2 * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README
4 /* Reiserfs block (de)allocator, bitmap-based. */
6 #include <linux/time.h>
7 #include <linux/reiserfs_fs.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/reiserfs_fs_sb.h>
13 #include <linux/reiserfs_fs_i.h>
14 #include <linux/quotaops.h>
16 #define PREALLOCATION_SIZE 9
18 /* different reiserfs block allocator options */
20 #define SB_ALLOC_OPTS(s) (REISERFS_SB(s)->s_alloc_options.bits)
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
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))
40 #define SET_OPTION(optname) \
42 reiserfs_warning(s, "reiserfs: option \"%s\" is set", #optname); \
43 set_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s)); \
45 #define TEST_OPTION(optname, s) \
46 test_bit(_ALLOC_ ## optname , &SB_ALLOC_OPTS(s))
48 static inline void get_bit_address(struct super_block *s,
49 b_blocknr_t block, int *bmap_nr, int *offset)
51 /* It is in the bitmap block number equal to the block
52 * number divided by the number of bits in a block. */
53 *bmap_nr = block / (s->s_blocksize << 3);
54 /* Within that bitmap block it is located at bit offset *offset. */
55 *offset = block & ((s->s_blocksize << 3) - 1);
59 #ifdef CONFIG_REISERFS_CHECK
60 int is_reusable(struct super_block *s, b_blocknr_t block, int bit_value)
64 if (block == 0 || block >= SB_BLOCK_COUNT(s)) {
66 "vs-4010: is_reusable: block number is out of range %lu (%u)",
67 block, SB_BLOCK_COUNT(s));
71 /* it can't be one of the bitmap blocks */
72 for (i = 0; i < SB_BMAP_NR(s); i++)
73 if (block == SB_AP_BITMAP(s)[i].bh->b_blocknr) {
74 reiserfs_warning(s, "vs: 4020: is_reusable: "
75 "bitmap block %lu(%u) can't be freed or reused",
76 block, SB_BMAP_NR(s));
80 get_bit_address(s, block, &i, &j);
82 if (i >= SB_BMAP_NR(s)) {
84 "vs-4030: is_reusable: there is no so many bitmap blocks: "
85 "block=%lu, bitmap_nr=%d", block, i);
89 if ((bit_value == 0 &&
90 reiserfs_test_le_bit(j, SB_AP_BITMAP(s)[i].bh->b_data)) ||
92 reiserfs_test_le_bit(j, SB_AP_BITMAP(s)[i].bh->b_data) == 0)) {
94 "vs-4040: is_reusable: corresponding bit of block %lu does not "
95 "match required value (i==%d, j==%d) test_bit==%d",
96 block, i, j, reiserfs_test_le_bit(j,
104 if (bit_value == 0 && block == SB_ROOT_BLOCK(s)) {
106 "vs-4050: is_reusable: this is root block (%u), "
107 "it must be busy", SB_ROOT_BLOCK(s));
113 #endif /* CONFIG_REISERFS_CHECK */
115 /* searches in journal structures for a given block number (bmap, off). If block
116 is found in reiserfs journal it suggests next free block candidate to test. */
117 static inline int is_block_in_journal(struct super_block *s, int bmap, int
122 if (reiserfs_in_journal(s, bmap, off, 1, &tmp)) {
123 if (tmp) { /* hint supplied */
125 PROC_INFO_INC(s, scan_bitmap.in_journal_hint);
127 (*next) = off + 1; /* inc offset to avoid looping. */
128 PROC_INFO_INC(s, scan_bitmap.in_journal_nohint);
130 PROC_INFO_INC(s, scan_bitmap.retry);
136 /* it searches for a window of zero bits with given minimum and maximum lengths in one bitmap
138 static int scan_bitmap_block(struct reiserfs_transaction_handle *th,
139 int bmap_n, int *beg, int boundary, int min,
142 struct super_block *s = th->t_super;
143 struct reiserfs_bitmap_info *bi = &SB_AP_BITMAP(s)[bmap_n];
147 BUG_ON(!th->t_trans_id);
149 RFALSE(bmap_n >= SB_BMAP_NR(s), "Bitmap %d is out of range (0..%d)",
150 bmap_n, SB_BMAP_NR(s) - 1);
151 PROC_INFO_INC(s, scan_bitmap.bmap);
152 /* this is unclear and lacks comments, explain how journal bitmaps
153 work here for the reader. Convey a sense of the design here. What
155 /* - I mean `a window of zero bits' as in description of this function - Zam. */
158 reiserfs_warning(s, "NULL bitmap info pointer for bitmap %d",
162 if (buffer_locked(bi->bh)) {
163 PROC_INFO_INC(s, scan_bitmap.wait);
164 __wait_on_buffer(bi->bh);
169 if (bi->free_count < min)
170 return 0; // No free blocks in this bitmap
172 /* search for a first zero bit -- beggining of a window */
173 *beg = reiserfs_find_next_zero_le_bit
174 ((unsigned long *)(bi->bh->b_data), boundary, *beg);
176 if (*beg + min > boundary) { /* search for a zero bit fails or the rest of bitmap block
177 * cannot contain a zero window of minimum size */
181 if (unfm && is_block_in_journal(s, bmap_n, *beg, beg))
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, bi->bh->b_data)) {
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))
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 */
200 reiserfs_prepare_for_journal(s, bi->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, bi->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' */
214 /* otherwise we clear all bit were set ... */
216 reiserfs_test_and_clear_le_bit
218 reiserfs_restore_prepared_buffer(s,
222 /* ... and search again in current block from beginning */
226 bi->free_count -= (end - *beg);
227 journal_mark_dirty(th, s, bi->bh);
229 /* free block count calculation */
230 reiserfs_prepare_for_journal(s, SB_BUFFER_WITH_SB(s),
232 PUT_SB_FREE_BLOCKS(s, SB_FREE_BLOCKS(s) - (end - *beg));
233 journal_mark_dirty(th, s, SB_BUFFER_WITH_SB(s));
242 static int bmap_hash_id(struct super_block *s, u32 id)
244 char *hash_in = NULL;
251 hash_in = (char *)(&id);
252 hash = keyed_hash(hash_in, 4);
253 bm = hash % SB_BMAP_NR(s);
257 /* this can only be true when SB_BMAP_NR = 1 */
258 if (bm >= SB_BMAP_NR(s))
264 * hashes the id and then returns > 0 if the block group for the
265 * corresponding hash is full
267 static inline int block_group_used(struct super_block *s, u32 id)
270 bm = bmap_hash_id(s, id);
271 if (SB_AP_BITMAP(s)[bm].free_count > ((s->s_blocksize << 3) * 60 / 100)) {
278 * the packing is returned in disk byte order
280 __le32 reiserfs_choose_packing(struct inode * dir)
283 if (TEST_OPTION(packing_groups, dir->i_sb)) {
284 u32 parent_dir = le32_to_cpu(INODE_PKEY(dir)->k_dir_id);
286 * some versions of reiserfsck expect packing locality 1 to be
289 if (parent_dir == 1 || block_group_used(dir->i_sb, parent_dir))
290 packing = INODE_PKEY(dir)->k_objectid;
292 packing = INODE_PKEY(dir)->k_dir_id;
294 packing = INODE_PKEY(dir)->k_objectid;
298 /* Tries to find contiguous zero bit window (given size) in given region of
299 * bitmap and place new blocks there. Returns number of allocated blocks. */
300 static int scan_bitmap(struct reiserfs_transaction_handle *th,
301 b_blocknr_t * start, b_blocknr_t finish,
302 int min, int max, int unfm, unsigned long file_block)
304 int nr_allocated = 0;
305 struct super_block *s = th->t_super;
306 /* find every bm and bmap and bmap_nr in this file, and change them all to bitmap_blocknr
307 * - Hans, it is not a block number - Zam. */
311 int off_max = s->s_blocksize << 3;
313 BUG_ON(!th->t_trans_id);
315 PROC_INFO_INC(s, scan_bitmap.call);
316 if (SB_FREE_BLOCKS(s) <= 0)
317 return 0; // No point in looking for more free blocks
319 get_bit_address(s, *start, &bm, &off);
320 get_bit_address(s, finish, &end_bm, &end_off);
321 if (bm > SB_BMAP_NR(s))
323 if (end_bm > SB_BMAP_NR(s))
324 end_bm = SB_BMAP_NR(s);
326 /* When the bitmap is more than 10% free, anyone can allocate.
327 * When it's less than 10% free, only files that already use the
328 * bitmap are allowed. Once we pass 80% full, this restriction
331 * We do this so that files that grow later still have space close to
332 * their original allocation. This improves locality, and presumably
333 * performance as a result.
335 * This is only an allocation policy and does not make up for getting a
336 * bad hint. Decent hinting must be implemented for this to work well.
338 if (TEST_OPTION(skip_busy, s)
339 && SB_FREE_BLOCKS(s) > SB_BLOCK_COUNT(s) / 20) {
340 for (; bm < end_bm; bm++, off = 0) {
341 if ((off && (!unfm || (file_block != 0)))
342 || SB_AP_BITMAP(s)[bm].free_count >
343 (s->s_blocksize << 3) / 10)
345 scan_bitmap_block(th, bm, &off, off_max,
350 /* we know from above that start is a reasonable number */
351 get_bit_address(s, *start, &bm, &off);
354 for (; bm < end_bm; bm++, off = 0) {
356 scan_bitmap_block(th, bm, &off, off_max, min, max, unfm);
362 scan_bitmap_block(th, bm, &off, end_off + 1, min, max, unfm);
365 *start = bm * off_max + off;
370 static void _reiserfs_free_block(struct reiserfs_transaction_handle *th,
371 struct inode *inode, b_blocknr_t block,
374 struct super_block *s = th->t_super;
375 struct reiserfs_super_block *rs;
376 struct buffer_head *sbh;
377 struct reiserfs_bitmap_info *apbi;
380 BUG_ON(!th->t_trans_id);
382 PROC_INFO_INC(s, free_block);
384 rs = SB_DISK_SUPER_BLOCK(s);
385 sbh = SB_BUFFER_WITH_SB(s);
386 apbi = SB_AP_BITMAP(s);
388 get_bit_address(s, block, &nr, &offset);
390 if (nr >= sb_bmap_nr(rs)) {
391 reiserfs_warning(s, "vs-4075: reiserfs_free_block: "
392 "block %lu is out of range on %s",
393 block, reiserfs_bdevname(s));
397 reiserfs_prepare_for_journal(s, apbi[nr].bh, 1);
399 /* clear bit for the given block in bit map */
400 if (!reiserfs_test_and_clear_le_bit(offset, apbi[nr].bh->b_data)) {
401 reiserfs_warning(s, "vs-4080: reiserfs_free_block: "
402 "free_block (%s:%lu)[dev:blocknr]: bit already cleared",
403 reiserfs_bdevname(s), block);
405 apbi[nr].free_count++;
406 journal_mark_dirty(th, s, apbi[nr].bh);
408 reiserfs_prepare_for_journal(s, sbh, 1);
409 /* update super block */
410 set_sb_free_blocks(rs, sb_free_blocks(rs) + 1);
412 journal_mark_dirty(th, s, sbh);
414 DQUOT_FREE_BLOCK_NODIRTY(inode, 1);
417 void reiserfs_free_block(struct reiserfs_transaction_handle *th,
418 struct inode *inode, b_blocknr_t block,
421 struct super_block *s = th->t_super;
423 BUG_ON(!th->t_trans_id);
425 RFALSE(!s, "vs-4061: trying to free block on nonexistent device");
426 RFALSE(is_reusable(s, block, 1) == 0,
427 "vs-4071: can not free such block");
428 /* mark it before we clear it, just in case */
429 journal_mark_freed(th, s, block);
430 _reiserfs_free_block(th, inode, block, for_unformatted);
433 /* preallocated blocks don't need to be run through journal_mark_freed */
434 static void reiserfs_free_prealloc_block(struct reiserfs_transaction_handle *th,
435 struct inode *inode, b_blocknr_t block)
438 "vs-4060: trying to free block on nonexistent device");
439 RFALSE(is_reusable(th->t_super, block, 1) == 0,
440 "vs-4070: can not free such block");
441 BUG_ON(!th->t_trans_id);
442 _reiserfs_free_block(th, inode, block, 1);
445 static void __discard_prealloc(struct reiserfs_transaction_handle *th,
446 struct reiserfs_inode_info *ei)
448 unsigned long save = ei->i_prealloc_block;
450 struct inode *inode = &ei->vfs_inode;
451 BUG_ON(!th->t_trans_id);
452 #ifdef CONFIG_REISERFS_CHECK
453 if (ei->i_prealloc_count < 0)
454 reiserfs_warning(th->t_super,
455 "zam-4001:%s: inode has negative prealloc blocks count.",
458 while (ei->i_prealloc_count > 0) {
459 reiserfs_free_prealloc_block(th, inode, ei->i_prealloc_block);
460 ei->i_prealloc_block++;
461 ei->i_prealloc_count--;
465 reiserfs_update_sd(th, inode);
466 ei->i_prealloc_block = save;
467 list_del_init(&(ei->i_prealloc_list));
470 /* FIXME: It should be inline function */
471 void reiserfs_discard_prealloc(struct reiserfs_transaction_handle *th,
474 struct reiserfs_inode_info *ei = REISERFS_I(inode);
475 BUG_ON(!th->t_trans_id);
476 if (ei->i_prealloc_count)
477 __discard_prealloc(th, ei);
480 void reiserfs_discard_all_prealloc(struct reiserfs_transaction_handle *th)
482 struct list_head *plist = &SB_JOURNAL(th->t_super)->j_prealloc_list;
484 BUG_ON(!th->t_trans_id);
486 while (!list_empty(plist)) {
487 struct reiserfs_inode_info *ei;
488 ei = list_entry(plist->next, struct reiserfs_inode_info,
490 #ifdef CONFIG_REISERFS_CHECK
491 if (!ei->i_prealloc_count) {
492 reiserfs_warning(th->t_super,
493 "zam-4001:%s: inode is in prealloc list but has no preallocated blocks.",
497 __discard_prealloc(th, ei);
501 void reiserfs_init_alloc_options(struct super_block *s)
503 set_bit(_ALLOC_skip_busy, &SB_ALLOC_OPTS(s));
504 set_bit(_ALLOC_dirid_groups, &SB_ALLOC_OPTS(s));
505 set_bit(_ALLOC_packing_groups, &SB_ALLOC_OPTS(s));
508 /* block allocator related options are parsed here */
509 int reiserfs_parse_alloc_options(struct super_block *s, char *options)
511 char *this_char, *value;
513 REISERFS_SB(s)->s_alloc_options.bits = 0; /* clear default settings */
515 while ((this_char = strsep(&options, ":")) != NULL) {
516 if ((value = strchr(this_char, '=')) != NULL)
519 if (!strcmp(this_char, "concentrating_formatted_nodes")) {
521 SET_OPTION(concentrating_formatted_nodes);
523 && *value) ? simple_strtoul(value, &value,
525 if (temp <= 0 || temp > 100) {
526 REISERFS_SB(s)->s_alloc_options.border = 10;
528 REISERFS_SB(s)->s_alloc_options.border =
533 if (!strcmp(this_char, "displacing_large_files")) {
534 SET_OPTION(displacing_large_files);
535 REISERFS_SB(s)->s_alloc_options.large_file_size =
537 && *value) ? simple_strtoul(value, &value, 0) : 16;
540 if (!strcmp(this_char, "displacing_new_packing_localities")) {
541 SET_OPTION(displacing_new_packing_localities);
545 if (!strcmp(this_char, "old_hashed_relocation")) {
546 SET_OPTION(old_hashed_relocation);
550 if (!strcmp(this_char, "new_hashed_relocation")) {
551 SET_OPTION(new_hashed_relocation);
555 if (!strcmp(this_char, "dirid_groups")) {
556 SET_OPTION(dirid_groups);
559 if (!strcmp(this_char, "oid_groups")) {
560 SET_OPTION(oid_groups);
563 if (!strcmp(this_char, "packing_groups")) {
564 SET_OPTION(packing_groups);
567 if (!strcmp(this_char, "hashed_formatted_nodes")) {
568 SET_OPTION(hashed_formatted_nodes);
572 if (!strcmp(this_char, "skip_busy")) {
573 SET_OPTION(skip_busy);
577 if (!strcmp(this_char, "hundredth_slices")) {
578 SET_OPTION(hundredth_slices);
582 if (!strcmp(this_char, "old_way")) {
587 if (!strcmp(this_char, "displace_based_on_dirid")) {
588 SET_OPTION(displace_based_on_dirid);
592 if (!strcmp(this_char, "preallocmin")) {
593 REISERFS_SB(s)->s_alloc_options.preallocmin =
595 && *value) ? simple_strtoul(value, &value, 0) : 4;
599 if (!strcmp(this_char, "preallocsize")) {
600 REISERFS_SB(s)->s_alloc_options.preallocsize =
602 && *value) ? simple_strtoul(value, &value,
608 reiserfs_warning(s, "zam-4001: %s : unknown option - %s",
609 __FUNCTION__, this_char);
613 reiserfs_warning(s, "allocator options = [%08x]\n", SB_ALLOC_OPTS(s));
617 static inline void new_hashed_relocation(reiserfs_blocknr_hint_t * hint)
620 if (hint->formatted_node) {
621 hash_in = (char *)&hint->key.k_dir_id;
624 //hint->search_start = hint->beg;
625 hash_in = (char *)&hint->key.k_dir_id;
627 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
628 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
631 (char *)(&INODE_PKEY(hint->inode)->k_objectid);
635 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
639 * Relocation based on dirid, hashing them into a given bitmap block
640 * files. Formatted nodes are unaffected, a seperate policy covers them
642 static void dirid_groups(reiserfs_blocknr_hint_t * hint)
647 struct super_block *sb = hint->th->t_super;
649 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
650 else if (hint->formatted_node)
651 dirid = hint->key.k_dir_id;
654 bm = bmap_hash_id(sb, dirid);
655 hash = bm * (sb->s_blocksize << 3);
656 /* give a portion of the block group to metadata */
658 hash += sb->s_blocksize / 2;
659 hint->search_start = hash;
664 * Relocation based on oid, hashing them into a given bitmap block
665 * files. Formatted nodes are unaffected, a seperate policy covers them
667 static void oid_groups(reiserfs_blocknr_hint_t * hint)
675 dirid = le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id);
677 /* keep the root dir and it's first set of subdirs close to
678 * the start of the disk
681 hash = (hint->inode->i_sb->s_blocksize << 3);
683 oid = le32_to_cpu(INODE_PKEY(hint->inode)->k_objectid);
684 bm = bmap_hash_id(hint->inode->i_sb, oid);
685 hash = bm * (hint->inode->i_sb->s_blocksize << 3);
687 hint->search_start = hash;
691 /* returns 1 if it finds an indirect item and gets valid hint info
692 * from it, otherwise 0
694 static int get_left_neighbor(reiserfs_blocknr_hint_t * hint)
697 struct buffer_head *bh;
698 struct item_head *ih;
703 if (!hint->path) /* reiserfs code can call this function w/o pointer to path
704 * structure supplied; then we rely on supplied search_start */
708 bh = get_last_bh(path);
709 RFALSE(!bh, "green-4002: Illegal path specified to get_left_neighbor");
711 pos_in_item = path->pos_in_item;
712 item = get_item(path);
714 hint->search_start = bh->b_blocknr;
716 if (!hint->formatted_node && is_indirect_le_ih(ih)) {
717 /* for indirect item: go to left and look for the first non-hole entry
718 in the indirect item */
719 if (pos_in_item == I_UNFM_NUM(ih))
721 // pos_in_item = I_UNFM_NUM (ih) - 1;
722 while (pos_in_item >= 0) {
723 int t = get_block_num(item, pos_in_item);
725 hint->search_start = t;
733 /* does result value fit into specified region? */
737 /* should be, if formatted node, then try to put on first part of the device
738 specified as number of percent with mount option device, else try to put
739 on last of device. This is not to say it is good code to do so,
740 but the effect should be measured. */
741 static inline void set_border_in_hint(struct super_block *s,
742 reiserfs_blocknr_hint_t * hint)
745 SB_BLOCK_COUNT(s) / REISERFS_SB(s)->s_alloc_options.border;
747 if (hint->formatted_node)
748 hint->end = border - 1;
753 static inline void displace_large_file(reiserfs_blocknr_hint_t * hint)
755 if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
758 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_dir_id),
759 4) % (hint->end - hint->beg);
763 keyed_hash((char *)(&INODE_PKEY(hint->inode)->k_objectid),
764 4) % (hint->end - hint->beg);
767 static inline void hash_formatted_node(reiserfs_blocknr_hint_t * hint)
772 hash_in = (char *)&hint->key.k_dir_id;
773 else if (TEST_OPTION(displace_based_on_dirid, hint->th->t_super))
774 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_dir_id);
776 hash_in = (char *)(&INODE_PKEY(hint->inode)->k_objectid);
779 hint->beg + keyed_hash(hash_in, 4) % (hint->end - hint->beg);
783 this_blocknr_allocation_would_make_it_a_large_file(reiserfs_blocknr_hint_t *
786 return hint->block ==
787 REISERFS_SB(hint->th->t_super)->s_alloc_options.large_file_size;
790 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
791 static inline void displace_new_packing_locality(reiserfs_blocknr_hint_t * hint)
793 struct in_core_key *key = &hint->key;
795 hint->th->displace_new_blocks = 0;
797 hint->beg + keyed_hash((char *)(&key->k_objectid),
798 4) % (hint->end - hint->beg);
802 static inline int old_hashed_relocation(reiserfs_blocknr_hint_t * hint)
807 if (hint->formatted_node || hint->inode == NULL) {
811 hash_in = le32_to_cpu((INODE_PKEY(hint->inode))->k_dir_id);
813 hint->beg + (u32) keyed_hash(((char *)(&hash_in)),
814 4) % (hint->end - hint->beg - 1);
815 if (border > hint->search_start)
816 hint->search_start = border;
821 static inline int old_way(reiserfs_blocknr_hint_t * hint)
825 if (hint->formatted_node || hint->inode == NULL) {
831 le32_to_cpu(INODE_PKEY(hint->inode)->k_dir_id) % (hint->end -
833 if (border > hint->search_start)
834 hint->search_start = border;
839 static inline void hundredth_slices(reiserfs_blocknr_hint_t * hint)
841 struct in_core_key *key = &hint->key;
842 b_blocknr_t slice_start;
845 (keyed_hash((char *)(&key->k_dir_id), 4) % 100) * (hint->end / 100);
846 if (slice_start > hint->search_start
847 || slice_start + (hint->end / 100) <= hint->search_start) {
848 hint->search_start = slice_start;
852 static void determine_search_start(reiserfs_blocknr_hint_t * hint,
855 struct super_block *s = hint->th->t_super;
859 hint->end = SB_BLOCK_COUNT(s) - 1;
861 /* This is former border algorithm. Now with tunable border offset */
862 if (concentrating_formatted_nodes(s))
863 set_border_in_hint(s, hint);
865 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
866 /* whenever we create a new directory, we displace it. At first we will
867 hash for location, later we might look for a moderately empty place for
869 if (displacing_new_packing_localities(s)
870 && hint->th->displace_new_blocks) {
871 displace_new_packing_locality(hint);
873 /* we do not continue determine_search_start,
874 * if new packing locality is being displaced */
879 /* all persons should feel encouraged to add more special cases here and
882 if (displacing_large_files(s) && !hint->formatted_node
883 && this_blocknr_allocation_would_make_it_a_large_file(hint)) {
884 displace_large_file(hint);
888 /* if none of our special cases is relevant, use the left neighbor in the
889 tree order of the new node we are allocating for */
890 if (hint->formatted_node && TEST_OPTION(hashed_formatted_nodes, s)) {
891 hash_formatted_node(hint);
895 unfm_hint = get_left_neighbor(hint);
897 /* Mimic old block allocator behaviour, that is if VFS allowed for preallocation,
898 new blocks are displaced based on directory ID. Also, if suggested search_start
899 is less than last preallocated block, we start searching from it, assuming that
900 HDD dataflow is faster in forward direction */
901 if (TEST_OPTION(old_way, s)) {
902 if (!hint->formatted_node) {
903 if (!reiserfs_hashed_relocation(s))
905 else if (!reiserfs_no_unhashed_relocation(s))
906 old_hashed_relocation(hint);
909 && hint->search_start <
910 REISERFS_I(hint->inode)->i_prealloc_block)
912 REISERFS_I(hint->inode)->i_prealloc_block;
917 /* This is an approach proposed by Hans */
918 if (TEST_OPTION(hundredth_slices, s)
919 && !(displacing_large_files(s) && !hint->formatted_node)) {
920 hundredth_slices(hint);
924 /* old_hashed_relocation only works on unformatted */
925 if (!unfm_hint && !hint->formatted_node &&
926 TEST_OPTION(old_hashed_relocation, s)) {
927 old_hashed_relocation(hint);
929 /* new_hashed_relocation works with both formatted/unformatted nodes */
930 if ((!unfm_hint || hint->formatted_node) &&
931 TEST_OPTION(new_hashed_relocation, s)) {
932 new_hashed_relocation(hint);
934 /* dirid grouping works only on unformatted nodes */
935 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
938 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
939 if (hint->formatted_node && TEST_OPTION(dirid_groups, s)) {
944 /* oid grouping works only on unformatted nodes */
945 if (!unfm_hint && !hint->formatted_node && TEST_OPTION(oid_groups, s)) {
951 static int determine_prealloc_size(reiserfs_blocknr_hint_t * hint)
953 /* make minimum size a mount option and benchmark both ways */
954 /* we preallocate blocks only for regular files, specific size */
955 /* benchmark preallocating always and see what happens */
957 hint->prealloc_size = 0;
959 if (!hint->formatted_node && hint->preallocate) {
960 if (S_ISREG(hint->inode->i_mode)
961 && hint->inode->i_size >=
962 REISERFS_SB(hint->th->t_super)->s_alloc_options.
963 preallocmin * hint->inode->i_sb->s_blocksize)
964 hint->prealloc_size =
965 REISERFS_SB(hint->th->t_super)->s_alloc_options.
971 /* XXX I know it could be merged with upper-level function;
972 but may be result function would be too complex. */
973 static inline int allocate_without_wrapping_disk(reiserfs_blocknr_hint_t * hint,
974 b_blocknr_t * new_blocknrs,
976 b_blocknr_t finish, int min,
980 int rest = amount_needed;
983 while (rest > 0 && start <= finish) {
984 nr_allocated = scan_bitmap(hint->th, &start, finish, min,
985 rest + prealloc_size,
986 !hint->formatted_node, hint->block);
988 if (nr_allocated == 0) /* no new blocks allocated, return */
991 /* fill free_blocknrs array first */
992 while (rest > 0 && nr_allocated > 0) {
993 *new_blocknrs++ = start++;
998 /* do we have something to fill prealloc. array also ? */
999 if (nr_allocated > 0) {
1000 /* it means prealloc_size was greater that 0 and we do preallocation */
1001 list_add(&REISERFS_I(hint->inode)->i_prealloc_list,
1002 &SB_JOURNAL(hint->th->t_super)->
1004 REISERFS_I(hint->inode)->i_prealloc_block = start;
1005 REISERFS_I(hint->inode)->i_prealloc_count =
1011 return (amount_needed - rest);
1014 static inline int blocknrs_and_prealloc_arrays_from_search_start
1015 (reiserfs_blocknr_hint_t * hint, b_blocknr_t * new_blocknrs,
1016 int amount_needed) {
1017 struct super_block *s = hint->th->t_super;
1018 b_blocknr_t start = hint->search_start;
1019 b_blocknr_t finish = SB_BLOCK_COUNT(s) - 1;
1021 int nr_allocated = 0;
1024 determine_prealloc_size(hint);
1025 if (!hint->formatted_node) {
1027 #ifdef REISERQUOTA_DEBUG
1028 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1029 "reiserquota: allocating %d blocks id=%u",
1030 amount_needed, hint->inode->i_uid);
1033 DQUOT_ALLOC_BLOCK_NODIRTY(hint->inode, amount_needed);
1034 if (quota_ret) /* Quota exceeded? */
1035 return QUOTA_EXCEEDED;
1036 if (hint->preallocate && hint->prealloc_size) {
1037 #ifdef REISERQUOTA_DEBUG
1038 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1039 "reiserquota: allocating (prealloc) %d blocks id=%u",
1040 hint->prealloc_size, hint->inode->i_uid);
1043 DQUOT_PREALLOC_BLOCK_NODIRTY(hint->inode,
1044 hint->prealloc_size);
1046 hint->preallocate = hint->prealloc_size = 0;
1048 /* for unformatted nodes, force large allocations */
1049 bigalloc = amount_needed;
1053 /* in bigalloc mode, nr_allocated should stay zero until
1054 * the entire allocation is filled
1056 if (unlikely(bigalloc && nr_allocated)) {
1057 reiserfs_warning(s, "bigalloc is %d, nr_allocated %d\n",
1058 bigalloc, nr_allocated);
1059 /* reset things to a sane value */
1060 bigalloc = amount_needed - nr_allocated;
1063 * try pass 0 and pass 1 looking for a nice big
1064 * contiguous allocation. Then reset and look
1065 * for anything you can find.
1067 if (passno == 2 && bigalloc) {
1072 case 0: /* Search from hint->search_start to end of disk */
1073 start = hint->search_start;
1074 finish = SB_BLOCK_COUNT(s) - 1;
1076 case 1: /* Search from hint->beg to hint->search_start */
1078 finish = hint->search_start;
1080 case 2: /* Last chance: Search from 0 to hint->beg */
1084 default: /* We've tried searching everywhere, not enough space */
1085 /* Free the blocks */
1086 if (!hint->formatted_node) {
1087 #ifdef REISERQUOTA_DEBUG
1088 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1089 "reiserquota: freeing (nospace) %d blocks id=%u",
1091 hint->prealloc_size -
1093 hint->inode->i_uid);
1095 DQUOT_FREE_BLOCK_NODIRTY(hint->inode, amount_needed + hint->prealloc_size - nr_allocated); /* Free not allocated blocks */
1097 while (nr_allocated--)
1098 reiserfs_free_block(hint->th, hint->inode,
1099 new_blocknrs[nr_allocated],
1100 !hint->formatted_node);
1102 return NO_DISK_SPACE;
1104 } while ((nr_allocated += allocate_without_wrapping_disk(hint,
1115 if (!hint->formatted_node &&
1116 amount_needed + hint->prealloc_size >
1117 nr_allocated + REISERFS_I(hint->inode)->i_prealloc_count) {
1118 /* Some of preallocation blocks were not allocated */
1119 #ifdef REISERQUOTA_DEBUG
1120 reiserfs_debug(s, REISERFS_DEBUG_CODE,
1121 "reiserquota: freeing (failed prealloc) %d blocks id=%u",
1122 amount_needed + hint->prealloc_size -
1124 REISERFS_I(hint->inode)->i_prealloc_count,
1125 hint->inode->i_uid);
1127 DQUOT_FREE_BLOCK_NODIRTY(hint->inode, amount_needed +
1128 hint->prealloc_size - nr_allocated -
1129 REISERFS_I(hint->inode)->
1136 /* grab new blocknrs from preallocated list */
1137 /* return amount still needed after using them */
1138 static int use_preallocated_list_if_available(reiserfs_blocknr_hint_t * hint,
1139 b_blocknr_t * new_blocknrs,
1142 struct inode *inode = hint->inode;
1144 if (REISERFS_I(inode)->i_prealloc_count > 0) {
1145 while (amount_needed) {
1147 *new_blocknrs++ = REISERFS_I(inode)->i_prealloc_block++;
1148 REISERFS_I(inode)->i_prealloc_count--;
1152 if (REISERFS_I(inode)->i_prealloc_count <= 0) {
1153 list_del(&REISERFS_I(inode)->i_prealloc_list);
1158 /* return amount still needed after using preallocated blocks */
1159 return amount_needed;
1162 int 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
1163 already reserved */ )
1165 int initial_amount_needed = amount_needed;
1167 struct super_block *s = hint->th->t_super;
1169 /* Check if there is enough space, taking into account reserved space */
1170 if (SB_FREE_BLOCKS(s) - REISERFS_SB(s)->reserved_blocks <
1171 amount_needed - reserved_by_us)
1172 return NO_DISK_SPACE;
1173 /* should this be if !hint->inode && hint->preallocate? */
1174 /* do you mean hint->formatted_node can be removed ? - Zam */
1175 /* hint->formatted_node cannot be removed because we try to access
1176 inode information here, and there is often no inode assotiated with
1177 metadata allocations - green */
1179 if (!hint->formatted_node && hint->preallocate) {
1180 amount_needed = use_preallocated_list_if_available
1181 (hint, new_blocknrs, amount_needed);
1182 if (amount_needed == 0) /* all blocknrs we need we got from
1185 new_blocknrs += (initial_amount_needed - amount_needed);
1188 /* find search start and save it in hint structure */
1189 determine_search_start(hint, amount_needed);
1190 if (hint->search_start >= SB_BLOCK_COUNT(s))
1191 hint->search_start = SB_BLOCK_COUNT(s) - 1;
1193 /* allocation itself; fill new_blocknrs and preallocation arrays */
1194 ret = blocknrs_and_prealloc_arrays_from_search_start
1195 (hint, new_blocknrs, amount_needed);
1197 /* we used prealloc. list to fill (partially) new_blocknrs array. If final allocation fails we
1198 * need to return blocks back to prealloc. list or just free them. -- Zam (I chose second
1201 if (ret != CARRY_ON) {
1202 while (amount_needed++ < initial_amount_needed) {
1203 reiserfs_free_block(hint->th, hint->inode,
1204 *(--new_blocknrs), 1);
1210 /* These 2 functions are here to provide blocks reservation to the rest of kernel */
1211 /* Reserve @blocks amount of blocks in fs pointed by @sb. Caller must make sure
1212 there are actually this much blocks on the FS available */
1213 void reiserfs_claim_blocks_to_be_allocated(struct super_block *sb, /* super block of
1217 int blocks /* How much to reserve */
1221 /* Fast case, if reservation is zero - exit immediately. */
1225 spin_lock(&REISERFS_SB(sb)->bitmap_lock);
1226 REISERFS_SB(sb)->reserved_blocks += blocks;
1227 spin_unlock(&REISERFS_SB(sb)->bitmap_lock);
1230 /* Unreserve @blocks amount of blocks in fs pointed by @sb */
1231 void reiserfs_release_claimed_blocks(struct super_block *sb, /* super block of
1235 int blocks /* How much to unreserve */
1239 /* Fast case, if unreservation is zero - exit immediately. */
1243 spin_lock(&REISERFS_SB(sb)->bitmap_lock);
1244 REISERFS_SB(sb)->reserved_blocks -= blocks;
1245 spin_unlock(&REISERFS_SB(sb)->bitmap_lock);
1246 RFALSE(REISERFS_SB(sb)->reserved_blocks < 0,
1247 "amount of blocks reserved became zero?");
1250 /* This function estimates how much pages we will be able to write to FS
1251 used for reiserfs_file_write() purposes for now. */
1252 int reiserfs_can_fit_pages(struct super_block *sb /* superblock of filesystem
1253 to estimate space */ )
1257 spin_lock(&REISERFS_SB(sb)->bitmap_lock);
1259 (SB_FREE_BLOCKS(sb) -
1260 REISERFS_SB(sb)->reserved_blocks) >> (PAGE_CACHE_SHIFT -
1261 sb->s_blocksize_bits);
1262 spin_unlock(&REISERFS_SB(sb)->bitmap_lock);
1264 return space > 0 ? space : 0;