
107 | * |
108 | * |
109 | * one block each for bitmap and buddy information. So for each group we |
110 | * take up 2 blocks. A page can contain blocks_per_page (PAGE_CACHE_SIZE / |
111 | * blocksize) blocks. So it can have information regarding groups_per_page |
112 | * which is blocks_per_page/2 |
113 | * |
114 | * The buddy cache inode is not stored on disk. The inode is thrown |
115 | * away when the filesystem is unmounted. |
116 | * |
117 | * We look for count number of blocks in the buddy cache. If we were able |
118 | * to locate that many free blocks we return with additional information |
119 | * regarding rest of the contiguous physical block available |
120 | * |
121 | * Before allocating blocks via buddy cache we normalize the request |
122 | * blocks. This ensure we ask for more blocks that we needed. The extra |
123 | * blocks that we get after allocation is added to the respective prealloc |
124 | * list. In case of inode preallocation we follow a list of heuristics |
125 | * based on file size. This can be found in ext4_mb_normalize_request. If |
126 | * we are doing a group prealloc we try to normalize the request to |
127 | * sbi->s_mb_group_prealloc. Default value of s_mb_group_prealloc is |
128 | * 512 blocks. This can be tuned via |
129 | * /sys/fs/ext4/<partition/mb_group_prealloc. The value is represented in |
130 | * terms of number of blocks. If we have mounted the file system with -O |
131 | * stripe=<value> option the group prealloc request is normalized to the |
132 | * stripe value (sbi->s_stripe) |
133 | * |
134 | * The regular allocator(using the buddy cache) supports few tunables. |
135 | * |
136 | * /sys/fs/ext4/<partition>/mb_min_to_scan |
137 | * /sys/fs/ext4/<partition>/mb_max_to_scan |
138 | * /sys/fs/ext4/<partition>/mb_order2_req |
139 | * |
140 | * The regular allocator uses buddy scan only if the request len is power of |
141 | * 2 blocks and the order of allocation is >= sbi->s_mb_order2_reqs. The |
142 | * value of s_mb_order2_reqs can be tuned via |
143 | * /sys/fs/ext4/<partition>/mb_order2_req. If the request len is equal to |
144 | * stripe size (sbi->s_stripe), we try to search for contigous block in |
145 | * stripe size. This should result in better allocation on RAID setups. If |
146 | * not, we search in the specific group using bitmap for best extents. The |
147 | * tunable min_to_scan and max_to_scan control the behaviour here. |
148 | * min_to_scan indicate how long the mballoc __must__ look for a best |
149 | * extent and max_to_scan indicates how long the mballoc __can__ look for a |
150 | * best extent in the found extents. Searching for the blocks starts with |
151 | * the group specified as the goal value in allocation context via |
152 | * ac_g_ex. Each group is first checked based on the criteria whether it |
153 | * can used for allocation. ext4_mb_good_group explains how the groups are |
154 | * checked. |
155 | * |
156 | * Both the prealloc space are getting populated as above. So for the first |
157 | * request we will hit the buddy cache which will result in this prealloc |
158 | * space getting filled. The prealloc space is then later used for the |
159 | * subsequent request. |
160 | */ |
161 | |
162 | /* |
163 | * mballoc operates on the following data: |
164 | * - on-disk bitmap |
165 | * - in-core buddy (actually includes buddy and bitmap) |
166 | * - preallocation descriptors (PAs) |
167 | * |
168 | * there are two types of preallocations: |
169 | * - inode |
170 | * assiged to specific inode and can be used for this inode only. |
171 | * it describes part of inode's space preallocated to specific |
172 | * physical blocks. any block from that preallocated can be used |
173 | * independent. the descriptor just tracks number of blocks left |
174 | * unused. so, before taking some block from descriptor, one must |
175 | * make sure corresponded logical block isn't allocated yet. this |
176 | * also means that freeing any block within descriptor's range |
177 | * must discard all preallocated blocks. |
178 | * - locality group |
179 | * assigned to specific locality group which does not translate to |
180 | * permanent set of inodes: inode can join and leave group. space |
181 | * from this type of preallocation can be used for any inode. thus |
182 | * it's consumed from the beginning to the end. |
183 | * |
184 | * relation between them can be expressed as: |
185 | * in-core buddy = on-disk bitmap + preallocation descriptors |
186 | * |
187 | * this mean blocks mballoc considers used are: |
188 | * - allocated blocks (persistent) |
189 | * - preallocated blocks (non-persistent) |
190 | * |
191 | * consistency in mballoc world means that at any time a block is either |
192 | * free or used in ALL structures. notice: "any time" should not be read |
193 | * literally -- time is discrete and delimited by locks. |
194 | * |
195 | * to keep it simple, we don't use block numbers, instead we count number of |
196 | * blocks: how many blocks marked used/free in on-disk bitmap, buddy and PA. |
197 | * |
198 | * all operations can be expressed as: |
199 | * - init buddy: buddy = on-disk + PAs |
200 | * - new PA: buddy += N; PA = N |
201 | * - use inode PA: on-disk += N; PA -= N |
202 | * - discard inode PA buddy -= on-disk - PA; PA = 0 |
203 | * - use locality group PA on-disk += N; PA -= N |
204 | * - discard locality group PA buddy -= PA; PA = 0 |
205 | * note: 'buddy -= on-disk - PA' is used to show that on-disk bitmap |
206 | * is used in real operation because we can't know actual used |
207 | * bits from PA, only from on-disk bitmap |
208 | * |
209 | * if we follow this strict logic, then all operations above should be atomic. |
210 | * given some of them can block, we'd have to use something like semaphores |
211 | * killing performance on high-end SMP hardware. let's try to relax it using |
212 | * the following knowledge: |
213 | * 1) if buddy is referenced, it's already initialized |
214 | * 2) while block is used in buddy and the buddy is referenced, |
215 | * nobody can re-allocate that block |
216 | * 3) we work on bitmaps and '+' actually means 'set bits'. if on-disk has |
217 | * bit set and PA claims same block, it's OK. IOW, one can set bit in |
218 | * on-disk bitmap if buddy has same bit set or/and PA covers corresponded |
219 | * block |
220 | * |
221 | * so, now we're building a concurrency table: |
222 | * - init buddy vs. |
223 | * - new PA |
224 | * blocks for PA are allocated in the buddy, buddy must be referenced |
225 | * until PA is linked to allocation group to avoid concurrent buddy init |
226 | * - use inode PA |
227 | * we need to make sure that either on-disk bitmap or PA has uptodate data |
228 | * given (3) we care that PA-=N operation doesn't interfere with init |
229 | * - discard inode PA |
230 | * the simplest way would be to have buddy initialized by the discard |
231 | * - use locality group PA |
232 | * again PA-=N must be serialized with init |
233 | * - discard locality group PA |
234 | * the simplest way would be to have buddy initialized by the discard |
235 | * - new PA vs. |
236 | * - use inode PA |
237 | * i_data_sem serializes them |
238 | * - discard inode PA |
239 | * discard process must wait until PA isn't used by another process |
240 | * - use locality group PA |
241 | * some mutex should serialize them |
242 | * - discard locality group PA |
243 | * discard process must wait until PA isn't used by another process |
244 | * - use inode PA |
245 | * - use inode PA |
246 | * i_data_sem or another mutex should serializes them |
247 | * - discard inode PA |
248 | * discard process must wait until PA isn't used by another process |
249 | * - use locality group PA |
250 | * nothing wrong here -- they're different PAs covering different blocks |
251 | * - discard locality group PA |
252 | * discard process must wait until PA isn't used by another process |
253 | * |
254 | * now we're ready to make few consequences: |
255 | * - PA is referenced and while it is no discard is possible |
256 | * - PA is referenced until block isn't marked in on-disk bitmap |
257 | * - PA changes only after on-disk bitmap |
258 | * - discard must not compete with init. either init is done before |
259 | * any discard or they're serialized somehow |
260 | * - buddy init as sum of on-disk bitmap and PAs is done atomically |
261 | * |
262 | * a special case when we've used PA to emptiness. no need to modify buddy |
263 | * in this case, but we should care about concurrent init |
264 | * |
265 | */ |
266 | |
267 | /* |
268 | * Logic in few words: |
269 | * |
270 | * - allocation: |
271 | * load group |
272 | * find blocks |
273 | * mark bits in on-disk bitmap |
274 | * release group |
275 | * |
276 | * - use preallocation: |
277 | * find proper PA (per-inode or group) |
278 | * load group |
279 | * mark bits in on-disk bitmap |
280 | * release group |
281 | * release PA |
282 | * |
283 | * - free: |
284 | * load group |
285 | * mark bits in on-disk bitmap |
286 | * release group |
287 | * |
288 | * - discard preallocations in group: |
289 | * mark PAs deleted |
290 | * move them onto local list |
291 | * load on-disk bitmap |
292 | * load group |
293 | * remove PA from object (inode or locality group) |
294 | * mark free blocks in-core |
295 | * |
296 | * - discard inode's preallocations: |
297 | */ |
298 | |
299 | /* |
300 | * Locking rules |
301 | * |
302 | * Locks: |
303 | * - bitlock on a group (group) |
304 | * - object (inode/locality) (object) |
305 | * - per-pa lock (pa) |
306 | * |
307 | * Paths: |
308 | * - new pa |
309 | * object |
310 | * group |
311 | * |
312 | * - find and use pa: |
313 | * pa |
314 | * |
315 | * - release consumed pa: |
316 | * pa |
317 | * group |
318 | * object |
319 | * |
320 | * - generate in-core bitmap: |
321 | * group |
322 | * pa |
323 | * |
324 | * - discard all for given object (inode, locality group): |
325 | * object |
326 | * pa |
327 | * group |
328 | * |
329 | * - discard all for given group: |
330 | * group |
331 | * pa |
332 | * group |
333 | * object |
334 | * |
335 | */ |
336 | static struct kmem_cache *ext4_pspace_cachep; |
337 | static struct kmem_cache *ext4_ac_cachep; |
338 | static struct kmem_cache *ext4_free_ext_cachep; |
339 | static void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, |
340 | ext4_group_t group); |
341 | static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, |
342 | ext4_group_t group); |
343 | static void release_blocks_on_commit(journal_t *journal, transaction_t *txn); |
344 | |
345 | static inline void *mb_correct_addr_and_bit(int *bit, void *addr) |
346 | { |
347 | #if BITS_PER_LONG == 64 |
348 | *bit += ((unsigned long) addr & 7UL) << 3; |
349 | addr = (void *) ((unsigned long) addr & ~7UL); |
350 | #elif BITS_PER_LONG == 32 |
351 | *bit += ((unsigned long) addr & 3UL) << 3; |
352 | addr = (void *) ((unsigned long) addr & ~3UL); |
353 | #else |
354 | #error "how many bits you are?!" |
355 | #endif |
356 | return addr; |
357 | } |
358 | |
359 | static inline int mb_test_bit(int bit, void *addr) |
360 | { |
361 | /* |
362 | * ext4_test_bit on architecture like powerpc |
363 | * needs unsigned long aligned address |
364 | */ |
365 | addr = mb_correct_addr_and_bit(&bit, addr); |
366 | return ext4_test_bit(bit, addr); |
367 | } |
368 | |
369 | static inline void mb_set_bit(int bit, void *addr) |
370 | { |
371 | addr = mb_correct_addr_and_bit(&bit, addr); |
372 | ext4_set_bit(bit, addr); |
373 | } |
374 | |
375 | static inline void mb_clear_bit(int bit, void *addr) |
376 | { |
377 | addr = mb_correct_addr_and_bit(&bit, addr); |
378 | ext4_clear_bit(bit, addr); |
379 | } |
380 | |
381 | static inline int mb_find_next_zero_bit(void *addr, int max, int start) |
382 | { |
383 | int fix = 0, ret, tmpmax; |
384 | addr = mb_correct_addr_and_bit(&fix, addr); |
385 | tmpmax = max + fix; |
386 | start += fix; |
387 | |
388 | ret = ext4_find_next_zero_bit(addr, tmpmax, start) - fix; |
389 | if (ret > max) |
390 | return max; |
391 | return ret; |
392 | } |
393 | |
394 | static inline int mb_find_next_bit(void *addr, int max, int start) |
395 | { |
396 | int fix = 0, ret, tmpmax; |
397 | addr = mb_correct_addr_and_bit(&fix, addr); |
398 | tmpmax = max + fix; |
399 | start += fix; |
400 | |
401 | ret = ext4_find_next_bit(addr, tmpmax, start) - fix; |
402 | if (ret > max) |
403 | return max; |
404 | return ret; |
405 | } |
406 | |
407 | static void *mb_find_buddy(struct ext4_buddy *e4b, int order, int *max) |
408 | { |
409 | char *bb; |
410 | |
411 | BUG_ON(EXT4_MB_BITMAP(e4b) == EXT4_MB_BUDDY(e4b)); |
412 | BUG_ON(max == NULL); |
413 | |
414 | if (order > e4b->bd_blkbits + 1) { |
415 | *max = 0; |
416 | return NULL; |
417 | } |
418 | |
419 | /* at order 0 we see each particular block */ |
420 | *max = 1 << (e4b->bd_blkbits + 3); |
421 | if (order == 0) |
422 | return EXT4_MB_BITMAP(e4b); |
423 | |
424 | bb = EXT4_MB_BUDDY(e4b) + EXT4_SB(e4b->bd_sb)->s_mb_offsets[order]; |
425 | *max = EXT4_SB(e4b->bd_sb)->s_mb_maxs[order]; |
426 | |
427 | return bb; |
428 | } |
429 | |
430 | #ifdef DOUBLE_CHECK |
431 | static void mb_free_blocks_double(struct inode *inode, struct ext4_buddy *e4b, |
432 | int first, int count) |
433 | { |
434 | int i; |
435 | struct super_block *sb = e4b->bd_sb; |
436 | |
437 | if (unlikely(e4b->bd_info->bb_bitmap == NULL)) |
438 | return; |
439 | assert_spin_locked(ext4_group_lock_ptr(sb, e4b->bd_group)); |
440 | for (i = 0; i < count; i++) { |
441 | if (!mb_test_bit(first + i, e4b->bd_info->bb_bitmap)) { |
442 | ext4_fsblk_t blocknr; |
443 | blocknr = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb); |
444 | blocknr += first + i; |
445 | blocknr += |
446 | le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); |
447 | ext4_grp_locked_error(sb, e4b->bd_group, |
448 | __func__, "double-free of inode" |
449 | " %lu's block %llu(bit %u in group %u)", |
450 | inode ? inode->i_ino : 0, blocknr, |
451 | first + i, e4b->bd_group); |
452 | } |
453 | mb_clear_bit(first + i, e4b->bd_info->bb_bitmap); |
454 | } |
455 | } |
456 | |
457 | static void mb_mark_used_double(struct ext4_buddy *e4b, int first, int count) |
458 | { |
459 | int i; |
460 | |
461 | if (unlikely(e4b->bd_info->bb_bitmap == NULL)) |
462 | return; |
463 | assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); |
464 | for (i = 0; i < count; i++) { |
465 | BUG_ON(mb_test_bit(first + i, e4b->bd_info->bb_bitmap)); |
466 | mb_set_bit(first + i, e4b->bd_info->bb_bitmap); |
467 | } |
468 | } |
469 | |
470 | static void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap) |
471 | { |
472 | if (memcmp(e4b->bd_info->bb_bitmap, bitmap, e4b->bd_sb->s_blocksize)) { |
473 | unsigned char *b1, *b2; |
474 | int i; |
475 | b1 = (unsigned char *) e4b->bd_info->bb_bitmap; |
476 | b2 = (unsigned char *) bitmap; |
477 | for (i = 0; i < e4b->bd_sb->s_blocksize; i++) { |
478 | if (b1[i] != b2[i]) { |
479 | printk(KERN_ERR "corruption in group %u " |
480 | "at byte %u(%u): %x in copy != %x " |
481 | "on disk/prealloc\n", |
482 | e4b->bd_group, i, i * 8, b1[i], b2[i]); |
483 | BUG(); |
484 | } |
485 | } |
486 | } |
487 | } |
488 | |
489 | #else |
490 | static inline void mb_free_blocks_double(struct inode *inode, |
491 | struct ext4_buddy *e4b, int first, int count) |
492 | { |
493 | return; |
494 | } |
495 | static inline void mb_mark_used_double(struct ext4_buddy *e4b, |
496 | int first, int count) |
497 | { |
498 | return; |
499 | } |
500 | static inline void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap) |
501 | { |
502 | return; |
503 | } |
504 | #endif |
505 | |
506 | #ifdef AGGRESSIVE_CHECK |
507 | |
508 | #define MB_CHECK_ASSERT(assert) \ |
509 | do { \ |
510 | if (!(assert)) { \ |
511 | printk(KERN_EMERG \ |
512 | "Assertion failure in %s() at %s:%d: \"%s\"\n", \ |
513 | function, file, line, # assert); \ |
514 | BUG(); \ |
515 | } \ |
516 | } while (0) |
517 | |
518 | static int __mb_check_buddy(struct ext4_buddy *e4b, char *file, |
519 | const char *function, int line) |
520 | { |
521 | struct super_block *sb = e4b->bd_sb; |
522 | int order = e4b->bd_blkbits + 1; |
523 | int max; |
524 | int max2; |
525 | int i; |
526 | int j; |
527 | int k; |
528 | int count; |
529 | struct ext4_group_info *grp; |
530 | int fragments = 0; |
531 | int fstart; |
532 | struct list_head *cur; |
533 | void *buddy; |
534 | void *buddy2; |
535 | |
536 | { |
537 | static int mb_check_counter; |
538 | if (mb_check_counter++ % 100 != 0) |
539 | return 0; |
540 | } |
541 | |
542 | while (order > 1) { |
543 | buddy = mb_find_buddy(e4b, order, &max); |
544 | MB_CHECK_ASSERT(buddy); |
545 | buddy2 = mb_find_buddy(e4b, order - 1, &max2); |
546 | MB_CHECK_ASSERT(buddy2); |
547 | MB_CHECK_ASSERT(buddy != buddy2); |
548 | MB_CHECK_ASSERT(max * 2 == max2); |
549 | |
550 | count = 0; |
551 | for (i = 0; i < max; i++) { |
552 | |
553 | if (mb_test_bit(i, buddy)) { |
554 | /* only single bit in buddy2 may be 1 */ |
555 | if (!mb_test_bit(i << 1, buddy2)) { |
556 | MB_CHECK_ASSERT( |
557 | mb_test_bit((i<<1)+1, buddy2)); |
558 | } else if (!mb_test_bit((i << 1) + 1, buddy2)) { |
559 | MB_CHECK_ASSERT( |
560 | mb_test_bit(i << 1, buddy2)); |
561 | } |
562 | continue; |
563 | } |
564 | |
565 | /* both bits in buddy2 must be 0 */ |
566 | MB_CHECK_ASSERT(mb_test_bit(i << 1, buddy2)); |
567 | MB_CHECK_ASSERT(mb_test_bit((i << 1) + 1, buddy2)); |
568 | |
569 | for (j = 0; j < (1 << order); j++) { |
570 | k = (i * (1 << order)) + j; |
571 | MB_CHECK_ASSERT( |
572 | !mb_test_bit(k, EXT4_MB_BITMAP(e4b))); |
573 | } |
574 | count++; |
575 | } |
576 | MB_CHECK_ASSERT(e4b->bd_info->bb_counters[order] == count); |
577 | order--; |
578 | } |
579 | |
580 | fstart = -1; |
581 | buddy = mb_find_buddy(e4b, 0, &max); |
582 | for (i = 0; i < max; i++) { |
583 | if (!mb_test_bit(i, buddy)) { |
584 | MB_CHECK_ASSERT(i >= e4b->bd_info->bb_first_free); |
585 | if (fstart == -1) { |
586 | fragments++; |
587 | fstart = i; |
588 | } |
589 | continue; |
590 | } |
591 | fstart = -1; |
592 | /* check used bits only */ |
593 | for (j = 0; j < e4b->bd_blkbits + 1; j++) { |
594 | buddy2 = mb_find_buddy(e4b, j, &max2); |
595 | k = i >> j; |
596 | MB_CHECK_ASSERT(k < max2); |
597 | MB_CHECK_ASSERT(mb_test_bit(k, buddy2)); |
598 | } |
599 | } |
600 | MB_CHECK_ASSERT(!EXT4_MB_GRP_NEED_INIT(e4b->bd_info)); |
601 | MB_CHECK_ASSERT(e4b->bd_info->bb_fragments == fragments); |
602 | |
603 | grp = ext4_get_group_info(sb, e4b->bd_group); |
604 | buddy = mb_find_buddy(e4b, 0, &max); |
605 | list_for_each(cur, &grp->bb_prealloc_list) { |
606 | ext4_group_t groupnr; |
607 | struct ext4_prealloc_space *pa; |
608 | pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); |
609 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, &groupnr, &k); |
610 | MB_CHECK_ASSERT(groupnr == e4b->bd_group); |
611 | for (i = 0; i < pa->pa_len; i++) |
612 | MB_CHECK_ASSERT(mb_test_bit(k + i, buddy)); |
613 | } |
614 | return 0; |
615 | } |
616 | #undef MB_CHECK_ASSERT |
617 | #define mb_check_buddy(e4b) __mb_check_buddy(e4b, \ |
618 | __FILE__, __func__, __LINE__) |
619 | #else |
620 | #define mb_check_buddy(e4b) |
621 | #endif |
622 | |
623 | /* FIXME!! need more doc */ |
624 | static void ext4_mb_mark_free_simple(struct super_block *sb, |
625 | void *buddy, unsigned first, int len, |
626 | struct ext4_group_info *grp) |
627 | { |
628 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
629 | unsigned short min; |
630 | unsigned short max; |
631 | unsigned short chunk; |
632 | unsigned short border; |
633 | |
634 | BUG_ON(len > EXT4_BLOCKS_PER_GROUP(sb)); |
635 | |
636 | border = 2 << sb->s_blocksize_bits; |
637 | |
638 | while (len > 0) { |
639 | /* find how many blocks can be covered since this position */ |
640 | max = ffs(first | border) - 1; |
641 | |
642 | /* find how many blocks of power 2 we need to mark */ |
643 | min = fls(len) - 1; |
644 | |
645 | if (max < min) |
646 | min = max; |
647 | chunk = 1 << min; |
648 | |
649 | /* mark multiblock chunks only */ |
650 | grp->bb_counters[min]++; |
651 | if (min > 0) |
652 | mb_clear_bit(first >> min, |
653 | buddy + sbi->s_mb_offsets[min]); |
654 | |
655 | len -= chunk; |
656 | first += chunk; |
657 | } |
658 | } |
659 | |
660 | static noinline_for_stack |
661 | void ext4_mb_generate_buddy(struct super_block *sb, |
662 | void *buddy, void *bitmap, ext4_group_t group) |
663 | { |
664 | struct ext4_group_info *grp = ext4_get_group_info(sb, group); |
665 | unsigned short max = EXT4_BLOCKS_PER_GROUP(sb); |
666 | unsigned short i = 0; |
667 | unsigned short first; |
668 | unsigned short len; |
669 | unsigned free = 0; |
670 | unsigned fragments = 0; |
671 | unsigned long long period = get_cycles(); |
672 | |
673 | /* initialize buddy from bitmap which is aggregation |
674 | * of on-disk bitmap and preallocations */ |
675 | i = mb_find_next_zero_bit(bitmap, max, 0); |
676 | grp->bb_first_free = i; |
677 | while (i < max) { |
678 | fragments++; |
679 | first = i; |
680 | i = mb_find_next_bit(bitmap, max, i); |
681 | len = i - first; |
682 | free += len; |
683 | if (len > 1) |
684 | ext4_mb_mark_free_simple(sb, buddy, first, len, grp); |
685 | else |
686 | grp->bb_counters[0]++; |
687 | if (i < max) |
688 | i = mb_find_next_zero_bit(bitmap, max, i); |
689 | } |
690 | grp->bb_fragments = fragments; |
691 | |
692 | if (free != grp->bb_free) { |
693 | ext4_grp_locked_error(sb, group, __func__, |
694 | "EXT4-fs: group %u: %u blocks in bitmap, %u in gd", |
695 | group, free, grp->bb_free); |
696 | /* |
697 | * If we intent to continue, we consider group descritor |
698 | * corrupt and update bb_free using bitmap value |
699 | */ |
700 | grp->bb_free = free; |
701 | } |
702 | |
703 | clear_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state)); |
704 | |
705 | period = get_cycles() - period; |
706 | spin_lock(&EXT4_SB(sb)->s_bal_lock); |
707 | EXT4_SB(sb)->s_mb_buddies_generated++; |
708 | EXT4_SB(sb)->s_mb_generation_time += period; |
709 | spin_unlock(&EXT4_SB(sb)->s_bal_lock); |
710 | } |
711 | |
712 | /* The buddy information is attached the buddy cache inode |
713 | * for convenience. The information regarding each group |
714 | * is loaded via ext4_mb_load_buddy. The information involve |
715 | * block bitmap and buddy information. The information are |
716 | * stored in the inode as |
717 | * |
718 | * { page } |
719 | * [ group 0 bitmap][ group 0 buddy] [group 1][ group 1]... |
720 | * |
721 | * |
722 | * one block each for bitmap and buddy information. |
723 | * So for each group we take up 2 blocks. A page can |
724 | * contain blocks_per_page (PAGE_CACHE_SIZE / blocksize) blocks. |
725 | * So it can have information regarding groups_per_page which |
726 | * is blocks_per_page/2 |
727 | */ |
728 | |
729 | static int ext4_mb_init_cache(struct page *page, char *incore) |
730 | { |
731 | ext4_group_t ngroups; |
732 | int blocksize; |
733 | int blocks_per_page; |
734 | int groups_per_page; |
735 | int err = 0; |
736 | int i; |
737 | ext4_group_t first_group; |
738 | int first_block; |
739 | struct super_block *sb; |
740 | struct buffer_head *bhs; |
741 | struct buffer_head **bh; |
742 | struct inode *inode; |
743 | char *data; |
744 | char *bitmap; |
745 | |
746 | mb_debug("init page %lu\n", page->index); |
747 | |
748 | inode = page->mapping->host; |
749 | sb = inode->i_sb; |
750 | ngroups = ext4_get_groups_count(sb); |
751 | blocksize = 1 << inode->i_blkbits; |
752 | blocks_per_page = PAGE_CACHE_SIZE / blocksize; |
753 | |
754 | groups_per_page = blocks_per_page >> 1; |
755 | if (groups_per_page == 0) |
756 | groups_per_page = 1; |
757 | |
758 | /* allocate buffer_heads to read bitmaps */ |
759 | if (groups_per_page > 1) { |
760 | err = -ENOMEM; |
761 | i = sizeof(struct buffer_head *) * groups_per_page; |
762 | bh = kzalloc(i, GFP_NOFS); |
763 | if (bh == NULL) |
764 | goto out; |
765 | } else |
766 | bh = &bhs; |
767 | |
768 | first_group = page->index * blocks_per_page / 2; |
769 | |
770 | /* read all groups the page covers into the cache */ |
771 | for (i = 0; i < groups_per_page; i++) { |
772 | struct ext4_group_desc *desc; |
773 | |
774 | if (first_group + i >= ngroups) |
775 | break; |
776 | |
777 | err = -EIO; |
778 | desc = ext4_get_group_desc(sb, first_group + i, NULL); |
779 | if (desc == NULL) |
780 | goto out; |
781 | |
782 | err = -ENOMEM; |
783 | bh[i] = sb_getblk(sb, ext4_block_bitmap(sb, desc)); |
784 | if (bh[i] == NULL) |
785 | goto out; |
786 | |
787 | if (bitmap_uptodate(bh[i])) |
788 | continue; |
789 | |
790 | lock_buffer(bh[i]); |
791 | if (bitmap_uptodate(bh[i])) { |
792 | unlock_buffer(bh[i]); |
793 | continue; |
794 | } |
795 | ext4_lock_group(sb, first_group + i); |
796 | if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) { |
797 | ext4_init_block_bitmap(sb, bh[i], |
798 | first_group + i, desc); |
799 | set_bitmap_uptodate(bh[i]); |
800 | set_buffer_uptodate(bh[i]); |
801 | ext4_unlock_group(sb, first_group + i); |
802 | unlock_buffer(bh[i]); |
803 | continue; |
804 | } |
805 | ext4_unlock_group(sb, first_group + i); |
806 | if (buffer_uptodate(bh[i])) { |
807 | /* |
808 | * if not uninit if bh is uptodate, |
809 | * bitmap is also uptodate |
810 | */ |
811 | set_bitmap_uptodate(bh[i]); |
812 | unlock_buffer(bh[i]); |
813 | continue; |
814 | } |
815 | get_bh(bh[i]); |
816 | /* |
817 | * submit the buffer_head for read. We can |
818 | * safely mark the bitmap as uptodate now. |
819 | * We do it here so the bitmap uptodate bit |
820 | * get set with buffer lock held. |
821 | */ |
822 | set_bitmap_uptodate(bh[i]); |
823 | bh[i]->b_end_io = end_buffer_read_sync; |
824 | submit_bh(READ, bh[i]); |
825 | mb_debug("read bitmap for group %u\n", first_group + i); |
826 | } |
827 | |
828 | /* wait for I/O completion */ |
829 | for (i = 0; i < groups_per_page && bh[i]; i++) |
830 | wait_on_buffer(bh[i]); |
831 | |
832 | err = -EIO; |
833 | for (i = 0; i < groups_per_page && bh[i]; i++) |
834 | if (!buffer_uptodate(bh[i])) |
835 | goto out; |
836 | |
837 | err = 0; |
838 | first_block = page->index * blocks_per_page; |
839 | /* init the page */ |
840 | memset(page_address(page), 0xff, PAGE_CACHE_SIZE); |
841 | for (i = 0; i < blocks_per_page; i++) { |
842 | int group; |
843 | struct ext4_group_info *grinfo; |
844 | |
845 | group = (first_block + i) >> 1; |
846 | if (group >= ngroups) |
847 | break; |
848 | |
849 | /* |
850 | * data carry information regarding this |
851 | * particular group in the format specified |
852 | * above |
853 | * |
854 | */ |
855 | data = page_address(page) + (i * blocksize); |
856 | bitmap = bh[group - first_group]->b_data; |
857 | |
858 | /* |
859 | * We place the buddy block and bitmap block |
860 | * close together |
861 | */ |
862 | if ((first_block + i) & 1) { |
863 | /* this is block of buddy */ |
864 | BUG_ON(incore == NULL); |
865 | mb_debug("put buddy for group %u in page %lu/%x\n", |
866 | group, page->index, i * blocksize); |
867 | grinfo = ext4_get_group_info(sb, group); |
868 | grinfo->bb_fragments = 0; |
869 | memset(grinfo->bb_counters, 0, |
870 | sizeof(unsigned short)*(sb->s_blocksize_bits+2)); |
871 | /* |
872 | * incore got set to the group block bitmap below |
873 | */ |
874 | ext4_lock_group(sb, group); |
875 | ext4_mb_generate_buddy(sb, data, incore, group); |
876 | ext4_unlock_group(sb, group); |
877 | incore = NULL; |
878 | } else { |
879 | /* this is block of bitmap */ |
880 | BUG_ON(incore != NULL); |
881 | mb_debug("put bitmap for group %u in page %lu/%x\n", |
882 | group, page->index, i * blocksize); |
883 | |
884 | /* see comments in ext4_mb_put_pa() */ |
885 | ext4_lock_group(sb, group); |
886 | memcpy(data, bitmap, blocksize); |
887 | |
888 | /* mark all preallocated blks used in in-core bitmap */ |
889 | ext4_mb_generate_from_pa(sb, data, group); |
890 | ext4_mb_generate_from_freelist(sb, data, group); |
891 | ext4_unlock_group(sb, group); |
892 | |
893 | /* set incore so that the buddy information can be |
894 | * generated using this |
895 | */ |
896 | incore = data; |
897 | } |
898 | } |
899 | SetPageUptodate(page); |
900 | |
901 | out: |
902 | if (bh) { |
903 | for (i = 0; i < groups_per_page && bh[i]; i++) |
904 | brelse(bh[i]); |
905 | if (bh != &bhs) |
906 | kfree(bh); |
907 | } |
908 | return err; |
909 | } |
910 | |
911 | static noinline_for_stack int |
912 | ext4_mb_load_buddy(struct super_block *sb, ext4_group_t group, |
913 | struct ext4_buddy *e4b) |
914 | { |
915 | int blocks_per_page; |
916 | int block; |
917 | int pnum; |
918 | int poff; |
919 | struct page *page; |
920 | int ret; |
921 | struct ext4_group_info *grp; |
922 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
923 | struct inode *inode = sbi->s_buddy_cache; |
924 | |
925 | mb_debug("load group %u\n", group); |
926 | |
927 | blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; |
928 | grp = ext4_get_group_info(sb, group); |
929 | |
930 | e4b->bd_blkbits = sb->s_blocksize_bits; |
931 | e4b->bd_info = ext4_get_group_info(sb, group); |
932 | e4b->bd_sb = sb; |
933 | e4b->bd_group = group; |
934 | e4b->bd_buddy_page = NULL; |
935 | e4b->bd_bitmap_page = NULL; |
936 | e4b->alloc_semp = &grp->alloc_sem; |
937 | |
938 | /* Take the read lock on the group alloc |
939 | * sem. This would make sure a parallel |
940 | * ext4_mb_init_group happening on other |
941 | * groups mapped by the page is blocked |
942 | * till we are done with allocation |
943 | */ |
944 | down_read(e4b->alloc_semp); |
945 | |
946 | /* |
947 | * the buddy cache inode stores the block bitmap |
948 | * and buddy information in consecutive blocks. |
949 | * So for each group we need two blocks. |
950 | */ |
951 | block = group * 2; |
952 | pnum = block / blocks_per_page; |
953 | poff = block % blocks_per_page; |
954 | |
955 | /* we could use find_or_create_page(), but it locks page |
956 | * what we'd like to avoid in fast path ... */ |
957 | page = find_get_page(inode->i_mapping, pnum); |
958 | if (page == NULL || !PageUptodate(page)) { |
959 | if (page) |
960 | /* |
961 | * drop the page reference and try |
962 | * to get the page with lock. If we |
963 | * are not uptodate that implies |
964 | * somebody just created the page but |
965 | * is yet to initialize the same. So |
966 | * wait for it to initialize. |
967 | */ |
968 | page_cache_release(page); |
969 | page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); |
970 | if (page) { |
971 | BUG_ON(page->mapping != inode->i_mapping); |
972 | if (!PageUptodate(page)) { |
973 | ret = ext4_mb_init_cache(page, NULL); |
974 | if (ret) { |
975 | unlock_page(page); |
976 | goto err; |
977 | } |
978 | mb_cmp_bitmaps(e4b, page_address(page) + |
979 | (poff * sb->s_blocksize)); |
980 | } |
981 | unlock_page(page); |
982 | } |
983 | } |
984 | if (page == NULL || !PageUptodate(page)) { |
985 | ret = -EIO; |
986 | goto err; |
987 | } |
988 | e4b->bd_bitmap_page = page; |
989 | e4b->bd_bitmap = page_address(page) + (poff * sb->s_blocksize); |
990 | mark_page_accessed(page); |
991 | |
992 | block++; |
993 | pnum = block / blocks_per_page; |
994 | poff = block % blocks_per_page; |
995 | |
996 | page = find_get_page(inode->i_mapping, pnum); |
997 | if (page == NULL || !PageUptodate(page)) { |
998 | if (page) |
999 | page_cache_release(page); |
1000 | page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); |
1001 | if (page) { |
1002 | BUG_ON(page->mapping != inode->i_mapping); |
1003 | if (!PageUptodate(page)) { |
1004 | ret = ext4_mb_init_cache(page, e4b->bd_bitmap); |
1005 | if (ret) { |
1006 | unlock_page(page); |
1007 | goto err; |
1008 | } |
1009 | } |
1010 | unlock_page(page); |
1011 | } |
1012 | } |
1013 | if (page == NULL || !PageUptodate(page)) { |
1014 | ret = -EIO; |
1015 | goto err; |
1016 | } |
1017 | e4b->bd_buddy_page = page; |
1018 | e4b->bd_buddy = page_address(page) + (poff * sb->s_blocksize); |
1019 | mark_page_accessed(page); |
1020 | |
1021 | BUG_ON(e4b->bd_bitmap_page == NULL); |
1022 | BUG_ON(e4b->bd_buddy_page == NULL); |
1023 | |
1024 | return 0; |
1025 | |
1026 | err: |
1027 | if (e4b->bd_bitmap_page) |
1028 | page_cache_release(e4b->bd_bitmap_page); |
1029 | if (e4b->bd_buddy_page) |
1030 | page_cache_release(e4b->bd_buddy_page); |
1031 | e4b->bd_buddy = NULL; |
1032 | e4b->bd_bitmap = NULL; |
1033 | |
1034 | /* Done with the buddy cache */ |
1035 | up_read(e4b->alloc_semp); |
1036 | return ret; |
1037 | } |
1038 | |
1039 | static void ext4_mb_release_desc(struct ext4_buddy *e4b) |
1040 | { |
1041 | if (e4b->bd_bitmap_page) |
1042 | page_cache_release(e4b->bd_bitmap_page); |
1043 | if (e4b->bd_buddy_page) |
1044 | page_cache_release(e4b->bd_buddy_page); |
1045 | /* Done with the buddy cache */ |
1046 | if (e4b->alloc_semp) |
1047 | up_read(e4b->alloc_semp); |
1048 | } |
1049 | |
1050 | |
1051 | static int mb_find_order_for_block(struct ext4_buddy *e4b, int block) |
1052 | { |
1053 | int order = 1; |
1054 | void *bb; |
1055 | |
1056 | BUG_ON(EXT4_MB_BITMAP(e4b) == EXT4_MB_BUDDY(e4b)); |
1057 | BUG_ON(block >= (1 << (e4b->bd_blkbits + 3))); |
1058 | |
1059 | bb = EXT4_MB_BUDDY(e4b); |
1060 | while (order <= e4b->bd_blkbits + 1) { |
1061 | block = block >> 1; |
1062 | if (!mb_test_bit(block, bb)) { |
1063 | /* this block is part of buddy of order 'order' */ |
1064 | return order; |
1065 | } |
1066 | bb += 1 << (e4b->bd_blkbits - order); |
1067 | order++; |
1068 | } |
1069 | return 0; |
1070 | } |
1071 | |
1072 | static void mb_clear_bits(void *bm, int cur, int len) |
1073 | { |
1074 | __u32 *addr; |
1075 | |
1076 | len = cur + len; |
1077 | while (cur < len) { |
1078 | if ((cur & 31) == 0 && (len - cur) >= 32) { |
1079 | /* fast path: clear whole word at once */ |
1080 | addr = bm + (cur >> 3); |
1081 | *addr = 0; |
1082 | cur += 32; |
1083 | continue; |
1084 | } |
1085 | mb_clear_bit(cur, bm); |
1086 | cur++; |
1087 | } |
1088 | } |
1089 | |
1090 | static void mb_set_bits(void *bm, int cur, int len) |
1091 | { |
1092 | __u32 *addr; |
1093 | |
1094 | len = cur + len; |
1095 | while (cur < len) { |
1096 | if ((cur & 31) == 0 && (len - cur) >= 32) { |
1097 | /* fast path: set whole word at once */ |
1098 | addr = bm + (cur >> 3); |
1099 | *addr = 0xffffffff; |
1100 | cur += 32; |
1101 | continue; |
1102 | } |
1103 | mb_set_bit(cur, bm); |
1104 | cur++; |
1105 | } |
1106 | } |
1107 | |
1108 | static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b, |
1109 | int first, int count) |
1110 | { |
1111 | int block = 0; |
1112 | int max = 0; |
1113 | int order; |
1114 | void *buddy; |
1115 | void *buddy2; |
1116 | struct super_block *sb = e4b->bd_sb; |
1117 | |
1118 | BUG_ON(first + count > (sb->s_blocksize << 3)); |
1119 | assert_spin_locked(ext4_group_lock_ptr(sb, e4b->bd_group)); |
1120 | mb_check_buddy(e4b); |
1121 | mb_free_blocks_double(inode, e4b, first, count); |
1122 | |
1123 | e4b->bd_info->bb_free += count; |
1124 | if (first < e4b->bd_info->bb_first_free) |
1125 | e4b->bd_info->bb_first_free = first; |
1126 | |
1127 | /* let's maintain fragments counter */ |
1128 | if (first != 0) |
1129 | block = !mb_test_bit(first - 1, EXT4_MB_BITMAP(e4b)); |
1130 | if (first + count < EXT4_SB(sb)->s_mb_maxs[0]) |
1131 | max = !mb_test_bit(first + count, EXT4_MB_BITMAP(e4b)); |
1132 | if (block && max) |
1133 | e4b->bd_info->bb_fragments--; |
1134 | else if (!block && !max) |
1135 | e4b->bd_info->bb_fragments++; |
1136 | |
1137 | /* let's maintain buddy itself */ |
1138 | while (count-- > 0) { |
1139 | block = first++; |
1140 | order = 0; |
1141 | |
1142 | if (!mb_test_bit(block, EXT4_MB_BITMAP(e4b))) { |
1143 | ext4_fsblk_t blocknr; |
1144 | blocknr = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb); |
1145 | blocknr += block; |
1146 | blocknr += |
1147 | le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); |
1148 | ext4_grp_locked_error(sb, e4b->bd_group, |
1149 | __func__, "double-free of inode" |
1150 | " %lu's block %llu(bit %u in group %u)", |
1151 | inode ? inode->i_ino : 0, blocknr, block, |
1152 | e4b->bd_group); |
1153 | } |
1154 | mb_clear_bit(block, EXT4_MB_BITMAP(e4b)); |
1155 | e4b->bd_info->bb_counters[order]++; |
1156 | |
1157 | /* start of the buddy */ |
1158 | buddy = mb_find_buddy(e4b, order, &max); |
1159 | |
1160 | do { |
1161 | block &= ~1UL; |
1162 | if (mb_test_bit(block, buddy) || |
1163 | mb_test_bit(block + 1, buddy)) |
1164 | break; |
1165 | |
1166 | /* both the buddies are free, try to coalesce them */ |
1167 | buddy2 = mb_find_buddy(e4b, order + 1, &max); |
1168 | |
1169 | if (!buddy2) |
1170 | break; |
1171 | |
1172 | if (order > 0) { |
1173 | /* for special purposes, we don't set |
1174 | * free bits in bitmap */ |
1175 | mb_set_bit(block, buddy); |
1176 | mb_set_bit(block + 1, buddy); |
1177 | } |
1178 | e4b->bd_info->bb_counters[order]--; |
1179 | e4b->bd_info->bb_counters[order]--; |
1180 | |
1181 | block = block >> 1; |
1182 | order++; |
1183 | e4b->bd_info->bb_counters[order]++; |
1184 | |
1185 | mb_clear_bit(block, buddy2); |
1186 | buddy = buddy2; |
1187 | } while (1); |
1188 | } |
1189 | mb_check_buddy(e4b); |
1190 | } |
1191 | |
1192 | static int mb_find_extent(struct ext4_buddy *e4b, int order, int block, |
1193 | int needed, struct ext4_free_extent *ex) |
1194 | { |
1195 | int next = block; |
1196 | int max; |
1197 | int ord; |
1198 | void *buddy; |
1199 | |
1200 | assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); |
1201 | BUG_ON(ex == NULL); |
1202 | |
1203 | buddy = mb_find_buddy(e4b, order, &max); |
1204 | BUG_ON(buddy == NULL); |
1205 | BUG_ON(block >= max); |
1206 | if (mb_test_bit(block, buddy)) { |
1207 | ex->fe_len = 0; |
1208 | ex->fe_start = 0; |
1209 | ex->fe_group = 0; |
1210 | return 0; |
1211 | } |
1212 | |
1213 | /* FIXME dorp order completely ? */ |
1214 | if (likely(order == 0)) { |
1215 | /* find actual order */ |
1216 | order = mb_find_order_for_block(e4b, block); |
1217 | block = block >> order; |
1218 | } |
1219 | |
1220 | ex->fe_len = 1 << order; |
1221 | ex->fe_start = block << order; |
1222 | ex->fe_group = e4b->bd_group; |
1223 | |
1224 | /* calc difference from given start */ |
1225 | next = next - ex->fe_start; |
1226 | ex->fe_len -= next; |
1227 | ex->fe_start += next; |
1228 | |
1229 | while (needed > ex->fe_len && |
1230 | (buddy = mb_find_buddy(e4b, order, &max))) { |
1231 | |
1232 | if (block + 1 >= max) |
1233 | break; |
1234 | |
1235 | next = (block + 1) * (1 << order); |
1236 | if (mb_test_bit(next, EXT4_MB_BITMAP(e4b))) |
1237 | break; |
1238 | |
1239 | ord = mb_find_order_for_block(e4b, next); |
1240 | |
1241 | order = ord; |
1242 | block = next >> order; |
1243 | ex->fe_len += 1 << order; |
1244 | } |
1245 | |
1246 | BUG_ON(ex->fe_start + ex->fe_len > (1 << (e4b->bd_blkbits + 3))); |
1247 | return ex->fe_len; |
1248 | } |
1249 | |
1250 | static int mb_mark_used(struct ext4_buddy *e4b, struct ext4_free_extent *ex) |
1251 | { |
1252 | int ord; |
1253 | int mlen = 0; |
1254 | int max = 0; |
1255 | int cur; |
1256 | int start = ex->fe_start; |
1257 | int len = ex->fe_len; |
1258 | unsigned ret = 0; |
1259 | int len0 = len; |
1260 | void *buddy; |
1261 | |
1262 | BUG_ON(start + len > (e4b->bd_sb->s_blocksize << 3)); |
1263 | BUG_ON(e4b->bd_group != ex->fe_group); |
1264 | assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); |
1265 | mb_check_buddy(e4b); |
1266 | mb_mark_used_double(e4b, start, len); |
1267 | |
1268 | e4b->bd_info->bb_free -= len; |
1269 | if (e4b->bd_info->bb_first_free == start) |
1270 | e4b->bd_info->bb_first_free += len; |
1271 | |
1272 | /* let's maintain fragments counter */ |
1273 | if (start != 0) |
1274 | mlen = !mb_test_bit(start - 1, EXT4_MB_BITMAP(e4b)); |
1275 | if (start + len < EXT4_SB(e4b->bd_sb)->s_mb_maxs[0]) |
1276 | max = !mb_test_bit(start + len, EXT4_MB_BITMAP(e4b)); |
1277 | if (mlen && max) |
1278 | e4b->bd_info->bb_fragments++; |
1279 | else if (!mlen && !max) |
1280 | e4b->bd_info->bb_fragments--; |
1281 | |
1282 | /* let's maintain buddy itself */ |
1283 | while (len) { |
1284 | ord = mb_find_order_for_block(e4b, start); |
1285 | |
1286 | if (((start >> ord) << ord) == start && len >= (1 << ord)) { |
1287 | /* the whole chunk may be allocated at once! */ |
1288 | mlen = 1 << ord; |
1289 | buddy = mb_find_buddy(e4b, ord, &max); |
1290 | BUG_ON((start >> ord) >= max); |
1291 | mb_set_bit(start >> ord, buddy); |
1292 | e4b->bd_info->bb_counters[ord]--; |
1293 | start += mlen; |
1294 | len -= mlen; |
1295 | BUG_ON(len < 0); |
1296 | continue; |
1297 | } |
1298 | |
1299 | /* store for history */ |
1300 | if (ret == 0) |
1301 | ret = len | (ord << 16); |
1302 | |
1303 | /* we have to split large buddy */ |
1304 | BUG_ON(ord <= 0); |
1305 | buddy = mb_find_buddy(e4b, ord, &max); |
1306 | mb_set_bit(start >> ord, buddy); |
1307 | e4b->bd_info->bb_counters[ord]--; |
1308 | |
1309 | ord--; |
1310 | cur = (start >> ord) & ~1U; |
1311 | buddy = mb_find_buddy(e4b, ord, &max); |
1312 | mb_clear_bit(cur, buddy); |
1313 | mb_clear_bit(cur + 1, buddy); |
1314 | e4b->bd_info->bb_counters[ord]++; |
1315 | e4b->bd_info->bb_counters[ord]++; |
1316 | } |
1317 | |
1318 | mb_set_bits(EXT4_MB_BITMAP(e4b), ex->fe_start, len0); |
1319 | mb_check_buddy(e4b); |
1320 | |
1321 | return ret; |
1322 | } |
1323 | |
1324 | /* |
1325 | * Must be called under group lock! |
1326 | */ |
1327 | static void ext4_mb_use_best_found(struct ext4_allocation_context *ac, |
1328 | struct ext4_buddy *e4b) |
1329 | { |
1330 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
1331 | int ret; |
1332 | |
1333 | BUG_ON(ac->ac_b_ex.fe_group != e4b->bd_group); |
1334 | BUG_ON(ac->ac_status == AC_STATUS_FOUND); |
1335 | |
1336 | ac->ac_b_ex.fe_len = min(ac->ac_b_ex.fe_len, ac->ac_g_ex.fe_len); |
1337 | ac->ac_b_ex.fe_logical = ac->ac_g_ex.fe_logical; |
1338 | ret = mb_mark_used(e4b, &ac->ac_b_ex); |
1339 | |
1340 | /* preallocation can change ac_b_ex, thus we store actually |
1341 | * allocated blocks for history */ |
1342 | ac->ac_f_ex = ac->ac_b_ex; |
1343 | |
1344 | ac->ac_status = AC_STATUS_FOUND; |
1345 | ac->ac_tail = ret & 0xffff; |
1346 | ac->ac_buddy = ret >> 16; |
1347 | |
1348 | /* |
1349 | * take the page reference. We want the page to be pinned |
1350 | * so that we don't get a ext4_mb_init_cache_call for this |
1351 | * group until we update the bitmap. That would mean we |
1352 | * double allocate blocks. The reference is dropped |
1353 | * in ext4_mb_release_context |
1354 | */ |
1355 | ac->ac_bitmap_page = e4b->bd_bitmap_page; |
1356 | get_page(ac->ac_bitmap_page); |
1357 | ac->ac_buddy_page = e4b->bd_buddy_page; |
1358 | get_page(ac->ac_buddy_page); |
1359 | /* on allocation we use ac to track the held semaphore */ |
1360 | ac->alloc_semp = e4b->alloc_semp; |
1361 | e4b->alloc_semp = NULL; |
1362 | /* store last allocated for subsequent stream allocation */ |
1363 | if ((ac->ac_flags & EXT4_MB_HINT_DATA)) { |
1364 | spin_lock(&sbi->s_md_lock); |
1365 | sbi->s_mb_last_group = ac->ac_f_ex.fe_group; |
1366 | sbi->s_mb_last_start = ac->ac_f_ex.fe_start; |
1367 | spin_unlock(&sbi->s_md_lock); |
1368 | } |
1369 | } |
1370 | |
1371 | /* |
1372 | * regular allocator, for general purposes allocation |
1373 | */ |
1374 | |
1375 | static void ext4_mb_check_limits(struct ext4_allocation_context *ac, |
1376 | struct ext4_buddy *e4b, |
1377 | int finish_group) |
1378 | { |
1379 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
1380 | struct ext4_free_extent *bex = &ac->ac_b_ex; |
1381 | struct ext4_free_extent *gex = &ac->ac_g_ex; |
1382 | struct ext4_free_extent ex; |
1383 | int max; |
1384 | |
1385 | if (ac->ac_status == AC_STATUS_FOUND) |
1386 | return; |
1387 | /* |
1388 | * We don't want to scan for a whole year |
1389 | */ |
1390 | if (ac->ac_found > sbi->s_mb_max_to_scan && |
1391 | !(ac->ac_flags & EXT4_MB_HINT_FIRST)) { |
1392 | ac->ac_status = AC_STATUS_BREAK; |
1393 | return; |
1394 | } |
1395 | |
1396 | /* |
1397 | * Haven't found good chunk so far, let's continue |
1398 | */ |
1399 | if (bex->fe_len < gex->fe_len) |
1400 | return; |
1401 | |
1402 | if ((finish_group || ac->ac_found > sbi->s_mb_min_to_scan) |
1403 | && bex->fe_group == e4b->bd_group) { |
1404 | /* recheck chunk's availability - we don't know |
1405 | * when it was found (within this lock-unlock |
1406 | * period or not) */ |
1407 | max = mb_find_extent(e4b, 0, bex->fe_start, gex->fe_len, &ex); |
1408 | if (max >= gex->fe_len) { |
1409 | ext4_mb_use_best_found(ac, e4b); |
1410 | return; |
1411 | } |
1412 | } |
1413 | } |
1414 | |
1415 | /* |
1416 | * The routine checks whether found extent is good enough. If it is, |
1417 | * then the extent gets marked used and flag is set to the context |
1418 | * to stop scanning. Otherwise, the extent is compared with the |
1419 | * previous found extent and if new one is better, then it's stored |
1420 | * in the context. Later, the best found extent will be used, if |
1421 | * mballoc can't find good enough extent. |
1422 | * |
1423 | * FIXME: real allocation policy is to be designed yet! |
1424 | */ |
1425 | static void ext4_mb_measure_extent(struct ext4_allocation_context *ac, |
1426 | struct ext4_free_extent *ex, |
1427 | struct ext4_buddy *e4b) |
1428 | { |
1429 | struct ext4_free_extent *bex = &ac->ac_b_ex; |
1430 | struct ext4_free_extent *gex = &ac->ac_g_ex; |
1431 | |
1432 | BUG_ON(ex->fe_len <= 0); |
1433 | BUG_ON(ex->fe_len > EXT4_BLOCKS_PER_GROUP(ac->ac_sb)); |
1434 | BUG_ON(ex->fe_start >= EXT4_BLOCKS_PER_GROUP(ac->ac_sb)); |
1435 | BUG_ON(ac->ac_status != AC_STATUS_CONTINUE); |
1436 | |
1437 | ac->ac_found++; |
1438 | |
1439 | /* |
1440 | * The special case - take what you catch first |
1441 | */ |
1442 | if (unlikely(ac->ac_flags & EXT4_MB_HINT_FIRST)) { |
1443 | *bex = *ex; |
1444 | ext4_mb_use_best_found(ac, e4b); |
1445 | return; |
1446 | } |
1447 | |
1448 | /* |
1449 | * Let's check whether the chuck is good enough |
1450 | */ |
1451 | if (ex->fe_len == gex->fe_len) { |
1452 | *bex = *ex; |
1453 | ext4_mb_use_best_found(ac, e4b); |
1454 | return; |
1455 | } |
1456 | |
1457 | /* |
1458 | * If this is first found extent, just store it in the context |
1459 | */ |
1460 | if (bex->fe_len == 0) { |
1461 | *bex = *ex; |
1462 | return; |
1463 | } |
1464 | |
1465 | /* |
1466 | * If new found extent is better, store it in the context |
1467 | */ |
1468 | if (bex->fe_len < gex->fe_len) { |
1469 | /* if the request isn't satisfied, any found extent |
1470 | * larger than previous best one is better */ |
1471 | if (ex->fe_len > bex->fe_len) |
1472 | *bex = *ex; |
1473 | } else if (ex->fe_len > gex->fe_len) { |
1474 | /* if the request is satisfied, then we try to find |
1475 | * an extent that still satisfy the request, but is |
1476 | * smaller than previous one */ |
1477 | if (ex->fe_len < bex->fe_len) |
1478 | *bex = *ex; |
1479 | } |
1480 | |
1481 | ext4_mb_check_limits(ac, e4b, 0); |
1482 | } |
1483 | |
1484 | static noinline_for_stack |
1485 | int ext4_mb_try_best_found(struct ext4_allocation_context *ac, |
1486 | struct ext4_buddy *e4b) |
1487 | { |
1488 | struct ext4_free_extent ex = ac->ac_b_ex; |
1489 | ext4_group_t group = ex.fe_group; |
1490 | int max; |
1491 | int err; |
1492 | |
1493 | BUG_ON(ex.fe_len <= 0); |
1494 | err = ext4_mb_load_buddy(ac->ac_sb, group, e4b); |
1495 | if (err) |
1496 | return err; |
1497 | |
1498 | ext4_lock_group(ac->ac_sb, group); |
1499 | max = mb_find_extent(e4b, 0, ex.fe_start, ex.fe_len, &ex); |
1500 | |
1501 | if (max > 0) { |
1502 | ac->ac_b_ex = ex; |
1503 | ext4_mb_use_best_found(ac, e4b); |
1504 | } |
1505 | |
1506 | ext4_unlock_group(ac->ac_sb, group); |
1507 | ext4_mb_release_desc(e4b); |
1508 | |
1509 | return 0; |
1510 | } |
1511 | |
1512 | static noinline_for_stack |
1513 | int ext4_mb_find_by_goal(struct ext4_allocation_context *ac, |
1514 | struct ext4_buddy *e4b) |
1515 | { |
1516 | ext4_group_t group = ac->ac_g_ex.fe_group; |
1517 | int max; |
1518 | int err; |
1519 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
1520 | struct ext4_super_block *es = sbi->s_es; |
1521 | struct ext4_free_extent ex; |
1522 | |
1523 | if (!(ac->ac_flags & EXT4_MB_HINT_TRY_GOAL)) |
1524 | return 0; |
1525 | |
1526 | err = ext4_mb_load_buddy(ac->ac_sb, group, e4b); |
1527 | if (err) |
1528 | return err; |
1529 | |
1530 | ext4_lock_group(ac->ac_sb, group); |
1531 | max = mb_find_extent(e4b, 0, ac->ac_g_ex.fe_start, |
1532 | ac->ac_g_ex.fe_len, &ex); |
1533 | |
1534 | if (max >= ac->ac_g_ex.fe_len && ac->ac_g_ex.fe_len == sbi->s_stripe) { |
1535 | ext4_fsblk_t start; |
1536 | |
1537 | start = (e4b->bd_group * EXT4_BLOCKS_PER_GROUP(ac->ac_sb)) + |
1538 | ex.fe_start + le32_to_cpu(es->s_first_data_block); |
1539 | /* use do_div to get remainder (would be 64-bit modulo) */ |
1540 | if (do_div(start, sbi->s_stripe) == 0) { |
1541 | ac->ac_found++; |
1542 | ac->ac_b_ex = ex; |
1543 | ext4_mb_use_best_found(ac, e4b); |
1544 | } |
1545 | } else if (max >= ac->ac_g_ex.fe_len) { |
1546 | BUG_ON(ex.fe_len <= 0); |
1547 | BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group); |
1548 | BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start); |
1549 | ac->ac_found++; |
1550 | ac->ac_b_ex = ex; |
1551 | ext4_mb_use_best_found(ac, e4b); |
1552 | } else if (max > 0 && (ac->ac_flags & EXT4_MB_HINT_MERGE)) { |
1553 | /* Sometimes, caller may want to merge even small |
1554 | * number of blocks to an existing extent */ |
1555 | BUG_ON(ex.fe_len <= 0); |
1556 | BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group); |
1557 | BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start); |
1558 | ac->ac_found++; |
1559 | ac->ac_b_ex = ex; |
1560 | ext4_mb_use_best_found(ac, e4b); |
1561 | } |
1562 | ext4_unlock_group(ac->ac_sb, group); |
1563 | ext4_mb_release_desc(e4b); |
1564 | |
1565 | return 0; |
1566 | } |
1567 | |
1568 | /* |
1569 | * The routine scans buddy structures (not bitmap!) from given order |
1570 | * to max order and tries to find big enough chunk to satisfy the req |
1571 | */ |
1572 | static noinline_for_stack |
1573 | void ext4_mb_simple_scan_group(struct ext4_allocation_context *ac, |
1574 | struct ext4_buddy *e4b) |
1575 | { |
1576 | struct super_block *sb = ac->ac_sb; |
1577 | struct ext4_group_info *grp = e4b->bd_info; |
1578 | void *buddy; |
1579 | int i; |
1580 | int k; |
1581 | int max; |
1582 | |
1583 | BUG_ON(ac->ac_2order <= 0); |
1584 | for (i = ac->ac_2order; i <= sb->s_blocksize_bits + 1; i++) { |
1585 | if (grp->bb_counters[i] == 0) |
1586 | continue; |
1587 | |
1588 | buddy = mb_find_buddy(e4b, i, &max); |
1589 | BUG_ON(buddy == NULL); |
1590 | |
1591 | k = mb_find_next_zero_bit(buddy, max, 0); |
1592 | BUG_ON(k >= max); |
1593 | |
1594 | ac->ac_found++; |
1595 | |
1596 | ac->ac_b_ex.fe_len = 1 << i; |
1597 | ac->ac_b_ex.fe_start = k << i; |
1598 | ac->ac_b_ex.fe_group = e4b->bd_group; |
1599 | |
1600 | ext4_mb_use_best_found(ac, e4b); |
1601 | |
1602 | BUG_ON(ac->ac_b_ex.fe_len != ac->ac_g_ex.fe_len); |
1603 | |
1604 | if (EXT4_SB(sb)->s_mb_stats) |
1605 | atomic_inc(&EXT4_SB(sb)->s_bal_2orders); |
1606 | |
1607 | break; |
1608 | } |
1609 | } |
1610 | |
1611 | /* |
1612 | * The routine scans the group and measures all found extents. |
1613 | * In order to optimize scanning, caller must pass number of |
1614 | * free blocks in the group, so the routine can know upper limit. |
1615 | */ |
1616 | static noinline_for_stack |
1617 | void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, |
1618 | struct ext4_buddy *e4b) |
1619 | { |
1620 | struct super_block *sb = ac->ac_sb; |
1621 | void *bitmap = EXT4_MB_BITMAP(e4b); |
1622 | struct ext4_free_extent ex; |
1623 | int i; |
1624 | int free; |
1625 | |
1626 | free = e4b->bd_info->bb_free; |
1627 | BUG_ON(free <= 0); |
1628 | |
1629 | i = e4b->bd_info->bb_first_free; |
1630 | |
1631 | while (free && ac->ac_status == AC_STATUS_CONTINUE) { |
1632 | i = mb_find_next_zero_bit(bitmap, |
1633 | EXT4_BLOCKS_PER_GROUP(sb), i); |
1634 | if (i >= EXT4_BLOCKS_PER_GROUP(sb)) { |
1635 | /* |
1636 | * IF we have corrupt bitmap, we won't find any |
1637 | * free blocks even though group info says we |
1638 | * we have free blocks |
1639 | */ |
1640 | ext4_grp_locked_error(sb, e4b->bd_group, |
1641 | __func__, "%d free blocks as per " |
1642 | "group info. But bitmap says 0", |
1643 | free); |
1644 | break; |
1645 | } |
1646 | |
1647 | mb_find_extent(e4b, 0, i, ac->ac_g_ex.fe_len, &ex); |
1648 | BUG_ON(ex.fe_len <= 0); |
1649 | if (free < ex.fe_len) { |
1650 | ext4_grp_locked_error(sb, e4b->bd_group, |
1651 | __func__, "%d free blocks as per " |
1652 | "group info. But got %d blocks", |
1653 | free, ex.fe_len); |
1654 | /* |
1655 | * The number of free blocks differs. This mostly |
1656 | * indicate that the bitmap is corrupt. So exit |
1657 | * without claiming the space. |
1658 | */ |
1659 | break; |
1660 | } |
1661 | |
1662 | ext4_mb_measure_extent(ac, &ex, e4b); |
1663 | |
1664 | i += ex.fe_len; |
1665 | free -= ex.fe_len; |
1666 | } |
1667 | |
1668 | ext4_mb_check_limits(ac, e4b, 1); |
1669 | } |
1670 | |
1671 | /* |
1672 | * This is a special case for storages like raid5 |
1673 | * we try to find stripe-aligned chunks for stripe-size requests |
1674 | * XXX should do so at least for multiples of stripe size as well |
1675 | */ |
1676 | static noinline_for_stack |
1677 | void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, |
1678 | struct ext4_buddy *e4b) |
1679 | { |
1680 | struct super_block *sb = ac->ac_sb; |
1681 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
1682 | void *bitmap = EXT4_MB_BITMAP(e4b); |
1683 | struct ext4_free_extent ex; |
1684 | ext4_fsblk_t first_group_block; |
1685 | ext4_fsblk_t a; |
1686 | ext4_grpblk_t i; |
1687 | int max; |
1688 | |
1689 | BUG_ON(sbi->s_stripe == 0); |
1690 | |
1691 | /* find first stripe-aligned block in group */ |
1692 | first_group_block = e4b->bd_group * EXT4_BLOCKS_PER_GROUP(sb) |
1693 | + le32_to_cpu(sbi->s_es->s_first_data_block); |
1694 | a = first_group_block + sbi->s_stripe - 1; |
1695 | do_div(a, sbi->s_stripe); |
1696 | i = (a * sbi->s_stripe) - first_group_block; |
1697 | |
1698 | while (i < EXT4_BLOCKS_PER_GROUP(sb)) { |
1699 | if (!mb_test_bit(i, bitmap)) { |
1700 | max = mb_find_extent(e4b, 0, i, sbi->s_stripe, &ex); |
1701 | if (max >= sbi->s_stripe) { |
1702 | ac->ac_found++; |
1703 | ac->ac_b_ex = ex; |
1704 | ext4_mb_use_best_found(ac, e4b); |
1705 | break; |
1706 | } |
1707 | } |
1708 | i += sbi->s_stripe; |
1709 | } |
1710 | } |
1711 | |
1712 | static int ext4_mb_good_group(struct ext4_allocation_context *ac, |
1713 | ext4_group_t group, int cr) |
1714 | { |
1715 | unsigned free, fragments; |
1716 | unsigned i, bits; |
1717 | int flex_size = ext4_flex_bg_size(EXT4_SB(ac->ac_sb)); |
1718 | struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group); |
1719 | |
1720 | BUG_ON(cr < 0 || cr >= 4); |
1721 | BUG_ON(EXT4_MB_GRP_NEED_INIT(grp)); |
1722 | |
1723 | free = grp->bb_free; |
1724 | fragments = grp->bb_fragments; |
1725 | if (free == 0) |
1726 | return 0; |
1727 | if (fragments == 0) |
1728 | return 0; |
1729 | |
1730 | switch (cr) { |
1731 | case 0: |
1732 | BUG_ON(ac->ac_2order == 0); |
1733 | |
1734 | /* Avoid using the first bg of a flexgroup for data files */ |
1735 | if ((ac->ac_flags & EXT4_MB_HINT_DATA) && |
1736 | (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) && |
1737 | ((group % flex_size) == 0)) |
1738 | return 0; |
1739 | |
1740 | bits = ac->ac_sb->s_blocksize_bits + 1; |
1741 | for (i = ac->ac_2order; i <= bits; i++) |
1742 | if (grp->bb_counters[i] > 0) |
1743 | return 1; |
1744 | break; |
1745 | case 1: |
1746 | if ((free / fragments) >= ac->ac_g_ex.fe_len) |
1747 | return 1; |
1748 | break; |
1749 | case 2: |
1750 | if (free >= ac->ac_g_ex.fe_len) |
1751 | return 1; |
1752 | break; |
1753 | case 3: |
1754 | return 1; |
1755 | default: |
1756 | BUG(); |
1757 | } |
1758 | |
1759 | return 0; |
1760 | } |
1761 | |
1762 | /* |
1763 | * lock the group_info alloc_sem of all the groups |
1764 | * belonging to the same buddy cache page. This |
1765 | * make sure other parallel operation on the buddy |
1766 | * cache doesn't happen whild holding the buddy cache |
1767 | * lock |
1768 | */ |
1769 | int ext4_mb_get_buddy_cache_lock(struct super_block *sb, ext4_group_t group) |
1770 | { |
1771 | int i; |
1772 | int block, pnum; |
1773 | int blocks_per_page; |
1774 | int groups_per_page; |
1775 | ext4_group_t ngroups = ext4_get_groups_count(sb); |
1776 | ext4_group_t first_group; |
1777 | struct ext4_group_info *grp; |
1778 | |
1779 | blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; |
1780 | /* |
1781 | * the buddy cache inode stores the block bitmap |
1782 | * and buddy information in consecutive blocks. |
1783 | * So for each group we need two blocks. |
1784 | */ |
1785 | block = group * 2; |
1786 | pnum = block / blocks_per_page; |
1787 | first_group = pnum * blocks_per_page / 2; |
1788 | |
1789 | groups_per_page = blocks_per_page >> 1; |
1790 | if (groups_per_page == 0) |
1791 | groups_per_page = 1; |
1792 | /* read all groups the page covers into the cache */ |
1793 | for (i = 0; i < groups_per_page; i++) { |
1794 | |
1795 | if ((first_group + i) >= ngroups) |
1796 | break; |
1797 | grp = ext4_get_group_info(sb, first_group + i); |
1798 | /* take all groups write allocation |
1799 | * semaphore. This make sure there is |
1800 | * no block allocation going on in any |
1801 | * of that groups |
1802 | */ |
1803 | down_write_nested(&grp->alloc_sem, i); |
1804 | } |
1805 | return i; |
1806 | } |
1807 | |
1808 | void ext4_mb_put_buddy_cache_lock(struct super_block *sb, |
1809 | ext4_group_t group, int locked_group) |
1810 | { |
1811 | int i; |
1812 | int block, pnum; |
1813 | int blocks_per_page; |
1814 | ext4_group_t first_group; |
1815 | struct ext4_group_info *grp; |
1816 | |
1817 | blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; |
1818 | /* |
1819 | * the buddy cache inode stores the block bitmap |
1820 | * and buddy information in consecutive blocks. |
1821 | * So for each group we need two blocks. |
1822 | */ |
1823 | block = group * 2; |
1824 | pnum = block / blocks_per_page; |
1825 | first_group = pnum * blocks_per_page / 2; |
1826 | /* release locks on all the groups */ |
1827 | for (i = 0; i < locked_group; i++) { |
1828 | |
1829 | grp = ext4_get_group_info(sb, first_group + i); |
1830 | /* take all groups write allocation |
1831 | * semaphore. This make sure there is |
1832 | * no block allocation going on in any |
1833 | * of that groups |
1834 | */ |
1835 | up_write(&grp->alloc_sem); |
1836 | } |
1837 | |
1838 | } |
1839 | |
1840 | static noinline_for_stack |
1841 | int ext4_mb_init_group(struct super_block *sb, ext4_group_t group) |
1842 | { |
1843 | |
1844 | int ret; |
1845 | void *bitmap; |
1846 | int blocks_per_page; |
1847 | int block, pnum, poff; |
1848 | int num_grp_locked = 0; |
1849 | struct ext4_group_info *this_grp; |
1850 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
1851 | struct inode *inode = sbi->s_buddy_cache; |
1852 | struct page *page = NULL, *bitmap_page = NULL; |
1853 | |
1854 | mb_debug("init group %lu\n", group); |
1855 | blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; |
1856 | this_grp = ext4_get_group_info(sb, group); |
1857 | /* |
1858 | * This ensures we don't add group |
1859 | * to this buddy cache via resize |
1860 | */ |
1861 | num_grp_locked = ext4_mb_get_buddy_cache_lock(sb, group); |
1862 | if (!EXT4_MB_GRP_NEED_INIT(this_grp)) { |
1863 | /* |
1864 | * somebody initialized the group |
1865 | * return without doing anything |
1866 | */ |
1867 | ret = 0; |
1868 | goto err; |
1869 | } |
1870 | /* |
1871 | * the buddy cache inode stores the block bitmap |
1872 | * and buddy information in consecutive blocks. |
1873 | * So for each group we need two blocks. |
1874 | */ |
1875 | block = group * 2; |
1876 | pnum = block / blocks_per_page; |
1877 | poff = block % blocks_per_page; |
1878 | page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); |
1879 | if (page) { |
1880 | BUG_ON(page->mapping != inode->i_mapping); |
1881 | ret = ext4_mb_init_cache(page, NULL); |
1882 | if (ret) { |
1883 | unlock_page(page); |
1884 | goto err; |
1885 | } |
1886 | unlock_page(page); |
1887 | } |
1888 | if (page == NULL || !PageUptodate(page)) { |
1889 | ret = -EIO; |
1890 | goto err; |
1891 | } |
1892 | mark_page_accessed(page); |
1893 | bitmap_page = page; |
1894 | bitmap = page_address(page) + (poff * sb->s_blocksize); |
1895 | |
1896 | /* init buddy cache */ |
1897 | block++; |
1898 | pnum = block / blocks_per_page; |
1899 | poff = block % blocks_per_page; |
1900 | page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); |
1901 | if (page == bitmap_page) { |
1902 | /* |
1903 | * If both the bitmap and buddy are in |
1904 | * the same page we don't need to force |
1905 | * init the buddy |
1906 | */ |
1907 | unlock_page(page); |
1908 | } else if (page) { |
1909 | BUG_ON(page->mapping != inode->i_mapping); |
1910 | ret = ext4_mb_init_cache(page, bitmap); |
1911 | if (ret) { |
1912 | unlock_page(page); |
1913 | goto err; |
1914 | } |
1915 | unlock_page(page); |
1916 | } |
1917 | if (page == NULL || !PageUptodate(page)) { |
1918 | ret = -EIO; |
1919 | goto err; |
1920 | } |
1921 | mark_page_accessed(page); |
1922 | err: |
1923 | ext4_mb_put_buddy_cache_lock(sb, group, num_grp_locked); |
1924 | if (bitmap_page) |
1925 | page_cache_release(bitmap_page); |
1926 | if (page) |
1927 | page_cache_release(page); |
1928 | return ret; |
1929 | } |
1930 | |
1931 | static noinline_for_stack int |
1932 | ext4_mb_regular_allocator(struct ext4_allocation_context *ac) |
1933 | { |
1934 | ext4_group_t ngroups, group, i; |
1935 | int cr; |
1936 | int err = 0; |
1937 | int bsbits; |
1938 | struct ext4_sb_info *sbi; |
1939 | struct super_block *sb; |
1940 | struct ext4_buddy e4b; |
1941 | loff_t size, isize; |
1942 | |
1943 | sb = ac->ac_sb; |
1944 | sbi = EXT4_SB(sb); |
1945 | ngroups = ext4_get_groups_count(sb); |
1946 | BUG_ON(ac->ac_status == AC_STATUS_FOUND); |
1947 | |
1948 | /* first, try the goal */ |
1949 | err = ext4_mb_find_by_goal(ac, &e4b); |
1950 | if (err || ac->ac_status == AC_STATUS_FOUND) |
1951 | goto out; |
1952 | |
1953 | if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) |
1954 | goto out; |
1955 | |
1956 | /* |
1957 | * ac->ac2_order is set only if the fe_len is a power of 2 |
1958 | * if ac2_order is set we also set criteria to 0 so that we |
1959 | * try exact allocation using buddy. |
1960 | */ |
1961 | i = fls(ac->ac_g_ex.fe_len); |
1962 | ac->ac_2order = 0; |
1963 | /* |
1964 | * We search using buddy data only if the order of the request |
1965 | * is greater than equal to the sbi_s_mb_order2_reqs |
1966 | * You can tune it via /sys/fs/ext4/<partition>/mb_order2_req |
1967 | */ |
1968 | if (i >= sbi->s_mb_order2_reqs) { |
1969 | /* |
1970 | * This should tell if fe_len is exactly power of 2 |
1971 | */ |
1972 | if ((ac->ac_g_ex.fe_len & (~(1 << (i - 1)))) == 0) |
1973 | ac->ac_2order = i - 1; |
1974 | } |
1975 | |
1976 | bsbits = ac->ac_sb->s_blocksize_bits; |
1977 | /* if stream allocation is enabled, use global goal */ |
1978 | size = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len; |
1979 | isize = i_size_read(ac->ac_inode) >> bsbits; |
1980 | if (size < isize) |
1981 | size = isize; |
1982 | |
1983 | if (size < sbi->s_mb_stream_request && |
1984 | (ac->ac_flags & EXT4_MB_HINT_DATA)) { |
1985 | /* TBD: may be hot point */ |
1986 | spin_lock(&sbi->s_md_lock); |
1987 | ac->ac_g_ex.fe_group = sbi->s_mb_last_group; |
1988 | ac->ac_g_ex.fe_start = sbi->s_mb_last_start; |
1989 | spin_unlock(&sbi->s_md_lock); |
1990 | } |
1991 | /* Let's just scan groups to find more-less suitable blocks */ |
1992 | cr = ac->ac_2order ? 0 : 1; |
1993 | /* |
1994 | * cr == 0 try to get exact allocation, |
1995 | * cr == 3 try to get anything |
1996 | */ |
1997 | repeat: |
1998 | for (; cr < 4 && ac->ac_status == AC_STATUS_CONTINUE; cr++) { |
1999 | ac->ac_criteria = cr; |
2000 | /* |
2001 | * searching for the right group start |
2002 | * from the goal value specified |
2003 | */ |
2004 | group = ac->ac_g_ex.fe_group; |
2005 | |
2006 | for (i = 0; i < ngroups; group++, i++) { |
2007 | struct ext4_group_info *grp; |
2008 | struct ext4_group_desc *desc; |
2009 | |
2010 | if (group == ngroups) |
2011 | group = 0; |
2012 | |
2013 | /* quick check to skip empty groups */ |
2014 | grp = ext4_get_group_info(sb, group); |
2015 | if (grp->bb_free == 0) |
2016 | continue; |
2017 | |
2018 | /* |
2019 | * if the group is already init we check whether it is |
2020 | * a good group and if not we don't load the buddy |
2021 | */ |
2022 | if (EXT4_MB_GRP_NEED_INIT(grp)) { |
2023 | /* |
2024 | * we need full data about the group |
2025 | * to make a good selection |
2026 | */ |
2027 | err = ext4_mb_init_group(sb, group); |
2028 | if (err) |
2029 | goto out; |
2030 | } |
2031 | |
2032 | /* |
2033 | * If the particular group doesn't satisfy our |
2034 | * criteria we continue with the next group |
2035 | */ |
2036 | if (!ext4_mb_good_group(ac, group, cr)) |
2037 | continue; |
2038 | |
2039 | err = ext4_mb_load_buddy(sb, group, &e4b); |
2040 | if (err) |
2041 | goto out; |
2042 | |
2043 | ext4_lock_group(sb, group); |
2044 | if (!ext4_mb_good_group(ac, group, cr)) { |
2045 | /* someone did allocation from this group */ |
2046 | ext4_unlock_group(sb, group); |
2047 | ext4_mb_release_desc(&e4b); |
2048 | continue; |
2049 | } |
2050 | |
2051 | ac->ac_groups_scanned++; |
2052 | desc = ext4_get_group_desc(sb, group, NULL); |
2053 | if (cr == 0) |
2054 | ext4_mb_simple_scan_group(ac, &e4b); |
2055 | else if (cr == 1 && |
2056 | ac->ac_g_ex.fe_len == sbi->s_stripe) |
2057 | ext4_mb_scan_aligned(ac, &e4b); |
2058 | else |
2059 | ext4_mb_complex_scan_group(ac, &e4b); |
2060 | |
2061 | ext4_unlock_group(sb, group); |
2062 | ext4_mb_release_desc(&e4b); |
2063 | |
2064 | if (ac->ac_status != AC_STATUS_CONTINUE) |
2065 | break; |
2066 | } |
2067 | } |
2068 | |
2069 | if (ac->ac_b_ex.fe_len > 0 && ac->ac_status != AC_STATUS_FOUND && |
2070 | !(ac->ac_flags & EXT4_MB_HINT_FIRST)) { |
2071 | /* |
2072 | * We've been searching too long. Let's try to allocate |
2073 | * the best chunk we've found so far |
2074 | */ |
2075 | |
2076 | ext4_mb_try_best_found(ac, &e4b); |
2077 | if (ac->ac_status != AC_STATUS_FOUND) { |
2078 | /* |
2079 | * Someone more lucky has already allocated it. |
2080 | * The only thing we can do is just take first |
2081 | * found block(s) |
2082 | printk(KERN_DEBUG "EXT4-fs: someone won our chunk\n"); |
2083 | */ |
2084 | ac->ac_b_ex.fe_group = 0; |
2085 | ac->ac_b_ex.fe_start = 0; |
2086 | ac->ac_b_ex.fe_len = 0; |
2087 | ac->ac_status = AC_STATUS_CONTINUE; |
2088 | ac->ac_flags |= EXT4_MB_HINT_FIRST; |
2089 | cr = 3; |
2090 | atomic_inc(&sbi->s_mb_lost_chunks); |
2091 | goto repeat; |
2092 | } |
2093 | } |
2094 | out: |
2095 | return err; |
2096 | } |
2097 | |
2098 | #ifdef EXT4_MB_HISTORY |
2099 | struct ext4_mb_proc_session { |
2100 | struct ext4_mb_history *history; |
2101 | struct super_block *sb; |
2102 | int start; |
2103 | int max; |
2104 | }; |
2105 | |
2106 | static void *ext4_mb_history_skip_empty(struct ext4_mb_proc_session *s, |
2107 | struct ext4_mb_history *hs, |
2108 | int first) |
2109 | { |
2110 | if (hs == s->history + s->max) |
2111 | hs = s->history; |
2112 | if (!first && hs == s->history + s->start) |
2113 | return NULL; |
2114 | while (hs->orig.fe_len == 0) { |
2115 | hs++; |
2116 | if (hs == s->history + s->max) |
2117 | hs = s->history; |
2118 | if (hs == s->history + s->start) |
2119 | return NULL; |
2120 | } |
2121 | return hs; |
2122 | } |
2123 | |
2124 | static void *ext4_mb_seq_history_start(struct seq_file *seq, loff_t *pos) |
2125 | { |
2126 | struct ext4_mb_proc_session *s = seq->private; |
2127 | struct ext4_mb_history *hs; |
2128 | int l = *pos; |
2129 | |
2130 | if (l == 0) |
2131 | return SEQ_START_TOKEN; |
2132 | hs = ext4_mb_history_skip_empty(s, s->history + s->start, 1); |
2133 | if (!hs) |
2134 | return NULL; |
2135 | while (--l && (hs = ext4_mb_history_skip_empty(s, ++hs, 0)) != NULL); |
2136 | return hs; |
2137 | } |
2138 | |
2139 | static void *ext4_mb_seq_history_next(struct seq_file *seq, void *v, |
2140 | loff_t *pos) |
2141 | { |
2142 | struct ext4_mb_proc_session *s = seq->private; |
2143 | struct ext4_mb_history *hs = v; |
2144 | |
2145 | ++*pos; |
2146 | if (v == SEQ_START_TOKEN) |
2147 | return ext4_mb_history_skip_empty(s, s->history + s->start, 1); |
2148 | else |
2149 | return ext4_mb_history_skip_empty(s, ++hs, 0); |
2150 | } |
2151 | |
2152 | static int ext4_mb_seq_history_show(struct seq_file *seq, void *v) |
2153 | { |
2154 | char buf[25], buf2[25], buf3[25], *fmt; |
2155 | struct ext4_mb_history *hs = v; |
2156 | |
2157 | if (v == SEQ_START_TOKEN) { |
2158 | seq_printf(seq, "%-5s %-8s %-23s %-23s %-23s %-5s " |
2159 | "%-5s %-2s %-5s %-5s %-5s %-6s\n", |
2160 | "pid", "inode", "original", "goal", "result", "found", |
2161 | "grps", "cr", "flags", "merge", "tail", "broken"); |
2162 | return 0; |
2163 | } |
2164 | |
2165 | if (hs->op == EXT4_MB_HISTORY_ALLOC) { |
2166 | fmt = "%-5u %-8u %-23s %-23s %-23s %-5u %-5u %-2u " |
2167 | "%-5u %-5s %-5u %-6u\n"; |
2168 | sprintf(buf2, "%u/%d/%u@%u", hs->result.fe_group, |
2169 | hs->result.fe_start, hs->result.fe_len, |
2170 | hs->result.fe_logical); |
2171 | sprintf(buf, "%u/%d/%u@%u", hs->orig.fe_group, |
2172 | hs->orig.fe_start, hs->orig.fe_len, |
2173 | hs->orig.fe_logical); |
2174 | sprintf(buf3, "%u/%d/%u@%u", hs->goal.fe_group, |
2175 | hs->goal.fe_start, hs->goal.fe_len, |
2176 | hs->goal.fe_logical); |
2177 | seq_printf(seq, fmt, hs->pid, hs->ino, buf, buf3, buf2, |
2178 | hs->found, hs->groups, hs->cr, hs->flags, |
2179 | hs->merged ? "M" : "", hs->tail, |
2180 | hs->buddy ? 1 << hs->buddy : 0); |
2181 | } else if (hs->op == EXT4_MB_HISTORY_PREALLOC) { |
2182 | fmt = "%-5u %-8u %-23s %-23s %-23s\n"; |
2183 | sprintf(buf2, "%u/%d/%u@%u", hs->result.fe_group, |
2184 | hs->result.fe_start, hs->result.fe_len, |
2185 | hs->result.fe_logical); |
2186 | sprintf(buf, "%u/%d/%u@%u", hs->orig.fe_group, |
2187 | hs->orig.fe_start, hs->orig.fe_len, |
2188 | hs->orig.fe_logical); |
2189 | seq_printf(seq, fmt, hs->pid, hs->ino, buf, "", buf2); |
2190 | } else if (hs->op == EXT4_MB_HISTORY_DISCARD) { |
2191 | sprintf(buf2, "%u/%d/%u", hs->result.fe_group, |
2192 | hs->result.fe_start, hs->result.fe_len); |
2193 | seq_printf(seq, "%-5u %-8u %-23s discard\n", |
2194 | hs->pid, hs->ino, buf2); |
2195 | } else if (hs->op == EXT4_MB_HISTORY_FREE) { |
2196 | sprintf(buf2, "%u/%d/%u", hs->result.fe_group, |
2197 | hs->result.fe_start, hs->result.fe_len); |
2198 | seq_printf(seq, "%-5u %-8u %-23s free\n", |
2199 | hs->pid, hs->ino, buf2); |
2200 | } |
2201 | return 0; |
2202 | } |
2203 | |
2204 | static void ext4_mb_seq_history_stop(struct seq_file *seq, void *v) |
2205 | { |
2206 | } |
2207 | |
2208 | static struct seq_operations ext4_mb_seq_history_ops = { |
2209 | .start = ext4_mb_seq_history_start, |
2210 | .next = ext4_mb_seq_history_next, |
2211 | .stop = ext4_mb_seq_history_stop, |
2212 | .show = ext4_mb_seq_history_show, |
2213 | }; |
2214 | |
2215 | static int ext4_mb_seq_history_open(struct inode *inode, struct file *file) |
2216 | { |
2217 | struct super_block *sb = PDE(inode)->data; |
2218 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2219 | struct ext4_mb_proc_session *s; |
2220 | int rc; |
2221 | int size; |
2222 | |
2223 | if (unlikely(sbi->s_mb_history == NULL)) |
2224 | return -ENOMEM; |
2225 | s = kmalloc(sizeof(*s), GFP_KERNEL); |
2226 | if (s == NULL) |
2227 | return -ENOMEM; |
2228 | s->sb = sb; |
2229 | size = sizeof(struct ext4_mb_history) * sbi->s_mb_history_max; |
2230 | s->history = kmalloc(size, GFP_KERNEL); |
2231 | if (s->history == NULL) { |
2232 | kfree(s); |
2233 | return -ENOMEM; |
2234 | } |
2235 | |
2236 | spin_lock(&sbi->s_mb_history_lock); |
2237 | memcpy(s->history, sbi->s_mb_history, size); |
2238 | s->max = sbi->s_mb_history_max; |
2239 | s->start = sbi->s_mb_history_cur % s->max; |
2240 | spin_unlock(&sbi->s_mb_history_lock); |
2241 | |
2242 | rc = seq_open(file, &ext4_mb_seq_history_ops); |
2243 | if (rc == 0) { |
2244 | struct seq_file *m = (struct seq_file *)file->private_data; |
2245 | m->private = s; |
2246 | } else { |
2247 | kfree(s->history); |
2248 | kfree(s); |
2249 | } |
2250 | return rc; |
2251 | |
2252 | } |
2253 | |
2254 | static int ext4_mb_seq_history_release(struct inode *inode, struct file *file) |
2255 | { |
2256 | struct seq_file *seq = (struct seq_file *)file->private_data; |
2257 | struct ext4_mb_proc_session *s = seq->private; |
2258 | kfree(s->history); |
2259 | kfree(s); |
2260 | return seq_release(inode, file); |
2261 | } |
2262 | |
2263 | static ssize_t ext4_mb_seq_history_write(struct file *file, |
2264 | const char __user *buffer, |
2265 | size_t count, loff_t *ppos) |
2266 | { |
2267 | struct seq_file *seq = (struct seq_file *)file->private_data; |
2268 | struct ext4_mb_proc_session *s = seq->private; |
2269 | struct super_block *sb = s->sb; |
2270 | char str[32]; |
2271 | int value; |
2272 | |
2273 | if (count >= sizeof(str)) { |
2274 | printk(KERN_ERR "EXT4-fs: %s string too long, max %u bytes\n", |
2275 | "mb_history", (int)sizeof(str)); |
2276 | return -EOVERFLOW; |
2277 | } |
2278 | |
2279 | if (copy_from_user(str, buffer, count)) |
2280 | return -EFAULT; |
2281 | |
2282 | value = simple_strtol(str, NULL, 0); |
2283 | if (value < 0) |
2284 | return -ERANGE; |
2285 | EXT4_SB(sb)->s_mb_history_filter = value; |
2286 | |
2287 | return count; |
2288 | } |
2289 | |
2290 | static struct file_operations ext4_mb_seq_history_fops = { |
2291 | .owner = THIS_MODULE, |
2292 | .open = ext4_mb_seq_history_open, |
2293 | .read = seq_read, |
2294 | .write = ext4_mb_seq_history_write, |
2295 | .llseek = seq_lseek, |
2296 | .release = ext4_mb_seq_history_release, |
2297 | }; |
2298 | |
2299 | static void *ext4_mb_seq_groups_start(struct seq_file *seq, loff_t *pos) |
2300 | { |
2301 | struct super_block *sb = seq->private; |
2302 | ext4_group_t group; |
2303 | |
2304 | if (*pos < 0 || *pos >= ext4_get_groups_count(sb)) |
2305 | return NULL; |
2306 | group = *pos + 1; |
2307 | return (void *) ((unsigned long) group); |
2308 | } |
2309 | |
2310 | static void *ext4_mb_seq_groups_next(struct seq_file *seq, void *v, loff_t *pos) |
2311 | { |
2312 | struct super_block *sb = seq->private; |
2313 | ext4_group_t group; |
2314 | |
2315 | ++*pos; |
2316 | if (*pos < 0 || *pos >= ext4_get_groups_count(sb)) |
2317 | return NULL; |
2318 | group = *pos + 1; |
2319 | return (void *) ((unsigned long) group); |
2320 | } |
2321 | |
2322 | static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) |
2323 | { |
2324 | struct super_block *sb = seq->private; |
2325 | ext4_group_t group = (ext4_group_t) ((unsigned long) v); |
2326 | int i; |
2327 | int err; |
2328 | struct ext4_buddy e4b; |
2329 | struct sg { |
2330 | struct ext4_group_info info; |
2331 | unsigned short counters[16]; |
2332 | } sg; |
2333 | |
2334 | group--; |
2335 | if (group == 0) |
2336 | seq_printf(seq, "#%-5s: %-5s %-5s %-5s " |
2337 | "[ %-5s %-5s %-5s %-5s %-5s %-5s %-5s " |
2338 | "%-5s %-5s %-5s %-5s %-5s %-5s %-5s ]\n", |
2339 | "group", "free", "frags", "first", |
2340 | "2^0", "2^1", "2^2", "2^3", "2^4", "2^5", "2^6", |
2341 | "2^7", "2^8", "2^9", "2^10", "2^11", "2^12", "2^13"); |
2342 | |
2343 | i = (sb->s_blocksize_bits + 2) * sizeof(sg.info.bb_counters[0]) + |
2344 | sizeof(struct ext4_group_info); |
2345 | err = ext4_mb_load_buddy(sb, group, &e4b); |
2346 | if (err) { |
2347 | seq_printf(seq, "#%-5u: I/O error\n", group); |
2348 | return 0; |
2349 | } |
2350 | ext4_lock_group(sb, group); |
2351 | memcpy(&sg, ext4_get_group_info(sb, group), i); |
2352 | ext4_unlock_group(sb, group); |
2353 | ext4_mb_release_desc(&e4b); |
2354 | |
2355 | seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg.info.bb_free, |
2356 | sg.info.bb_fragments, sg.info.bb_first_free); |
2357 | for (i = 0; i <= 13; i++) |
2358 | seq_printf(seq, " %-5u", i <= sb->s_blocksize_bits + 1 ? |
2359 | sg.info.bb_counters[i] : 0); |
2360 | seq_printf(seq, " ]\n"); |
2361 | |
2362 | return 0; |
2363 | } |
2364 | |
2365 | static void ext4_mb_seq_groups_stop(struct seq_file *seq, void *v) |
2366 | { |
2367 | } |
2368 | |
2369 | static struct seq_operations ext4_mb_seq_groups_ops = { |
2370 | .start = ext4_mb_seq_groups_start, |
2371 | .next = ext4_mb_seq_groups_next, |
2372 | .stop = ext4_mb_seq_groups_stop, |
2373 | .show = ext4_mb_seq_groups_show, |
2374 | }; |
2375 | |
2376 | static int ext4_mb_seq_groups_open(struct inode *inode, struct file *file) |
2377 | { |
2378 | struct super_block *sb = PDE(inode)->data; |
2379 | int rc; |
2380 | |
2381 | rc = seq_open(file, &ext4_mb_seq_groups_ops); |
2382 | if (rc == 0) { |
2383 | struct seq_file *m = (struct seq_file *)file->private_data; |
2384 | m->private = sb; |
2385 | } |
2386 | return rc; |
2387 | |
2388 | } |
2389 | |
2390 | static struct file_operations ext4_mb_seq_groups_fops = { |
2391 | .owner = THIS_MODULE, |
2392 | .open = ext4_mb_seq_groups_open, |
2393 | .read = seq_read, |
2394 | .llseek = seq_lseek, |
2395 | .release = seq_release, |
2396 | }; |
2397 | |
2398 | static void ext4_mb_history_release(struct super_block *sb) |
2399 | { |
2400 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2401 | |
2402 | if (sbi->s_proc != NULL) { |
2403 | remove_proc_entry("mb_groups", sbi->s_proc); |
2404 | if (sbi->s_mb_history_max) |
2405 | remove_proc_entry("mb_history", sbi->s_proc); |
2406 | } |
2407 | kfree(sbi->s_mb_history); |
2408 | } |
2409 | |
2410 | static void ext4_mb_history_init(struct super_block *sb) |
2411 | { |
2412 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2413 | int i; |
2414 | |
2415 | if (sbi->s_proc != NULL) { |
2416 | if (sbi->s_mb_history_max) |
2417 | proc_create_data("mb_history", S_IRUGO, sbi->s_proc, |
2418 | &ext4_mb_seq_history_fops, sb); |
2419 | proc_create_data("mb_groups", S_IRUGO, sbi->s_proc, |
2420 | &ext4_mb_seq_groups_fops, sb); |
2421 | } |
2422 | |
2423 | sbi->s_mb_history_cur = 0; |
2424 | spin_lock_init(&sbi->s_mb_history_lock); |
2425 | i = sbi->s_mb_history_max * sizeof(struct ext4_mb_history); |
2426 | sbi->s_mb_history = i ? kzalloc(i, GFP_KERNEL) : NULL; |
2427 | /* if we can't allocate history, then we simple won't use it */ |
2428 | } |
2429 | |
2430 | static noinline_for_stack void |
2431 | ext4_mb_store_history(struct ext4_allocation_context *ac) |
2432 | { |
2433 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
2434 | struct ext4_mb_history h; |
2435 | |
2436 | if (sbi->s_mb_history == NULL) |
2437 | return; |
2438 | |
2439 | if (!(ac->ac_op & sbi->s_mb_history_filter)) |
2440 | return; |
2441 | |
2442 | h.op = ac->ac_op; |
2443 | h.pid = current->pid; |
2444 | h.ino = ac->ac_inode ? ac->ac_inode->i_ino : 0; |
2445 | h.orig = ac->ac_o_ex; |
2446 | h.result = ac->ac_b_ex; |
2447 | h.flags = ac->ac_flags; |
2448 | h.found = ac->ac_found; |
2449 | h.groups = ac->ac_groups_scanned; |
2450 | h.cr = ac->ac_criteria; |
2451 | h.tail = ac->ac_tail; |
2452 | h.buddy = ac->ac_buddy; |
2453 | h.merged = 0; |
2454 | if (ac->ac_op == EXT4_MB_HISTORY_ALLOC) { |
2455 | if (ac->ac_g_ex.fe_start == ac->ac_b_ex.fe_start && |
2456 | ac->ac_g_ex.fe_group == ac->ac_b_ex.fe_group) |
2457 | h.merged = 1; |
2458 | h.goal = ac->ac_g_ex; |
2459 | h.result = ac->ac_f_ex; |
2460 | } |
2461 | |
2462 | spin_lock(&sbi->s_mb_history_lock); |
2463 | memcpy(sbi->s_mb_history + sbi->s_mb_history_cur, &h, sizeof(h)); |
2464 | if (++sbi->s_mb_history_cur >= sbi->s_mb_history_max) |
2465 | sbi->s_mb_history_cur = 0; |
2466 | spin_unlock(&sbi->s_mb_history_lock); |
2467 | } |
2468 | |
2469 | #else |
2470 | #define ext4_mb_history_release(sb) |
2471 | #define ext4_mb_history_init(sb) |
2472 | #endif |
2473 | |
2474 | |
2475 | /* Create and initialize ext4_group_info data for the given group. */ |
2476 | int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, |
2477 | struct ext4_group_desc *desc) |
2478 | { |
2479 | int i, len; |
2480 | int metalen = 0; |
2481 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2482 | struct ext4_group_info **meta_group_info; |
2483 | |
2484 | /* |
2485 | * First check if this group is the first of a reserved block. |
2486 | * If it's true, we have to allocate a new table of pointers |
2487 | * to ext4_group_info structures |
2488 | */ |
2489 | if (group % EXT4_DESC_PER_BLOCK(sb) == 0) { |
2490 | metalen = sizeof(*meta_group_info) << |
2491 | EXT4_DESC_PER_BLOCK_BITS(sb); |
2492 | meta_group_info = kmalloc(metalen, GFP_KERNEL); |
2493 | if (meta_group_info == NULL) { |
2494 | printk(KERN_ERR "EXT4-fs: can't allocate mem for a " |
2495 | "buddy group\n"); |
2496 | goto exit_meta_group_info; |
2497 | } |
2498 | sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] = |
2499 | meta_group_info; |
2500 | } |
2501 | |
2502 | /* |
2503 | * calculate needed size. if change bb_counters size, |
2504 | * don't forget about ext4_mb_generate_buddy() |
2505 | */ |
2506 | len = offsetof(typeof(**meta_group_info), |
2507 | bb_counters[sb->s_blocksize_bits + 2]); |
2508 | |
2509 | meta_group_info = |
2510 | sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)]; |
2511 | i = group & (EXT4_DESC_PER_BLOCK(sb) - 1); |
2512 | |
2513 | meta_group_info[i] = kzalloc(len, GFP_KERNEL); |
2514 | if (meta_group_info[i] == NULL) { |
2515 | printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n"); |
2516 | goto exit_group_info; |
2517 | } |
2518 | set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, |
2519 | &(meta_group_info[i]->bb_state)); |
2520 | |
2521 | /* |
2522 | * initialize bb_free to be able to skip |
2523 | * empty groups without initialization |
2524 | */ |
2525 | if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) { |
2526 | meta_group_info[i]->bb_free = |
2527 | ext4_free_blocks_after_init(sb, group, desc); |
2528 | } else { |
2529 | meta_group_info[i]->bb_free = |
2530 | ext4_free_blks_count(sb, desc); |
2531 | } |
2532 | |
2533 | INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list); |
2534 | init_rwsem(&meta_group_info[i]->alloc_sem); |
2535 | meta_group_info[i]->bb_free_root.rb_node = NULL;; |
2536 | |
2537 | #ifdef DOUBLE_CHECK |
2538 | { |
2539 | struct buffer_head *bh; |
2540 | meta_group_info[i]->bb_bitmap = |
2541 | kmalloc(sb->s_blocksize, GFP_KERNEL); |
2542 | BUG_ON(meta_group_info[i]->bb_bitmap == NULL); |
2543 | bh = ext4_read_block_bitmap(sb, group); |
2544 | BUG_ON(bh == NULL); |
2545 | memcpy(meta_group_info[i]->bb_bitmap, bh->b_data, |
2546 | sb->s_blocksize); |
2547 | put_bh(bh); |
2548 | } |
2549 | #endif |
2550 | |
2551 | return 0; |
2552 | |
2553 | exit_group_info: |
2554 | /* If a meta_group_info table has been allocated, release it now */ |
2555 | if (group % EXT4_DESC_PER_BLOCK(sb) == 0) |
2556 | kfree(sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)]); |
2557 | exit_meta_group_info: |
2558 | return -ENOMEM; |
2559 | } /* ext4_mb_add_groupinfo */ |
2560 | |
2561 | /* |
2562 | * Update an existing group. |
2563 | * This function is used for online resize |
2564 | */ |
2565 | void ext4_mb_update_group_info(struct ext4_group_info *grp, ext4_grpblk_t add) |
2566 | { |
2567 | grp->bb_free += add; |
2568 | } |
2569 | |
2570 | static int ext4_mb_init_backend(struct super_block *sb) |
2571 | { |
2572 | ext4_group_t ngroups = ext4_get_groups_count(sb); |
2573 | ext4_group_t i; |
2574 | int metalen; |
2575 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2576 | struct ext4_super_block *es = sbi->s_es; |
2577 | int num_meta_group_infos; |
2578 | int num_meta_group_infos_max; |
2579 | int array_size; |
2580 | struct ext4_group_info **meta_group_info; |
2581 | struct ext4_group_desc *desc; |
2582 | |
2583 | /* This is the number of blocks used by GDT */ |
2584 | num_meta_group_infos = (ngroups + EXT4_DESC_PER_BLOCK(sb) - |
2585 | 1) >> EXT4_DESC_PER_BLOCK_BITS(sb); |
2586 | |
2587 | /* |
2588 | * This is the total number of blocks used by GDT including |
2589 | * the number of reserved blocks for GDT. |
2590 | * The s_group_info array is allocated with this value |
2591 | * to allow a clean online resize without a complex |
2592 | * manipulation of pointer. |
2593 | * The drawback is the unused memory when no resize |
2594 | * occurs but it's very low in terms of pages |
2595 | * (see comments below) |
2596 | * Need to handle this properly when META_BG resizing is allowed |
2597 | */ |
2598 | num_meta_group_infos_max = num_meta_group_infos + |
2599 | le16_to_cpu(es->s_reserved_gdt_blocks); |
2600 | |
2601 | /* |
2602 | * array_size is the size of s_group_info array. We round it |
2603 | * to the next power of two because this approximation is done |
2604 | * internally by kmalloc so we can have some more memory |
2605 | * for free here (e.g. may be used for META_BG resize). |
2606 | */ |
2607 | array_size = 1; |
2608 | while (array_size < sizeof(*sbi->s_group_info) * |
2609 | num_meta_group_infos_max) |
2610 | array_size = array_size << 1; |
2611 | /* An 8TB filesystem with 64-bit pointers requires a 4096 byte |
2612 | * kmalloc. A 128kb malloc should suffice for a 256TB filesystem. |
2613 | * So a two level scheme suffices for now. */ |
2614 | sbi->s_group_info = kmalloc(array_size, GFP_KERNEL); |
2615 | if (sbi->s_group_info == NULL) { |
2616 | printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n"); |
2617 | return -ENOMEM; |
2618 | } |
2619 | sbi->s_buddy_cache = new_inode(sb); |
2620 | if (sbi->s_buddy_cache == NULL) { |
2621 | printk(KERN_ERR "EXT4-fs: can't get new inode\n"); |
2622 | goto err_freesgi; |
2623 | } |
2624 | EXT4_I(sbi->s_buddy_cache)->i_disksize = 0; |
2625 | |
2626 | metalen = sizeof(*meta_group_info) << EXT4_DESC_PER_BLOCK_BITS(sb); |
2627 | for (i = 0; i < num_meta_group_infos; i++) { |
2628 | if ((i + 1) == num_meta_group_infos) |
2629 | metalen = sizeof(*meta_group_info) * |
2630 | (ngroups - |
2631 | (i << EXT4_DESC_PER_BLOCK_BITS(sb))); |
2632 | meta_group_info = kmalloc(metalen, GFP_KERNEL); |
2633 | if (meta_group_info == NULL) { |
2634 | printk(KERN_ERR "EXT4-fs: can't allocate mem for a " |
2635 | "buddy group\n"); |
2636 | goto err_freemeta; |
2637 | } |
2638 | sbi->s_group_info[i] = meta_group_info; |
2639 | } |
2640 | |
2641 | for (i = 0; i < ngroups; i++) { |
2642 | desc = ext4_get_group_desc(sb, i, NULL); |
2643 | if (desc == NULL) { |
2644 | printk(KERN_ERR |
2645 | "EXT4-fs: can't read descriptor %u\n", i); |
2646 | goto err_freebuddy; |
2647 | } |
2648 | if (ext4_mb_add_groupinfo(sb, i, desc) != 0) |
2649 | goto err_freebuddy; |
2650 | } |
2651 | |
2652 | return 0; |
2653 | |
2654 | err_freebuddy: |
2655 | while (i-- > 0) |
2656 | kfree(ext4_get_group_info(sb, i)); |
2657 | i = num_meta_group_infos; |
2658 | err_freemeta: |
2659 | while (i-- > 0) |
2660 | kfree(sbi->s_group_info[i]); |
2661 | iput(sbi->s_buddy_cache); |
2662 | err_freesgi: |
2663 | kfree(sbi->s_group_info); |
2664 | return -ENOMEM; |
2665 | } |
2666 | |
2667 | int ext4_mb_init(struct super_block *sb, int needs_recovery) |
2668 | { |
2669 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2670 | unsigned i, j; |
2671 | unsigned offset; |
2672 | unsigned max; |
2673 | int ret; |
2674 | |
2675 | i = (sb->s_blocksize_bits + 2) * sizeof(unsigned short); |
2676 | |
2677 | sbi->s_mb_offsets = kmalloc(i, GFP_KERNEL); |
2678 | if (sbi->s_mb_offsets == NULL) { |
2679 | return -ENOMEM; |
2680 | } |
2681 | |
2682 | i = (sb->s_blocksize_bits + 2) * sizeof(unsigned int); |
2683 | sbi->s_mb_maxs = kmalloc(i, GFP_KERNEL); |
2684 | if (sbi->s_mb_maxs == NULL) { |
2685 | kfree(sbi->s_mb_offsets); |
2686 | return -ENOMEM; |
2687 | } |
2688 | |
2689 | /* order 0 is regular bitmap */ |
2690 | sbi->s_mb_maxs[0] = sb->s_blocksize << 3; |
2691 | sbi->s_mb_offsets[0] = 0; |
2692 | |
2693 | i = 1; |
2694 | offset = 0; |
2695 | max = sb->s_blocksize << 2; |
2696 | do { |
2697 | sbi->s_mb_offsets[i] = offset; |
2698 | sbi->s_mb_maxs[i] = max; |
2699 | offset += 1 << (sb->s_blocksize_bits - i); |
2700 | max = max >> 1; |
2701 | i++; |
2702 | } while (i <= sb->s_blocksize_bits + 1); |
2703 | |
2704 | /* init file for buddy data */ |
2705 | ret = ext4_mb_init_backend(sb); |
2706 | if (ret != 0) { |
2707 | kfree(sbi->s_mb_offsets); |
2708 | kfree(sbi->s_mb_maxs); |
2709 | return ret; |
2710 | } |
2711 | |
2712 | spin_lock_init(&sbi->s_md_lock); |
2713 | spin_lock_init(&sbi->s_bal_lock); |
2714 | |
2715 | sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN; |
2716 | sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN; |
2717 | sbi->s_mb_stats = MB_DEFAULT_STATS; |
2718 | sbi->s_mb_stream_request = MB_DEFAULT_STREAM_THRESHOLD; |
2719 | sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS; |
2720 | sbi->s_mb_history_filter = EXT4_MB_HISTORY_DEFAULT; |
2721 | sbi->s_mb_group_prealloc = MB_DEFAULT_GROUP_PREALLOC; |
2722 | |
2723 | sbi->s_locality_groups = alloc_percpu(struct ext4_locality_group); |
2724 | if (sbi->s_locality_groups == NULL) { |
2725 | kfree(sbi->s_mb_offsets); |
2726 | kfree(sbi->s_mb_maxs); |
2727 | return -ENOMEM; |
2728 | } |
2729 | for_each_possible_cpu(i) { |
2730 | struct ext4_locality_group *lg; |
2731 | lg = per_cpu_ptr(sbi->s_locality_groups, i); |
2732 | mutex_init(&lg->lg_mutex); |
2733 | for (j = 0; j < PREALLOC_TB_SIZE; j++) |
2734 | INIT_LIST_HEAD(&lg->lg_prealloc_list[j]); |
2735 | spin_lock_init(&lg->lg_prealloc_lock); |
2736 | } |
2737 | |
2738 | ext4_mb_history_init(sb); |
2739 | |
2740 | if (sbi->s_journal) |
2741 | sbi->s_journal->j_commit_callback = release_blocks_on_commit; |
2742 | |
2743 | printk(KERN_INFO "EXT4-fs: mballoc enabled\n"); |
2744 | return 0; |
2745 | } |
2746 | |
2747 | /* need to called with the ext4 group lock held */ |
2748 | static void ext4_mb_cleanup_pa(struct ext4_group_info *grp) |
2749 | { |
2750 | struct ext4_prealloc_space *pa; |
2751 | struct list_head *cur, *tmp; |
2752 | int count = 0; |
2753 | |
2754 | list_for_each_safe(cur, tmp, &grp->bb_prealloc_list) { |
2755 | pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); |
2756 | list_del(&pa->pa_group_list); |
2757 | count++; |
2758 | kmem_cache_free(ext4_pspace_cachep, pa); |
2759 | } |
2760 | if (count) |
2761 | mb_debug("mballoc: %u PAs left\n", count); |
2762 | |
2763 | } |
2764 | |
2765 | int ext4_mb_release(struct super_block *sb) |
2766 | { |
2767 | ext4_group_t ngroups = ext4_get_groups_count(sb); |
2768 | ext4_group_t i; |
2769 | int num_meta_group_infos; |
2770 | struct ext4_group_info *grinfo; |
2771 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
2772 | |
2773 | if (sbi->s_group_info) { |
2774 | for (i = 0; i < ngroups; i++) { |
2775 | grinfo = ext4_get_group_info(sb, i); |
2776 | #ifdef DOUBLE_CHECK |
2777 | kfree(grinfo->bb_bitmap); |
2778 | #endif |
2779 | ext4_lock_group(sb, i); |
2780 | ext4_mb_cleanup_pa(grinfo); |
2781 | ext4_unlock_group(sb, i); |
2782 | kfree(grinfo); |
2783 | } |
2784 | num_meta_group_infos = (ngroups + |
2785 | EXT4_DESC_PER_BLOCK(sb) - 1) >> |
2786 | EXT4_DESC_PER_BLOCK_BITS(sb); |
2787 | for (i = 0; i < num_meta_group_infos; i++) |
2788 | kfree(sbi->s_group_info[i]); |
2789 | kfree(sbi->s_group_info); |
2790 | } |
2791 | kfree(sbi->s_mb_offsets); |
2792 | kfree(sbi->s_mb_maxs); |
2793 | if (sbi->s_buddy_cache) |
2794 | iput(sbi->s_buddy_cache); |
2795 | if (sbi->s_mb_stats) { |
2796 | printk(KERN_INFO |
2797 | "EXT4-fs: mballoc: %u blocks %u reqs (%u success)\n", |
2798 | atomic_read(&sbi->s_bal_allocated), |
2799 | atomic_read(&sbi->s_bal_reqs), |
2800 | atomic_read(&sbi->s_bal_success)); |
2801 | printk(KERN_INFO |
2802 | "EXT4-fs: mballoc: %u extents scanned, %u goal hits, " |
2803 | "%u 2^N hits, %u breaks, %u lost\n", |
2804 | atomic_read(&sbi->s_bal_ex_scanned), |
2805 | atomic_read(&sbi->s_bal_goals), |
2806 | atomic_read(&sbi->s_bal_2orders), |
2807 | atomic_read(&sbi->s_bal_breaks), |
2808 | atomic_read(&sbi->s_mb_lost_chunks)); |
2809 | printk(KERN_INFO |
2810 | "EXT4-fs: mballoc: %lu generated and it took %Lu\n", |
2811 | sbi->s_mb_buddies_generated++, |
2812 | sbi->s_mb_generation_time); |
2813 | printk(KERN_INFO |
2814 | "EXT4-fs: mballoc: %u preallocated, %u discarded\n", |
2815 | atomic_read(&sbi->s_mb_preallocated), |
2816 | atomic_read(&sbi->s_mb_discarded)); |
2817 | } |
2818 | |
2819 | free_percpu(sbi->s_locality_groups); |
2820 | ext4_mb_history_release(sb); |
2821 | |
2822 | return 0; |
2823 | } |
2824 | |
2825 | /* |
2826 | * This function is called by the jbd2 layer once the commit has finished, |
2827 | * so we know we can free the blocks that were released with that commit. |
2828 | */ |
2829 | static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) |
2830 | { |
2831 | struct super_block *sb = journal->j_private; |
2832 | struct ext4_buddy e4b; |
2833 | struct ext4_group_info *db; |
2834 | int err, count = 0, count2 = 0; |
2835 | struct ext4_free_data *entry; |
2836 | ext4_fsblk_t discard_block; |
2837 | struct list_head *l, *ltmp; |
2838 | |
2839 | list_for_each_safe(l, ltmp, &txn->t_private_list) { |
2840 | entry = list_entry(l, struct ext4_free_data, list); |
2841 | |
2842 | mb_debug("gonna free %u blocks in group %u (0x%p):", |
2843 | entry->count, entry->group, entry); |
2844 | |
2845 | err = ext4_mb_load_buddy(sb, entry->group, &e4b); |
2846 | /* we expect to find existing buddy because it's pinned */ |
2847 | BUG_ON(err != 0); |
2848 | |
2849 | db = e4b.bd_info; |
2850 | /* there are blocks to put in buddy to make them really free */ |
2851 | count += entry->count; |
2852 | count2++; |
2853 | ext4_lock_group(sb, entry->group); |
2854 | /* Take it out of per group rb tree */ |
2855 | rb_erase(&entry->node, &(db->bb_free_root)); |
2856 | mb_free_blocks(NULL, &e4b, entry->start_blk, entry->count); |
2857 | |
2858 | if (!db->bb_free_root.rb_node) { |
2859 | /* No more items in the per group rb tree |
2860 | * balance refcounts from ext4_mb_free_metadata() |
2861 | */ |
2862 | page_cache_release(e4b.bd_buddy_page); |
2863 | page_cache_release(e4b.bd_bitmap_page); |
2864 | } |
2865 | ext4_unlock_group(sb, entry->group); |
2866 | discard_block = (ext4_fsblk_t) entry->group * EXT4_BLOCKS_PER_GROUP(sb) |
2867 | + entry->start_blk |
2868 | + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); |
2869 | trace_ext4_discard_blocks(sb, (unsigned long long)discard_block, |
2870 | entry->count); |
2871 | sb_issue_discard(sb, discard_block, entry->count); |
2872 | |
2873 | kmem_cache_free(ext4_free_ext_cachep, entry); |
2874 | ext4_mb_release_desc(&e4b); |
2875 | } |
2876 | |
2877 | mb_debug("freed %u blocks in %u structures\n", count, count2); |
2878 | } |
2879 | |
2880 | int __init init_ext4_mballoc(void) |
2881 | { |
2882 | ext4_pspace_cachep = |
2883 | kmem_cache_create("ext4_prealloc_space", |
2884 | sizeof(struct ext4_prealloc_space), |
2885 | 0, SLAB_RECLAIM_ACCOUNT, NULL); |
2886 | if (ext4_pspace_cachep == NULL) |
2887 | return -ENOMEM; |
2888 | |
2889 | ext4_ac_cachep = |
2890 | kmem_cache_create("ext4_alloc_context", |
2891 | sizeof(struct ext4_allocation_context), |
2892 | 0, SLAB_RECLAIM_ACCOUNT, NULL); |
2893 | if (ext4_ac_cachep == NULL) { |
2894 | kmem_cache_destroy(ext4_pspace_cachep); |
2895 | return -ENOMEM; |
2896 | } |
2897 | |
2898 | ext4_free_ext_cachep = |
2899 | kmem_cache_create("ext4_free_block_extents", |
2900 | sizeof(struct ext4_free_data), |
2901 | 0, SLAB_RECLAIM_ACCOUNT, NULL); |
2902 | if (ext4_free_ext_cachep == NULL) { |
2903 | kmem_cache_destroy(ext4_pspace_cachep); |
2904 | kmem_cache_destroy(ext4_ac_cachep); |
2905 | return -ENOMEM; |
2906 | } |
2907 | return 0; |
2908 | } |
2909 | |
2910 | void exit_ext4_mballoc(void) |
2911 | { |
2912 | /* |
2913 | * Wait for completion of call_rcu()'s on ext4_pspace_cachep |
2914 | * before destroying the slab cache. |
2915 | */ |
2916 | rcu_barrier(); |
2917 | kmem_cache_destroy(ext4_pspace_cachep); |
2918 | kmem_cache_destroy(ext4_ac_cachep); |
2919 | kmem_cache_destroy(ext4_free_ext_cachep); |
2920 | } |
2921 | |
2922 | |
2923 | /* |
2924 | * Check quota and mark choosed space (ac->ac_b_ex) non-free in bitmaps |
2925 | * Returns 0 if success or error code |
2926 | */ |
2927 | static noinline_for_stack int |
2928 | ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, |
2929 | handle_t *handle, unsigned int reserv_blks) |
2930 | { |
2931 | struct buffer_head *bitmap_bh = NULL; |
2932 | struct ext4_super_block *es; |
2933 | struct ext4_group_desc *gdp; |
2934 | struct buffer_head *gdp_bh; |
2935 | struct ext4_sb_info *sbi; |
2936 | struct super_block *sb; |
2937 | ext4_fsblk_t block; |
2938 | int err, len; |
2939 | |
2940 | BUG_ON(ac->ac_status != AC_STATUS_FOUND); |
2941 | BUG_ON(ac->ac_b_ex.fe_len <= 0); |
2942 | |
2943 | sb = ac->ac_sb; |
2944 | sbi = EXT4_SB(sb); |
2945 | es = sbi->s_es; |
2946 | |
2947 | |
2948 | err = -EIO; |
2949 | bitmap_bh = ext4_read_block_bitmap(sb, ac->ac_b_ex.fe_group); |
2950 | if (!bitmap_bh) |
2951 | goto out_err; |
2952 | |
2953 | err = ext4_journal_get_write_access(handle, bitmap_bh); |
2954 | if (err) |
2955 | goto out_err; |
2956 | |
2957 | err = -EIO; |
2958 | gdp = ext4_get_group_desc(sb, ac->ac_b_ex.fe_group, &gdp_bh); |
2959 | if (!gdp) |
2960 | goto out_err; |
2961 | |
2962 | ext4_debug("using block group %u(%d)\n", ac->ac_b_ex.fe_group, |
2963 | ext4_free_blks_count(sb, gdp)); |
2964 | |
2965 | err = ext4_journal_get_write_access(handle, gdp_bh); |
2966 | if (err) |
2967 | goto out_err; |
2968 | |
2969 | block = ac->ac_b_ex.fe_group * EXT4_BLOCKS_PER_GROUP(sb) |
2970 | + ac->ac_b_ex.fe_start |
2971 | + le32_to_cpu(es->s_first_data_block); |
2972 | |
2973 | len = ac->ac_b_ex.fe_len; |
2974 | if (!ext4_data_block_valid(sbi, block, len)) { |
2975 | ext4_error(sb, __func__, |
2976 | "Allocating blocks %llu-%llu which overlap " |
2977 | "fs metadata\n", block, block+len); |
2978 | /* File system mounted not to panic on error |
2979 | * Fix the bitmap and repeat the block allocation |
2980 | * We leak some of the blocks here. |
2981 | */ |
2982 | ext4_lock_group(sb, ac->ac_b_ex.fe_group); |
2983 | mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start, |
2984 | ac->ac_b_ex.fe_len); |
2985 | ext4_unlock_group(sb, ac->ac_b_ex.fe_group); |
2986 | err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); |
2987 | if (!err) |
2988 | err = -EAGAIN; |
2989 | goto out_err; |
2990 | } |
2991 | |
2992 | ext4_lock_group(sb, ac->ac_b_ex.fe_group); |
2993 | #ifdef AGGRESSIVE_CHECK |
2994 | { |
2995 | int i; |
2996 | for (i = 0; i < ac->ac_b_ex.fe_len; i++) { |
2997 | BUG_ON(mb_test_bit(ac->ac_b_ex.fe_start + i, |
2998 | bitmap_bh->b_data)); |
2999 | } |
3000 | } |
3001 | #endif |
3002 | mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start,ac->ac_b_ex.fe_len); |
3003 | if (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) { |
3004 | gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT); |
3005 | ext4_free_blks_set(sb, gdp, |
3006 | ext4_free_blocks_after_init(sb, |
3007 | ac->ac_b_ex.fe_group, gdp)); |
3008 | } |
3009 | len = ext4_free_blks_count(sb, gdp) - ac->ac_b_ex.fe_len; |
3010 | ext4_free_blks_set(sb, gdp, len); |
3011 | gdp->bg_checksum = ext4_group_desc_csum(sbi, ac->ac_b_ex.fe_group, gdp); |
3012 | |
3013 | ext4_unlock_group(sb, ac->ac_b_ex.fe_group); |
3014 | percpu_counter_sub(&sbi->s_freeblocks_counter, ac->ac_b_ex.fe_len); |
3015 | /* |
3016 | * Now reduce the dirty block count also. Should not go negative |
3017 | */ |
3018 | if (!(ac->ac_flags & EXT4_MB_DELALLOC_RESERVED)) |
3019 | /* release all the reserved blocks if non delalloc */ |
3020 | percpu_counter_sub(&sbi->s_dirtyblocks_counter, reserv_blks); |
3021 | else { |
3022 | percpu_counter_sub(&sbi->s_dirtyblocks_counter, |
3023 | ac->ac_b_ex.fe_len); |
3024 | /* convert reserved quota blocks to real quota blocks */ |
3025 | vfs_dq_claim_block(ac->ac_inode, ac->ac_b_ex.fe_len); |
3026 | } |
3027 | |
3028 | if (sbi->s_log_groups_per_flex) { |
3029 | ext4_group_t flex_group = ext4_flex_group(sbi, |
3030 | ac->ac_b_ex.fe_group); |
3031 | atomic_sub(ac->ac_b_ex.fe_len, |
3032 | &sbi->s_flex_groups[flex_group].free_blocks); |
3033 | } |
3034 | |
3035 | err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); |
3036 | if (err) |
3037 | goto out_err; |
3038 | err = ext4_handle_dirty_metadata(handle, NULL, gdp_bh); |
3039 | |
3040 | out_err: |
3041 | sb->s_dirt = 1; |
3042 | brelse(bitmap_bh); |
3043 | return err; |
3044 | } |
3045 | |
3046 | /* |
3047 | * here we normalize request for locality group |
3048 | * Group request are normalized to s_strip size if we set the same via mount |
3049 | * option. If not we set it to s_mb_group_prealloc which can be configured via |
3050 | * /sys/fs/ext4/<partition>/mb_group_prealloc |
3051 | * |
3052 | * XXX: should we try to preallocate more than the group has now? |
3053 | */ |
3054 | static void ext4_mb_normalize_group_request(struct ext4_allocation_context *ac) |
3055 | { |
3056 | struct super_block *sb = ac->ac_sb; |
3057 | struct ext4_locality_group *lg = ac->ac_lg; |
3058 | |
3059 | BUG_ON(lg == NULL); |
3060 | if (EXT4_SB(sb)->s_stripe) |
3061 | ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_stripe; |
3062 | else |
3063 | ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_mb_group_prealloc; |
3064 | mb_debug("#%u: goal %u blocks for locality group\n", |
3065 | current->pid, ac->ac_g_ex.fe_len); |
3066 | } |
3067 | |
3068 | /* |
3069 | * Normalization means making request better in terms of |
3070 | * size and alignment |
3071 | */ |
3072 | static noinline_for_stack void |
3073 | ext4_mb_normalize_request(struct ext4_allocation_context *ac, |
3074 | struct ext4_allocation_request *ar) |
3075 | { |
3076 | int bsbits, max; |
3077 | ext4_lblk_t end; |
3078 | loff_t size, orig_size, start_off; |
3079 | ext4_lblk_t start, orig_start; |
3080 | struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); |
3081 | struct ext4_prealloc_space *pa; |
3082 | |
3083 | /* do normalize only data requests, metadata requests |
3084 | do not need preallocation */ |
3085 | if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) |
3086 | return; |
3087 | |
3088 | /* sometime caller may want exact blocks */ |
3089 | if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) |
3090 | return; |
3091 | |
3092 | /* caller may indicate that preallocation isn't |
3093 | * required (it's a tail, for example) */ |
3094 | if (ac->ac_flags & EXT4_MB_HINT_NOPREALLOC) |
3095 | return; |
3096 | |
3097 | if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) { |
3098 | ext4_mb_normalize_group_request(ac); |
3099 | return ; |
3100 | } |
3101 | |
3102 | bsbits = ac->ac_sb->s_blocksize_bits; |
3103 | |
3104 | /* first, let's learn actual file size |
3105 | * given current request is allocated */ |
3106 | size = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len; |
3107 | size = size << bsbits; |
3108 | if (size < i_size_read(ac->ac_inode)) |
3109 | size = i_size_read(ac->ac_inode); |
3110 | |
3111 | /* max size of free chunks */ |
3112 | max = 2 << bsbits; |
3113 | |
3114 | #define NRL_CHECK_SIZE(req, size, max, chunk_size) \ |
3115 | (req <= (size) || max <= (chunk_size)) |
3116 | |
3117 | /* first, try to predict filesize */ |
3118 | /* XXX: should this table be tunable? */ |
3119 | start_off = 0; |
3120 | if (size <= 16 * 1024) { |
3121 | size = 16 * 1024; |
3122 | } else if (size <= 32 * 1024) { |
3123 | size = 32 * 1024; |
3124 | } else if (size <= 64 * 1024) { |
3125 | size = 64 * 1024; |
3126 | } else if (size <= 128 * 1024) { |
3127 | size = 128 * 1024; |
3128 | } else if (size <= 256 * 1024) { |
3129 | size = 256 * 1024; |
3130 | } else if (size <= 512 * 1024) { |
3131 | size = 512 * 1024; |
3132 | } else if (size <= 1024 * 1024) { |
3133 | size = 1024 * 1024; |
3134 | } else if (NRL_CHECK_SIZE(size, 4 * 1024 * 1024, max, 2 * 1024)) { |
3135 | start_off = ((loff_t)ac->ac_o_ex.fe_logical >> |
3136 | (21 - bsbits)) << 21; |
3137 | size = 2 * 1024 * 1024; |
3138 | } else if (NRL_CHECK_SIZE(size, 8 * 1024 * 1024, max, 4 * 1024)) { |
3139 | start_off = ((loff_t)ac->ac_o_ex.fe_logical >> |
3140 | (22 - bsbits)) << 22; |
3141 | size = 4 * 1024 * 1024; |
3142 | } else if (NRL_CHECK_SIZE(ac->ac_o_ex.fe_len, |
3143 | (8<<20)>>bsbits, max, 8 * 1024)) { |
3144 | start_off = ((loff_t)ac->ac_o_ex.fe_logical >> |
3145 | (23 - bsbits)) << 23; |
3146 | size = 8 * 1024 * 1024; |
3147 | } else { |
3148 | start_off = (loff_t)ac->ac_o_ex.fe_logical << bsbits; |
3149 | size = ac->ac_o_ex.fe_len << bsbits; |
3150 | } |
3151 | orig_size = size = size >> bsbits; |
3152 | orig_start = start = start_off >> bsbits; |
3153 | |
3154 | /* don't cover already allocated blocks in selected range */ |
3155 | if (ar->pleft && start <= ar->lleft) { |
3156 | size -= ar->lleft + 1 - start; |
3157 | start = ar->lleft + 1; |
3158 | } |
3159 | if (ar->pright && start + size - 1 >= ar->lright) |
3160 | size -= start + size - ar->lright; |
3161 | |
3162 | end = start + size; |
3163 | |
3164 | /* check we don't cross already preallocated blocks */ |
3165 | rcu_read_lock(); |
3166 | list_for_each_entry_rcu(pa, &ei->i_prealloc_list, pa_inode_list) { |
3167 | ext4_lblk_t pa_end; |
3168 | |
3169 | if (pa->pa_deleted) |
3170 | continue; |
3171 | spin_lock(&pa->pa_lock); |
3172 | if (pa->pa_deleted) { |
3173 | spin_unlock(&pa->pa_lock); |
3174 | continue; |
3175 | } |
3176 | |
3177 | pa_end = pa->pa_lstart + pa->pa_len; |
3178 | |
3179 | /* PA must not overlap original request */ |
3180 | BUG_ON(!(ac->ac_o_ex.fe_logical >= pa_end || |
3181 | ac->ac_o_ex.fe_logical < pa->pa_lstart)); |
3182 | |
3183 | /* skip PA normalized request doesn't overlap with */ |
3184 | if (pa->pa_lstart >= end) { |
3185 | spin_unlock(&pa->pa_lock); |
3186 | continue; |
3187 | } |
3188 | if (pa_end <= start) { |
3189 | spin_unlock(&pa->pa_lock); |
3190 | continue; |
3191 | } |
3192 | BUG_ON(pa->pa_lstart <= start && pa_end >= end); |
3193 | |
3194 | if (pa_end <= ac->ac_o_ex.fe_logical) { |
3195 | BUG_ON(pa_end < start); |
3196 | start = pa_end; |
3197 | } |
3198 | |
3199 | if (pa->pa_lstart > ac->ac_o_ex.fe_logical) { |
3200 | BUG_ON(pa->pa_lstart > end); |
3201 | end = pa->pa_lstart; |
3202 | } |
3203 | spin_unlock(&pa->pa_lock); |
3204 | } |
3205 | rcu_read_unlock(); |
3206 | size = end - start; |
3207 | |
3208 | /* XXX: extra loop to check we really don't overlap preallocations */ |
3209 | rcu_read_lock(); |
3210 | list_for_each_entry_rcu(pa, &ei->i_prealloc_list, pa_inode_list) { |
3211 | ext4_lblk_t pa_end; |
3212 | spin_lock(&pa->pa_lock); |
3213 | if (pa->pa_deleted == 0) { |
3214 | pa_end = pa->pa_lstart + pa->pa_len; |
3215 | BUG_ON(!(start >= pa_end || end <= pa->pa_lstart)); |
3216 | } |
3217 | spin_unlock(&pa->pa_lock); |
3218 | } |
3219 | rcu_read_unlock(); |
3220 | |
3221 | if (start + size <= ac->ac_o_ex.fe_logical && |
3222 | start > ac->ac_o_ex.fe_logical) { |
3223 | printk(KERN_ERR "start %lu, size %lu, fe_logical %lu\n", |
3224 | (unsigned long) start, (unsigned long) size, |
3225 | (unsigned long) ac->ac_o_ex.fe_logical); |
3226 | } |
3227 | BUG_ON(start + size <= ac->ac_o_ex.fe_logical && |
3228 | start > ac->ac_o_ex.fe_logical); |
3229 | BUG_ON(size <= 0 || size > EXT4_BLOCKS_PER_GROUP(ac->ac_sb)); |
3230 | |
3231 | /* now prepare goal request */ |
3232 | |
3233 | /* XXX: is it better to align blocks WRT to logical |
3234 | * placement or satisfy big request as is */ |
3235 | ac->ac_g_ex.fe_logical = start; |
3236 | ac->ac_g_ex.fe_len = size; |
3237 | |
3238 | /* define goal start in order to merge */ |
3239 | if (ar->pright && (ar->lright == (start + size))) { |
3240 | /* merge to the right */ |
3241 | ext4_get_group_no_and_offset(ac->ac_sb, ar->pright - size, |
3242 | &ac->ac_f_ex.fe_group, |
3243 | &ac->ac_f_ex.fe_start); |
3244 | ac->ac_flags |= EXT4_MB_HINT_TRY_GOAL; |
3245 | } |
3246 | if (ar->pleft && (ar->lleft + 1 == start)) { |
3247 | /* merge to the left */ |
3248 | ext4_get_group_no_and_offset(ac->ac_sb, ar->pleft + 1, |
3249 | &ac->ac_f_ex.fe_group, |
3250 | &ac->ac_f_ex.fe_start); |
3251 | ac->ac_flags |= EXT4_MB_HINT_TRY_GOAL; |
3252 | } |
3253 | |
3254 | mb_debug("goal: %u(was %u) blocks at %u\n", (unsigned) size, |
3255 | (unsigned) orig_size, (unsigned) start); |
3256 | } |
3257 | |
3258 | static void ext4_mb_collect_stats(struct ext4_allocation_context *ac) |
3259 | { |
3260 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
3261 | |
3262 | if (sbi->s_mb_stats && ac->ac_g_ex.fe_len > 1) { |
3263 | atomic_inc(&sbi->s_bal_reqs); |
3264 | atomic_add(ac->ac_b_ex.fe_len, &sbi->s_bal_allocated); |
3265 | if (ac->ac_o_ex.fe_len >= ac->ac_g_ex.fe_len) |
3266 | atomic_inc(&sbi->s_bal_success); |
3267 | atomic_add(ac->ac_found, &sbi->s_bal_ex_scanned); |
3268 | if (ac->ac_g_ex.fe_start == ac->ac_b_ex.fe_start && |
3269 | ac->ac_g_ex.fe_group == ac->ac_b_ex.fe_group) |
3270 | atomic_inc(&sbi->s_bal_goals); |
3271 | if (ac->ac_found > sbi->s_mb_max_to_scan) |
3272 | atomic_inc(&sbi->s_bal_breaks); |
3273 | } |
3274 | |
3275 | ext4_mb_store_history(ac); |
3276 | } |
3277 | |
3278 | /* |
3279 | * use blocks preallocated to inode |
3280 | */ |
3281 | static void ext4_mb_use_inode_pa(struct ext4_allocation_context *ac, |
3282 | struct ext4_prealloc_space *pa) |
3283 | { |
3284 | ext4_fsblk_t start; |
3285 | ext4_fsblk_t end; |
3286 | int len; |
3287 | |
3288 | /* found preallocated blocks, use them */ |
3289 | start = pa->pa_pstart + (ac->ac_o_ex.fe_logical - pa->pa_lstart); |
3290 | end = min(pa->pa_pstart + pa->pa_len, start + ac->ac_o_ex.fe_len); |
3291 | len = end - start; |
3292 | ext4_get_group_no_and_offset(ac->ac_sb, start, &ac->ac_b_ex.fe_group, |
3293 | &ac->ac_b_ex.fe_start); |
3294 | ac->ac_b_ex.fe_len = len; |
3295 | ac->ac_status = AC_STATUS_FOUND; |
3296 | ac->ac_pa = pa; |
3297 | |
3298 | BUG_ON(start < pa->pa_pstart); |
3299 | BUG_ON(start + len > pa->pa_pstart + pa->pa_len); |
3300 | BUG_ON(pa->pa_free < len); |
3301 | pa->pa_free -= len; |
3302 | |
3303 | mb_debug("use %llu/%u from inode pa %p\n", start, len, pa); |
3304 | } |
3305 | |
3306 | /* |
3307 | * use blocks preallocated to locality group |
3308 | */ |
3309 | static void ext4_mb_use_group_pa(struct ext4_allocation_context *ac, |
3310 | struct ext4_prealloc_space *pa) |
3311 | { |
3312 | unsigned int len = ac->ac_o_ex.fe_len; |
3313 | |
3314 | ext4_get_group_no_and_offset(ac->ac_sb, pa->pa_pstart, |
3315 | &ac->ac_b_ex.fe_group, |
3316 | &ac->ac_b_ex.fe_start); |
3317 | ac->ac_b_ex.fe_len = len; |
3318 | ac->ac_status = AC_STATUS_FOUND; |
3319 | ac->ac_pa = pa; |
3320 | |
3321 | /* we don't correct pa_pstart or pa_plen here to avoid |
3322 | * possible race when the group is being loaded concurrently |
3323 | * instead we correct pa later, after blocks are marked |
3324 | * in on-disk bitmap -- see ext4_mb_release_context() |
3325 | * Other CPUs are prevented from allocating from this pa by lg_mutex |
3326 | */ |
3327 | mb_debug("use %u/%u from group pa %p\n", pa->pa_lstart-len, len, pa); |
3328 | } |
3329 | |
3330 | /* |
3331 | * Return the prealloc space that have minimal distance |
3332 | * from the goal block. @cpa is the prealloc |
3333 | * space that is having currently known minimal distance |
3334 | * from the goal block. |
3335 | */ |
3336 | static struct ext4_prealloc_space * |
3337 | ext4_mb_check_group_pa(ext4_fsblk_t goal_block, |
3338 | struct ext4_prealloc_space *pa, |
3339 | struct ext4_prealloc_space *cpa) |
3340 | { |
3341 | ext4_fsblk_t cur_distance, new_distance; |
3342 | |
3343 | if (cpa == NULL) { |
3344 | atomic_inc(&pa->pa_count); |
3345 | return pa; |
3346 | } |
3347 | cur_distance = abs(goal_block - cpa->pa_pstart); |
3348 | new_distance = abs(goal_block - pa->pa_pstart); |
3349 | |
3350 | if (cur_distance < new_distance) |
3351 | return cpa; |
3352 | |
3353 | /* drop the previous reference */ |
3354 | atomic_dec(&cpa->pa_count); |
3355 | atomic_inc(&pa->pa_count); |
3356 | return pa; |
3357 | } |
3358 | |
3359 | /* |
3360 | * search goal blocks in preallocated space |
3361 | */ |
3362 | static noinline_for_stack int |
3363 | ext4_mb_use_preallocated(struct ext4_allocation_context *ac) |
3364 | { |
3365 | int order, i; |
3366 | struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); |
3367 | struct ext4_locality_group *lg; |
3368 | struct ext4_prealloc_space *pa, *cpa = NULL; |
3369 | ext4_fsblk_t goal_block; |
3370 | |
3371 | /* only data can be preallocated */ |
3372 | if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) |
3373 | return 0; |
3374 | |
3375 | /* first, try per-file preallocation */ |
3376 | rcu_read_lock(); |
3377 | list_for_each_entry_rcu(pa, &ei->i_prealloc_list, pa_inode_list) { |
3378 | |
3379 | /* all fields in this condition don't change, |
3380 | * so we can skip locking for them */ |
3381 | if (ac->ac_o_ex.fe_logical < pa->pa_lstart || |
3382 | ac->ac_o_ex.fe_logical >= pa->pa_lstart + pa->pa_len) |
3383 | continue; |
3384 | |
3385 | /* found preallocated blocks, use them */ |
3386 | spin_lock(&pa->pa_lock); |
3387 | if (pa->pa_deleted == 0 && pa->pa_free) { |
3388 | atomic_inc(&pa->pa_count); |
3389 | ext4_mb_use_inode_pa(ac, pa); |
3390 | spin_unlock(&pa->pa_lock); |
3391 | ac->ac_criteria = 10; |
3392 | rcu_read_unlock(); |
3393 | return 1; |
3394 | } |
3395 | spin_unlock(&pa->pa_lock); |
3396 | } |
3397 | rcu_read_unlock(); |
3398 | |
3399 | /* can we use group allocation? */ |
3400 | if (!(ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC)) |
3401 | return 0; |
3402 | |
3403 | /* inode may have no locality group for some reason */ |
3404 | lg = ac->ac_lg; |
3405 | if (lg == NULL) |
3406 | return 0; |
3407 | order = fls(ac->ac_o_ex.fe_len) - 1; |
3408 | if (order > PREALLOC_TB_SIZE - 1) |
3409 | /* The max size of hash table is PREALLOC_TB_SIZE */ |
3410 | order = PREALLOC_TB_SIZE - 1; |
3411 | |
3412 | goal_block = ac->ac_g_ex.fe_group * EXT4_BLOCKS_PER_GROUP(ac->ac_sb) + |
3413 | ac->ac_g_ex.fe_start + |
3414 | le32_to_cpu(EXT4_SB(ac->ac_sb)->s_es->s_first_data_block); |
3415 | /* |
3416 | * search for the prealloc space that is having |
3417 | * minimal distance from the goal block. |
3418 | */ |
3419 | for (i = order; i < PREALLOC_TB_SIZE; i++) { |
3420 | rcu_read_lock(); |
3421 | list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[i], |
3422 | pa_inode_list) { |
3423 | spin_lock(&pa->pa_lock); |
3424 | if (pa->pa_deleted == 0 && |
3425 | pa->pa_free >= ac->ac_o_ex.fe_len) { |
3426 | |
3427 | cpa = ext4_mb_check_group_pa(goal_block, |
3428 | pa, cpa); |
3429 | } |
3430 | spin_unlock(&pa->pa_lock); |
3431 | } |
3432 | rcu_read_unlock(); |
3433 | } |
3434 | if (cpa) { |
3435 | ext4_mb_use_group_pa(ac, cpa); |
3436 | ac->ac_criteria = 20; |
3437 | return 1; |
3438 | } |
3439 | return 0; |
3440 | } |
3441 | |
3442 | /* |
3443 | * the function goes through all block freed in the group |
3444 | * but not yet committed and marks them used in in-core bitmap. |
3445 | * buddy must be generated from this bitmap |
3446 | * Need to be called with the ext4 group lock held |
3447 | */ |
3448 | static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, |
3449 | ext4_group_t group) |
3450 | { |
3451 | struct rb_node *n; |
3452 | struct ext4_group_info *grp; |
3453 | struct ext4_free_data *entry; |
3454 | |
3455 | grp = ext4_get_group_info(sb, group); |
3456 | n = rb_first(&(grp->bb_free_root)); |
3457 | |
3458 | while (n) { |
3459 | entry = rb_entry(n, struct ext4_free_data, node); |
3460 | mb_set_bits(bitmap, entry->start_blk, entry->count); |
3461 | n = rb_next(n); |
3462 | } |
3463 | return; |
3464 | } |
3465 | |
3466 | /* |
3467 | * the function goes through all preallocation in this group and marks them |
3468 | * used in in-core bitmap. buddy must be generated from this bitmap |
3469 | * Need to be called with ext4 group lock held |
3470 | */ |
3471 | static noinline_for_stack |
3472 | void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, |
3473 | ext4_group_t group) |
3474 | { |
3475 | struct ext4_group_info *grp = ext4_get_group_info(sb, group); |
3476 | struct ext4_prealloc_space *pa; |
3477 | struct list_head *cur; |
3478 | ext4_group_t groupnr; |
3479 | ext4_grpblk_t start; |
3480 | int preallocated = 0; |
3481 | int count = 0; |
3482 | int len; |
3483 | |
3484 | /* all form of preallocation discards first load group, |
3485 | * so the only competing code is preallocation use. |
3486 | * we don't need any locking here |
3487 | * notice we do NOT ignore preallocations with pa_deleted |
3488 | * otherwise we could leave used blocks available for |
3489 | * allocation in buddy when concurrent ext4_mb_put_pa() |
3490 | * is dropping preallocation |
3491 | */ |
3492 | list_for_each(cur, &grp->bb_prealloc_list) { |
3493 | pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); |
3494 | spin_lock(&pa->pa_lock); |
3495 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, |
3496 | &groupnr, &start); |
3497 | len = pa->pa_len; |
3498 | spin_unlock(&pa->pa_lock); |
3499 | if (unlikely(len == 0)) |
3500 | continue; |
3501 | BUG_ON(groupnr != group); |
3502 | mb_set_bits(bitmap, start, len); |
3503 | preallocated += len; |
3504 | count++; |
3505 | } |
3506 | mb_debug("prellocated %u for group %u\n", preallocated, group); |
3507 | } |
3508 | |
3509 | static void ext4_mb_pa_callback(struct rcu_head *head) |
3510 | { |
3511 | struct ext4_prealloc_space *pa; |
3512 | pa = container_of(head, struct ext4_prealloc_space, u.pa_rcu); |
3513 | kmem_cache_free(ext4_pspace_cachep, pa); |
3514 | } |
3515 | |
3516 | /* |
3517 | * drops a reference to preallocated space descriptor |
3518 | * if this was the last reference and the space is consumed |
3519 | */ |
3520 | static void ext4_mb_put_pa(struct ext4_allocation_context *ac, |
3521 | struct super_block *sb, struct ext4_prealloc_space *pa) |
3522 | { |
3523 | ext4_group_t grp; |
3524 | ext4_fsblk_t grp_blk; |
3525 | |
3526 | if (!atomic_dec_and_test(&pa->pa_count) || pa->pa_free != 0) |
3527 | return; |
3528 | |
3529 | /* in this short window concurrent discard can set pa_deleted */ |
3530 | spin_lock(&pa->pa_lock); |
3531 | if (pa->pa_deleted == 1) { |
3532 | spin_unlock(&pa->pa_lock); |
3533 | return; |
3534 | } |
3535 | |
3536 | pa->pa_deleted = 1; |
3537 | spin_unlock(&pa->pa_lock); |
3538 | |
3539 | grp_blk = pa->pa_pstart; |
3540 | /* |
3541 | * If doing group-based preallocation, pa_pstart may be in the |
3542 | * next group when pa is used up |
3543 | */ |
3544 | if (pa->pa_type == MB_GROUP_PA) |
3545 | grp_blk--; |
3546 | |
3547 | ext4_get_group_no_and_offset(sb, grp_blk, &grp, NULL); |
3548 | |
3549 | /* |
3550 | * possible race: |
3551 | * |
3552 | * P1 (buddy init) P2 (regular allocation) |
3553 | * find block B in PA |
3554 | * copy on-disk bitmap to buddy |
3555 | * mark B in on-disk bitmap |
3556 | * drop PA from group |
3557 | * mark all PAs in buddy |
3558 | * |
3559 | * thus, P1 initializes buddy with B available. to prevent this |
3560 | * we make "copy" and "mark all PAs" atomic and serialize "drop PA" |
3561 | * against that pair |
3562 | */ |
3563 | ext4_lock_group(sb, grp); |
3564 | list_del(&pa->pa_group_list); |
3565 | ext4_unlock_group(sb, grp); |
3566 | |
3567 | spin_lock(pa->pa_obj_lock); |
3568 | list_del_rcu(&pa->pa_inode_list); |
3569 | spin_unlock(pa->pa_obj_lock); |
3570 | |
3571 | call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); |
3572 | } |
3573 | |
3574 | /* |
3575 | * creates new preallocated space for given inode |
3576 | */ |
3577 | static noinline_for_stack int |
3578 | ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) |
3579 | { |
3580 | struct super_block *sb = ac->ac_sb; |
3581 | struct ext4_prealloc_space *pa; |
3582 | struct ext4_group_info *grp; |
3583 | struct ext4_inode_info *ei; |
3584 | |
3585 | /* preallocate only when found space is larger then requested */ |
3586 | BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len); |
3587 | BUG_ON(ac->ac_status != AC_STATUS_FOUND); |
3588 | BUG_ON(!S_ISREG(ac->ac_inode->i_mode)); |
3589 | |
3590 | pa = kmem_cache_alloc(ext4_pspace_cachep, GFP_NOFS); |
3591 | if (pa == NULL) |
3592 | return -ENOMEM; |
3593 | |
3594 | if (ac->ac_b_ex.fe_len < ac->ac_g_ex.fe_len) { |
3595 | int winl; |
3596 | int wins; |
3597 | int win; |
3598 | int offs; |
3599 | |
3600 | /* we can't allocate as much as normalizer wants. |
3601 | * so, found space must get proper lstart |
3602 | * to cover original request */ |
3603 | BUG_ON(ac->ac_g_ex.fe_logical > ac->ac_o_ex.fe_logical); |
3604 | BUG_ON(ac->ac_g_ex.fe_len < ac->ac_o_ex.fe_len); |
3605 | |
3606 | /* we're limited by original request in that |
3607 | * logical block must be covered any way |
3608 | * winl is window we can move our chunk within */ |
3609 | winl = ac->ac_o_ex.fe_logical - ac->ac_g_ex.fe_logical; |
3610 | |
3611 | /* also, we should cover whole original request */ |
3612 | wins = ac->ac_b_ex.fe_len - ac->ac_o_ex.fe_len; |
3613 | |
3614 | /* the smallest one defines real window */ |
3615 | win = min(winl, wins); |
3616 | |
3617 | offs = ac->ac_o_ex.fe_logical % ac->ac_b_ex.fe_len; |
3618 | if (offs && offs < win) |
3619 | win = offs; |
3620 | |
3621 | ac->ac_b_ex.fe_logical = ac->ac_o_ex.fe_logical - win; |
3622 | BUG_ON(ac->ac_o_ex.fe_logical < ac->ac_b_ex.fe_logical); |
3623 | BUG_ON(ac->ac_o_ex.fe_len > ac->ac_b_ex.fe_len); |
3624 | } |
3625 | |
3626 | /* preallocation can change ac_b_ex, thus we store actually |
3627 | * allocated blocks for history */ |
3628 | ac->ac_f_ex = ac->ac_b_ex; |
3629 | |
3630 | pa->pa_lstart = ac->ac_b_ex.fe_logical; |
3631 | pa->pa_pstart = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); |
3632 | pa->pa_len = ac->ac_b_ex.fe_len; |
3633 | pa->pa_free = pa->pa_len; |
3634 | atomic_set(&pa->pa_count, 1); |
3635 | spin_lock_init(&pa->pa_lock); |
3636 | INIT_LIST_HEAD(&pa->pa_inode_list); |
3637 | INIT_LIST_HEAD(&pa->pa_group_list); |
3638 | pa->pa_deleted = 0; |
3639 | pa->pa_type = MB_INODE_PA; |
3640 | |
3641 | mb_debug("new inode pa %p: %llu/%u for %u\n", pa, |
3642 | pa->pa_pstart, pa->pa_len, pa->pa_lstart); |
3643 | trace_ext4_mb_new_inode_pa(ac, pa); |
3644 | |
3645 | ext4_mb_use_inode_pa(ac, pa); |
3646 | atomic_add(pa->pa_free, &EXT4_SB(sb)->s_mb_preallocated); |
3647 | |
3648 | ei = EXT4_I(ac->ac_inode); |
3649 | grp = ext4_get_group_info(sb, ac->ac_b_ex.fe_group); |
3650 | |
3651 | pa->pa_obj_lock = &ei->i_prealloc_lock; |
3652 | pa->pa_inode = ac->ac_inode; |
3653 | |
3654 | ext4_lock_group(sb, ac->ac_b_ex.fe_group); |
3655 | list_add(&pa->pa_group_list, &grp->bb_prealloc_list); |
3656 | ext4_unlock_group(sb, ac->ac_b_ex.fe_group); |
3657 | |
3658 | spin_lock(pa->pa_obj_lock); |
3659 | list_add_rcu(&pa->pa_inode_list, &ei->i_prealloc_list); |
3660 | spin_unlock(pa->pa_obj_lock); |
3661 | |
3662 | return 0; |
3663 | } |
3664 | |
3665 | /* |
3666 | * creates new preallocated space for locality group inodes belongs to |
3667 | */ |
3668 | static noinline_for_stack int |
3669 | ext4_mb_new_group_pa(struct ext4_allocation_context *ac) |
3670 | { |
3671 | struct super_block *sb = ac->ac_sb; |
3672 | struct ext4_locality_group *lg; |
3673 | struct ext4_prealloc_space *pa; |
3674 | struct ext4_group_info *grp; |
3675 | |
3676 | /* preallocate only when found space is larger then requested */ |
3677 | BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len); |
3678 | BUG_ON(ac->ac_status != AC_STATUS_FOUND); |
3679 | BUG_ON(!S_ISREG(ac->ac_inode->i_mode)); |
3680 | |
3681 | BUG_ON(ext4_pspace_cachep == NULL); |
3682 | pa = kmem_cache_alloc(ext4_pspace_cachep, GFP_NOFS); |
3683 | if (pa == NULL) |
3684 | return -ENOMEM; |
3685 | |
3686 | /* preallocation can change ac_b_ex, thus we store actually |
3687 | * allocated blocks for history */ |
3688 | ac->ac_f_ex = ac->ac_b_ex; |
3689 | |
3690 | pa->pa_pstart = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); |
3691 | pa->pa_lstart = pa->pa_pstart; |
3692 | pa->pa_len = ac->ac_b_ex.fe_len; |
3693 | pa->pa_free = pa->pa_len; |
3694 | atomic_set(&pa->pa_count, 1); |
3695 | spin_lock_init(&pa->pa_lock); |
3696 | INIT_LIST_HEAD(&pa->pa_inode_list); |
3697 | INIT_LIST_HEAD(&pa->pa_group_list); |
3698 | pa->pa_deleted = 0; |
3699 | pa->pa_type = MB_GROUP_PA; |
3700 | |
3701 | mb_debug("new group pa %p: %llu/%u for %u\n", pa, |
3702 | pa->pa_pstart, pa->pa_len, pa->pa_lstart); |
3703 | trace_ext4_mb_new_group_pa(ac, pa); |
3704 | |
3705 | ext4_mb_use_group_pa(ac, pa); |
3706 | atomic_add(pa->pa_free, &EXT4_SB(sb)->s_mb_preallocated); |
3707 | |
3708 | grp = ext4_get_group_info(sb, ac->ac_b_ex.fe_group); |
3709 | lg = ac->ac_lg; |
3710 | BUG_ON(lg == NULL); |
3711 | |
3712 | pa->pa_obj_lock = &lg->lg_prealloc_lock; |
3713 | pa->pa_inode = NULL; |
3714 | |
3715 | ext4_lock_group(sb, ac->ac_b_ex.fe_group); |
3716 | list_add(&pa->pa_group_list, &grp->bb_prealloc_list); |
3717 | ext4_unlock_group(sb, ac->ac_b_ex.fe_group); |
3718 | |
3719 | /* |
3720 | * We will later add the new pa to the right bucket |
3721 | * after updating the pa_free in ext4_mb_release_context |
3722 | */ |
3723 | return 0; |
3724 | } |
3725 | |
3726 | static int ext4_mb_new_preallocation(struct ext4_allocation_context *ac) |
3727 | { |
3728 | int err; |
3729 | |
3730 | if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) |
3731 | err = ext4_mb_new_group_pa(ac); |
3732 | else |
3733 | err = ext4_mb_new_inode_pa(ac); |
3734 | return err; |
3735 | } |
3736 | |
3737 | /* |
3738 | * finds all unused blocks in on-disk bitmap, frees them in |
3739 | * in-core bitmap and buddy. |
3740 | * @pa must be unlinked from inode and group lists, so that |
3741 | * nobody else can find/use it. |
3742 | * the caller MUST hold group/inode locks. |
3743 | * TODO: optimize the case when there are no in-core structures yet |
3744 | */ |
3745 | static noinline_for_stack int |
3746 | ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, |
3747 | struct ext4_prealloc_space *pa, |
3748 | struct ext4_allocation_context *ac) |
3749 | { |
3750 | struct super_block *sb = e4b->bd_sb; |
3751 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
3752 | unsigned int end; |
3753 | unsigned int next; |
3754 | ext4_group_t group; |
3755 | ext4_grpblk_t bit; |
3756 | unsigned long long grp_blk_start; |
3757 | sector_t start; |
3758 | int err = 0; |
3759 | int free = 0; |
3760 | |
3761 | BUG_ON(pa->pa_deleted == 0); |
3762 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); |
3763 | grp_blk_start = pa->pa_pstart - bit; |
3764 | BUG_ON(group != e4b->bd_group && pa->pa_len != 0); |
3765 | end = bit + pa->pa_len; |
3766 | |
3767 | if (ac) { |
3768 | ac->ac_sb = sb; |
3769 | ac->ac_inode = pa->pa_inode; |
3770 | ac->ac_op = EXT4_MB_HISTORY_DISCARD; |
3771 | } |
3772 | |
3773 | while (bit < end) { |
3774 | bit = mb_find_next_zero_bit(bitmap_bh->b_data, end, bit); |
3775 | if (bit >= end) |
3776 | break; |
3777 | next = mb_find_next_bit(bitmap_bh->b_data, end, bit); |
3778 | start = group * EXT4_BLOCKS_PER_GROUP(sb) + bit + |
3779 | le32_to_cpu(sbi->s_es->s_first_data_block); |
3780 | mb_debug(" free preallocated %u/%u in group %u\n", |
3781 | (unsigned) start, (unsigned) next - bit, |
3782 | (unsigned) group); |
3783 | free += next - bit; |
3784 | |
3785 | if (ac) { |
3786 | ac->ac_b_ex.fe_group = group; |
3787 | ac->ac_b_ex.fe_start = bit; |
3788 | ac->ac_b_ex.fe_len = next - bit; |
3789 | ac->ac_b_ex.fe_logical = 0; |
3790 | ext4_mb_store_history(ac); |
3791 | } |
3792 | |
3793 | trace_ext4_mb_release_inode_pa(ac, pa, grp_blk_start + bit, |
3794 | next - bit); |
3795 | mb_free_blocks(pa->pa_inode, e4b, bit, next - bit); |
3796 | bit = next + 1; |
3797 | } |
3798 | if (free != pa->pa_free) { |
3799 | printk(KERN_CRIT "pa %p: logic %lu, phys. %lu, len %lu\n", |
3800 | pa, (unsigned long) pa->pa_lstart, |
3801 | (unsigned long) pa->pa_pstart, |
3802 | (unsigned long) pa->pa_len); |
3803 | ext4_grp_locked_error(sb, group, |
3804 | __func__, "free %u, pa_free %u", |
3805 | free, pa->pa_free); |
3806 | /* |
3807 | * pa is already deleted so we use the value obtained |
3808 | * from the bitmap and continue. |
3809 | */ |
3810 | } |
3811 | atomic_add(free, &sbi->s_mb_discarded); |
3812 | |
3813 | return err; |
3814 | } |
3815 | |
3816 | static noinline_for_stack int |
3817 | ext4_mb_release_group_pa(struct ext4_buddy *e4b, |
3818 | struct ext4_prealloc_space *pa, |
3819 | struct ext4_allocation_context *ac) |
3820 | { |
3821 | struct super_block *sb = e4b->bd_sb; |
3822 | ext4_group_t group; |
3823 | ext4_grpblk_t bit; |
3824 | |
3825 | if (ac) |
3826 | ac->ac_op = EXT4_MB_HISTORY_DISCARD; |
3827 | |
3828 | trace_ext4_mb_release_group_pa(ac, pa); |
3829 | BUG_ON(pa->pa_deleted == 0); |
3830 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); |
3831 | BUG_ON(group != e4b->bd_group && pa->pa_len != 0); |
3832 | mb_free_blocks(pa->pa_inode, e4b, bit, pa->pa_len); |
3833 | atomic_add(pa->pa_len, &EXT4_SB(sb)->s_mb_discarded); |
3834 | |
3835 | if (ac) { |
3836 | ac->ac_sb = sb; |
3837 | ac->ac_inode = NULL; |
3838 | ac->ac_b_ex.fe_group = group; |
3839 | ac->ac_b_ex.fe_start = bit; |
3840 | ac->ac_b_ex.fe_len = pa->pa_len; |
3841 | ac->ac_b_ex.fe_logical = 0; |
3842 | ext4_mb_store_history(ac); |
3843 | } |
3844 | |
3845 | return 0; |
3846 | } |
3847 | |
3848 | /* |
3849 | * releases all preallocations in given group |
3850 | * |
3851 | * first, we need to decide discard policy: |
3852 | * - when do we discard |
3853 | * 1) ENOSPC |
3854 | * - how many do we discard |
3855 | * 1) how many requested |
3856 | */ |
3857 | static noinline_for_stack int |
3858 | ext4_mb_discard_group_preallocations(struct super_block *sb, |
3859 | ext4_group_t group, int needed) |
3860 | { |
3861 | struct ext4_group_info *grp = ext4_get_group_info(sb, group); |
3862 | struct buffer_head *bitmap_bh = NULL; |
3863 | struct ext4_prealloc_space *pa, *tmp; |
3864 | struct ext4_allocation_context *ac; |
3865 | struct list_head list; |
3866 | struct ext4_buddy e4b; |
3867 | int err; |
3868 | int busy = 0; |
3869 | int free = 0; |
3870 | |
3871 | mb_debug("discard preallocation for group %u\n", group); |
3872 | |
3873 | if (list_empty(&grp->bb_prealloc_list)) |
3874 | return 0; |
3875 | |
3876 | bitmap_bh = ext4_read_block_bitmap(sb, group); |
3877 | if (bitmap_bh == NULL) { |
3878 | ext4_error(sb, __func__, "Error in reading block " |
3879 | "bitmap for %u", group); |
3880 | return 0; |
3881 | } |
3882 | |
3883 | err = ext4_mb_load_buddy(sb, group, &e4b); |
3884 | if (err) { |
3885 | ext4_error(sb, __func__, "Error in loading buddy " |
3886 | "information for %u", group); |
3887 | put_bh(bitmap_bh); |
3888 | return 0; |
3889 | } |
3890 | |
3891 | if (needed == 0) |
3892 | needed = EXT4_BLOCKS_PER_GROUP(sb) + 1; |
3893 | |
3894 | INIT_LIST_HEAD(&list); |
3895 | ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); |
3896 | if (ac) |
3897 | ac->ac_sb = sb; |
3898 | repeat: |
3899 | ext4_lock_group(sb, group); |
3900 | list_for_each_entry_safe(pa, tmp, |
3901 | &grp->bb_prealloc_list, pa_group_list) { |
3902 | spin_lock(&pa->pa_lock); |
3903 | if (atomic_read(&pa->pa_count)) { |
3904 | spin_unlock(&pa->pa_lock); |
3905 | busy = 1; |
3906 | continue; |
3907 | } |
3908 | if (pa->pa_deleted) { |
3909 | spin_unlock(&pa->pa_lock); |
3910 | continue; |
3911 | } |
3912 | |
3913 | /* seems this one can be freed ... */ |
3914 | pa->pa_deleted = 1; |
3915 | |
3916 | /* we can trust pa_free ... */ |
3917 | free += pa->pa_free; |
3918 | |
3919 | spin_unlock(&pa->pa_lock); |
3920 | |
3921 | list_del(&pa->pa_group_list); |
3922 | list_add(&pa->u.pa_tmp_list, &list); |
3923 | } |
3924 | |
3925 | /* if we still need more blocks and some PAs were used, try again */ |
3926 | if (free < needed && busy) { |
3927 | busy = 0; |
3928 | ext4_unlock_group(sb, group); |
3929 | /* |
3930 | * Yield the CPU here so that we don't get soft lockup |
3931 | * in non preempt case. |
3932 | */ |
3933 | yield(); |
3934 | goto repeat; |
3935 | } |
3936 | |
3937 | /* found anything to free? */ |
3938 | if (list_empty(&list)) { |
3939 | BUG_ON(free != 0); |
3940 | goto out; |
3941 | } |
3942 | |
3943 | /* now free all selected PAs */ |
3944 | list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) { |
3945 | |
3946 | /* remove from object (inode or locality group) */ |
3947 | spin_lock(pa->pa_obj_lock); |
3948 | list_del_rcu(&pa->pa_inode_list); |
3949 | spin_unlock(pa->pa_obj_lock); |
3950 | |
3951 | if (pa->pa_type == MB_GROUP_PA) |
3952 | ext4_mb_release_group_pa(&e4b, pa, ac); |
3953 | else |
3954 | ext4_mb_release_inode_pa(&e4b, bitmap_bh, pa, ac); |
3955 | |
3956 | list_del(&pa->u.pa_tmp_list); |
3957 | call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); |
3958 | } |
3959 | |
3960 | out: |
3961 | ext4_unlock_group(sb, group); |
3962 | if (ac) |
3963 | kmem_cache_free(ext4_ac_cachep, ac); |
3964 | ext4_mb_release_desc(&e4b); |
3965 | put_bh(bitmap_bh); |
3966 | return free; |
3967 | } |
3968 | |
3969 | /* |
3970 | * releases all non-used preallocated blocks for given inode |
3971 | * |
3972 | * It's important to discard preallocations under i_data_sem |
3973 | * We don't want another block to be served from the prealloc |
3974 | * space when we are discarding the inode prealloc space. |
3975 | * |
3976 | * FIXME!! Make sure it is valid at all the call sites |
3977 | */ |
3978 | void ext4_discard_preallocations(struct inode *inode) |
3979 | { |
3980 | struct ext4_inode_info *ei = EXT4_I(inode); |
3981 | struct super_block *sb = inode->i_sb; |
3982 | struct buffer_head *bitmap_bh = NULL; |
3983 | struct ext4_prealloc_space *pa, *tmp; |
3984 | struct ext4_allocation_context *ac; |
3985 | ext4_group_t group = 0; |
3986 | struct list_head list; |
3987 | struct ext4_buddy e4b; |
3988 | int err; |
3989 | |
3990 | if (!S_ISREG(inode->i_mode)) { |
3991 | /*BUG_ON(!list_empty(&ei->i_prealloc_list));*/ |
3992 | return; |
3993 | } |
3994 | |
3995 | mb_debug("discard preallocation for inode %lu\n", inode->i_ino); |
3996 | trace_ext4_discard_preallocations(inode); |
3997 | |
3998 | INIT_LIST_HEAD(&list); |
3999 | |
4000 | ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); |
4001 | if (ac) { |
4002 | ac->ac_sb = sb; |
4003 | ac->ac_inode = inode; |
4004 | } |
4005 | repeat: |
4006 | /* first, collect all pa's in the inode */ |
4007 | spin_lock(&ei->i_prealloc_lock); |
4008 | while (!list_empty(&ei->i_prealloc_list)) { |
4009 | pa = list_entry(ei->i_prealloc_list.next, |
4010 | struct ext4_prealloc_space, pa_inode_list); |
4011 | BUG_ON(pa->pa_obj_lock != &ei->i_prealloc_lock); |
4012 | spin_lock(&pa->pa_lock); |
4013 | if (atomic_read(&pa->pa_count)) { |
4014 | /* this shouldn't happen often - nobody should |
4015 | * use preallocation while we're discarding it */ |
4016 | spin_unlock(&pa->pa_lock); |
4017 | spin_unlock(&ei->i_prealloc_lock); |
4018 | printk(KERN_ERR "uh-oh! used pa while discarding\n"); |
4019 | WARN_ON(1); |
4020 | schedule_timeout_uninterruptible(HZ); |
4021 | goto repeat; |
4022 | |
4023 | } |
4024 | if (pa->pa_deleted == 0) { |
4025 | pa->pa_deleted = 1; |
4026 | spin_unlock(&pa->pa_lock); |
4027 | list_del_rcu(&pa->pa_inode_list); |
4028 | list_add(&pa->u.pa_tmp_list, &list); |
4029 | continue; |
4030 | } |
4031 | |
4032 | /* someone is deleting pa right now */ |
4033 | spin_unlock(&pa->pa_lock); |
4034 | spin_unlock(&ei->i_prealloc_lock); |
4035 | |
4036 | /* we have to wait here because pa_deleted |
4037 | * doesn't mean pa is already unlinked from |
4038 | * the list. as we might be called from |
4039 | * ->clear_inode() the inode will get freed |
4040 | * and concurrent thread which is unlinking |
4041 | * pa from inode's list may access already |
4042 | * freed memory, bad-bad-bad */ |
4043 | |
4044 | /* XXX: if this happens too often, we can |
4045 | * add a flag to force wait only in case |
4046 | * of ->clear_inode(), but not in case of |
4047 | * regular truncate */ |
4048 | schedule_timeout_uninterruptible(HZ); |
4049 | goto repeat; |
4050 | } |
4051 | spin_unlock(&ei->i_prealloc_lock); |
4052 | |
4053 | list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) { |
4054 | BUG_ON(pa->pa_type != MB_INODE_PA); |
4055 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL); |
4056 | |
4057 | err = ext4_mb_load_buddy(sb, group, &e4b); |
4058 | if (err) { |
4059 | ext4_error(sb, __func__, "Error in loading buddy " |
4060 | "information for %u", group); |
4061 | continue; |
4062 | } |
4063 | |
4064 | bitmap_bh = ext4_read_block_bitmap(sb, group); |
4065 | if (bitmap_bh == NULL) { |
4066 | ext4_error(sb, __func__, "Error in reading block " |
4067 | "bitmap for %u", group); |
4068 | ext4_mb_release_desc(&e4b); |
4069 | continue; |
4070 | } |
4071 | |
4072 | ext4_lock_group(sb, group); |
4073 | list_del(&pa->pa_group_list); |
4074 | ext4_mb_release_inode_pa(&e4b, bitmap_bh, pa, ac); |
4075 | ext4_unlock_group(sb, group); |
4076 | |
4077 | ext4_mb_release_desc(&e4b); |
4078 | put_bh(bitmap_bh); |
4079 | |
4080 | list_del(&pa->u.pa_tmp_list); |
4081 | call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); |
4082 | } |
4083 | if (ac) |
4084 | kmem_cache_free(ext4_ac_cachep, ac); |
4085 | } |
4086 | |
4087 | /* |
4088 | * finds all preallocated spaces and return blocks being freed to them |
4089 | * if preallocated space becomes full (no block is used from the space) |
4090 | * then the function frees space in buddy |
4091 | * XXX: at the moment, truncate (which is the only way to free blocks) |
4092 | * discards all preallocations |
4093 | */ |
4094 | static void ext4_mb_return_to_preallocation(struct inode *inode, |
4095 | struct ext4_buddy *e4b, |
4096 | sector_t block, int count) |
4097 | { |
4098 | BUG_ON(!list_empty(&EXT4_I(inode)->i_prealloc_list)); |
4099 | } |
4100 | #ifdef MB_DEBUG |
4101 | static void ext4_mb_show_ac(struct ext4_allocation_context *ac) |
4102 | { |
4103 | struct super_block *sb = ac->ac_sb; |
4104 | ext4_group_t ngroups, i; |
4105 | |
4106 | printk(KERN_ERR "EXT4-fs: Can't allocate:" |
4107 | " Allocation context details:\n"); |
4108 | printk(KERN_ERR "EXT4-fs: status %d flags %d\n", |
4109 | ac->ac_status, ac->ac_flags); |
4110 | printk(KERN_ERR "EXT4-fs: orig %lu/%lu/%lu@%lu, goal %lu/%lu/%lu@%lu, " |
4111 | "best %lu/%lu/%lu@%lu cr %d\n", |
4112 | (unsigned long)ac->ac_o_ex.fe_group, |
4113 | (unsigned long)ac->ac_o_ex.fe_start, |
4114 | (unsigned long)ac->ac_o_ex.fe_len, |
4115 | (unsigned long)ac->ac_o_ex.fe_logical, |
4116 | (unsigned long)ac->ac_g_ex.fe_group, |
4117 | (unsigned long)ac->ac_g_ex.fe_start, |
4118 | (unsigned long)ac->ac_g_ex.fe_len, |
4119 | (unsigned long)ac->ac_g_ex.fe_logical, |
4120 | (unsigned long)ac->ac_b_ex.fe_group, |
4121 | (unsigned long)ac->ac_b_ex.fe_start, |
4122 | (unsigned long)ac->ac_b_ex.fe_len, |
4123 | (unsigned long)ac->ac_b_ex.fe_logical, |
4124 | (int)ac->ac_criteria); |
4125 | printk(KERN_ERR "EXT4-fs: %lu scanned, %d found\n", ac->ac_ex_scanned, |
4126 | ac->ac_found); |
4127 | printk(KERN_ERR "EXT4-fs: groups: \n"); |
4128 | ngroups = ext4_get_groups_count(sb); |
4129 | for (i = 0; i < ngroups; i++) { |
4130 | struct ext4_group_info *grp = ext4_get_group_info(sb, i); |
4131 | struct ext4_prealloc_space *pa; |
4132 | ext4_grpblk_t start; |
4133 | struct list_head *cur; |
4134 | ext4_lock_group(sb, i); |
4135 | list_for_each(cur, &grp->bb_prealloc_list) { |
4136 | pa = list_entry(cur, struct ext4_prealloc_space, |
4137 | pa_group_list); |
4138 | spin_lock(&pa->pa_lock); |
4139 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, |
4140 | NULL, &start); |
4141 | spin_unlock(&pa->pa_lock); |
4142 | printk(KERN_ERR "PA:%lu:%d:%u \n", i, |
4143 | start, pa->pa_len); |
4144 | } |
4145 | ext4_unlock_group(sb, i); |
4146 | |
4147 | if (grp->bb_free == 0) |
4148 | continue; |
4149 | printk(KERN_ERR "%lu: %d/%d \n", |
4150 | i, grp->bb_free, grp->bb_fragments); |
4151 | } |
4152 | printk(KERN_ERR "\n"); |
4153 | } |
4154 | #else |
4155 | static inline void ext4_mb_show_ac(struct ext4_allocation_context *ac) |
4156 | { |
4157 | return; |
4158 | } |
4159 | #endif |
4160 | |
4161 | /* |
4162 | * We use locality group preallocation for small size file. The size of the |
4163 | * file is determined by the current size or the resulting size after |
4164 | * allocation which ever is larger |
4165 | * |
4166 | * One can tune this size via /sys/fs/ext4/<partition>/mb_stream_req |
4167 | */ |
4168 | static void ext4_mb_group_or_file(struct ext4_allocation_context *ac) |
4169 | { |
4170 | struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); |
4171 | int bsbits = ac->ac_sb->s_blocksize_bits; |
4172 | loff_t size, isize; |
4173 | |
4174 | if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) |
4175 | return; |
4176 | |
4177 | size = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len; |
4178 | isize = i_size_read(ac->ac_inode) >> bsbits; |
4179 | size = max(size, isize); |
4180 | |
4181 | /* don't use group allocation for large files */ |
4182 | if (size >= sbi->s_mb_stream_request) |
4183 | return; |
4184 | |
4185 | if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) |
4186 | return; |
4187 | |
4188 | BUG_ON(ac->ac_lg != NULL); |
4189 | /* |
4190 | * locality group prealloc space are per cpu. The reason for having |
4191 | * per cpu locality group is to reduce the contention between block |
4192 | * request from multiple CPUs. |
4193 | */ |
4194 | ac->ac_lg = per_cpu_ptr(sbi->s_locality_groups, raw_smp_processor_id()); |
4195 | |
4196 | /* we're going to use group allocation */ |
4197 | ac->ac_flags |= EXT4_MB_HINT_GROUP_ALLOC; |
4198 | |
4199 | /* serialize all allocations in the group */ |
4200 | mutex_lock(&ac->ac_lg->lg_mutex); |
4201 | } |
4202 | |
4203 | static noinline_for_stack int |
4204 | ext4_mb_initialize_context(struct ext4_allocation_context *ac, |
4205 | struct ext4_allocation_request *ar) |
4206 | { |
4207 | struct super_block *sb = ar->inode->i_sb; |
4208 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
4209 | struct ext4_super_block *es = sbi->s_es; |
4210 | ext4_group_t group; |
4211 | unsigned int len; |
4212 | ext4_fsblk_t goal; |
4213 | ext4_grpblk_t block; |
4214 | |
4215 | /* we can't allocate > group size */ |
4216 | len = ar->len; |
4217 | |
4218 | /* just a dirty hack to filter too big requests */ |
4219 | if (len >= EXT4_BLOCKS_PER_GROUP(sb) - 10) |
4220 | len = EXT4_BLOCKS_PER_GROUP(sb) - 10; |
4221 | |
4222 | /* start searching from the goal */ |
4223 | goal = ar->goal; |
4224 | if (goal < le32_to_cpu(es->s_first_data_block) || |
4225 | goal >= ext4_blocks_count(es)) |
4226 | goal = le32_to_cpu(es->s_first_data_block); |
4227 | ext4_get_group_no_and_offset(sb, goal, &group, &block); |
4228 | |
4229 | /* set up allocation goals */ |
4230 | memset(ac, 0, sizeof(struct ext4_allocation_context)); |
4231 | ac->ac_b_ex.fe_logical = ar->logical; |
4232 | ac->ac_status = AC_STATUS_CONTINUE; |
4233 | ac->ac_sb = sb; |
4234 | ac->ac_inode = ar->inode; |
4235 | ac->ac_o_ex.fe_logical = ar->logical; |
4236 | ac->ac_o_ex.fe_group = group; |
4237 | ac->ac_o_ex.fe_start = block; |
4238 | ac->ac_o_ex.fe_len = len; |
4239 | ac->ac_g_ex.fe_logical = ar->logical; |
4240 | ac->ac_g_ex.fe_group = group; |
4241 | ac->ac_g_ex.fe_start = block; |
4242 | ac->ac_g_ex.fe_len = len; |
4243 | ac->ac_flags = ar->flags; |
4244 | |
4245 | /* we have to define context: we'll we work with a file or |
4246 | * locality group. this is a policy, actually */ |
4247 | ext4_mb_group_or_file(ac); |
4248 | |
4249 | mb_debug("init ac: %u blocks @ %u, goal %u, flags %x, 2^%d, " |
4250 | "left: %u/%u, right %u/%u to %swritable\n", |
4251 | (unsigned) ar->len, (unsigned) ar->logical, |
4252 | (unsigned) ar->goal, ac->ac_flags, ac->ac_2order, |
4253 | (unsigned) ar->lleft, (unsigned) ar->pleft, |
4254 | (unsigned) ar->lright, (unsigned) ar->pright, |
4255 | atomic_read(&ar->inode->i_writecount) ? "" : "non-"); |
4256 | return 0; |
4257 | |
4258 | } |
4259 | |
4260 | static noinline_for_stack void |
4261 | ext4_mb_discard_lg_preallocations(struct super_block *sb, |
4262 | struct ext4_locality_group *lg, |
4263 | int order, int total_entries) |
4264 | { |
4265 | ext4_group_t group = 0; |
4266 | struct ext4_buddy e4b; |
4267 | struct list_head discard_list; |
4268 | struct ext4_prealloc_space *pa, *tmp; |
4269 | struct ext4_allocation_context *ac; |
4270 | |
4271 | mb_debug("discard locality group preallocation\n"); |
4272 | |
4273 | INIT_LIST_HEAD(&discard_list); |
4274 | ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); |
4275 | if (ac) |
4276 | ac->ac_sb = sb; |
4277 | |
4278 | spin_lock(&lg->lg_prealloc_lock); |
4279 | list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[order], |
4280 | pa_inode_list) { |
4281 | spin_lock(&pa->pa_lock); |
4282 | if (atomic_read(&pa->pa_count)) { |
4283 | /* |
4284 | * This is the pa that we just used |
4285 | * for block allocation. So don't |
4286 | * free that |
4287 | */ |
4288 | spin_unlock(&pa->pa_lock); |
4289 | continue; |
4290 | } |
4291 | if (pa->pa_deleted) { |
4292 | spin_unlock(&pa->pa_lock); |
4293 | continue; |
4294 | } |
4295 | /* only lg prealloc space */ |
4296 | BUG_ON(pa->pa_type != MB_GROUP_PA); |
4297 | |
4298 | /* seems this one can be freed ... */ |
4299 | pa->pa_deleted = 1; |
4300 | spin_unlock(&pa->pa_lock); |
4301 | |
4302 | list_del_rcu(&pa->pa_inode_list); |
4303 | list_add(&pa->u.pa_tmp_list, &discard_list); |
4304 | |
4305 | total_entries--; |
4306 | if (total_entries <= 5) { |
4307 | /* |
4308 | * we want to keep only 5 entries |
4309 | * allowing it to grow to 8. This |
4310 | * mak sure we don't call discard |
4311 | * soon for this list. |
4312 | */ |
4313 | break; |
4314 | } |
4315 | } |
4316 | spin_unlock(&lg->lg_prealloc_lock); |
4317 | |
4318 | list_for_each_entry_safe(pa, tmp, &discard_list, u.pa_tmp_list) { |
4319 | |
4320 | ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL); |
4321 | if (ext4_mb_load_buddy(sb, group, &e4b)) { |
4322 | ext4_error(sb, __func__, "Error in loading buddy " |
4323 | "information for %u", group); |
4324 | continue; |
4325 | } |
4326 | ext4_lock_group(sb, group); |
4327 | list_del(&pa->pa_group_list); |
4328 | ext4_mb_release_group_pa(&e4b, pa, ac); |
4329 | ext4_unlock_group(sb, group); |
4330 | |
4331 | ext4_mb_release_desc(&e4b); |
4332 | list_del(&pa->u.pa_tmp_list); |
4333 | call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); |
4334 | } |
4335 | if (ac) |
4336 | kmem_cache_free(ext4_ac_cachep, ac); |
4337 | } |
4338 | |
4339 | /* |
4340 | * We have incremented pa_count. So it cannot be freed at this |
4341 | * point. Also we hold lg_mutex. So no parallel allocation is |
4342 | * possible from this lg. That means pa_free cannot be updated. |
4343 | * |
4344 | * A parallel ext4_mb_discard_group_preallocations is possible. |
4345 | * which can cause the lg_prealloc_list to be updated. |
4346 | */ |
4347 | |
4348 | static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) |
4349 | { |
4350 | int order, added = 0, lg_prealloc_count = 1; |
4351 | struct super_block *sb = ac->ac_sb; |
4352 | struct ext4_locality_group *lg = ac->ac_lg; |
4353 | struct ext4_prealloc_space *tmp_pa, *pa = ac->ac_pa; |
4354 | |
4355 | order = fls(pa->pa_free) - 1; |
4356 | if (order > PREALLOC_TB_SIZE - 1) |
4357 | /* The max size of hash table is PREALLOC_TB_SIZE */ |
4358 | order = PREALLOC_TB_SIZE - 1; |
4359 | /* Add the prealloc space to lg */ |
4360 | rcu_read_lock(); |
4361 | list_for_each_entry_rcu(tmp_pa, &lg->lg_prealloc_list[order], |
4362 | pa_inode_list) { |
4363 | spin_lock(&tmp_pa->pa_lock); |
4364 | if (tmp_pa->pa_deleted) { |
4365 | spin_unlock(&tmp_pa->pa_lock); |
4366 | continue; |
4367 | } |
4368 | if (!added && pa->pa_free < tmp_pa->pa_free) { |
4369 | /* Add to the tail of the previous entry */ |
4370 | list_add_tail_rcu(&pa->pa_inode_list, |
4371 | &tmp_pa->pa_inode_list); |
4372 | added = 1; |
4373 | /* |
4374 | * we want to count the total |
4375 | * number of entries in the list |
4376 | */ |
4377 | } |
4378 | spin_unlock(&tmp_pa->pa_lock); |
4379 | lg_prealloc_count++; |
4380 | } |
4381 | if (!added) |
4382 | list_add_tail_rcu(&pa->pa_inode_list, |
4383 | &lg->lg_prealloc_list[order]); |
4384 | rcu_read_unlock(); |
4385 | |
4386 | /* Now trim the list to be not more than 8 elements */ |
4387 | if (lg_prealloc_count > 8) { |
4388 | ext4_mb_discard_lg_preallocations(sb, lg, |
4389 | order, lg_prealloc_count); |
4390 | return; |
4391 | } |
4392 | return ; |
4393 | } |
4394 | |
4395 | /* |
4396 | * release all resource we used in allocation |
4397 | */ |
4398 | static int ext4_mb_release_context(struct ext4_allocation_context *ac) |
4399 | { |
4400 | struct ext4_prealloc_space *pa = ac->ac_pa; |
4401 | if (pa) { |
4402 | if (pa->pa_type == MB_GROUP_PA) { |
4403 | /* see comment in ext4_mb_use_group_pa() */ |
4404 | spin_lock(&pa->pa_lock); |
4405 | pa->pa_pstart += ac->ac_b_ex.fe_len; |
4406 | pa->pa_lstart += ac->ac_b_ex.fe_len; |
4407 | pa->pa_free -= ac->ac_b_ex.fe_len; |
4408 | pa->pa_len -= ac->ac_b_ex.fe_len; |
4409 | spin_unlock(&pa->pa_lock); |
4410 | } |
4411 | } |
4412 | if (ac->alloc_semp) |
4413 | up_read(ac->alloc_semp); |
4414 | if (pa) { |
4415 | /* |
4416 | * We want to add the pa to the right bucket. |
4417 | * Remove it from the list and while adding |
4418 | * make sure the list to which we are adding |
4419 | * doesn't grow big. We need to release |
4420 | * alloc_semp before calling ext4_mb_add_n_trim() |
4421 | */ |
4422 | if ((pa->pa_type == MB_GROUP_PA) && likely(pa->pa_free)) { |
4423 | spin_lock(pa->pa_obj_lock); |
4424 | list_del_rcu(&pa->pa_inode_list); |
4425 | spin_unlock(pa->pa_obj_lock); |
4426 | ext4_mb_add_n_trim(ac); |
4427 | } |
4428 | ext4_mb_put_pa(ac, ac->ac_sb, pa); |
4429 | } |
4430 | if (ac->ac_bitmap_page) |
4431 | page_cache_release(ac->ac_bitmap_page); |
4432 | if (ac->ac_buddy_page) |
4433 | page_cache_release(ac->ac_buddy_page); |
4434 | if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) |
4435 | mutex_unlock(&ac->ac_lg->lg_mutex); |
4436 | ext4_mb_collect_stats(ac); |
4437 | return 0; |
4438 | } |
4439 | |
4440 | static int ext4_mb_discard_preallocations(struct super_block *sb, int needed) |
4441 | { |
4442 | ext4_group_t i, ngroups = ext4_get_groups_count(sb); |
4443 | int ret; |
4444 | int freed = 0; |
4445 | |
4446 | trace_ext4_mb_discard_preallocations(sb, needed); |
4447 | for (i = 0; i < ngroups && needed > 0; i++) { |
4448 | ret = ext4_mb_discard_group_preallocations(sb, i, needed); |
4449 | freed += ret; |
4450 | needed -= ret; |
4451 | } |
4452 | |
4453 | return freed; |
4454 | } |
4455 | |
4456 | /* |
4457 | * Main entry point into mballoc to allocate blocks |
4458 | * it tries to use preallocation first, then falls back |
4459 | * to usual allocation |
4460 | */ |
4461 | ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, |
4462 | struct ext4_allocation_request *ar, int *errp) |
4463 | { |
4464 | int freed; |
4465 | struct ext4_allocation_context *ac = NULL; |
4466 | struct ext4_sb_info *sbi; |
4467 | struct super_block *sb; |
4468 | ext4_fsblk_t block = 0; |
4469 | unsigned int inquota = 0; |
4470 | unsigned int reserv_blks = 0; |
4471 | |
4472 | sb = ar->inode->i_sb; |
4473 | sbi = EXT4_SB(sb); |
4474 | |
4475 | trace_ext4_request_blocks(ar); |
4476 | |
4477 | /* |
4478 | * For delayed allocation, we could skip the ENOSPC and |
4479 | * EDQUOT check, as blocks and quotas have been already |
4480 | * reserved when data being copied into pagecache. |
4481 | */ |
4482 | if (EXT4_I(ar->inode)->i_delalloc_reserved_flag) |
4483 | ar->flags |= EXT4_MB_DELALLOC_RESERVED; |
4484 | else { |
4485 | /* Without delayed allocation we need to verify |
4486 | * there is enough free blocks to do block allocation |
4487 | * and verify allocation doesn't exceed the quota limits. |
4488 | */ |
4489 | while (ar->len && ext4_claim_free_blocks(sbi, ar->len)) { |
4490 | /* let others to free the space */ |
4491 | yield(); |
4492 | ar->len = ar->len >> 1; |
4493 | } |
4494 | if (!ar->len) { |
4495 | *errp = -ENOSPC; |
4496 | return 0; |
4497 | } |
4498 | reserv_blks = ar->len; |
4499 | while (ar->len && vfs_dq_alloc_block(ar->inode, ar->len)) { |
4500 | ar->flags |= EXT4_MB_HINT_NOPREALLOC; |
4501 | ar->len--; |
4502 | } |
4503 | inquota = ar->len; |
4504 | if (ar->len == 0) { |
4505 | *errp = -EDQUOT; |
4506 | goto out3; |
4507 | } |
4508 | } |
4509 | |
4510 | ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); |
4511 | if (!ac) { |
4512 | ar->len = 0; |
4513 | *errp = -ENOMEM; |
4514 | goto out1; |
4515 | } |
4516 | |
4517 | *errp = ext4_mb_initialize_context(ac, ar); |
4518 | if (*errp) { |
4519 | ar->len = 0; |
4520 | goto out2; |
4521 | } |
4522 | |
4523 | ac->ac_op = EXT4_MB_HISTORY_PREALLOC; |
4524 | if (!ext4_mb_use_preallocated(ac)) { |
4525 | ac->ac_op = EXT4_MB_HISTORY_ALLOC; |
4526 | ext4_mb_normalize_request(ac, ar); |
4527 | repeat: |
4528 | /* allocate space in core */ |
4529 | ext4_mb_regular_allocator(ac); |
4530 | |
4531 | /* as we've just preallocated more space than |
4532 | * user requested orinally, we store allocated |
4533 | * space in a special descriptor */ |
4534 | if (ac->ac_status == AC_STATUS_FOUND && |
4535 | ac->ac_o_ex.fe_len < ac->ac_b_ex.fe_len) |
4536 | ext4_mb_new_preallocation(ac); |
4537 | } |
4538 | if (likely(ac->ac_status == AC_STATUS_FOUND)) { |
4539 | *errp = ext4_mb_mark_diskspace_used(ac, handle, reserv_blks); |
4540 | if (*errp == -EAGAIN) { |
4541 | /* |
4542 | * drop the reference that we took |
4543 | * in ext4_mb_use_best_found |
4544 | */ |
4545 | ext4_mb_release_context(ac); |
4546 | ac->ac_b_ex.fe_group = 0; |
4547 | ac->ac_b_ex.fe_start = 0; |
4548 | ac->ac_b_ex.fe_len = 0; |
4549 | ac->ac_status = AC_STATUS_CONTINUE; |
4550 | goto repeat; |
4551 | } else if (*errp) { |
4552 | ac->ac_b_ex.fe_len = 0; |
4553 | ar->len = 0; |
4554 | ext4_mb_show_ac(ac); |
4555 | } else { |
4556 | block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); |
4557 | ar->len = ac->ac_b_ex.fe_len; |
4558 | } |
4559 | } else { |
4560 | freed = ext4_mb_discard_preallocations(sb, ac->ac_o_ex.fe_len); |
4561 | if (freed) |
4562 | goto repeat; |
4563 | *errp = -ENOSPC; |
4564 | ac->ac_b_ex.fe_len = 0; |
4565 | ar->len = 0; |
4566 | ext4_mb_show_ac(ac); |
4567 | } |
4568 | |
4569 | ext4_mb_release_context(ac); |
4570 | |
4571 | out2: |
4572 | kmem_cache_free(ext4_ac_cachep, ac); |
4573 | out1: |
4574 | if (inquota && ar->len < inquota) |
4575 | vfs_dq_free_block(ar->inode, inquota - ar->len); |
4576 | out3: |
4577 | if (!ar->len) { |
4578 | if (!EXT4_I(ar->inode)->i_delalloc_reserved_flag) |
4579 | /* release all the reserved blocks if non delalloc */ |
4580 | percpu_counter_sub(&sbi->s_dirtyblocks_counter, |
4581 | reserv_blks); |
4582 | } |
4583 | |
4584 | trace_ext4_allocate_blocks(ar, (unsigned long long)block); |
4585 | |
4586 | return block; |
4587 | } |
4588 | |
4589 | /* |
4590 | * We can merge two free data extents only if the physical blocks |
4591 | * are contiguous, AND the extents were freed by the same transaction, |
4592 | * AND the blocks are associated with the same group. |
4593 | */ |
4594 | static int can_merge(struct ext4_free_data *entry1, |
4595 | struct ext4_free_data *entry2) |
4596 | { |
4597 | if ((entry1->t_tid == entry2->t_tid) && |
4598 | (entry1->group == entry2->group) && |
4599 | ((entry1->start_blk + entry1->count) == entry2->start_blk)) |
4600 | return 1; |
4601 | return 0; |
4602 | } |
4603 | |
4604 | static noinline_for_stack int |
4605 | ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, |
4606 | struct ext4_free_data *new_entry) |
4607 | { |
4608 | ext4_grpblk_t block; |
4609 | struct ext4_free_data *entry; |
4610 | struct ext4_group_info *db = e4b->bd_info; |
4611 | struct super_block *sb = e4b->bd_sb; |
4612 | struct ext4_sb_info *sbi = EXT4_SB(sb); |
4613 | struct rb_node **n = &db->bb_free_root.rb_node, *node; |
4614 | struct rb_node *parent = NULL, *new_node; |
4615 | |
4616 | BUG_ON(!ext4_handle_valid(handle)); |
4617 | BUG_ON(e4b->bd_bitmap_page == NULL); |
4618 | BUG_ON(e4b->bd_buddy_page == NULL); |
4619 | |
4620 | new_node = &new_entry->node; |
4621 | block = new_entry->start_blk; |
4622 | |
4623 | if (!*n) { |
4624 | /* first free block exent. We need to |
4625 | protect buddy cache from being freed, |
4626 | * otherwise we'll refresh it from |
4627 | * on-disk bitmap and lose not-yet-available |
4628 | * blocks */ |
4629 | page_cache_get(e4b->bd_buddy_page); |
4630 | page_cache_get(e4b->bd_bitmap_page); |
4631 | } |
4632 | while (*n) { |
4633 | parent = *n; |
4634 | entry = rb_entry(parent, struct ext4_free_data, node); |
4635 | if (block < entry->start_blk) |
4636 | n = &(*n)->rb_left; |
4637 | else if (block >= (entry->start_blk + entry->count)) |
4638 | n = &(*n)->rb_right; |
4639 | else { |
4640 | ext4_grp_locked_error(sb, e4b->bd_group, __func__, |
4641 | "Double free of blocks %d (%d %d)", |
4642 | block, entry->start_blk, entry->count); |
4643 | return 0; |
4644 | } |
4645 | } |
4646 | |
4647 | rb_link_node(new_node, parent, n); |
4648 | rb_insert_color(new_node, &db->bb_free_root); |
4649 | |
4650 | /* Now try to see the extent can be merged to left and right */ |
4651 | node = rb_prev(new_node); |
4652 | if (node) { |
4653 | entry = rb_entry(node, struct ext4_free_data, node); |
4654 | if (can_merge(entry, new_entry)) { |
4655 | new_entry->start_blk = entry->start_blk; |
4656 | new_entry->count += entry->count; |
4657 | rb_erase(node, &(db->bb_free_root)); |
4658 | spin_lock(&sbi->s_md_lock); |
4659 | list_del(&entry->list); |
4660 | spin_unlock(&sbi->s_md_lock); |
4661 | kmem_cache_free(ext4_free_ext_cachep, entry); |
4662 | } |
4663 | } |
4664 | |
4665 | node = rb_next(new_node); |
4666 | if (node) { |
4667 | entry = rb_entry(node, struct ext4_free_data, node); |
4668 | if (can_merge(new_entry, entry)) { |
4669 | new_entry->count += entry->count; |
4670 | rb_erase(node, &(db->bb_free_root)); |
4671 | spin_lock(&sbi->s_md_lock); |
4672 | list_del(&entry->list); |
4673 | spin_unlock(&sbi->s_md_lock); |
4674 | kmem_cache_free(ext4_free_ext_cachep, entry); |
4675 | } |
4676 | } |
4677 | /* Add the extent to transaction's private list */ |
4678 | spin_lock(&sbi->s_md_lock); |
4679 | list_add(&new_entry->list, &handle->h_transaction->t_private_list); |
4680 | spin_unlock(&sbi->s_md_lock); |
4681 | return 0; |
4682 | } |
4683 | |
4684 | /* |
4685 | * Main entry point into mballoc to free blocks |
4686 | */ |
4687 | void ext4_mb_free_blocks(handle_t *handle, struct inode *inode, |
4688 | ext4_fsblk_t block, unsigned long count, |
4689 | int metadata, unsigned long *freed) |
4690 | { |
4691 | struct buffer_head *bitmap_bh = NULL; |
4692 | struct super_block *sb = inode->i_sb; |
4693 | struct ext4_allocation_context *ac = NULL; |
4694 | struct ext4_group_desc *gdp; |
4695 | struct ext4_super_block *es; |
4696 | unsigned int overflow; |
4697 | ext4_grpblk_t bit; |
4698 | struct buffer_head *gd_bh; |
4699 | ext4_group_t block_group; |
4700 | struct ext4_sb_info *sbi; |
4701 | struct ext4_buddy e4b; |
4702 | int err = 0; |
4703 | int ret; |
4704 | |
4705 | *freed = 0; |
4706 | |
4707 | sbi = EXT4_SB(sb); |
4708 | es = EXT4_SB(sb)->s_es; |
4709 | if (block < le32_to_cpu(es->s_first_data_block) || |
4710 | block + count < block || |
4711 | block + count > ext4_blocks_count(es)) { |
4712 | ext4_error(sb, __func__, |
4713 | "Freeing blocks not in datazone - " |
4714 | "block = %llu, count = %lu", block, count); |
4715 | goto error_return; |
4716 | } |
4717 | |
4718 | ext4_debug("freeing block %llu\n", block); |
4719 | trace_ext4_free_blocks(inode, block, count, metadata); |
4720 | |
4721 | ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); |
4722 | if (ac) { |
4723 | ac->ac_op = EXT4_MB_HISTORY_FREE; |
4724 | ac->ac_inode = inode; |
4725 | ac->ac_sb = sb; |
4726 | } |
4727 | |
4728 | do_more: |
4729 | overflow = 0; |
4730 | ext4_get_group_no_and_offset(sb, block, &block_group, &bit); |
4731 | |
4732 | /* |
4733 | * Check to see if we are freeing blocks across a group |
4734 | * boundary. |
4735 | */ |
4736 | if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) { |
4737 | overflow = bit + count - EXT4_BLOCKS_PER_GROUP(sb); |
4738 | count -= overflow; |
4739 | } |
4740 | bitmap_bh = ext4_read_block_bitmap(sb, block_group); |
4741 | if (!bitmap_bh) { |
4742 | err = -EIO; |
4743 | goto error_return; |
4744 | } |
4745 | gdp = ext4_get_group_desc(sb, block_group, &gd_bh); |
4746 | if (!gdp) { |
4747 | err = -EIO; |
4748 | goto error_return; |
4749 | } |
4750 | |
4751 | if (in_range(ext4_block_bitmap(sb, gdp), block, count) || |
4752 | in_range(ext4_inode_bitmap(sb, gdp), block, count) || |
4753 | in_range(block, ext4_inode_table(sb, gdp), |
4754 | EXT4_SB(sb)->s_itb_per_group) || |
4755 | in_range(block + count - 1, ext4_inode_table(sb, gdp), |
4756 | EXT4_SB(sb)->s_itb_per_group)) { |
4757 | |
4758 | ext4_error(sb, __func__, |
4759 | "Freeing blocks in system zone - " |
4760 | "Block = %llu, count = %lu", block, count); |
4761 | /* err = 0. ext4_std_error should be a no op */ |
4762 | goto error_return; |
4763 | } |
4764 | |
4765 | BUFFER_TRACE(bitmap_bh, "getting write access"); |
4766 | err = ext4_journal_get_write_access(handle, bitmap_bh); |
4767 | if (err) |
4768 | goto error_return; |
4769 | |
4770 | /* |
4771 | * We are about to modify some metadata. Call the journal APIs |
4772 | * to unshare ->b_data if a currently-committing transaction is |
4773 | * using it |
4774 | */ |
4775 | BUFFER_TRACE(gd_bh, "get_write_access"); |
4776 | err = ext4_journal_get_write_access(handle, gd_bh); |
4777 | if (err) |
4778 | goto error_return; |
4779 | #ifdef AGGRESSIVE_CHECK |
4780 | { |
4781 | int i; |
4782 | for (i = 0; i < count; i++) |
4783 | BUG_ON(!mb_test_bit(bit + i, bitmap_bh->b_data)); |
4784 | } |
4785 | #endif |
4786 | if (ac) { |
4787 | ac->ac_b_ex.fe_group = block_group; |
4788 | ac->ac_b_ex.fe_start = bit; |
4789 | ac->ac_b_ex.fe_len = count; |
4790 | ext4_mb_store_history(ac); |
4791 | } |
4792 | |
4793 | err = ext4_mb_load_buddy(sb, block_group, &e4b); |
4794 | if (err) |
4795 | goto error_return; |
4796 | if (metadata && ext4_handle_valid(handle)) { |
4797 | struct ext4_free_data *new_entry; |
4798 | /* |
4799 | * blocks being freed are metadata. these blocks shouldn't |
4800 | * be used until this transaction is committed |
4801 | */ |
4802 | new_entry = kmem_cache_alloc(ext4_free_ext_cachep, GFP_NOFS); |
4803 | new_entry->start_blk = bit; |
4804 | new_entry->group = block_group; |
4805 | new_entry->count = count; |
4806 | new_entry->t_tid = handle->h_transaction->t_tid; |
4807 | |
4808 | ext4_lock_group(sb, block_group); |
4809 | mb_clear_bits(bitmap_bh->b_data, bit, count); |
4810 | ext4_mb_free_metadata(handle, &e4b, new_entry); |
4811 | } else { |
4812 | /* need to update group_info->bb_free and bitmap |
4813 | * with group lock held. generate_buddy look at |
4814 | * them with group lock_held |
4815 | */ |
4816 | ext4_lock_group(sb, block_group); |
4817 | mb_clear_bits(bitmap_bh->b_data, bit, count); |
4818 | mb_free_blocks(inode, &e4b, bit, count); |
4819 | ext4_mb_return_to_preallocation(inode, &e4b, block, count); |
4820 | } |
4821 | |
4822 | ret = ext4_free_blks_count(sb, gdp) + count; |
4823 | ext4_free_blks_set(sb, gdp, ret); |
4824 | gdp->bg_checksum = ext4_group_desc_csum(sbi, block_group, gdp); |
4825 | ext4_unlock_group(sb, block_group); |
4826 | percpu_counter_add(&sbi->s_freeblocks_counter, count); |
4827 | |
4828 | if (sbi->s_log_groups_per_flex) { |
4829 | ext4_group_t flex_group = ext4_flex_group(sbi, block_group); |
4830 | atomic_add(count, &sbi->s_flex_groups[flex_group].free_blocks); |
4831 | } |
4832 | |
4833 | ext4_mb_release_desc(&e4b); |
4834 | |
4835 | *freed += count; |
4836 | |
4837 | /* We dirtied the bitmap block */ |
4838 | BUFFER_TRACE(bitmap_bh, "dirtied bitmap block"); |
4839 | err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); |
4840 | |
4841 | /* And the group descriptor block */ |
4842 | BUFFER_TRACE(gd_bh, "dirtied group descriptor block"); |
4843 | ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh); |
4844 | if (!err) |
4845 | err = ret; |
4846 | |
4847 | if (overflow && !err) { |
4848 | block += count; |
4849 | count = overflow; |
4850 | put_bh(bitmap_bh); |
4851 | goto do_more; |
4852 | } |
4853 | sb->s_dirt = 1; |
4854 | error_return: |
4855 | brelse(bitmap_bh); |
4856 | ext4_std_error(sb, err); |
4857 | if (ac) |
4858 | kmem_cache_free(ext4_ac_cachep, ac); |
4859 | return; |
4860 | } |
4861 |
Branches:
ben-wpan
ben-wpan-stefan
javiroman/ks7010
jz-2.6.34
jz-2.6.34-rc5
jz-2.6.34-rc6
jz-2.6.34-rc7
jz-2.6.35
jz-2.6.36
jz-2.6.37
jz-2.6.38
jz-2.6.39
jz-3.0
jz-3.1
jz-3.11
jz-3.12
jz-3.13
jz-3.15
jz-3.16
jz-3.18-dt
jz-3.2
jz-3.3
jz-3.4
jz-3.5
jz-3.6
jz-3.6-rc2-pwm
jz-3.9
jz-3.9-clk
jz-3.9-rc8
jz47xx
jz47xx-2.6.38
master
Tags:
od-2011-09-04
od-2011-09-18
v2.6.34-rc5
v2.6.34-rc6
v2.6.34-rc7
v3.9