Loading...
1/*
2 * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README
3 */
4/* Reiserfs block (de)allocator, bitmap-based. */
5
6#include <linux/time.h>
7#include "reiserfs.h"
8#include <linux/errno.h>
9#include <linux/buffer_head.h>
10#include <linux/kernel.h>
11#include <linux/pagemap.h>
12#include <linux/vmalloc.h>
13#include <linux/quotaops.h>
14#include <linux/seq_file.h>
15
16#define PREALLOCATION_SIZE 9
17
18/* different reiserfs block allocator options */
19
20#define SB_ALLOC_OPTS(s) (REISERFS_SB(s)->s_alloc_options.bits)
21
22#define _ALLOC_concentrating_formatted_nodes 0
23#define _ALLOC_displacing_large_files 1
24#define _ALLOC_displacing_new_packing_localities 2
25#define _ALLOC_old_hashed_relocation 3
26#define _ALLOC_new_hashed_relocation 4
27#define _ALLOC_skip_busy 5
28#define _ALLOC_displace_based_on_dirid 6
29#define _ALLOC_hashed_formatted_nodes 7
30#define _ALLOC_old_way 8
31#define _ALLOC_hundredth_slices 9
32#define _ALLOC_dirid_groups 10
33#define _ALLOC_oid_groups 11
34#define _ALLOC_packing_groups 12
35
36#define concentrating_formatted_nodes(s) test_bit(_ALLOC_concentrating_formatted_nodes, &SB_ALLOC_OPTS(s))
37#define displacing_large_files(s) test_bit(_ALLOC_displacing_large_files, &SB_ALLOC_OPTS(s))
38#define displacing_new_packing_localities(s) test_bit(_ALLOC_displacing_new_packing_localities, &SB_ALLOC_OPTS(s))
39
40#define SET_OPTION(optname) \
41 do { \
42 reiserfs_info(s, "block allocator option \"%s\" is set", #optname); \
43 set_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s)); \
44 } while(0)
45#define TEST_OPTION(optname, s) \
46 test_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s))
47
48static inline void get_bit_address(struct super_block *s,
49 b_blocknr_t block,
50 unsigned int *bmap_nr,
51 unsigned int *offset)
52{
53 /*
54 * It is in the bitmap block number equal to the block
55 * number divided by the number of bits in a block.
56 */
57 *bmap_nr = block >> (s->s_blocksize_bits + 3);
58 /* Within that bitmap block it is located at bit offset *offset. */
59 *offset = block & ((s->s_blocksize << 3) - 1);
60}
61
62int is_reusable(struct super_block *s, b_blocknr_t block, int bit_value)
63{
64 unsigned int bmap, offset;
65 unsigned int bmap_count = reiserfs_bmap_count(s);
66
67 if (block == 0 || block >= SB_BLOCK_COUNT(s)) {
68 reiserfs_error(s, "vs-4010",
69 "block number is out of range %lu (%u)",
70 block, SB_BLOCK_COUNT(s));
71 return 0;
72 }
73
74 get_bit_address(s, block, &bmap, &offset);
75
76 /*
77 * Old format filesystem? Unlikely, but the bitmaps are all
78 * up front so we need to account for it.
79 */
80 if (unlikely(test_bit(REISERFS_OLD_FORMAT,
81 &REISERFS_SB(s)->s_properties))) {
82 b_blocknr_t bmap1 = REISERFS_SB(s)->s_sbh->b_blocknr + 1;
83 if (block >= bmap1 &&
84 block <= bmap1 + bmap_count) {
85 reiserfs_error(s, "vs-4019", "bitmap block %lu(%u) "
86 "can't be freed or reused",
87 block, bmap_count);
88 return 0;
89 }
90 } else {
91 if (offset == 0) {
92 reiserfs_error(s, "vs-4020", "bitmap block %lu(%u) "
93 "can't be freed or reused",
94 block, bmap_count);
95 return 0;
96 }
97 }
98
99 if (bmap >= bmap_count) {
100 reiserfs_error(s, "vs-4030", "bitmap for requested block "
101 "is out of range: block=%lu, bitmap_nr=%u",
102 block, bmap);
103 return 0;
104 }
105
106 if (bit_value == 0 && block == SB_ROOT_BLOCK(s)) {
107 reiserfs_error(s, "vs-4050", "this is root block (%u), "
108 "it must be busy", SB_ROOT_BLOCK(s));
109 return 0;
110 }
111
112 return 1;
113}
114
115/*
116 * Searches in journal structures for a given block number (bmap, off).
117 * If block is found in reiserfs journal it suggests next free block
118 * candidate to test.
119 */
120static inline int is_block_in_journal(struct super_block *s, unsigned int bmap,
121 int off, int *next)
122{
123 b_blocknr_t tmp;
124
125 if (reiserfs_in_journal(s, bmap, off, 1, &tmp)) {
126 if (tmp) { /* hint supplied */
127 *next = tmp;
128 PROC_INFO_INC(s, scan_bitmap.in_journal_hint);
129 } else {
130 (*next) = off + 1; /* inc offset to avoid looping. */
131 PROC_INFO_INC(s, scan_bitmap.in_journal_nohint);
132 }
133 PROC_INFO_INC(s, scan_bitmap.retry);
134 return 1;
135 }
136 return 0;
137}
138
139/*
140 * Searches for a window of zero bits with given minimum and maximum
141 * lengths in one bitmap block
142 */
143static int scan_bitmap_block(struct reiserfs_transaction_handle *th,
144 unsigned int bmap_n, int *beg, int boundary,
145 int min, int max, int unfm)
146{
147 struct super_block *s = th->t_super;
148 struct reiserfs_bitmap_info *bi = &SB_AP_BITMAP(s)[bmap_n];
149 struct buffer_head *bh;
150 int end, next;
151 int org = *beg;
152
153 BUG_ON(!th->t_trans_id);
154 RFALSE(bmap_n >= reiserfs_bmap_count(s), "Bitmap %u is out of "
155 "range (0..%u)", bmap_n, reiserfs_bmap_count(s) - 1);
156 PROC_INFO_INC(s, scan_bitmap.bmap);
157
158 if (!bi) {
159 reiserfs_error(s, "jdm-4055", "NULL bitmap info pointer "
160 "for bitmap %d", bmap_n);
161 return 0;
162 }
163
164 bh = reiserfs_read_bitmap_block(s, bmap_n);
165 if (bh == NULL)
166 return 0;
167
168 while (1) {
169cont:
170 if (bi->free_count < min) {
171 brelse(bh);
172 return 0; /* No free blocks in this bitmap */
173 }
174
175 /* search for a first zero bit -- beginning of a window */
176 *beg = reiserfs_find_next_zero_le_bit
177 ((unsigned long *)(bh->b_data), boundary, *beg);
178
179 /*
180 * search for a zero bit fails or the rest of bitmap block
181 * cannot contain a zero window of minimum size
182 */
183 if (*beg + min > boundary) {
184 brelse(bh);
185 return 0;
186 }
187
188 if (unfm && is_block_in_journal(s, bmap_n, *beg, beg))
189 continue;
190 /* first zero bit found; we check next bits */
191 for (end = *beg + 1;; end++) {
192 if (end >= *beg + max || end >= boundary
193 || reiserfs_test_le_bit(end, bh->b_data)) {
194 next = end;
195 break;
196 }
197
198 /*
199 * finding the other end of zero bit window requires
200 * looking into journal structures (in case of
201 * searching for free blocks for unformatted nodes)
202 */
203 if (unfm && is_block_in_journal(s, bmap_n, end, &next))
204 break;
205 }
206
207 /*
208 * now (*beg) points to beginning of zero bits window,
209 * (end) points to one bit after the window end
210 */
211
212 /* found window of proper size */
213 if (end - *beg >= min) {
214 int i;
215 reiserfs_prepare_for_journal(s, bh, 1);
216 /*
217 * try to set all blocks used checking are
218 * they still free
219 */
220 for (i = *beg; i < end; i++) {
221 /* Don't check in journal again. */
222 if (reiserfs_test_and_set_le_bit
223 (i, bh->b_data)) {
224 /*
225 * bit was set by another process while
226 * we slept in prepare_for_journal()
227 */
228 PROC_INFO_INC(s, scan_bitmap.stolen);
229
230 /*
231 * we can continue with smaller set
232 * of allocated blocks, if length of
233 * this set is more or equal to `min'
234 */
235 if (i >= *beg + min) {
236 end = i;
237 break;
238 }
239
240 /*
241 * otherwise we clear all bit
242 * were set ...
243 */
244 while (--i >= *beg)
245 reiserfs_clear_le_bit
246 (i, bh->b_data);
247 reiserfs_restore_prepared_buffer(s, bh);
248 *beg = org;
249
250 /*
251 * Search again in current block
252 * from beginning
253 */
254 goto cont;
255 }
256 }
257 bi->free_count -= (end - *beg);
258 journal_mark_dirty(th, bh);
259 brelse(bh);
260
261 /* free block count calculation */
262 reiserfs_prepare_for_journal(s, SB_BUFFER_WITH_SB(s),
263 1);
264 PUT_SB_FREE_BLOCKS(s, SB_FREE_BLOCKS(s) - (end - *beg));
265 journal_mark_dirty(th, SB_BUFFER_WITH_SB(s));
266
267 return end - (*beg);
268 } else {
269 *beg = next;
270 }
271 }
272}
273
274static int bmap_hash_id(struct super_block *s, u32 id)
275{
276 char *hash_in = NULL;
277 unsigned long hash;
278 unsigned bm;
279
280 if (id <= 2) {
281 bm = 1;
282 } else {
283 hash_in = (char *)(&id);
284 hash = keyed_hash(hash_in, 4);
285 bm = hash % reiserfs_bmap_count(s);
286 if (!bm)
287 bm = 1;
288 }
289 /* this can only be true when SB_BMAP_NR = 1 */
290 if (bm >= reiserfs_bmap_count(s))
291 bm = 0;
292 return bm;
293}
294
295/*
296 * hashes the id and then returns > 0 if the block group for the
297 * corresponding hash is full
298 */
299static inline int block_group_used(struct super_block *s, u32 id)
300{
301 int bm = bmap_hash_id(s, id);
302 struct reiserfs_bitmap_info *info = &SB_AP_BITMAP(s)[bm];
303
304 /*
305 * If we don't have cached information on this bitmap block, we're
306 * going to have to load it later anyway. Loading it here allows us
307 * to make a better decision. This favors long-term performance gain
308 * with a better on-disk layout vs. a short term gain of skipping the
309 * read and potentially having a bad placement.
310 */
311 if (info->free_count == UINT_MAX) {
312 struct buffer_head *bh = reiserfs_read_bitmap_block(s, bm);
313 brelse(bh);
314 }
315
316 if (info->free_count > ((s->s_blocksize << 3) * 60 / 100)) {
317 return 0;
318 }
319 return 1;
320}
321
322/*
323 * the packing is returned in disk byte order
324 */
325__le32 reiserfs_choose_packing(struct inode * dir)
326{
327 __le32 packing;
328 if (TEST_OPTION(packing_groups, dir->i_sb)) {
329 u32 parent_dir = le32_to_cpu(INODE_PKEY(dir)->k_dir_id);
330 /*
331 * some versions of reiserfsck expect packing locality 1 to be
332 * special
333 */
334 if (parent_dir == 1 || block_group_used(dir->i_sb, parent_dir))
335 packing = INODE_PKEY(dir)->k_objectid;
336 else
337 packing = INODE_PKEY(dir)->k_dir_id;
338 } else
339 packing = INODE_PKEY(dir)->k_objectid;
340 return packing;
341}
342
343/*
344 * Tries to find contiguous zero bit window (given size) in given region of
345 * bitmap and place new blocks there. Returns number of allocated blocks.
346 */
347static int scan_bitmap(struct reiserfs_transaction_handle *th,
348 b_blocknr_t * start, b_blocknr_t finish,
349 int min, int max, int unfm, sector_t file_block)
350{
351 int nr_allocated = 0;
352 struct super_block *s = th->t_super;
353 unsigned int bm, off;
354 unsigned int end_bm, end_off;
355 unsigned int off_max = s->s_blocksize << 3;
356
357 BUG_ON(!th->t_trans_id);
358 PROC_INFO_INC(s, scan_bitmap.call);
359
360 /* No point in looking for more free blocks */
361 if (SB_FREE_BLOCKS(s) <= 0)
362 return 0;
363
364 get_bit_address(s, *start, &bm, &off);
365 get_bit_address(s, finish, &end_bm, &end_off);
366 if (bm > reiserfs_bmap_count(s))
367 return 0;
368 if (end_bm > reiserfs_bmap_count(s))
369 end_bm = reiserfs_bmap_count(s);
370
371 /*
372 * When the bitmap is more than 10% free, anyone can allocate.
373 * When it's less than 10% free, only files that already use the
374 * bitmap are allowed. Once we pass 80% full, this restriction
375 * is lifted.
376 *
377 * We do this so that files that grow later still have space close to
378 * their original allocation. This improves locality, and presumably
379 * performance as a result.
380 *
381 * This is only an allocation policy and does not make up for getting a
382 * bad hint. Decent hinting must be implemented for this to work well.
383 */
384 if (TEST_OPTION(skip_busy, s)
385 && SB_FREE_BLOCKS(s) > SB_BLOCK_COUNT(s) / 20) {
386 for (; bm < end_bm; bm++, off = 0) {
387 if ((off && (!unfm || (file_block != 0)))
388 || SB_AP_BITMAP(s)[bm].free_count >
389 (s->s_blocksize << 3) / 10)
390 nr_allocated =
391 scan_bitmap_block(th, bm, &off, off_max,
392 min, max, unfm);
393 if (nr_allocated)
394 goto ret;
395 }
396 /* we know from above that start is a reasonable number */
397 get_bit_address(s, *start, &bm, &off);
398 }
399
400 for (; bm < end_bm; bm++, off = 0) {
401 nr_allocated =
402 scan_bitmap_block(th, bm, &off, off_max, min, max, unfm);
403 if (nr_allocated)
404 goto ret;
405 }
406
407 nr_allocated =
408 scan_bitmap_block(th, bm, &off, end_off + 1, min, max, unfm);
409
410ret:
411 *start = bm * off_max + off;
412 return nr_allocated;
413
414}
415
416static void _reiserfs_free_block(struct reiserfs_transaction_handle *th,
417 struct inode *inode, b_blocknr_t block,
418 int for_unformatted)
419{
420 struct super_block *s = th->t_super;
421 struct reiserfs_super_block *rs;
422 struct buffer_head *sbh, *bmbh;
423 struct reiserfs_bitmap_info *apbi;
424 unsigned int nr, offset;
425
426 BUG_ON(!th->t_trans_id);
427 PROC_INFO_INC(s, free_block);
428 rs = SB_DISK_SUPER_BLOCK(s);
429 sbh = SB_BUFFER_WITH_SB(s);
430 apbi = SB_AP_BITMAP(s);
431
432 get_bit_address(s, block, &nr, &offset);
433
434 if (nr >= reiserfs_bmap_count(s)) {
435 reiserfs_error(s, "vs-4075", "block %lu is out of range",
436 block);
437 return;
438 }
439
440 bmbh = reiserfs_read_bitmap_block(s, nr);
441 if (!bmbh)
442 return;
443
444 reiserfs_prepare_for_journal(s, bmbh, 1);
445
446 /* clear bit for the given block in bit map */
447 if (!reiserfs_test_and_clear_le_bit(offset, bmbh->b_data)) {
448 reiserfs_error(s, "vs-4080",
449 "block %lu: bit already cleared", block);
450 }
451 apbi[nr].free_count++;
452 journal_mark_dirty(th, bmbh);
453 brelse(bmbh);
454
455 reiserfs_prepare_for_journal(s, sbh, 1);
456 /* update super block */
457 set_sb_free_blocks(rs, sb_free_blocks(rs) + 1);
458
459 journal_mark_dirty(th, sbh);
460 if (for_unformatted) {
461 int depth = reiserfs_write_unlock_nested(s);
462 dquot_free_block_nodirty(inode, 1);
463 reiserfs_write_lock_nested(s, depth);
464 }
465}
466
467void reiserfs_free_block(struct reiserfs_transaction_handle *th,
468 struct inode *inode, b_blocknr_t block,
469 int for_unformatted)
470{
471 struct super_block *s = th->t_super;
472
473 BUG_ON(!th->t_trans_id);
474 RFALSE(!s, "vs-4061: trying to free block on nonexistent device");
475 if (!is_reusable(s, block, 1))
476 return;
477
478 if (block > sb_block_count(REISERFS_SB(s)->s_rs)) {
479 reiserfs_error(th->t_super, "bitmap-4072",
480 "Trying to free block outside file system "
481 "boundaries (%lu > %lu)",
482 block, sb_block_count(REISERFS_SB(s)->s_rs));
483 return;
484 }
485 /* mark it before we clear it, just in case */
486 journal_mark_freed(th, s, block);
487 _reiserfs_free_block(th, inode, block, for_unformatted);
488}
489
490/* preallocated blocks don't need to be run through journal_mark_freed */
491static void reiserfs_free_prealloc_block(struct reiserfs_transaction_handle *th,
492 struct inode *inode, b_blocknr_t block)
493{
494 BUG_ON(!th->t_trans_id);
495 RFALSE(!th->t_super,
496 "vs-4060: trying to free block on nonexistent device");
497 if (!is_reusable(th->t_super, block, 1))
498 return;
499 _reiserfs_free_block(th, inode, block, 1);
500}
501
502static void __discard_prealloc(struct reiserfs_transaction_handle *th,
503 struct reiserfs_inode_info *ei)
504{
505 unsigned long save = ei->i_prealloc_block;
506 int dirty = 0;
507 struct inode *inode = &ei->vfs_inode;
508
509 BUG_ON(!th->t_trans_id);
510#ifdef CONFIG_REISERFS_CHECK
511 if (ei->i_prealloc_count < 0)
512 reiserfs_error(th->t_super, "zam-4001",
513 "inode has negative prealloc blocks count.");
514#endif
515 while (ei->i_prealloc_count > 0) {
516 b_blocknr_t block_to_free;
517
518 /*
519 * reiserfs_free_prealloc_block can drop the write lock,
520 * which could allow another caller to free the same block.
521 * We can protect against it by modifying the prealloc
522 * state before calling it.
523 */
524 block_to_free = ei->i_prealloc_block++;
525 ei->i_prealloc_count--;
526 reiserfs_free_prealloc_block(th, inode, block_to_free);
527 dirty = 1;
528 }
529 if (dirty)
530 reiserfs_update_sd(th, inode);
531 ei->i_prealloc_block = save;
532 list_del_init(&ei->i_prealloc_list);
533}
534
535/* FIXME: It should be inline function */
536void reiserfs_discard_prealloc(struct reiserfs_transaction_handle *th,
537 struct inode *inode)
538{
539 struct reiserfs_inode_info *ei = REISERFS_I(inode);
540
541 BUG_ON(!th->t_trans_id);
542 if (ei->i_prealloc_count)
543 __discard_prealloc(th, ei);
544}
545
546void reiserfs_discard_all_prealloc(struct reiserfs_transaction_handle *th)
547{
548 struct list_head *plist = &SB_JOURNAL(th->t_super)->j_prealloc_list;
549
550 BUG_ON(!th->t_trans_id);
551 while (!list_empty(plist)) {
552 struct reiserfs_inode_info *ei;
553 ei = list_entry(plist->next, struct reiserfs_inode_info,
554 i_prealloc_list);
555#ifdef CONFIG_REISERFS_CHECK
556 if (!ei->i_prealloc_count) {
557 reiserfs_error(th->t_super, "zam-4001",
558 "inode is in prealloc list but has "
559 "no preallocated blocks.");
560 }
561#endif
562 __discard_prealloc(th, ei);
563 }
564}
565
566void reiserfs_init_alloc_options(struct super_block *s)
567{
568 set_bit(_ALLOC_skip_busy, &SB_ALLOC_OPTS(s));
569 set_bit(_ALLOC_dirid_groups, &SB_ALLOC_OPTS(s));
570 set_bit(_ALLOC_packing_groups, &SB_ALLOC_OPTS(s));
571}
572
573/* block allocator related options are parsed here */
574int reiserfs_parse_alloc_options(struct super_block *s, char *options)
575{
576 char *this_char, *value;
577
578 /* clear default settings */
579 REISERFS_SB(s)->s_alloc_options.bits = 0;
580
581 while ((this_char = strsep(&options, ":")) != NULL) {
582 if ((value = strchr(this_char, '=')) != NULL)
583 *value++ = 0;
584
585 if (!strcmp(this_char, "concentrating_formatted_nodes")) {
586 int temp;
587 SET_OPTION(concentrating_formatted_nodes);
588 temp = (value
589 && *value) ? simple_strtoul(value, &value,
590 0) : 10;
591 if (temp <= 0 || temp > 100) {
592 REISERFS_SB(s)->s_alloc_options.border = 10;
593 } else {
594 REISERFS_SB(s)->s_alloc_options.border =
595 100 / temp;
596 }
597 continue;
598 }
599 if (!strcmp(this_char, "displacing_large_files")) {
600 SET_OPTION(displacing_large_files);
601 REISERFS_SB(s)->s_alloc_options.large_file_size =
602 (value
603 && *value) ? simple_strtoul(value, &value, 0) : 16;
604 continue;
605 }
606 if (!strcmp(this_char, "displacing_new_packing_localities")) {
607 SET_OPTION(displacing_new_packing_localities);
608 continue;
609 }
610
611 if (!strcmp(this_char, "old_hashed_relocation")) {
612 SET_OPTION(old_hashed_relocation);
613 continue;
614 }
615
616 if (!strcmp(this_char, "new_hashed_relocation")) {
617 SET_OPTION(new_hashed_relocation);
618 continue;
619 }
620
621 if (!strcmp(this_char, "dirid_groups")) {
622 SET_OPTION(dirid_groups);
623 continue;
624 }
625 if (!strcmp(this_char, "oid_groups")) {
626 SET_OPTION(oid_groups);
627 continue;
628 }
629 if (!strcmp(this_char, "packing_groups")) {
630 SET_OPTION(packing_groups);
631 continue;
632 }
633 if (!strcmp(this_char, "hashed_formatted_nodes")) {
634 SET_OPTION(hashed_formatted_nodes);
635 continue;
636 }
637
638 if (!strcmp(this_char, "skip_busy")) {
639 SET_OPTION(skip_busy);
640 continue;
641 }
642
643 if (!strcmp(this_char, "hundredth_slices")) {
644 SET_OPTION(hundredth_slices);
645 continue;
646 }
647
648 if (!strcmp(this_char, "old_way")) {
649 SET_OPTION(old_way);
650 continue;
651 }
652
653 if (!strcmp(this_char, "displace_based_on_dirid")) {
654 SET_OPTION(displace_based_on_dirid);
655 continue;
656 }
657
658 if (!strcmp(this_char, "preallocmin")) {
659 REISERFS_SB(s)->s_alloc_options.preallocmin =
660 (value
661 && *value) ? simple_strtoul(value, &value, 0) : 4;
662 continue;
663 }
664
665 if (!strcmp(this_char, "preallocsize")) {
666 REISERFS_SB(s)->s_alloc_options.preallocsize =
667 (value
668 && *value) ? simple_strtoul(value, &value,
669 0) :
670 PREALLOCATION_SIZE;
671 continue;
672 }
673
674 reiserfs_warning(s, "zam-4001", "unknown option - %s",
675 this_char);
676 return 1;
677 }
678
679 reiserfs_info(s, "allocator options = [%08x]\n", SB_ALLOC_OPTS(s));
680 return 0;
681}
682
683static void print_sep(struct seq_file *seq, int *first)
684{
685 if (!*first)
686 seq_puts(seq, ":");
687 else
688 *first = 0;
689}
690
691void show_alloc_options(struct seq_file *seq, struct super_block *s)
692{
693 int first = 1;
694
695 if (SB_ALLOC_OPTS(s) == ((1 << _ALLOC_skip_busy) |
696 (1 << _ALLOC_dirid_groups) | (1 << _ALLOC_packing_groups)))
697 return;
698
699 seq_puts(seq, ",alloc=");
700
701 if (TEST_OPTION(concentrating_formatted_nodes, s)) {
702 print_sep(seq, &first);
703 if (REISERFS_SB(s)->s_alloc_options.border != 10) {
704 seq_printf(seq, "concentrating_formatted_nodes=%d",
705 100 / REISERFS_SB(s)->s_alloc_options.border);
706 } else
707 seq_puts(seq, "concentrating_formatted_nodes");
708 }
709 if (TEST_OPTION(displacing_large_files, s)) {
710 print_sep(seq, &first);
711 if (REISERFS_SB(s)->s_alloc_options.large_file_size != 16) {
712 seq_printf(seq, "displacing_large_files=%lu",
713 REISERFS_SB(s)->s_alloc_options.large_file_size);
714 } else
715 seq_puts(seq, "displacing_large_files");
716 }
717 if (TEST_OPTION(displacing_new_packing_localities, s)) {
718 print_sep(seq, &first);
719 seq_puts(seq, "displacing_new_packing_localities");
720 }
721 if (TEST_OPTION(old_hashed_relocation, s)) {
722 print_sep(seq, &first);
723 seq_puts(seq, "old_hashed_relocation");
724 }
725 if (TEST_OPTION(new_hashed_relocation, s)) {
726 print_sep(seq, &first);
727 seq_puts(seq, "new_hashed_relocation");
728 }
729 if (TEST_OPTION(dirid_groups, s)) {
730 print_sep(seq, &first);
731 seq_puts(seq, "dirid_groups");
732 }
733 if (TEST_OPTION(oid_groups, s)) {
734 print_sep(seq, &first);
735 seq_puts(seq, "oid_groups");
736 }
737 if (TEST_OPTION(packing_groups, s)) {
738 print_sep(seq, &first);
739 seq_puts(seq, "packing_groups");
740 }
741 if (TEST_OPTION(hashed_formatted_nodes, s)) {
742 print_sep(seq, &first);
743 seq_puts(seq, "hashed_formatted_nodes");
744 }
745 if (TEST_OPTION(skip_busy, s)) {
746 print_sep(seq, &first);
747 seq_puts(seq, "skip_busy");
748 }
749 if (TEST_OPTION(hundredth_slices, s)) {
750 print_sep(seq, &first);
751 seq_puts(seq, "hundredth_slices");
752 }
753 if (TEST_OPTION(old_way, s)) {
754 print_sep(seq, &first);
755 seq_puts(seq, "old_way");
756 }
757 if (TEST_OPTION(displace_based_on_dirid, s)) {
758 print_sep(seq, &first);
759 seq_puts(seq, "displace_based_on_dirid");
760 }
761 if (REISERFS_SB(s)->s_alloc_options.preallocmin != 0) {
762 print_sep(seq, &first);
763 seq_printf(seq, "preallocmin=%d",
764 REISERFS_SB(s)->s_alloc_options.preallocmin);
765 }
766 if (REISERFS_SB(s)->s_alloc_options.preallocsize != 17) {
767 print_sep(seq, &first);
768 seq_printf(seq, "preallocsize=%d",
769 REISERFS_SB(s)->s_alloc_options.preallocsize);
770 }
771}
772
773static inline void new_hashed_relocation(reiserfs_blocknr_hint_t * hint)
774{
775 char *hash_in;
776
777 if (hint->formatted_node) {
778 hash_in = (char *)&hint->key.k_dir_id;
779 } else {
780 if (!hint->inode) {
781 /*hint->search_start = hint->beg;*/
782 hash_in = (char *)&hint->key.k_dir_id;
783 } else
784 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
785 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
786 else
787 hash_in =
788 (char *)(&INODE_PKEY(hint->inode)->k_objectid);
789 }
790
791 hint->search_start =
792 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
793}
794
795/*
796 * Relocation based on dirid, hashing them into a given bitmap block
797 * files. Formatted nodes are unaffected, a separate policy covers them
798 */
799static void dirid_groups(reiserfs_blocknr_hint_t * hint)
800{
801 unsigned long hash;
802 __u32 dirid = 0;
803 int bm = 0;
804 struct super_block *sb = hint->th->t_super;
805
806 if (hint->inode)
807 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
808 else if (hint->formatted_node)
809 dirid = hint->key.k_dir_id;
810
811 if (dirid) {
812 bm = bmap_hash_id(sb, dirid);
813 hash = bm * (sb->s_blocksize << 3);
814 /* give a portion of the block group to metadata */
815 if (hint->inode)
816 hash += sb->s_blocksize / 2;
817 hint->search_start = hash;
818 }
819}
820
821/*
822 * Relocation based on oid, hashing them into a given bitmap block
823 * files. Formatted nodes are unaffected, a separate policy covers them
824 */
825static void oid_groups(reiserfs_blocknr_hint_t * hint)
826{
827 if (hint->inode) {
828 unsigned long hash;
829 __u32 oid;
830 __u32 dirid;
831 int bm;
832
833 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
834
835 /*
836 * keep the root dir and it's first set of subdirs close to
837 * the start of the disk
838 */
839 if (dirid <= 2)
840 hash = (hint->inode->i_sb->s_blocksize << 3);
841 else {
842 oid = le32_to_cpu(INODE_PKEY(hint->inode)->k_objectid);
843 bm = bmap_hash_id(hint->inode->i_sb, oid);
844 hash = bm * (hint->inode->i_sb->s_blocksize << 3);
845 }
846 hint->search_start = hash;
847 }
848}
849
850/*
851 * returns 1 if it finds an indirect item and gets valid hint info
852 * from it, otherwise 0
853 */
854static int get_left_neighbor(reiserfs_blocknr_hint_t * hint)
855{
856 struct treepath *path;
857 struct buffer_head *bh;
858 struct item_head *ih;
859 int pos_in_item;
860 __le32 *item;
861 int ret = 0;
862
863 /*
864 * reiserfs code can call this function w/o pointer to path
865 * structure supplied; then we rely on supplied search_start
866 */
867 if (!hint->path)
868 return 0;
869
870 path = hint->path;
871 bh = get_last_bh(path);
872 RFALSE(!bh, "green-4002: Illegal path specified to get_left_neighbor");
873 ih = tp_item_head(path);
874 pos_in_item = path->pos_in_item;
875 item = tp_item_body(path);
876
877 hint->search_start = bh->b_blocknr;
878
879 /*
880 * for indirect item: go to left and look for the first non-hole entry
881 * in the indirect item
882 */
883 if (!hint->formatted_node && is_indirect_le_ih(ih)) {
884 if (pos_in_item == I_UNFM_NUM(ih))
885 pos_in_item--;
886 while (pos_in_item >= 0) {
887 int t = get_block_num(item, pos_in_item);
888 if (t) {
889 hint->search_start = t;
890 ret = 1;
891 break;
892 }
893 pos_in_item--;
894 }
895 }
896
897 /* does result value fit into specified region? */
898 return ret;
899}
900
901/*
902 * should be, if formatted node, then try to put on first part of the device
903 * specified as number of percent with mount option device, else try to put
904 * on last of device. This is not to say it is good code to do so,
905 * but the effect should be measured.
906 */
907static inline void set_border_in_hint(struct super_block *s,
908 reiserfs_blocknr_hint_t * hint)
909{
910 b_blocknr_t border =
911 SB_BLOCK_COUNT(s) / REISERFS_SB(s)->s_alloc_options.border;
912
913 if (hint->formatted_node)
914 hint->end = border - 1;
915 else
916 hint->beg = border;
917}
918
919static inline void displace_large_file(reiserfs_blocknr_hint_t * hint)
920{
921 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
922 hint->search_start =
923 hint->beg +
924 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_dir_id),
925 4) % (hint->end - hint->beg);
926 else
927 hint->search_start =
928 hint->beg +
929 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_objectid),
930 4) % (hint->end - hint->beg);
931}
932
933static inline void hash_formatted_node(reiserfs_blocknr_hint_t * hint)
934{
935 char *hash_in;
936
937 if (!hint->inode)
938 hash_in = (char *)&hint->key.k_dir_id;
939 else if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
940 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
941 else
942 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_objectid);
943
944 hint->search_start =
945 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
946}
947
948static inline int
949this_blocknr_allocation_would_make_it_a_large_file(reiserfs_blocknr_hint_t *
950 hint)
951{
952 return hint->block ==
953 REISERFS_SB(hint->th->t_super)->s_alloc_options.large_file_size;
954}
955
956#ifdef DISPLACE_NEW_PACKING_LOCALITIES
957static inline void displace_new_packing_locality(reiserfs_blocknr_hint_t * hint)
958{
959 struct in_core_key *key = &hint->key;
960
961 hint->th->displace_new_blocks = 0;
962 hint->search_start =
963 hint->beg + keyed_hash((char *)(&key->k_objectid),
964 4) % (hint->end - hint->beg);
965}
966#endif
967
968static inline int old_hashed_relocation(reiserfs_blocknr_hint_t * hint)
969{
970 b_blocknr_t border;
971 u32 hash_in;
972
973 if (hint->formatted_node || hint->inode == NULL) {
974 return 0;
975 }
976
977 hash_in = le32_to_cpu((INODE_PKEY(hint->inode))->k_dir_id);
978 border =
979 hint->beg + (u32) keyed_hash(((char *)(&hash_in)),
980 4) % (hint->end - hint->beg - 1);
981 if (border > hint->search_start)
982 hint->search_start = border;
983
984 return 1;
985}
986
987static inline int old_way(reiserfs_blocknr_hint_t * hint)
988{
989 b_blocknr_t border;
990
991 if (hint->formatted_node || hint->inode == NULL) {
992 return 0;
993 }
994
995 border =
996 hint->beg +
997 le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id) % (hint->end -
998 hint->beg);
999 if (border > hint->search_start)
1000 hint->search_start = border;
1001
1002 return 1;
1003}
1004
1005static inline void hundredth_slices(reiserfs_blocknr_hint_t * hint)
1006{
1007 struct in_core_key *key = &hint->key;
1008 b_blocknr_t slice_start;
1009
1010 slice_start =
1011 (keyed_hash((char *)(&key->k_dir_id), 4) % 100) * (hint->end / 100);
1012 if (slice_start > hint->search_start
1013 || slice_start + (hint->end / 100) <= hint->search_start) {
1014 hint->search_start = slice_start;
1015 }
1016}
1017
1018static void determine_search_start(reiserfs_blocknr_hint_t * hint,
1019 int amount_needed)
1020{
1021 struct super_block *s = hint->th->t_super;
1022 int unfm_hint;
1023
1024 hint->beg = 0;
1025 hint->end = SB_BLOCK_COUNT(s) - 1;
1026
1027 /* This is former border algorithm. Now with tunable border offset */
1028 if (concentrating_formatted_nodes(s))
1029 set_border_in_hint(s, hint);
1030
1031#ifdef DISPLACE_NEW_PACKING_LOCALITIES
1032 /*
1033 * whenever we create a new directory, we displace it. At first
1034 * we will hash for location, later we might look for a moderately
1035 * empty place for it
1036 */
1037 if (displacing_new_packing_localities(s)
1038 && hint->th->displace_new_blocks) {
1039 displace_new_packing_locality(hint);
1040
1041 /*
1042 * we do not continue determine_search_start,
1043 * if new packing locality is being displaced
1044 */
1045 return;
1046 }
1047#endif
1048
1049 /*
1050 * all persons should feel encouraged to add more special cases
1051 * here and test them
1052 */
1053
1054 if (displacing_large_files(s) && !hint->formatted_node
1055 && this_blocknr_allocation_would_make_it_a_large_file(hint)) {
1056 displace_large_file(hint);
1057 return;
1058 }
1059
1060 /*
1061 * if none of our special cases is relevant, use the left
1062 * neighbor in the tree order of the new node we are allocating for
1063 */
1064 if (hint->formatted_node && TEST_OPTION(hashed_formatted_nodes, s)) {
1065 hash_formatted_node(hint);
1066 return;
1067 }
1068
1069 unfm_hint = get_left_neighbor(hint);
1070
1071 /*
1072 * Mimic old block allocator behaviour, that is if VFS allowed for
1073 * preallocation, new blocks are displaced based on directory ID.
1074 * Also, if suggested search_start is less than last preallocated
1075 * block, we start searching from it, assuming that HDD dataflow
1076 * is faster in forward direction
1077 */
1078 if (TEST_OPTION(old_way, s)) {
1079 if (!hint->formatted_node) {
1080 if (!reiserfs_hashed_relocation(s))
1081 old_way(hint);
1082 else if (!reiserfs_no_unhashed_relocation(s))
1083 old_hashed_relocation(hint);
1084
1085 if (hint->inode
1086 && hint->search_start <
1087 REISERFS_I(hint->inode)->i_prealloc_block)
1088 hint->search_start =
1089 REISERFS_I(hint->inode)->i_prealloc_block;
1090 }
1091 return;
1092 }
1093
1094 /* This is an approach proposed by Hans */
1095 if (TEST_OPTION(hundredth_slices, s)
1096 && !(displacing_large_files(s) && !hint->formatted_node)) {
1097 hundredth_slices(hint);
1098 return;
1099 }
1100
1101 /* old_hashed_relocation only works on unformatted */
1102 if (!unfm_hint && !hint->formatted_node &&
1103 TEST_OPTION(old_hashed_relocation, s)) {
1104 old_hashed_relocation(hint);
1105 }
1106
1107 /* new_hashed_relocation works with both formatted/unformatted nodes */
1108 if ((!unfm_hint || hint->formatted_node) &&
1109 TEST_OPTION(new_hashed_relocation, s)) {
1110 new_hashed_relocation(hint);
1111 }
1112
1113 /* dirid grouping works only on unformatted nodes */
1114 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
1115 dirid_groups(hint);
1116 }
1117#ifdef DISPLACE_NEW_PACKING_LOCALITIES
1118 if (hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
1119 dirid_groups(hint);
1120 }
1121#endif
1122
1123 /* oid grouping works only on unformatted nodes */
1124 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(oid_groups, s)) {
1125 oid_groups(hint);
1126 }
1127 return;
1128}
1129
1130static int determine_prealloc_size(reiserfs_blocknr_hint_t * hint)
1131{
1132 /* make minimum size a mount option and benchmark both ways */
1133 /* we preallocate blocks only for regular files, specific size */
1134 /* benchmark preallocating always and see what happens */
1135
1136 hint->prealloc_size = 0;
1137
1138 if (!hint->formatted_node && hint->preallocate) {
1139 if (S_ISREG(hint->inode->i_mode) && !IS_PRIVATE(hint->inode)
1140 && hint->inode->i_size >=
1141 REISERFS_SB(hint->th->t_super)->s_alloc_options.
1142 preallocmin * hint->inode->i_sb->s_blocksize)
1143 hint->prealloc_size =
1144 REISERFS_SB(hint->th->t_super)->s_alloc_options.
1145 preallocsize - 1;
1146 }
1147 return CARRY_ON;
1148}
1149
1150static inline int allocate_without_wrapping_disk(reiserfs_blocknr_hint_t * hint,
1151 b_blocknr_t * new_blocknrs,
1152 b_blocknr_t start,
1153 b_blocknr_t finish, int min,
1154 int amount_needed,
1155 int prealloc_size)
1156{
1157 int rest = amount_needed;
1158 int nr_allocated;
1159
1160 while (rest > 0 && start <= finish) {
1161 nr_allocated = scan_bitmap(hint->th, &start, finish, min,
1162 rest + prealloc_size,
1163 !hint->formatted_node, hint->block);
1164
1165 if (nr_allocated == 0) /* no new blocks allocated, return */
1166 break;
1167
1168 /* fill free_blocknrs array first */
1169 while (rest > 0 && nr_allocated > 0) {
1170 *new_blocknrs++ = start++;
1171 rest--;
1172 nr_allocated--;
1173 }
1174
1175 /* do we have something to fill prealloc. array also ? */
1176 if (nr_allocated > 0) {
1177 /*
1178 * it means prealloc_size was greater that 0 and
1179 * we do preallocation
1180 */
1181 list_add(&REISERFS_I(hint->inode)->i_prealloc_list,
1182 &SB_JOURNAL(hint->th->t_super)->
1183 j_prealloc_list);
1184 REISERFS_I(hint->inode)->i_prealloc_block = start;
1185 REISERFS_I(hint->inode)->i_prealloc_count =
1186 nr_allocated;
1187 break;
1188 }
1189 }
1190
1191 return (amount_needed - rest);
1192}
1193
1194static inline int blocknrs_and_prealloc_arrays_from_search_start
1195 (reiserfs_blocknr_hint_t * hint, b_blocknr_t * new_blocknrs,
1196 int amount_needed) {
1197 struct super_block *s = hint->th->t_super;
1198 b_blocknr_t start = hint->search_start;
1199 b_blocknr_t finish = SB_BLOCK_COUNT(s) - 1;
1200 int passno = 0;
1201 int nr_allocated = 0;
1202 int depth;
1203
1204 determine_prealloc_size(hint);
1205 if (!hint->formatted_node) {
1206 int quota_ret;
1207#ifdef REISERQUOTA_DEBUG
1208 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1209 "reiserquota: allocating %d blocks id=%u",
1210 amount_needed, hint->inode->i_uid);
1211#endif
1212 depth = reiserfs_write_unlock_nested(s);
1213 quota_ret =
1214 dquot_alloc_block_nodirty(hint->inode, amount_needed);
1215 if (quota_ret) { /* Quota exceeded? */
1216 reiserfs_write_lock_nested(s, depth);
1217 return QUOTA_EXCEEDED;
1218 }
1219 if (hint->preallocate && hint->prealloc_size) {
1220#ifdef REISERQUOTA_DEBUG
1221 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1222 "reiserquota: allocating (prealloc) %d blocks id=%u",
1223 hint->prealloc_size, hint->inode->i_uid);
1224#endif
1225 quota_ret = dquot_prealloc_block_nodirty(hint->inode,
1226 hint->prealloc_size);
1227 if (quota_ret)
1228 hint->preallocate = hint->prealloc_size = 0;
1229 }
1230 /* for unformatted nodes, force large allocations */
1231 reiserfs_write_lock_nested(s, depth);
1232 }
1233
1234 do {
1235 switch (passno++) {
1236 case 0: /* Search from hint->search_start to end of disk */
1237 start = hint->search_start;
1238 finish = SB_BLOCK_COUNT(s) - 1;
1239 break;
1240 case 1: /* Search from hint->beg to hint->search_start */
1241 start = hint->beg;
1242 finish = hint->search_start;
1243 break;
1244 case 2: /* Last chance: Search from 0 to hint->beg */
1245 start = 0;
1246 finish = hint->beg;
1247 break;
1248 default:
1249 /* We've tried searching everywhere, not enough space */
1250 /* Free the blocks */
1251 if (!hint->formatted_node) {
1252#ifdef REISERQUOTA_DEBUG
1253 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1254 "reiserquota: freeing (nospace) %d blocks id=%u",
1255 amount_needed +
1256 hint->prealloc_size -
1257 nr_allocated,
1258 hint->inode->i_uid);
1259#endif
1260 /* Free not allocated blocks */
1261 depth = reiserfs_write_unlock_nested(s);
1262 dquot_free_block_nodirty(hint->inode,
1263 amount_needed + hint->prealloc_size -
1264 nr_allocated);
1265 reiserfs_write_lock_nested(s, depth);
1266 }
1267 while (nr_allocated--)
1268 reiserfs_free_block(hint->th, hint->inode,
1269 new_blocknrs[nr_allocated],
1270 !hint->formatted_node);
1271
1272 return NO_DISK_SPACE;
1273 }
1274 } while ((nr_allocated += allocate_without_wrapping_disk(hint,
1275 new_blocknrs +
1276 nr_allocated,
1277 start, finish,
1278 1,
1279 amount_needed -
1280 nr_allocated,
1281 hint->
1282 prealloc_size))
1283 < amount_needed);
1284 if (!hint->formatted_node &&
1285 amount_needed + hint->prealloc_size >
1286 nr_allocated + REISERFS_I(hint->inode)->i_prealloc_count) {
1287 /* Some of preallocation blocks were not allocated */
1288#ifdef REISERQUOTA_DEBUG
1289 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1290 "reiserquota: freeing (failed prealloc) %d blocks id=%u",
1291 amount_needed + hint->prealloc_size -
1292 nr_allocated -
1293 REISERFS_I(hint->inode)->i_prealloc_count,
1294 hint->inode->i_uid);
1295#endif
1296
1297 depth = reiserfs_write_unlock_nested(s);
1298 dquot_free_block_nodirty(hint->inode, amount_needed +
1299 hint->prealloc_size - nr_allocated -
1300 REISERFS_I(hint->inode)->
1301 i_prealloc_count);
1302 reiserfs_write_lock_nested(s, depth);
1303 }
1304
1305 return CARRY_ON;
1306}
1307
1308/* grab new blocknrs from preallocated list */
1309/* return amount still needed after using them */
1310static int use_preallocated_list_if_available(reiserfs_blocknr_hint_t * hint,
1311 b_blocknr_t * new_blocknrs,
1312 int amount_needed)
1313{
1314 struct inode *inode = hint->inode;
1315
1316 if (REISERFS_I(inode)->i_prealloc_count > 0) {
1317 while (amount_needed) {
1318
1319 *new_blocknrs++ = REISERFS_I(inode)->i_prealloc_block++;
1320 REISERFS_I(inode)->i_prealloc_count--;
1321
1322 amount_needed--;
1323
1324 if (REISERFS_I(inode)->i_prealloc_count <= 0) {
1325 list_del(&REISERFS_I(inode)->i_prealloc_list);
1326 break;
1327 }
1328 }
1329 }
1330 /* return amount still needed after using preallocated blocks */
1331 return amount_needed;
1332}
1333
1334int reiserfs_allocate_blocknrs(reiserfs_blocknr_hint_t *hint,
1335 b_blocknr_t *new_blocknrs,
1336 int amount_needed,
1337 /* Amount of blocks we have already reserved */
1338 int reserved_by_us)
1339{
1340 int initial_amount_needed = amount_needed;
1341 int ret;
1342 struct super_block *s = hint->th->t_super;
1343
1344 /* Check if there is enough space, taking into account reserved space */
1345 if (SB_FREE_BLOCKS(s) - REISERFS_SB(s)->reserved_blocks <
1346 amount_needed - reserved_by_us)
1347 return NO_DISK_SPACE;
1348 /* should this be if !hint->inode && hint->preallocate? */
1349 /* do you mean hint->formatted_node can be removed ? - Zam */
1350 /*
1351 * hint->formatted_node cannot be removed because we try to access
1352 * inode information here, and there is often no inode associated with
1353 * metadata allocations - green
1354 */
1355
1356 if (!hint->formatted_node && hint->preallocate) {
1357 amount_needed = use_preallocated_list_if_available
1358 (hint, new_blocknrs, amount_needed);
1359
1360 /*
1361 * We have all the block numbers we need from the
1362 * prealloc list
1363 */
1364 if (amount_needed == 0)
1365 return CARRY_ON;
1366 new_blocknrs += (initial_amount_needed - amount_needed);
1367 }
1368
1369 /* find search start and save it in hint structure */
1370 determine_search_start(hint, amount_needed);
1371 if (hint->search_start >= SB_BLOCK_COUNT(s))
1372 hint->search_start = SB_BLOCK_COUNT(s) - 1;
1373
1374 /* allocation itself; fill new_blocknrs and preallocation arrays */
1375 ret = blocknrs_and_prealloc_arrays_from_search_start
1376 (hint, new_blocknrs, amount_needed);
1377
1378 /*
1379 * We used prealloc. list to fill (partially) new_blocknrs array.
1380 * If final allocation fails we need to return blocks back to
1381 * prealloc. list or just free them. -- Zam (I chose second
1382 * variant)
1383 */
1384 if (ret != CARRY_ON) {
1385 while (amount_needed++ < initial_amount_needed) {
1386 reiserfs_free_block(hint->th, hint->inode,
1387 *(--new_blocknrs), 1);
1388 }
1389 }
1390 return ret;
1391}
1392
1393void reiserfs_cache_bitmap_metadata(struct super_block *sb,
1394 struct buffer_head *bh,
1395 struct reiserfs_bitmap_info *info)
1396{
1397 unsigned long *cur = (unsigned long *)(bh->b_data + bh->b_size);
1398
1399 /* The first bit must ALWAYS be 1 */
1400 if (!reiserfs_test_le_bit(0, (unsigned long *)bh->b_data))
1401 reiserfs_error(sb, "reiserfs-2025", "bitmap block %lu is "
1402 "corrupted: first bit must be 1", bh->b_blocknr);
1403
1404 info->free_count = 0;
1405
1406 while (--cur >= (unsigned long *)bh->b_data) {
1407 /* 0 and ~0 are special, we can optimize for them */
1408 if (*cur == 0)
1409 info->free_count += BITS_PER_LONG;
1410 else if (*cur != ~0L) /* A mix, investigate */
1411 info->free_count += BITS_PER_LONG - hweight_long(*cur);
1412 }
1413}
1414
1415struct buffer_head *reiserfs_read_bitmap_block(struct super_block *sb,
1416 unsigned int bitmap)
1417{
1418 b_blocknr_t block = (sb->s_blocksize << 3) * bitmap;
1419 struct reiserfs_bitmap_info *info = SB_AP_BITMAP(sb) + bitmap;
1420 struct buffer_head *bh;
1421
1422 /*
1423 * Way old format filesystems had the bitmaps packed up front.
1424 * I doubt there are any of these left, but just in case...
1425 */
1426 if (unlikely(test_bit(REISERFS_OLD_FORMAT,
1427 &REISERFS_SB(sb)->s_properties)))
1428 block = REISERFS_SB(sb)->s_sbh->b_blocknr + 1 + bitmap;
1429 else if (bitmap == 0)
1430 block = (REISERFS_DISK_OFFSET_IN_BYTES >> sb->s_blocksize_bits) + 1;
1431
1432 bh = sb_bread(sb, block);
1433 if (bh == NULL)
1434 reiserfs_warning(sb, "sh-2029: %s: bitmap block (#%u) "
1435 "reading failed", __func__, block);
1436 else {
1437 if (buffer_locked(bh)) {
1438 int depth;
1439 PROC_INFO_INC(sb, scan_bitmap.wait);
1440 depth = reiserfs_write_unlock_nested(sb);
1441 __wait_on_buffer(bh);
1442 reiserfs_write_lock_nested(sb, depth);
1443 }
1444 BUG_ON(!buffer_uptodate(bh));
1445 BUG_ON(atomic_read(&bh->b_count) == 0);
1446
1447 if (info->free_count == UINT_MAX)
1448 reiserfs_cache_bitmap_metadata(sb, bh, info);
1449 }
1450
1451 return bh;
1452}
1453
1454int reiserfs_init_bitmap_cache(struct super_block *sb)
1455{
1456 struct reiserfs_bitmap_info *bitmap;
1457 unsigned int bmap_nr = reiserfs_bmap_count(sb);
1458
1459 bitmap = vmalloc(array_size(bmap_nr, sizeof(*bitmap)));
1460 if (bitmap == NULL)
1461 return -ENOMEM;
1462
1463 memset(bitmap, 0xff, sizeof(*bitmap) * bmap_nr);
1464
1465 SB_AP_BITMAP(sb) = bitmap;
1466
1467 return 0;
1468}
1469
1470void reiserfs_free_bitmap_cache(struct super_block *sb)
1471{
1472 if (SB_AP_BITMAP(sb)) {
1473 vfree(SB_AP_BITMAP(sb));
1474 SB_AP_BITMAP(sb) = NULL;
1475 }
1476}
1/*
2 * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README
3 */
4/* Reiserfs block (de)allocator, bitmap-based. */
5
6#include <linux/time.h>
7#include "reiserfs.h"
8#include <linux/errno.h>
9#include <linux/buffer_head.h>
10#include <linux/kernel.h>
11#include <linux/pagemap.h>
12#include <linux/vmalloc.h>
13#include <linux/quotaops.h>
14#include <linux/seq_file.h>
15
16#define PREALLOCATION_SIZE 9
17
18/* different reiserfs block allocator options */
19
20#define SB_ALLOC_OPTS(s) (REISERFS_SB(s)->s_alloc_options.bits)
21
22#define _ALLOC_concentrating_formatted_nodes 0
23#define _ALLOC_displacing_large_files 1
24#define _ALLOC_displacing_new_packing_localities 2
25#define _ALLOC_old_hashed_relocation 3
26#define _ALLOC_new_hashed_relocation 4
27#define _ALLOC_skip_busy 5
28#define _ALLOC_displace_based_on_dirid 6
29#define _ALLOC_hashed_formatted_nodes 7
30#define _ALLOC_old_way 8
31#define _ALLOC_hundredth_slices 9
32#define _ALLOC_dirid_groups 10
33#define _ALLOC_oid_groups 11
34#define _ALLOC_packing_groups 12
35
36#define concentrating_formatted_nodes(s) test_bit(_ALLOC_concentrating_formatted_nodes, &SB_ALLOC_OPTS(s))
37#define displacing_large_files(s) test_bit(_ALLOC_displacing_large_files, &SB_ALLOC_OPTS(s))
38#define displacing_new_packing_localities(s) test_bit(_ALLOC_displacing_new_packing_localities, &SB_ALLOC_OPTS(s))
39
40#define SET_OPTION(optname) \
41 do { \
42 reiserfs_info(s, "block allocator option \"%s\" is set", #optname); \
43 set_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s)); \
44 } while(0)
45#define TEST_OPTION(optname, s) \
46 test_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s))
47
48static inline void get_bit_address(struct super_block *s,
49 b_blocknr_t block,
50 unsigned int *bmap_nr,
51 unsigned int *offset)
52{
53 /* It is in the bitmap block number equal to the block
54 * number divided by the number of bits in a block. */
55 *bmap_nr = block >> (s->s_blocksize_bits + 3);
56 /* Within that bitmap block it is located at bit offset *offset. */
57 *offset = block & ((s->s_blocksize << 3) - 1);
58}
59
60int is_reusable(struct super_block *s, b_blocknr_t block, int bit_value)
61{
62 unsigned int bmap, offset;
63 unsigned int bmap_count = reiserfs_bmap_count(s);
64
65 if (block == 0 || block >= SB_BLOCK_COUNT(s)) {
66 reiserfs_error(s, "vs-4010",
67 "block number is out of range %lu (%u)",
68 block, SB_BLOCK_COUNT(s));
69 return 0;
70 }
71
72 get_bit_address(s, block, &bmap, &offset);
73
74 /* Old format filesystem? Unlikely, but the bitmaps are all up front so
75 * we need to account for it. */
76 if (unlikely(test_bit(REISERFS_OLD_FORMAT,
77 &(REISERFS_SB(s)->s_properties)))) {
78 b_blocknr_t bmap1 = REISERFS_SB(s)->s_sbh->b_blocknr + 1;
79 if (block >= bmap1 &&
80 block <= bmap1 + bmap_count) {
81 reiserfs_error(s, "vs-4019", "bitmap block %lu(%u) "
82 "can't be freed or reused",
83 block, bmap_count);
84 return 0;
85 }
86 } else {
87 if (offset == 0) {
88 reiserfs_error(s, "vs-4020", "bitmap block %lu(%u) "
89 "can't be freed or reused",
90 block, bmap_count);
91 return 0;
92 }
93 }
94
95 if (bmap >= bmap_count) {
96 reiserfs_error(s, "vs-4030", "bitmap for requested block "
97 "is out of range: block=%lu, bitmap_nr=%u",
98 block, bmap);
99 return 0;
100 }
101
102 if (bit_value == 0 && block == SB_ROOT_BLOCK(s)) {
103 reiserfs_error(s, "vs-4050", "this is root block (%u), "
104 "it must be busy", SB_ROOT_BLOCK(s));
105 return 0;
106 }
107
108 return 1;
109}
110
111/* searches in journal structures for a given block number (bmap, off). If block
112 is found in reiserfs journal it suggests next free block candidate to test. */
113static inline int is_block_in_journal(struct super_block *s, unsigned int bmap,
114 int off, int *next)
115{
116 b_blocknr_t tmp;
117
118 if (reiserfs_in_journal(s, bmap, off, 1, &tmp)) {
119 if (tmp) { /* hint supplied */
120 *next = tmp;
121 PROC_INFO_INC(s, scan_bitmap.in_journal_hint);
122 } else {
123 (*next) = off + 1; /* inc offset to avoid looping. */
124 PROC_INFO_INC(s, scan_bitmap.in_journal_nohint);
125 }
126 PROC_INFO_INC(s, scan_bitmap.retry);
127 return 1;
128 }
129 return 0;
130}
131
132/* it searches for a window of zero bits with given minimum and maximum lengths in one bitmap
133 * block; */
134static int scan_bitmap_block(struct reiserfs_transaction_handle *th,
135 unsigned int bmap_n, int *beg, int boundary,
136 int min, int max, int unfm)
137{
138 struct super_block *s = th->t_super;
139 struct reiserfs_bitmap_info *bi = &SB_AP_BITMAP(s)[bmap_n];
140 struct buffer_head *bh;
141 int end, next;
142 int org = *beg;
143
144 BUG_ON(!th->t_trans_id);
145
146 RFALSE(bmap_n >= reiserfs_bmap_count(s), "Bitmap %u is out of "
147 "range (0..%u)", bmap_n, reiserfs_bmap_count(s) - 1);
148 PROC_INFO_INC(s, scan_bitmap.bmap);
149/* this is unclear and lacks comments, explain how journal bitmaps
150 work here for the reader. Convey a sense of the design here. What
151 is a window? */
152/* - I mean `a window of zero bits' as in description of this function - Zam. */
153
154 if (!bi) {
155 reiserfs_error(s, "jdm-4055", "NULL bitmap info pointer "
156 "for bitmap %d", bmap_n);
157 return 0;
158 }
159
160 bh = reiserfs_read_bitmap_block(s, bmap_n);
161 if (bh == NULL)
162 return 0;
163
164 while (1) {
165 cont:
166 if (bi->free_count < min) {
167 brelse(bh);
168 return 0; // No free blocks in this bitmap
169 }
170
171 /* search for a first zero bit -- beginning of a window */
172 *beg = reiserfs_find_next_zero_le_bit
173 ((unsigned long *)(bh->b_data), boundary, *beg);
174
175 if (*beg + min > boundary) { /* search for a zero bit fails or the rest of bitmap block
176 * cannot contain a zero window of minimum size */
177 brelse(bh);
178 return 0;
179 }
180
181 if (unfm && is_block_in_journal(s, bmap_n, *beg, beg))
182 continue;
183 /* first zero bit found; we check next bits */
184 for (end = *beg + 1;; end++) {
185 if (end >= *beg + max || end >= boundary
186 || reiserfs_test_le_bit(end, bh->b_data)) {
187 next = end;
188 break;
189 }
190 /* finding the other end of zero bit window requires looking into journal structures (in
191 * case of searching for free blocks for unformatted nodes) */
192 if (unfm && is_block_in_journal(s, bmap_n, end, &next))
193 break;
194 }
195
196 /* now (*beg) points to beginning of zero bits window,
197 * (end) points to one bit after the window end */
198 if (end - *beg >= min) { /* it seems we have found window of proper size */
199 int i;
200 reiserfs_prepare_for_journal(s, bh, 1);
201 /* try to set all blocks used checking are they still free */
202 for (i = *beg; i < end; i++) {
203 /* It seems that we should not check in journal again. */
204 if (reiserfs_test_and_set_le_bit
205 (i, bh->b_data)) {
206 /* bit was set by another process
207 * while we slept in prepare_for_journal() */
208 PROC_INFO_INC(s, scan_bitmap.stolen);
209 if (i >= *beg + min) { /* we can continue with smaller set of allocated blocks,
210 * if length of this set is more or equal to `min' */
211 end = i;
212 break;
213 }
214 /* otherwise we clear all bit were set ... */
215 while (--i >= *beg)
216 reiserfs_clear_le_bit
217 (i, bh->b_data);
218 reiserfs_restore_prepared_buffer(s, bh);
219 *beg = org;
220 /* ... and search again in current block from beginning */
221 goto cont;
222 }
223 }
224 bi->free_count -= (end - *beg);
225 journal_mark_dirty(th, s, bh);
226 brelse(bh);
227
228 /* free block count calculation */
229 reiserfs_prepare_for_journal(s, SB_BUFFER_WITH_SB(s),
230 1);
231 PUT_SB_FREE_BLOCKS(s, SB_FREE_BLOCKS(s) - (end - *beg));
232 journal_mark_dirty(th, s, SB_BUFFER_WITH_SB(s));
233
234 return end - (*beg);
235 } else {
236 *beg = next;
237 }
238 }
239}
240
241static int bmap_hash_id(struct super_block *s, u32 id)
242{
243 char *hash_in = NULL;
244 unsigned long hash;
245 unsigned bm;
246
247 if (id <= 2) {
248 bm = 1;
249 } else {
250 hash_in = (char *)(&id);
251 hash = keyed_hash(hash_in, 4);
252 bm = hash % reiserfs_bmap_count(s);
253 if (!bm)
254 bm = 1;
255 }
256 /* this can only be true when SB_BMAP_NR = 1 */
257 if (bm >= reiserfs_bmap_count(s))
258 bm = 0;
259 return bm;
260}
261
262/*
263 * hashes the id and then returns > 0 if the block group for the
264 * corresponding hash is full
265 */
266static inline int block_group_used(struct super_block *s, u32 id)
267{
268 int bm = bmap_hash_id(s, id);
269 struct reiserfs_bitmap_info *info = &SB_AP_BITMAP(s)[bm];
270
271 /* If we don't have cached information on this bitmap block, we're
272 * going to have to load it later anyway. Loading it here allows us
273 * to make a better decision. This favors long-term performance gain
274 * with a better on-disk layout vs. a short term gain of skipping the
275 * read and potentially having a bad placement. */
276 if (info->free_count == UINT_MAX) {
277 struct buffer_head *bh = reiserfs_read_bitmap_block(s, bm);
278 brelse(bh);
279 }
280
281 if (info->free_count > ((s->s_blocksize << 3) * 60 / 100)) {
282 return 0;
283 }
284 return 1;
285}
286
287/*
288 * the packing is returned in disk byte order
289 */
290__le32 reiserfs_choose_packing(struct inode * dir)
291{
292 __le32 packing;
293 if (TEST_OPTION(packing_groups, dir->i_sb)) {
294 u32 parent_dir = le32_to_cpu(INODE_PKEY(dir)->k_dir_id);
295 /*
296 * some versions of reiserfsck expect packing locality 1 to be
297 * special
298 */
299 if (parent_dir == 1 || block_group_used(dir->i_sb, parent_dir))
300 packing = INODE_PKEY(dir)->k_objectid;
301 else
302 packing = INODE_PKEY(dir)->k_dir_id;
303 } else
304 packing = INODE_PKEY(dir)->k_objectid;
305 return packing;
306}
307
308/* Tries to find contiguous zero bit window (given size) in given region of
309 * bitmap and place new blocks there. Returns number of allocated blocks. */
310static int scan_bitmap(struct reiserfs_transaction_handle *th,
311 b_blocknr_t * start, b_blocknr_t finish,
312 int min, int max, int unfm, sector_t file_block)
313{
314 int nr_allocated = 0;
315 struct super_block *s = th->t_super;
316 /* find every bm and bmap and bmap_nr in this file, and change them all to bitmap_blocknr
317 * - Hans, it is not a block number - Zam. */
318
319 unsigned int bm, off;
320 unsigned int end_bm, end_off;
321 unsigned int off_max = s->s_blocksize << 3;
322
323 BUG_ON(!th->t_trans_id);
324
325 PROC_INFO_INC(s, scan_bitmap.call);
326 if (SB_FREE_BLOCKS(s) <= 0)
327 return 0; // No point in looking for more free blocks
328
329 get_bit_address(s, *start, &bm, &off);
330 get_bit_address(s, finish, &end_bm, &end_off);
331 if (bm > reiserfs_bmap_count(s))
332 return 0;
333 if (end_bm > reiserfs_bmap_count(s))
334 end_bm = reiserfs_bmap_count(s);
335
336 /* When the bitmap is more than 10% free, anyone can allocate.
337 * When it's less than 10% free, only files that already use the
338 * bitmap are allowed. Once we pass 80% full, this restriction
339 * is lifted.
340 *
341 * We do this so that files that grow later still have space close to
342 * their original allocation. This improves locality, and presumably
343 * performance as a result.
344 *
345 * This is only an allocation policy and does not make up for getting a
346 * bad hint. Decent hinting must be implemented for this to work well.
347 */
348 if (TEST_OPTION(skip_busy, s)
349 && SB_FREE_BLOCKS(s) > SB_BLOCK_COUNT(s) / 20) {
350 for (; bm < end_bm; bm++, off = 0) {
351 if ((off && (!unfm || (file_block != 0)))
352 || SB_AP_BITMAP(s)[bm].free_count >
353 (s->s_blocksize << 3) / 10)
354 nr_allocated =
355 scan_bitmap_block(th, bm, &off, off_max,
356 min, max, unfm);
357 if (nr_allocated)
358 goto ret;
359 }
360 /* we know from above that start is a reasonable number */
361 get_bit_address(s, *start, &bm, &off);
362 }
363
364 for (; bm < end_bm; bm++, off = 0) {
365 nr_allocated =
366 scan_bitmap_block(th, bm, &off, off_max, min, max, unfm);
367 if (nr_allocated)
368 goto ret;
369 }
370
371 nr_allocated =
372 scan_bitmap_block(th, bm, &off, end_off + 1, min, max, unfm);
373
374 ret:
375 *start = bm * off_max + off;
376 return nr_allocated;
377
378}
379
380static void _reiserfs_free_block(struct reiserfs_transaction_handle *th,
381 struct inode *inode, b_blocknr_t block,
382 int for_unformatted)
383{
384 struct super_block *s = th->t_super;
385 struct reiserfs_super_block *rs;
386 struct buffer_head *sbh, *bmbh;
387 struct reiserfs_bitmap_info *apbi;
388 unsigned int nr, offset;
389
390 BUG_ON(!th->t_trans_id);
391
392 PROC_INFO_INC(s, free_block);
393
394 rs = SB_DISK_SUPER_BLOCK(s);
395 sbh = SB_BUFFER_WITH_SB(s);
396 apbi = SB_AP_BITMAP(s);
397
398 get_bit_address(s, block, &nr, &offset);
399
400 if (nr >= reiserfs_bmap_count(s)) {
401 reiserfs_error(s, "vs-4075", "block %lu is out of range",
402 block);
403 return;
404 }
405
406 bmbh = reiserfs_read_bitmap_block(s, nr);
407 if (!bmbh)
408 return;
409
410 reiserfs_prepare_for_journal(s, bmbh, 1);
411
412 /* clear bit for the given block in bit map */
413 if (!reiserfs_test_and_clear_le_bit(offset, bmbh->b_data)) {
414 reiserfs_error(s, "vs-4080",
415 "block %lu: bit already cleared", block);
416 }
417 apbi[nr].free_count++;
418 journal_mark_dirty(th, s, bmbh);
419 brelse(bmbh);
420
421 reiserfs_prepare_for_journal(s, sbh, 1);
422 /* update super block */
423 set_sb_free_blocks(rs, sb_free_blocks(rs) + 1);
424
425 journal_mark_dirty(th, s, sbh);
426 if (for_unformatted)
427 dquot_free_block_nodirty(inode, 1);
428}
429
430void reiserfs_free_block(struct reiserfs_transaction_handle *th,
431 struct inode *inode, b_blocknr_t block,
432 int for_unformatted)
433{
434 struct super_block *s = th->t_super;
435 BUG_ON(!th->t_trans_id);
436
437 RFALSE(!s, "vs-4061: trying to free block on nonexistent device");
438 if (!is_reusable(s, block, 1))
439 return;
440
441 if (block > sb_block_count(REISERFS_SB(s)->s_rs)) {
442 reiserfs_error(th->t_super, "bitmap-4072",
443 "Trying to free block outside file system "
444 "boundaries (%lu > %lu)",
445 block, sb_block_count(REISERFS_SB(s)->s_rs));
446 return;
447 }
448 /* mark it before we clear it, just in case */
449 journal_mark_freed(th, s, block);
450 _reiserfs_free_block(th, inode, block, for_unformatted);
451}
452
453/* preallocated blocks don't need to be run through journal_mark_freed */
454static void reiserfs_free_prealloc_block(struct reiserfs_transaction_handle *th,
455 struct inode *inode, b_blocknr_t block)
456{
457 BUG_ON(!th->t_trans_id);
458 RFALSE(!th->t_super,
459 "vs-4060: trying to free block on nonexistent device");
460 if (!is_reusable(th->t_super, block, 1))
461 return;
462 _reiserfs_free_block(th, inode, block, 1);
463}
464
465static void __discard_prealloc(struct reiserfs_transaction_handle *th,
466 struct reiserfs_inode_info *ei)
467{
468 unsigned long save = ei->i_prealloc_block;
469 int dirty = 0;
470 struct inode *inode = &ei->vfs_inode;
471 BUG_ON(!th->t_trans_id);
472#ifdef CONFIG_REISERFS_CHECK
473 if (ei->i_prealloc_count < 0)
474 reiserfs_error(th->t_super, "zam-4001",
475 "inode has negative prealloc blocks count.");
476#endif
477 while (ei->i_prealloc_count > 0) {
478 reiserfs_free_prealloc_block(th, inode, ei->i_prealloc_block);
479 ei->i_prealloc_block++;
480 ei->i_prealloc_count--;
481 dirty = 1;
482 }
483 if (dirty)
484 reiserfs_update_sd(th, inode);
485 ei->i_prealloc_block = save;
486 list_del_init(&(ei->i_prealloc_list));
487}
488
489/* FIXME: It should be inline function */
490void reiserfs_discard_prealloc(struct reiserfs_transaction_handle *th,
491 struct inode *inode)
492{
493 struct reiserfs_inode_info *ei = REISERFS_I(inode);
494 BUG_ON(!th->t_trans_id);
495 if (ei->i_prealloc_count)
496 __discard_prealloc(th, ei);
497}
498
499void reiserfs_discard_all_prealloc(struct reiserfs_transaction_handle *th)
500{
501 struct list_head *plist = &SB_JOURNAL(th->t_super)->j_prealloc_list;
502
503 BUG_ON(!th->t_trans_id);
504
505 while (!list_empty(plist)) {
506 struct reiserfs_inode_info *ei;
507 ei = list_entry(plist->next, struct reiserfs_inode_info,
508 i_prealloc_list);
509#ifdef CONFIG_REISERFS_CHECK
510 if (!ei->i_prealloc_count) {
511 reiserfs_error(th->t_super, "zam-4001",
512 "inode is in prealloc list but has "
513 "no preallocated blocks.");
514 }
515#endif
516 __discard_prealloc(th, ei);
517 }
518}
519
520void reiserfs_init_alloc_options(struct super_block *s)
521{
522 set_bit(_ALLOC_skip_busy, &SB_ALLOC_OPTS(s));
523 set_bit(_ALLOC_dirid_groups, &SB_ALLOC_OPTS(s));
524 set_bit(_ALLOC_packing_groups, &SB_ALLOC_OPTS(s));
525}
526
527/* block allocator related options are parsed here */
528int reiserfs_parse_alloc_options(struct super_block *s, char *options)
529{
530 char *this_char, *value;
531
532 REISERFS_SB(s)->s_alloc_options.bits = 0; /* clear default settings */
533
534 while ((this_char = strsep(&options, ":")) != NULL) {
535 if ((value = strchr(this_char, '=')) != NULL)
536 *value++ = 0;
537
538 if (!strcmp(this_char, "concentrating_formatted_nodes")) {
539 int temp;
540 SET_OPTION(concentrating_formatted_nodes);
541 temp = (value
542 && *value) ? simple_strtoul(value, &value,
543 0) : 10;
544 if (temp <= 0 || temp > 100) {
545 REISERFS_SB(s)->s_alloc_options.border = 10;
546 } else {
547 REISERFS_SB(s)->s_alloc_options.border =
548 100 / temp;
549 }
550 continue;
551 }
552 if (!strcmp(this_char, "displacing_large_files")) {
553 SET_OPTION(displacing_large_files);
554 REISERFS_SB(s)->s_alloc_options.large_file_size =
555 (value
556 && *value) ? simple_strtoul(value, &value, 0) : 16;
557 continue;
558 }
559 if (!strcmp(this_char, "displacing_new_packing_localities")) {
560 SET_OPTION(displacing_new_packing_localities);
561 continue;
562 };
563
564 if (!strcmp(this_char, "old_hashed_relocation")) {
565 SET_OPTION(old_hashed_relocation);
566 continue;
567 }
568
569 if (!strcmp(this_char, "new_hashed_relocation")) {
570 SET_OPTION(new_hashed_relocation);
571 continue;
572 }
573
574 if (!strcmp(this_char, "dirid_groups")) {
575 SET_OPTION(dirid_groups);
576 continue;
577 }
578 if (!strcmp(this_char, "oid_groups")) {
579 SET_OPTION(oid_groups);
580 continue;
581 }
582 if (!strcmp(this_char, "packing_groups")) {
583 SET_OPTION(packing_groups);
584 continue;
585 }
586 if (!strcmp(this_char, "hashed_formatted_nodes")) {
587 SET_OPTION(hashed_formatted_nodes);
588 continue;
589 }
590
591 if (!strcmp(this_char, "skip_busy")) {
592 SET_OPTION(skip_busy);
593 continue;
594 }
595
596 if (!strcmp(this_char, "hundredth_slices")) {
597 SET_OPTION(hundredth_slices);
598 continue;
599 }
600
601 if (!strcmp(this_char, "old_way")) {
602 SET_OPTION(old_way);
603 continue;
604 }
605
606 if (!strcmp(this_char, "displace_based_on_dirid")) {
607 SET_OPTION(displace_based_on_dirid);
608 continue;
609 }
610
611 if (!strcmp(this_char, "preallocmin")) {
612 REISERFS_SB(s)->s_alloc_options.preallocmin =
613 (value
614 && *value) ? simple_strtoul(value, &value, 0) : 4;
615 continue;
616 }
617
618 if (!strcmp(this_char, "preallocsize")) {
619 REISERFS_SB(s)->s_alloc_options.preallocsize =
620 (value
621 && *value) ? simple_strtoul(value, &value,
622 0) :
623 PREALLOCATION_SIZE;
624 continue;
625 }
626
627 reiserfs_warning(s, "zam-4001", "unknown option - %s",
628 this_char);
629 return 1;
630 }
631
632 reiserfs_info(s, "allocator options = [%08x]\n", SB_ALLOC_OPTS(s));
633 return 0;
634}
635
636static void print_sep(struct seq_file *seq, int *first)
637{
638 if (!*first)
639 seq_puts(seq, ":");
640 else
641 *first = 0;
642}
643
644void show_alloc_options(struct seq_file *seq, struct super_block *s)
645{
646 int first = 1;
647
648 if (SB_ALLOC_OPTS(s) == ((1 << _ALLOC_skip_busy) |
649 (1 << _ALLOC_dirid_groups) | (1 << _ALLOC_packing_groups)))
650 return;
651
652 seq_puts(seq, ",alloc=");
653
654 if (TEST_OPTION(concentrating_formatted_nodes, s)) {
655 print_sep(seq, &first);
656 if (REISERFS_SB(s)->s_alloc_options.border != 10) {
657 seq_printf(seq, "concentrating_formatted_nodes=%d",
658 100 / REISERFS_SB(s)->s_alloc_options.border);
659 } else
660 seq_puts(seq, "concentrating_formatted_nodes");
661 }
662 if (TEST_OPTION(displacing_large_files, s)) {
663 print_sep(seq, &first);
664 if (REISERFS_SB(s)->s_alloc_options.large_file_size != 16) {
665 seq_printf(seq, "displacing_large_files=%lu",
666 REISERFS_SB(s)->s_alloc_options.large_file_size);
667 } else
668 seq_puts(seq, "displacing_large_files");
669 }
670 if (TEST_OPTION(displacing_new_packing_localities, s)) {
671 print_sep(seq, &first);
672 seq_puts(seq, "displacing_new_packing_localities");
673 }
674 if (TEST_OPTION(old_hashed_relocation, s)) {
675 print_sep(seq, &first);
676 seq_puts(seq, "old_hashed_relocation");
677 }
678 if (TEST_OPTION(new_hashed_relocation, s)) {
679 print_sep(seq, &first);
680 seq_puts(seq, "new_hashed_relocation");
681 }
682 if (TEST_OPTION(dirid_groups, s)) {
683 print_sep(seq, &first);
684 seq_puts(seq, "dirid_groups");
685 }
686 if (TEST_OPTION(oid_groups, s)) {
687 print_sep(seq, &first);
688 seq_puts(seq, "oid_groups");
689 }
690 if (TEST_OPTION(packing_groups, s)) {
691 print_sep(seq, &first);
692 seq_puts(seq, "packing_groups");
693 }
694 if (TEST_OPTION(hashed_formatted_nodes, s)) {
695 print_sep(seq, &first);
696 seq_puts(seq, "hashed_formatted_nodes");
697 }
698 if (TEST_OPTION(skip_busy, s)) {
699 print_sep(seq, &first);
700 seq_puts(seq, "skip_busy");
701 }
702 if (TEST_OPTION(hundredth_slices, s)) {
703 print_sep(seq, &first);
704 seq_puts(seq, "hundredth_slices");
705 }
706 if (TEST_OPTION(old_way, s)) {
707 print_sep(seq, &first);
708 seq_puts(seq, "old_way");
709 }
710 if (TEST_OPTION(displace_based_on_dirid, s)) {
711 print_sep(seq, &first);
712 seq_puts(seq, "displace_based_on_dirid");
713 }
714 if (REISERFS_SB(s)->s_alloc_options.preallocmin != 0) {
715 print_sep(seq, &first);
716 seq_printf(seq, "preallocmin=%d",
717 REISERFS_SB(s)->s_alloc_options.preallocmin);
718 }
719 if (REISERFS_SB(s)->s_alloc_options.preallocsize != 17) {
720 print_sep(seq, &first);
721 seq_printf(seq, "preallocsize=%d",
722 REISERFS_SB(s)->s_alloc_options.preallocsize);
723 }
724}
725
726static inline void new_hashed_relocation(reiserfs_blocknr_hint_t * hint)
727{
728 char *hash_in;
729 if (hint->formatted_node) {
730 hash_in = (char *)&hint->key.k_dir_id;
731 } else {
732 if (!hint->inode) {
733 //hint->search_start = hint->beg;
734 hash_in = (char *)&hint->key.k_dir_id;
735 } else
736 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
737 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
738 else
739 hash_in =
740 (char *)(&INODE_PKEY(hint->inode)->k_objectid);
741 }
742
743 hint->search_start =
744 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
745}
746
747/*
748 * Relocation based on dirid, hashing them into a given bitmap block
749 * files. Formatted nodes are unaffected, a separate policy covers them
750 */
751static void dirid_groups(reiserfs_blocknr_hint_t * hint)
752{
753 unsigned long hash;
754 __u32 dirid = 0;
755 int bm = 0;
756 struct super_block *sb = hint->th->t_super;
757 if (hint->inode)
758 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
759 else if (hint->formatted_node)
760 dirid = hint->key.k_dir_id;
761
762 if (dirid) {
763 bm = bmap_hash_id(sb, dirid);
764 hash = bm * (sb->s_blocksize << 3);
765 /* give a portion of the block group to metadata */
766 if (hint->inode)
767 hash += sb->s_blocksize / 2;
768 hint->search_start = hash;
769 }
770}
771
772/*
773 * Relocation based on oid, hashing them into a given bitmap block
774 * files. Formatted nodes are unaffected, a separate policy covers them
775 */
776static void oid_groups(reiserfs_blocknr_hint_t * hint)
777{
778 if (hint->inode) {
779 unsigned long hash;
780 __u32 oid;
781 __u32 dirid;
782 int bm;
783
784 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
785
786 /* keep the root dir and it's first set of subdirs close to
787 * the start of the disk
788 */
789 if (dirid <= 2)
790 hash = (hint->inode->i_sb->s_blocksize << 3);
791 else {
792 oid = le32_to_cpu(INODE_PKEY(hint->inode)->k_objectid);
793 bm = bmap_hash_id(hint->inode->i_sb, oid);
794 hash = bm * (hint->inode->i_sb->s_blocksize << 3);
795 }
796 hint->search_start = hash;
797 }
798}
799
800/* returns 1 if it finds an indirect item and gets valid hint info
801 * from it, otherwise 0
802 */
803static int get_left_neighbor(reiserfs_blocknr_hint_t * hint)
804{
805 struct treepath *path;
806 struct buffer_head *bh;
807 struct item_head *ih;
808 int pos_in_item;
809 __le32 *item;
810 int ret = 0;
811
812 if (!hint->path) /* reiserfs code can call this function w/o pointer to path
813 * structure supplied; then we rely on supplied search_start */
814 return 0;
815
816 path = hint->path;
817 bh = get_last_bh(path);
818 RFALSE(!bh, "green-4002: Illegal path specified to get_left_neighbor");
819 ih = get_ih(path);
820 pos_in_item = path->pos_in_item;
821 item = get_item(path);
822
823 hint->search_start = bh->b_blocknr;
824
825 if (!hint->formatted_node && is_indirect_le_ih(ih)) {
826 /* for indirect item: go to left and look for the first non-hole entry
827 in the indirect item */
828 if (pos_in_item == I_UNFM_NUM(ih))
829 pos_in_item--;
830// pos_in_item = I_UNFM_NUM (ih) - 1;
831 while (pos_in_item >= 0) {
832 int t = get_block_num(item, pos_in_item);
833 if (t) {
834 hint->search_start = t;
835 ret = 1;
836 break;
837 }
838 pos_in_item--;
839 }
840 }
841
842 /* does result value fit into specified region? */
843 return ret;
844}
845
846/* should be, if formatted node, then try to put on first part of the device
847 specified as number of percent with mount option device, else try to put
848 on last of device. This is not to say it is good code to do so,
849 but the effect should be measured. */
850static inline void set_border_in_hint(struct super_block *s,
851 reiserfs_blocknr_hint_t * hint)
852{
853 b_blocknr_t border =
854 SB_BLOCK_COUNT(s) / REISERFS_SB(s)->s_alloc_options.border;
855
856 if (hint->formatted_node)
857 hint->end = border - 1;
858 else
859 hint->beg = border;
860}
861
862static inline void displace_large_file(reiserfs_blocknr_hint_t * hint)
863{
864 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
865 hint->search_start =
866 hint->beg +
867 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_dir_id),
868 4) % (hint->end - hint->beg);
869 else
870 hint->search_start =
871 hint->beg +
872 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_objectid),
873 4) % (hint->end - hint->beg);
874}
875
876static inline void hash_formatted_node(reiserfs_blocknr_hint_t * hint)
877{
878 char *hash_in;
879
880 if (!hint->inode)
881 hash_in = (char *)&hint->key.k_dir_id;
882 else if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
883 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
884 else
885 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_objectid);
886
887 hint->search_start =
888 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
889}
890
891static inline int
892this_blocknr_allocation_would_make_it_a_large_file(reiserfs_blocknr_hint_t *
893 hint)
894{
895 return hint->block ==
896 REISERFS_SB(hint->th->t_super)->s_alloc_options.large_file_size;
897}
898
899#ifdef DISPLACE_NEW_PACKING_LOCALITIES
900static inline void displace_new_packing_locality(reiserfs_blocknr_hint_t * hint)
901{
902 struct in_core_key *key = &hint->key;
903
904 hint->th->displace_new_blocks = 0;
905 hint->search_start =
906 hint->beg + keyed_hash((char *)(&key->k_objectid),
907 4) % (hint->end - hint->beg);
908}
909#endif
910
911static inline int old_hashed_relocation(reiserfs_blocknr_hint_t * hint)
912{
913 b_blocknr_t border;
914 u32 hash_in;
915
916 if (hint->formatted_node || hint->inode == NULL) {
917 return 0;
918 }
919
920 hash_in = le32_to_cpu((INODE_PKEY(hint->inode))->k_dir_id);
921 border =
922 hint->beg + (u32) keyed_hash(((char *)(&hash_in)),
923 4) % (hint->end - hint->beg - 1);
924 if (border > hint->search_start)
925 hint->search_start = border;
926
927 return 1;
928}
929
930static inline int old_way(reiserfs_blocknr_hint_t * hint)
931{
932 b_blocknr_t border;
933
934 if (hint->formatted_node || hint->inode == NULL) {
935 return 0;
936 }
937
938 border =
939 hint->beg +
940 le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id) % (hint->end -
941 hint->beg);
942 if (border > hint->search_start)
943 hint->search_start = border;
944
945 return 1;
946}
947
948static inline void hundredth_slices(reiserfs_blocknr_hint_t * hint)
949{
950 struct in_core_key *key = &hint->key;
951 b_blocknr_t slice_start;
952
953 slice_start =
954 (keyed_hash((char *)(&key->k_dir_id), 4) % 100) * (hint->end / 100);
955 if (slice_start > hint->search_start
956 || slice_start + (hint->end / 100) <= hint->search_start) {
957 hint->search_start = slice_start;
958 }
959}
960
961static void determine_search_start(reiserfs_blocknr_hint_t * hint,
962 int amount_needed)
963{
964 struct super_block *s = hint->th->t_super;
965 int unfm_hint;
966
967 hint->beg = 0;
968 hint->end = SB_BLOCK_COUNT(s) - 1;
969
970 /* This is former border algorithm. Now with tunable border offset */
971 if (concentrating_formatted_nodes(s))
972 set_border_in_hint(s, hint);
973
974#ifdef DISPLACE_NEW_PACKING_LOCALITIES
975 /* whenever we create a new directory, we displace it. At first we will
976 hash for location, later we might look for a moderately empty place for
977 it */
978 if (displacing_new_packing_localities(s)
979 && hint->th->displace_new_blocks) {
980 displace_new_packing_locality(hint);
981
982 /* we do not continue determine_search_start,
983 * if new packing locality is being displaced */
984 return;
985 }
986#endif
987
988 /* all persons should feel encouraged to add more special cases here and
989 * test them */
990
991 if (displacing_large_files(s) && !hint->formatted_node
992 && this_blocknr_allocation_would_make_it_a_large_file(hint)) {
993 displace_large_file(hint);
994 return;
995 }
996
997 /* if none of our special cases is relevant, use the left neighbor in the
998 tree order of the new node we are allocating for */
999 if (hint->formatted_node && TEST_OPTION(hashed_formatted_nodes, s)) {
1000 hash_formatted_node(hint);
1001 return;
1002 }
1003
1004 unfm_hint = get_left_neighbor(hint);
1005
1006 /* Mimic old block allocator behaviour, that is if VFS allowed for preallocation,
1007 new blocks are displaced based on directory ID. Also, if suggested search_start
1008 is less than last preallocated block, we start searching from it, assuming that
1009 HDD dataflow is faster in forward direction */
1010 if (TEST_OPTION(old_way, s)) {
1011 if (!hint->formatted_node) {
1012 if (!reiserfs_hashed_relocation(s))
1013 old_way(hint);
1014 else if (!reiserfs_no_unhashed_relocation(s))
1015 old_hashed_relocation(hint);
1016
1017 if (hint->inode
1018 && hint->search_start <
1019 REISERFS_I(hint->inode)->i_prealloc_block)
1020 hint->search_start =
1021 REISERFS_I(hint->inode)->i_prealloc_block;
1022 }
1023 return;
1024 }
1025
1026 /* This is an approach proposed by Hans */
1027 if (TEST_OPTION(hundredth_slices, s)
1028 && !(displacing_large_files(s) && !hint->formatted_node)) {
1029 hundredth_slices(hint);
1030 return;
1031 }
1032
1033 /* old_hashed_relocation only works on unformatted */
1034 if (!unfm_hint && !hint->formatted_node &&
1035 TEST_OPTION(old_hashed_relocation, s)) {
1036 old_hashed_relocation(hint);
1037 }
1038 /* new_hashed_relocation works with both formatted/unformatted nodes */
1039 if ((!unfm_hint || hint->formatted_node) &&
1040 TEST_OPTION(new_hashed_relocation, s)) {
1041 new_hashed_relocation(hint);
1042 }
1043 /* dirid grouping works only on unformatted nodes */
1044 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
1045 dirid_groups(hint);
1046 }
1047#ifdef DISPLACE_NEW_PACKING_LOCALITIES
1048 if (hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
1049 dirid_groups(hint);
1050 }
1051#endif
1052
1053 /* oid grouping works only on unformatted nodes */
1054 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(oid_groups, s)) {
1055 oid_groups(hint);
1056 }
1057 return;
1058}
1059
1060static int determine_prealloc_size(reiserfs_blocknr_hint_t * hint)
1061{
1062 /* make minimum size a mount option and benchmark both ways */
1063 /* we preallocate blocks only for regular files, specific size */
1064 /* benchmark preallocating always and see what happens */
1065
1066 hint->prealloc_size = 0;
1067
1068 if (!hint->formatted_node && hint->preallocate) {
1069 if (S_ISREG(hint->inode->i_mode)
1070 && hint->inode->i_size >=
1071 REISERFS_SB(hint->th->t_super)->s_alloc_options.
1072 preallocmin * hint->inode->i_sb->s_blocksize)
1073 hint->prealloc_size =
1074 REISERFS_SB(hint->th->t_super)->s_alloc_options.
1075 preallocsize - 1;
1076 }
1077 return CARRY_ON;
1078}
1079
1080/* XXX I know it could be merged with upper-level function;
1081 but may be result function would be too complex. */
1082static inline int allocate_without_wrapping_disk(reiserfs_blocknr_hint_t * hint,
1083 b_blocknr_t * new_blocknrs,
1084 b_blocknr_t start,
1085 b_blocknr_t finish, int min,
1086 int amount_needed,
1087 int prealloc_size)
1088{
1089 int rest = amount_needed;
1090 int nr_allocated;
1091
1092 while (rest > 0 && start <= finish) {
1093 nr_allocated = scan_bitmap(hint->th, &start, finish, min,
1094 rest + prealloc_size,
1095 !hint->formatted_node, hint->block);
1096
1097 if (nr_allocated == 0) /* no new blocks allocated, return */
1098 break;
1099
1100 /* fill free_blocknrs array first */
1101 while (rest > 0 && nr_allocated > 0) {
1102 *new_blocknrs++ = start++;
1103 rest--;
1104 nr_allocated--;
1105 }
1106
1107 /* do we have something to fill prealloc. array also ? */
1108 if (nr_allocated > 0) {
1109 /* it means prealloc_size was greater that 0 and we do preallocation */
1110 list_add(&REISERFS_I(hint->inode)->i_prealloc_list,
1111 &SB_JOURNAL(hint->th->t_super)->
1112 j_prealloc_list);
1113 REISERFS_I(hint->inode)->i_prealloc_block = start;
1114 REISERFS_I(hint->inode)->i_prealloc_count =
1115 nr_allocated;
1116 break;
1117 }
1118 }
1119
1120 return (amount_needed - rest);
1121}
1122
1123static inline int blocknrs_and_prealloc_arrays_from_search_start
1124 (reiserfs_blocknr_hint_t * hint, b_blocknr_t * new_blocknrs,
1125 int amount_needed) {
1126 struct super_block *s = hint->th->t_super;
1127 b_blocknr_t start = hint->search_start;
1128 b_blocknr_t finish = SB_BLOCK_COUNT(s) - 1;
1129 int passno = 0;
1130 int nr_allocated = 0;
1131
1132 determine_prealloc_size(hint);
1133 if (!hint->formatted_node) {
1134 int quota_ret;
1135#ifdef REISERQUOTA_DEBUG
1136 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1137 "reiserquota: allocating %d blocks id=%u",
1138 amount_needed, hint->inode->i_uid);
1139#endif
1140 quota_ret =
1141 dquot_alloc_block_nodirty(hint->inode, amount_needed);
1142 if (quota_ret) /* Quota exceeded? */
1143 return QUOTA_EXCEEDED;
1144 if (hint->preallocate && hint->prealloc_size) {
1145#ifdef REISERQUOTA_DEBUG
1146 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1147 "reiserquota: allocating (prealloc) %d blocks id=%u",
1148 hint->prealloc_size, hint->inode->i_uid);
1149#endif
1150 quota_ret = dquot_prealloc_block_nodirty(hint->inode,
1151 hint->prealloc_size);
1152 if (quota_ret)
1153 hint->preallocate = hint->prealloc_size = 0;
1154 }
1155 /* for unformatted nodes, force large allocations */
1156 }
1157
1158 do {
1159 switch (passno++) {
1160 case 0: /* Search from hint->search_start to end of disk */
1161 start = hint->search_start;
1162 finish = SB_BLOCK_COUNT(s) - 1;
1163 break;
1164 case 1: /* Search from hint->beg to hint->search_start */
1165 start = hint->beg;
1166 finish = hint->search_start;
1167 break;
1168 case 2: /* Last chance: Search from 0 to hint->beg */
1169 start = 0;
1170 finish = hint->beg;
1171 break;
1172 default: /* We've tried searching everywhere, not enough space */
1173 /* Free the blocks */
1174 if (!hint->formatted_node) {
1175#ifdef REISERQUOTA_DEBUG
1176 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1177 "reiserquota: freeing (nospace) %d blocks id=%u",
1178 amount_needed +
1179 hint->prealloc_size -
1180 nr_allocated,
1181 hint->inode->i_uid);
1182#endif
1183 /* Free not allocated blocks */
1184 dquot_free_block_nodirty(hint->inode,
1185 amount_needed + hint->prealloc_size -
1186 nr_allocated);
1187 }
1188 while (nr_allocated--)
1189 reiserfs_free_block(hint->th, hint->inode,
1190 new_blocknrs[nr_allocated],
1191 !hint->formatted_node);
1192
1193 return NO_DISK_SPACE;
1194 }
1195 } while ((nr_allocated += allocate_without_wrapping_disk(hint,
1196 new_blocknrs +
1197 nr_allocated,
1198 start, finish,
1199 1,
1200 amount_needed -
1201 nr_allocated,
1202 hint->
1203 prealloc_size))
1204 < amount_needed);
1205 if (!hint->formatted_node &&
1206 amount_needed + hint->prealloc_size >
1207 nr_allocated + REISERFS_I(hint->inode)->i_prealloc_count) {
1208 /* Some of preallocation blocks were not allocated */
1209#ifdef REISERQUOTA_DEBUG
1210 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1211 "reiserquota: freeing (failed prealloc) %d blocks id=%u",
1212 amount_needed + hint->prealloc_size -
1213 nr_allocated -
1214 REISERFS_I(hint->inode)->i_prealloc_count,
1215 hint->inode->i_uid);
1216#endif
1217 dquot_free_block_nodirty(hint->inode, amount_needed +
1218 hint->prealloc_size - nr_allocated -
1219 REISERFS_I(hint->inode)->
1220 i_prealloc_count);
1221 }
1222
1223 return CARRY_ON;
1224}
1225
1226/* grab new blocknrs from preallocated list */
1227/* return amount still needed after using them */
1228static int use_preallocated_list_if_available(reiserfs_blocknr_hint_t * hint,
1229 b_blocknr_t * new_blocknrs,
1230 int amount_needed)
1231{
1232 struct inode *inode = hint->inode;
1233
1234 if (REISERFS_I(inode)->i_prealloc_count > 0) {
1235 while (amount_needed) {
1236
1237 *new_blocknrs++ = REISERFS_I(inode)->i_prealloc_block++;
1238 REISERFS_I(inode)->i_prealloc_count--;
1239
1240 amount_needed--;
1241
1242 if (REISERFS_I(inode)->i_prealloc_count <= 0) {
1243 list_del(&REISERFS_I(inode)->i_prealloc_list);
1244 break;
1245 }
1246 }
1247 }
1248 /* return amount still needed after using preallocated blocks */
1249 return amount_needed;
1250}
1251
1252int reiserfs_allocate_blocknrs(reiserfs_blocknr_hint_t * hint, b_blocknr_t * new_blocknrs, int amount_needed, int reserved_by_us /* Amount of blocks we have
1253 already reserved */ )
1254{
1255 int initial_amount_needed = amount_needed;
1256 int ret;
1257 struct super_block *s = hint->th->t_super;
1258
1259 /* Check if there is enough space, taking into account reserved space */
1260 if (SB_FREE_BLOCKS(s) - REISERFS_SB(s)->reserved_blocks <
1261 amount_needed - reserved_by_us)
1262 return NO_DISK_SPACE;
1263 /* should this be if !hint->inode && hint->preallocate? */
1264 /* do you mean hint->formatted_node can be removed ? - Zam */
1265 /* hint->formatted_node cannot be removed because we try to access
1266 inode information here, and there is often no inode assotiated with
1267 metadata allocations - green */
1268
1269 if (!hint->formatted_node && hint->preallocate) {
1270 amount_needed = use_preallocated_list_if_available
1271 (hint, new_blocknrs, amount_needed);
1272 if (amount_needed == 0) /* all blocknrs we need we got from
1273 prealloc. list */
1274 return CARRY_ON;
1275 new_blocknrs += (initial_amount_needed - amount_needed);
1276 }
1277
1278 /* find search start and save it in hint structure */
1279 determine_search_start(hint, amount_needed);
1280 if (hint->search_start >= SB_BLOCK_COUNT(s))
1281 hint->search_start = SB_BLOCK_COUNT(s) - 1;
1282
1283 /* allocation itself; fill new_blocknrs and preallocation arrays */
1284 ret = blocknrs_and_prealloc_arrays_from_search_start
1285 (hint, new_blocknrs, amount_needed);
1286
1287 /* we used prealloc. list to fill (partially) new_blocknrs array. If final allocation fails we
1288 * need to return blocks back to prealloc. list or just free them. -- Zam (I chose second
1289 * variant) */
1290
1291 if (ret != CARRY_ON) {
1292 while (amount_needed++ < initial_amount_needed) {
1293 reiserfs_free_block(hint->th, hint->inode,
1294 *(--new_blocknrs), 1);
1295 }
1296 }
1297 return ret;
1298}
1299
1300void reiserfs_cache_bitmap_metadata(struct super_block *sb,
1301 struct buffer_head *bh,
1302 struct reiserfs_bitmap_info *info)
1303{
1304 unsigned long *cur = (unsigned long *)(bh->b_data + bh->b_size);
1305
1306 /* The first bit must ALWAYS be 1 */
1307 if (!reiserfs_test_le_bit(0, (unsigned long *)bh->b_data))
1308 reiserfs_error(sb, "reiserfs-2025", "bitmap block %lu is "
1309 "corrupted: first bit must be 1", bh->b_blocknr);
1310
1311 info->free_count = 0;
1312
1313 while (--cur >= (unsigned long *)bh->b_data) {
1314 /* 0 and ~0 are special, we can optimize for them */
1315 if (*cur == 0)
1316 info->free_count += BITS_PER_LONG;
1317 else if (*cur != ~0L) /* A mix, investigate */
1318 info->free_count += BITS_PER_LONG - hweight_long(*cur);
1319 }
1320}
1321
1322struct buffer_head *reiserfs_read_bitmap_block(struct super_block *sb,
1323 unsigned int bitmap)
1324{
1325 b_blocknr_t block = (sb->s_blocksize << 3) * bitmap;
1326 struct reiserfs_bitmap_info *info = SB_AP_BITMAP(sb) + bitmap;
1327 struct buffer_head *bh;
1328
1329 /* Way old format filesystems had the bitmaps packed up front.
1330 * I doubt there are any of these left, but just in case... */
1331 if (unlikely(test_bit(REISERFS_OLD_FORMAT,
1332 &(REISERFS_SB(sb)->s_properties))))
1333 block = REISERFS_SB(sb)->s_sbh->b_blocknr + 1 + bitmap;
1334 else if (bitmap == 0)
1335 block = (REISERFS_DISK_OFFSET_IN_BYTES >> sb->s_blocksize_bits) + 1;
1336
1337 reiserfs_write_unlock(sb);
1338 bh = sb_bread(sb, block);
1339 reiserfs_write_lock(sb);
1340 if (bh == NULL)
1341 reiserfs_warning(sb, "sh-2029: %s: bitmap block (#%u) "
1342 "reading failed", __func__, block);
1343 else {
1344 if (buffer_locked(bh)) {
1345 PROC_INFO_INC(sb, scan_bitmap.wait);
1346 reiserfs_write_unlock(sb);
1347 __wait_on_buffer(bh);
1348 reiserfs_write_lock(sb);
1349 }
1350 BUG_ON(!buffer_uptodate(bh));
1351 BUG_ON(atomic_read(&bh->b_count) == 0);
1352
1353 if (info->free_count == UINT_MAX)
1354 reiserfs_cache_bitmap_metadata(sb, bh, info);
1355 }
1356
1357 return bh;
1358}
1359
1360int reiserfs_init_bitmap_cache(struct super_block *sb)
1361{
1362 struct reiserfs_bitmap_info *bitmap;
1363 unsigned int bmap_nr = reiserfs_bmap_count(sb);
1364
1365 bitmap = vmalloc(sizeof(*bitmap) * bmap_nr);
1366 if (bitmap == NULL)
1367 return -ENOMEM;
1368
1369 memset(bitmap, 0xff, sizeof(*bitmap) * bmap_nr);
1370
1371 SB_AP_BITMAP(sb) = bitmap;
1372
1373 return 0;
1374}
1375
1376void reiserfs_free_bitmap_cache(struct super_block *sb)
1377{
1378 if (SB_AP_BITMAP(sb)) {
1379 vfree(SB_AP_BITMAP(sb));
1380 SB_AP_BITMAP(sb) = NULL;
1381 }
1382}