1 /*
2  *  Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README
3  */
4 
5 /*
6  *  Written by Anatoly P. Pinchuk pap@namesys.botik.ru
7  *  Programm System Institute
8  *  Pereslavl-Zalessky Russia
9  */
10 
11 /*
12  *  This file contains functions dealing with S+tree
13  *
14  * B_IS_IN_TREE
15  * copy_item_head
16  * comp_short_keys
17  * comp_keys
18  * comp_short_le_keys
19  * le_key2cpu_key
20  * comp_le_keys
21  * bin_search
22  * get_lkey
23  * get_rkey
24  * key_in_buffer
25  * decrement_bcount
26  * reiserfs_check_path
27  * pathrelse_and_restore
28  * pathrelse
29  * search_by_key_reada
30  * search_by_key
31  * search_for_position_by_key
32  * comp_items
33  * prepare_for_direct_item
34  * prepare_for_direntry_item
35  * prepare_for_delete_or_cut
36  * calc_deleted_bytes_number
37  * init_tb_struct
38  * padd_item
39  * reiserfs_delete_item
40  * reiserfs_delete_solid_item
41  * reiserfs_delete_object
42  * maybe_indirect_to_direct
43  * indirect_to_direct_roll_back
44  * reiserfs_cut_from_item
45  * truncate_directory
46  * reiserfs_do_truncate
47  * reiserfs_paste_into_item
48  * reiserfs_insert_item
49  */
50 
51 #include <linux/time.h>
52 #include <linux/string.h>
53 #include <linux/pagemap.h>
54 #include "reiserfs.h"
55 #include <linux/buffer_head.h>
56 #include <linux/quotaops.h>
57 
58 /* Does the buffer contain a disk block which is in the tree. */
B_IS_IN_TREE(const struct buffer_head * bh)59 inline int B_IS_IN_TREE(const struct buffer_head *bh)
60 {
61 
62 	RFALSE(B_LEVEL(bh) > MAX_HEIGHT,
63 	       "PAP-1010: block (%b) has too big level (%z)", bh, bh);
64 
65 	return (B_LEVEL(bh) != FREE_LEVEL);
66 }
67 
68 //
69 // to gets item head in le form
70 //
copy_item_head(struct item_head * to,const struct item_head * from)71 inline void copy_item_head(struct item_head *to,
72 			   const struct item_head *from)
73 {
74 	memcpy(to, from, IH_SIZE);
75 }
76 
77 /* k1 is pointer to on-disk structure which is stored in little-endian
78    form. k2 is pointer to cpu variable. For key of items of the same
79    object this returns 0.
80    Returns: -1 if key1 < key2
81    0 if key1 == key2
82    1 if key1 > key2 */
comp_short_keys(const struct reiserfs_key * le_key,const struct cpu_key * cpu_key)83 inline int comp_short_keys(const struct reiserfs_key *le_key,
84 			   const struct cpu_key *cpu_key)
85 {
86 	__u32 n;
87 	n = le32_to_cpu(le_key->k_dir_id);
88 	if (n < cpu_key->on_disk_key.k_dir_id)
89 		return -1;
90 	if (n > cpu_key->on_disk_key.k_dir_id)
91 		return 1;
92 	n = le32_to_cpu(le_key->k_objectid);
93 	if (n < cpu_key->on_disk_key.k_objectid)
94 		return -1;
95 	if (n > cpu_key->on_disk_key.k_objectid)
96 		return 1;
97 	return 0;
98 }
99 
100 /* k1 is pointer to on-disk structure which is stored in little-endian
101    form. k2 is pointer to cpu variable.
102    Compare keys using all 4 key fields.
103    Returns: -1 if key1 < key2 0
104    if key1 = key2 1 if key1 > key2 */
comp_keys(const struct reiserfs_key * le_key,const struct cpu_key * cpu_key)105 static inline int comp_keys(const struct reiserfs_key *le_key,
106 			    const struct cpu_key *cpu_key)
107 {
108 	int retval;
109 
110 	retval = comp_short_keys(le_key, cpu_key);
111 	if (retval)
112 		return retval;
113 	if (le_key_k_offset(le_key_version(le_key), le_key) <
114 	    cpu_key_k_offset(cpu_key))
115 		return -1;
116 	if (le_key_k_offset(le_key_version(le_key), le_key) >
117 	    cpu_key_k_offset(cpu_key))
118 		return 1;
119 
120 	if (cpu_key->key_length == 3)
121 		return 0;
122 
123 	/* this part is needed only when tail conversion is in progress */
124 	if (le_key_k_type(le_key_version(le_key), le_key) <
125 	    cpu_key_k_type(cpu_key))
126 		return -1;
127 
128 	if (le_key_k_type(le_key_version(le_key), le_key) >
129 	    cpu_key_k_type(cpu_key))
130 		return 1;
131 
132 	return 0;
133 }
134 
comp_short_le_keys(const struct reiserfs_key * key1,const struct reiserfs_key * key2)135 inline int comp_short_le_keys(const struct reiserfs_key *key1,
136 			      const struct reiserfs_key *key2)
137 {
138 	__u32 *k1_u32, *k2_u32;
139 	int key_length = REISERFS_SHORT_KEY_LEN;
140 
141 	k1_u32 = (__u32 *) key1;
142 	k2_u32 = (__u32 *) key2;
143 	for (; key_length--; ++k1_u32, ++k2_u32) {
144 		if (le32_to_cpu(*k1_u32) < le32_to_cpu(*k2_u32))
145 			return -1;
146 		if (le32_to_cpu(*k1_u32) > le32_to_cpu(*k2_u32))
147 			return 1;
148 	}
149 	return 0;
150 }
151 
le_key2cpu_key(struct cpu_key * to,const struct reiserfs_key * from)152 inline void le_key2cpu_key(struct cpu_key *to, const struct reiserfs_key *from)
153 {
154 	int version;
155 	to->on_disk_key.k_dir_id = le32_to_cpu(from->k_dir_id);
156 	to->on_disk_key.k_objectid = le32_to_cpu(from->k_objectid);
157 
158 	// find out version of the key
159 	version = le_key_version(from);
160 	to->version = version;
161 	to->on_disk_key.k_offset = le_key_k_offset(version, from);
162 	to->on_disk_key.k_type = le_key_k_type(version, from);
163 }
164 
165 // this does not say which one is bigger, it only returns 1 if keys
166 // are not equal, 0 otherwise
comp_le_keys(const struct reiserfs_key * k1,const struct reiserfs_key * k2)167 inline int comp_le_keys(const struct reiserfs_key *k1,
168 			const struct reiserfs_key *k2)
169 {
170 	return memcmp(k1, k2, sizeof(struct reiserfs_key));
171 }
172 
173 /**************************************************************************
174  *  Binary search toolkit function                                        *
175  *  Search for an item in the array by the item key                       *
176  *  Returns:    1 if found,  0 if not found;                              *
177  *        *pos = number of the searched element if found, else the        *
178  *        number of the first element that is larger than key.            *
179  **************************************************************************/
180 /* For those not familiar with binary search: lbound is the leftmost item that it
181  could be, rbound the rightmost item that it could be.  We examine the item
182  halfway between lbound and rbound, and that tells us either that we can increase
183  lbound, or decrease rbound, or that we have found it, or if lbound <= rbound that
184  there are no possible items, and we have not found it. With each examination we
185  cut the number of possible items it could be by one more than half rounded down,
186  or we find it. */
bin_search(const void * key,const void * base,int num,int width,int * pos)187 static inline int bin_search(const void *key,	/* Key to search for. */
188 			     const void *base,	/* First item in the array. */
189 			     int num,	/* Number of items in the array. */
190 			     int width,	/* Item size in the array.
191 					   searched. Lest the reader be
192 					   confused, note that this is crafted
193 					   as a general function, and when it
194 					   is applied specifically to the array
195 					   of item headers in a node, width
196 					   is actually the item header size not
197 					   the item size. */
198 			     int *pos /* Number of the searched for element. */
199     )
200 {
201 	int rbound, lbound, j;
202 
203 	for (j = ((rbound = num - 1) + (lbound = 0)) / 2;
204 	     lbound <= rbound; j = (rbound + lbound) / 2)
205 		switch (comp_keys
206 			((struct reiserfs_key *)((char *)base + j * width),
207 			 (struct cpu_key *)key)) {
208 		case -1:
209 			lbound = j + 1;
210 			continue;
211 		case 1:
212 			rbound = j - 1;
213 			continue;
214 		case 0:
215 			*pos = j;
216 			return ITEM_FOUND;	/* Key found in the array.  */
217 		}
218 
219 	/* bin_search did not find given key, it returns position of key,
220 	   that is minimal and greater than the given one. */
221 	*pos = lbound;
222 	return ITEM_NOT_FOUND;
223 }
224 
225 
226 /* Minimal possible key. It is never in the tree. */
227 const struct reiserfs_key MIN_KEY = { 0, 0, {{0, 0},} };
228 
229 /* Maximal possible key. It is never in the tree. */
230 static const struct reiserfs_key MAX_KEY = {
231 	__constant_cpu_to_le32(0xffffffff),
232 	__constant_cpu_to_le32(0xffffffff),
233 	{{__constant_cpu_to_le32(0xffffffff),
234 	  __constant_cpu_to_le32(0xffffffff)},}
235 };
236 
237 /* Get delimiting key of the buffer by looking for it in the buffers in the path, starting from the bottom
238    of the path, and going upwards.  We must check the path's validity at each step.  If the key is not in
239    the path, there is no delimiting key in the tree (buffer is first or last buffer in tree), and in this
240    case we return a special key, either MIN_KEY or MAX_KEY. */
get_lkey(const struct treepath * chk_path,const struct super_block * sb)241 static inline const struct reiserfs_key *get_lkey(const struct treepath *chk_path,
242 						  const struct super_block *sb)
243 {
244 	int position, path_offset = chk_path->path_length;
245 	struct buffer_head *parent;
246 
247 	RFALSE(path_offset < FIRST_PATH_ELEMENT_OFFSET,
248 	       "PAP-5010: invalid offset in the path");
249 
250 	/* While not higher in path than first element. */
251 	while (path_offset-- > FIRST_PATH_ELEMENT_OFFSET) {
252 
253 		RFALSE(!buffer_uptodate
254 		       (PATH_OFFSET_PBUFFER(chk_path, path_offset)),
255 		       "PAP-5020: parent is not uptodate");
256 
257 		/* Parent at the path is not in the tree now. */
258 		if (!B_IS_IN_TREE
259 		    (parent =
260 		     PATH_OFFSET_PBUFFER(chk_path, path_offset)))
261 			return &MAX_KEY;
262 		/* Check whether position in the parent is correct. */
263 		if ((position =
264 		     PATH_OFFSET_POSITION(chk_path,
265 					  path_offset)) >
266 		    B_NR_ITEMS(parent))
267 			return &MAX_KEY;
268 		/* Check whether parent at the path really points to the child. */
269 		if (B_N_CHILD_NUM(parent, position) !=
270 		    PATH_OFFSET_PBUFFER(chk_path,
271 					path_offset + 1)->b_blocknr)
272 			return &MAX_KEY;
273 		/* Return delimiting key if position in the parent is not equal to zero. */
274 		if (position)
275 			return B_N_PDELIM_KEY(parent, position - 1);
276 	}
277 	/* Return MIN_KEY if we are in the root of the buffer tree. */
278 	if (PATH_OFFSET_PBUFFER(chk_path, FIRST_PATH_ELEMENT_OFFSET)->
279 	    b_blocknr == SB_ROOT_BLOCK(sb))
280 		return &MIN_KEY;
281 	return &MAX_KEY;
282 }
283 
284 /* Get delimiting key of the buffer at the path and its right neighbor. */
get_rkey(const struct treepath * chk_path,const struct super_block * sb)285 inline const struct reiserfs_key *get_rkey(const struct treepath *chk_path,
286 					   const struct super_block *sb)
287 {
288 	int position, path_offset = chk_path->path_length;
289 	struct buffer_head *parent;
290 
291 	RFALSE(path_offset < FIRST_PATH_ELEMENT_OFFSET,
292 	       "PAP-5030: invalid offset in the path");
293 
294 	while (path_offset-- > FIRST_PATH_ELEMENT_OFFSET) {
295 
296 		RFALSE(!buffer_uptodate
297 		       (PATH_OFFSET_PBUFFER(chk_path, path_offset)),
298 		       "PAP-5040: parent is not uptodate");
299 
300 		/* Parent at the path is not in the tree now. */
301 		if (!B_IS_IN_TREE
302 		    (parent =
303 		     PATH_OFFSET_PBUFFER(chk_path, path_offset)))
304 			return &MIN_KEY;
305 		/* Check whether position in the parent is correct. */
306 		if ((position =
307 		     PATH_OFFSET_POSITION(chk_path,
308 					  path_offset)) >
309 		    B_NR_ITEMS(parent))
310 			return &MIN_KEY;
311 		/* Check whether parent at the path really points to the child. */
312 		if (B_N_CHILD_NUM(parent, position) !=
313 		    PATH_OFFSET_PBUFFER(chk_path,
314 					path_offset + 1)->b_blocknr)
315 			return &MIN_KEY;
316 		/* Return delimiting key if position in the parent is not the last one. */
317 		if (position != B_NR_ITEMS(parent))
318 			return B_N_PDELIM_KEY(parent, position);
319 	}
320 	/* Return MAX_KEY if we are in the root of the buffer tree. */
321 	if (PATH_OFFSET_PBUFFER(chk_path, FIRST_PATH_ELEMENT_OFFSET)->
322 	    b_blocknr == SB_ROOT_BLOCK(sb))
323 		return &MAX_KEY;
324 	return &MIN_KEY;
325 }
326 
327 /* Check whether a key is contained in the tree rooted from a buffer at a path. */
328 /* This works by looking at the left and right delimiting keys for the buffer in the last path_element in
329    the path.  These delimiting keys are stored at least one level above that buffer in the tree. If the
330    buffer is the first or last node in the tree order then one of the delimiting keys may be absent, and in
331    this case get_lkey and get_rkey return a special key which is MIN_KEY or MAX_KEY. */
key_in_buffer(struct treepath * chk_path,const struct cpu_key * key,struct super_block * sb)332 static inline int key_in_buffer(struct treepath *chk_path,	/* Path which should be checked.  */
333 				const struct cpu_key *key,	/* Key which should be checked.   */
334 				struct super_block *sb
335     )
336 {
337 
338 	RFALSE(!key || chk_path->path_length < FIRST_PATH_ELEMENT_OFFSET
339 	       || chk_path->path_length > MAX_HEIGHT,
340 	       "PAP-5050: pointer to the key(%p) is NULL or invalid path length(%d)",
341 	       key, chk_path->path_length);
342 	RFALSE(!PATH_PLAST_BUFFER(chk_path)->b_bdev,
343 	       "PAP-5060: device must not be NODEV");
344 
345 	if (comp_keys(get_lkey(chk_path, sb), key) == 1)
346 		/* left delimiting key is bigger, that the key we look for */
347 		return 0;
348 	/*  if ( comp_keys(key, get_rkey(chk_path, sb)) != -1 ) */
349 	if (comp_keys(get_rkey(chk_path, sb), key) != 1)
350 		/* key must be less than right delimitiing key */
351 		return 0;
352 	return 1;
353 }
354 
reiserfs_check_path(struct treepath * p)355 int reiserfs_check_path(struct treepath *p)
356 {
357 	RFALSE(p->path_length != ILLEGAL_PATH_ELEMENT_OFFSET,
358 	       "path not properly relsed");
359 	return 0;
360 }
361 
362 /* Drop the reference to each buffer in a path and restore
363  * dirty bits clean when preparing the buffer for the log.
364  * This version should only be called from fix_nodes() */
pathrelse_and_restore(struct super_block * sb,struct treepath * search_path)365 void pathrelse_and_restore(struct super_block *sb,
366 			   struct treepath *search_path)
367 {
368 	int path_offset = search_path->path_length;
369 
370 	RFALSE(path_offset < ILLEGAL_PATH_ELEMENT_OFFSET,
371 	       "clm-4000: invalid path offset");
372 
373 	while (path_offset > ILLEGAL_PATH_ELEMENT_OFFSET) {
374 		struct buffer_head *bh;
375 		bh = PATH_OFFSET_PBUFFER(search_path, path_offset--);
376 		reiserfs_restore_prepared_buffer(sb, bh);
377 		brelse(bh);
378 	}
379 	search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET;
380 }
381 
382 /* Drop the reference to each buffer in a path */
pathrelse(struct treepath * search_path)383 void pathrelse(struct treepath *search_path)
384 {
385 	int path_offset = search_path->path_length;
386 
387 	RFALSE(path_offset < ILLEGAL_PATH_ELEMENT_OFFSET,
388 	       "PAP-5090: invalid path offset");
389 
390 	while (path_offset > ILLEGAL_PATH_ELEMENT_OFFSET)
391 		brelse(PATH_OFFSET_PBUFFER(search_path, path_offset--));
392 
393 	search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET;
394 }
395 
is_leaf(char * buf,int blocksize,struct buffer_head * bh)396 static int is_leaf(char *buf, int blocksize, struct buffer_head *bh)
397 {
398 	struct block_head *blkh;
399 	struct item_head *ih;
400 	int used_space;
401 	int prev_location;
402 	int i;
403 	int nr;
404 
405 	blkh = (struct block_head *)buf;
406 	if (blkh_level(blkh) != DISK_LEAF_NODE_LEVEL) {
407 		reiserfs_warning(NULL, "reiserfs-5080",
408 				 "this should be caught earlier");
409 		return 0;
410 	}
411 
412 	nr = blkh_nr_item(blkh);
413 	if (nr < 1 || nr > ((blocksize - BLKH_SIZE) / (IH_SIZE + MIN_ITEM_LEN))) {
414 		/* item number is too big or too small */
415 		reiserfs_warning(NULL, "reiserfs-5081",
416 				 "nr_item seems wrong: %z", bh);
417 		return 0;
418 	}
419 	ih = (struct item_head *)(buf + BLKH_SIZE) + nr - 1;
420 	used_space = BLKH_SIZE + IH_SIZE * nr + (blocksize - ih_location(ih));
421 	if (used_space != blocksize - blkh_free_space(blkh)) {
422 		/* free space does not match to calculated amount of use space */
423 		reiserfs_warning(NULL, "reiserfs-5082",
424 				 "free space seems wrong: %z", bh);
425 		return 0;
426 	}
427 	// FIXME: it is_leaf will hit performance too much - we may have
428 	// return 1 here
429 
430 	/* check tables of item heads */
431 	ih = (struct item_head *)(buf + BLKH_SIZE);
432 	prev_location = blocksize;
433 	for (i = 0; i < nr; i++, ih++) {
434 		if (le_ih_k_type(ih) == TYPE_ANY) {
435 			reiserfs_warning(NULL, "reiserfs-5083",
436 					 "wrong item type for item %h",
437 					 ih);
438 			return 0;
439 		}
440 		if (ih_location(ih) >= blocksize
441 		    || ih_location(ih) < IH_SIZE * nr) {
442 			reiserfs_warning(NULL, "reiserfs-5084",
443 					 "item location seems wrong: %h",
444 					 ih);
445 			return 0;
446 		}
447 		if (ih_item_len(ih) < 1
448 		    || ih_item_len(ih) > MAX_ITEM_LEN(blocksize)) {
449 			reiserfs_warning(NULL, "reiserfs-5085",
450 					 "item length seems wrong: %h",
451 					 ih);
452 			return 0;
453 		}
454 		if (prev_location - ih_location(ih) != ih_item_len(ih)) {
455 			reiserfs_warning(NULL, "reiserfs-5086",
456 					 "item location seems wrong "
457 					 "(second one): %h", ih);
458 			return 0;
459 		}
460 		prev_location = ih_location(ih);
461 	}
462 
463 	// one may imagine much more checks
464 	return 1;
465 }
466 
467 /* returns 1 if buf looks like an internal node, 0 otherwise */
is_internal(char * buf,int blocksize,struct buffer_head * bh)468 static int is_internal(char *buf, int blocksize, struct buffer_head *bh)
469 {
470 	struct block_head *blkh;
471 	int nr;
472 	int used_space;
473 
474 	blkh = (struct block_head *)buf;
475 	nr = blkh_level(blkh);
476 	if (nr <= DISK_LEAF_NODE_LEVEL || nr > MAX_HEIGHT) {
477 		/* this level is not possible for internal nodes */
478 		reiserfs_warning(NULL, "reiserfs-5087",
479 				 "this should be caught earlier");
480 		return 0;
481 	}
482 
483 	nr = blkh_nr_item(blkh);
484 	if (nr > (blocksize - BLKH_SIZE - DC_SIZE) / (KEY_SIZE + DC_SIZE)) {
485 		/* for internal which is not root we might check min number of keys */
486 		reiserfs_warning(NULL, "reiserfs-5088",
487 				 "number of key seems wrong: %z", bh);
488 		return 0;
489 	}
490 
491 	used_space = BLKH_SIZE + KEY_SIZE * nr + DC_SIZE * (nr + 1);
492 	if (used_space != blocksize - blkh_free_space(blkh)) {
493 		reiserfs_warning(NULL, "reiserfs-5089",
494 				 "free space seems wrong: %z", bh);
495 		return 0;
496 	}
497 	// one may imagine much more checks
498 	return 1;
499 }
500 
501 // make sure that bh contains formatted node of reiserfs tree of
502 // 'level'-th level
is_tree_node(struct buffer_head * bh,int level)503 static int is_tree_node(struct buffer_head *bh, int level)
504 {
505 	if (B_LEVEL(bh) != level) {
506 		reiserfs_warning(NULL, "reiserfs-5090", "node level %d does "
507 				 "not match to the expected one %d",
508 				 B_LEVEL(bh), level);
509 		return 0;
510 	}
511 	if (level == DISK_LEAF_NODE_LEVEL)
512 		return is_leaf(bh->b_data, bh->b_size, bh);
513 
514 	return is_internal(bh->b_data, bh->b_size, bh);
515 }
516 
517 #define SEARCH_BY_KEY_READA 16
518 
519 /*
520  * The function is NOT SCHEDULE-SAFE!
521  * It might unlock the write lock if we needed to wait for a block
522  * to be read. Note that in this case it won't recover the lock to avoid
523  * high contention resulting from too much lock requests, especially
524  * the caller (search_by_key) will perform other schedule-unsafe
525  * operations just after calling this function.
526  *
527  * @return true if we have unlocked
528  */
search_by_key_reada(struct super_block * s,struct buffer_head ** bh,b_blocknr_t * b,int num)529 static bool search_by_key_reada(struct super_block *s,
530 				struct buffer_head **bh,
531 				b_blocknr_t *b, int num)
532 {
533 	int i, j;
534 	bool unlocked = false;
535 
536 	for (i = 0; i < num; i++) {
537 		bh[i] = sb_getblk(s, b[i]);
538 	}
539 	/*
540 	 * We are going to read some blocks on which we
541 	 * have a reference. It's safe, though we might be
542 	 * reading blocks concurrently changed if we release
543 	 * the lock. But it's still fine because we check later
544 	 * if the tree changed
545 	 */
546 	for (j = 0; j < i; j++) {
547 		/*
548 		 * note, this needs attention if we are getting rid of the BKL
549 		 * you have to make sure the prepared bit isn't set on this buffer
550 		 */
551 		if (!buffer_uptodate(bh[j])) {
552 			if (!unlocked) {
553 				reiserfs_write_unlock(s);
554 				unlocked = true;
555 			}
556 			ll_rw_block(READA, 1, bh + j);
557 		}
558 		brelse(bh[j]);
559 	}
560 	return unlocked;
561 }
562 
563 /**************************************************************************
564  * Algorithm   SearchByKey                                                *
565  *             look for item in the Disk S+Tree by its key                *
566  * Input:  sb   -  super block                                            *
567  *         key  - pointer to the key to search                            *
568  * Output: ITEM_FOUND, ITEM_NOT_FOUND or IO_ERROR                         *
569  *         search_path - path from the root to the needed leaf            *
570  **************************************************************************/
571 
572 /* This function fills up the path from the root to the leaf as it
573    descends the tree looking for the key.  It uses reiserfs_bread to
574    try to find buffers in the cache given their block number.  If it
575    does not find them in the cache it reads them from disk.  For each
576    node search_by_key finds using reiserfs_bread it then uses
577    bin_search to look through that node.  bin_search will find the
578    position of the block_number of the next node if it is looking
579    through an internal node.  If it is looking through a leaf node
580    bin_search will find the position of the item which has key either
581    equal to given key, or which is the maximal key less than the given
582    key.  search_by_key returns a path that must be checked for the
583    correctness of the top of the path but need not be checked for the
584    correctness of the bottom of the path */
585 /* The function is NOT SCHEDULE-SAFE! */
search_by_key(struct super_block * sb,const struct cpu_key * key,struct treepath * search_path,int stop_level)586 int search_by_key(struct super_block *sb, const struct cpu_key *key,	/* Key to search. */
587 		  struct treepath *search_path,/* This structure was
588 						   allocated and initialized
589 						   by the calling
590 						   function. It is filled up
591 						   by this function.  */
592 		  int stop_level	/* How far down the tree to search. To
593 					   stop at leaf level - set to
594 					   DISK_LEAF_NODE_LEVEL */
595     )
596 {
597 	b_blocknr_t block_number;
598 	int expected_level;
599 	struct buffer_head *bh;
600 	struct path_element *last_element;
601 	int node_level, retval;
602 	int right_neighbor_of_leaf_node;
603 	int fs_gen;
604 	struct buffer_head *reada_bh[SEARCH_BY_KEY_READA];
605 	b_blocknr_t reada_blocks[SEARCH_BY_KEY_READA];
606 	int reada_count = 0;
607 
608 #ifdef CONFIG_REISERFS_CHECK
609 	int repeat_counter = 0;
610 #endif
611 
612 	PROC_INFO_INC(sb, search_by_key);
613 
614 	/* As we add each node to a path we increase its count.  This means that
615 	   we must be careful to release all nodes in a path before we either
616 	   discard the path struct or re-use the path struct, as we do here. */
617 
618 	pathrelse(search_path);
619 
620 	right_neighbor_of_leaf_node = 0;
621 
622 	/* With each iteration of this loop we search through the items in the
623 	   current node, and calculate the next current node(next path element)
624 	   for the next iteration of this loop.. */
625 	block_number = SB_ROOT_BLOCK(sb);
626 	expected_level = -1;
627 	while (1) {
628 
629 #ifdef CONFIG_REISERFS_CHECK
630 		if (!(++repeat_counter % 50000))
631 			reiserfs_warning(sb, "PAP-5100",
632 					 "%s: there were %d iterations of "
633 					 "while loop looking for key %K",
634 					 current->comm, repeat_counter,
635 					 key);
636 #endif
637 
638 		/* prep path to have another element added to it. */
639 		last_element =
640 		    PATH_OFFSET_PELEMENT(search_path,
641 					 ++search_path->path_length);
642 		fs_gen = get_generation(sb);
643 
644 		/* Read the next tree node, and set the last element in the path to
645 		   have a pointer to it. */
646 		if ((bh = last_element->pe_buffer =
647 		     sb_getblk(sb, block_number))) {
648 			bool unlocked = false;
649 
650 			if (!buffer_uptodate(bh) && reada_count > 1)
651 				/* may unlock the write lock */
652 				unlocked = search_by_key_reada(sb, reada_bh,
653 						    reada_blocks, reada_count);
654 			/*
655 			 * If we haven't already unlocked the write lock,
656 			 * then we need to do that here before reading
657 			 * the current block
658 			 */
659 			if (!buffer_uptodate(bh) && !unlocked) {
660 				reiserfs_write_unlock(sb);
661 				unlocked = true;
662 			}
663 			ll_rw_block(READ, 1, &bh);
664 			wait_on_buffer(bh);
665 
666 			if (unlocked)
667 				reiserfs_write_lock(sb);
668 			if (!buffer_uptodate(bh))
669 				goto io_error;
670 		} else {
671 		      io_error:
672 			search_path->path_length--;
673 			pathrelse(search_path);
674 			return IO_ERROR;
675 		}
676 		reada_count = 0;
677 		if (expected_level == -1)
678 			expected_level = SB_TREE_HEIGHT(sb);
679 		expected_level--;
680 
681 		/* It is possible that schedule occurred. We must check whether the key
682 		   to search is still in the tree rooted from the current buffer. If
683 		   not then repeat search from the root. */
684 		if (fs_changed(fs_gen, sb) &&
685 		    (!B_IS_IN_TREE(bh) ||
686 		     B_LEVEL(bh) != expected_level ||
687 		     !key_in_buffer(search_path, key, sb))) {
688 			PROC_INFO_INC(sb, search_by_key_fs_changed);
689 			PROC_INFO_INC(sb, search_by_key_restarted);
690 			PROC_INFO_INC(sb,
691 				      sbk_restarted[expected_level - 1]);
692 			pathrelse(search_path);
693 
694 			/* Get the root block number so that we can repeat the search
695 			   starting from the root. */
696 			block_number = SB_ROOT_BLOCK(sb);
697 			expected_level = -1;
698 			right_neighbor_of_leaf_node = 0;
699 
700 			/* repeat search from the root */
701 			continue;
702 		}
703 
704 		/* only check that the key is in the buffer if key is not
705 		   equal to the MAX_KEY. Latter case is only possible in
706 		   "finish_unfinished()" processing during mount. */
707 		RFALSE(comp_keys(&MAX_KEY, key) &&
708 		       !key_in_buffer(search_path, key, sb),
709 		       "PAP-5130: key is not in the buffer");
710 #ifdef CONFIG_REISERFS_CHECK
711 		if (REISERFS_SB(sb)->cur_tb) {
712 			print_cur_tb("5140");
713 			reiserfs_panic(sb, "PAP-5140",
714 				       "schedule occurred in do_balance!");
715 		}
716 #endif
717 
718 		// make sure, that the node contents look like a node of
719 		// certain level
720 		if (!is_tree_node(bh, expected_level)) {
721 			reiserfs_error(sb, "vs-5150",
722 				       "invalid format found in block %ld. "
723 				       "Fsck?", bh->b_blocknr);
724 			pathrelse(search_path);
725 			return IO_ERROR;
726 		}
727 
728 		/* ok, we have acquired next formatted node in the tree */
729 		node_level = B_LEVEL(bh);
730 
731 		PROC_INFO_BH_STAT(sb, bh, node_level - 1);
732 
733 		RFALSE(node_level < stop_level,
734 		       "vs-5152: tree level (%d) is less than stop level (%d)",
735 		       node_level, stop_level);
736 
737 		retval = bin_search(key, B_N_PITEM_HEAD(bh, 0),
738 				      B_NR_ITEMS(bh),
739 				      (node_level ==
740 				       DISK_LEAF_NODE_LEVEL) ? IH_SIZE :
741 				      KEY_SIZE,
742 				      &(last_element->pe_position));
743 		if (node_level == stop_level) {
744 			return retval;
745 		}
746 
747 		/* we are not in the stop level */
748 		if (retval == ITEM_FOUND)
749 			/* item has been found, so we choose the pointer which is to the right of the found one */
750 			last_element->pe_position++;
751 
752 		/* if item was not found we choose the position which is to
753 		   the left of the found item. This requires no code,
754 		   bin_search did it already. */
755 
756 		/* So we have chosen a position in the current node which is
757 		   an internal node.  Now we calculate child block number by
758 		   position in the node. */
759 		block_number =
760 		    B_N_CHILD_NUM(bh, last_element->pe_position);
761 
762 		/* if we are going to read leaf nodes, try for read ahead as well */
763 		if ((search_path->reada & PATH_READA) &&
764 		    node_level == DISK_LEAF_NODE_LEVEL + 1) {
765 			int pos = last_element->pe_position;
766 			int limit = B_NR_ITEMS(bh);
767 			struct reiserfs_key *le_key;
768 
769 			if (search_path->reada & PATH_READA_BACK)
770 				limit = 0;
771 			while (reada_count < SEARCH_BY_KEY_READA) {
772 				if (pos == limit)
773 					break;
774 				reada_blocks[reada_count++] =
775 				    B_N_CHILD_NUM(bh, pos);
776 				if (search_path->reada & PATH_READA_BACK)
777 					pos--;
778 				else
779 					pos++;
780 
781 				/*
782 				 * check to make sure we're in the same object
783 				 */
784 				le_key = B_N_PDELIM_KEY(bh, pos);
785 				if (le32_to_cpu(le_key->k_objectid) !=
786 				    key->on_disk_key.k_objectid) {
787 					break;
788 				}
789 			}
790 		}
791 	}
792 }
793 
794 /* Form the path to an item and position in this item which contains
795    file byte defined by key. If there is no such item
796    corresponding to the key, we point the path to the item with
797    maximal key less than key, and *pos_in_item is set to one
798    past the last entry/byte in the item.  If searching for entry in a
799    directory item, and it is not found, *pos_in_item is set to one
800    entry more than the entry with maximal key which is less than the
801    sought key.
802 
803    Note that if there is no entry in this same node which is one more,
804    then we point to an imaginary entry.  for direct items, the
805    position is in units of bytes, for indirect items the position is
806    in units of blocknr entries, for directory items the position is in
807    units of directory entries.  */
808 
809 /* The function is NOT SCHEDULE-SAFE! */
search_for_position_by_key(struct super_block * sb,const struct cpu_key * p_cpu_key,struct treepath * search_path)810 int search_for_position_by_key(struct super_block *sb,	/* Pointer to the super block.          */
811 			       const struct cpu_key *p_cpu_key,	/* Key to search (cpu variable)         */
812 			       struct treepath *search_path	/* Filled up by this function.          */
813     )
814 {
815 	struct item_head *p_le_ih;	/* pointer to on-disk structure */
816 	int blk_size;
817 	loff_t item_offset, offset;
818 	struct reiserfs_dir_entry de;
819 	int retval;
820 
821 	/* If searching for directory entry. */
822 	if (is_direntry_cpu_key(p_cpu_key))
823 		return search_by_entry_key(sb, p_cpu_key, search_path,
824 					   &de);
825 
826 	/* If not searching for directory entry. */
827 
828 	/* If item is found. */
829 	retval = search_item(sb, p_cpu_key, search_path);
830 	if (retval == IO_ERROR)
831 		return retval;
832 	if (retval == ITEM_FOUND) {
833 
834 		RFALSE(!ih_item_len
835 		       (B_N_PITEM_HEAD
836 			(PATH_PLAST_BUFFER(search_path),
837 			 PATH_LAST_POSITION(search_path))),
838 		       "PAP-5165: item length equals zero");
839 
840 		pos_in_item(search_path) = 0;
841 		return POSITION_FOUND;
842 	}
843 
844 	RFALSE(!PATH_LAST_POSITION(search_path),
845 	       "PAP-5170: position equals zero");
846 
847 	/* Item is not found. Set path to the previous item. */
848 	p_le_ih =
849 	    B_N_PITEM_HEAD(PATH_PLAST_BUFFER(search_path),
850 			   --PATH_LAST_POSITION(search_path));
851 	blk_size = sb->s_blocksize;
852 
853 	if (comp_short_keys(&(p_le_ih->ih_key), p_cpu_key)) {
854 		return FILE_NOT_FOUND;
855 	}
856 	// FIXME: quite ugly this far
857 
858 	item_offset = le_ih_k_offset(p_le_ih);
859 	offset = cpu_key_k_offset(p_cpu_key);
860 
861 	/* Needed byte is contained in the item pointed to by the path. */
862 	if (item_offset <= offset &&
863 	    item_offset + op_bytes_number(p_le_ih, blk_size) > offset) {
864 		pos_in_item(search_path) = offset - item_offset;
865 		if (is_indirect_le_ih(p_le_ih)) {
866 			pos_in_item(search_path) /= blk_size;
867 		}
868 		return POSITION_FOUND;
869 	}
870 
871 	/* Needed byte is not contained in the item pointed to by the
872 	   path. Set pos_in_item out of the item. */
873 	if (is_indirect_le_ih(p_le_ih))
874 		pos_in_item(search_path) =
875 		    ih_item_len(p_le_ih) / UNFM_P_SIZE;
876 	else
877 		pos_in_item(search_path) = ih_item_len(p_le_ih);
878 
879 	return POSITION_NOT_FOUND;
880 }
881 
882 /* Compare given item and item pointed to by the path. */
comp_items(const struct item_head * stored_ih,const struct treepath * path)883 int comp_items(const struct item_head *stored_ih, const struct treepath *path)
884 {
885 	struct buffer_head *bh = PATH_PLAST_BUFFER(path);
886 	struct item_head *ih;
887 
888 	/* Last buffer at the path is not in the tree. */
889 	if (!B_IS_IN_TREE(bh))
890 		return 1;
891 
892 	/* Last path position is invalid. */
893 	if (PATH_LAST_POSITION(path) >= B_NR_ITEMS(bh))
894 		return 1;
895 
896 	/* we need only to know, whether it is the same item */
897 	ih = get_ih(path);
898 	return memcmp(stored_ih, ih, IH_SIZE);
899 }
900 
901 /* unformatted nodes are not logged anymore, ever.  This is safe
902 ** now
903 */
904 #define held_by_others(bh) (atomic_read(&(bh)->b_count) > 1)
905 
906 // block can not be forgotten as it is in I/O or held by someone
907 #define block_in_use(bh) (buffer_locked(bh) || (held_by_others(bh)))
908 
909 // prepare for delete or cut of direct item
prepare_for_direct_item(struct treepath * path,struct item_head * le_ih,struct inode * inode,loff_t new_file_length,int * cut_size)910 static inline int prepare_for_direct_item(struct treepath *path,
911 					  struct item_head *le_ih,
912 					  struct inode *inode,
913 					  loff_t new_file_length, int *cut_size)
914 {
915 	loff_t round_len;
916 
917 	if (new_file_length == max_reiserfs_offset(inode)) {
918 		/* item has to be deleted */
919 		*cut_size = -(IH_SIZE + ih_item_len(le_ih));
920 		return M_DELETE;
921 	}
922 	// new file gets truncated
923 	if (get_inode_item_key_version(inode) == KEY_FORMAT_3_6) {
924 		//
925 		round_len = ROUND_UP(new_file_length);
926 		/* this was new_file_length < le_ih ... */
927 		if (round_len < le_ih_k_offset(le_ih)) {
928 			*cut_size = -(IH_SIZE + ih_item_len(le_ih));
929 			return M_DELETE;	/* Delete this item. */
930 		}
931 		/* Calculate first position and size for cutting from item. */
932 		pos_in_item(path) = round_len - (le_ih_k_offset(le_ih) - 1);
933 		*cut_size = -(ih_item_len(le_ih) - pos_in_item(path));
934 
935 		return M_CUT;	/* Cut from this item. */
936 	}
937 
938 	// old file: items may have any length
939 
940 	if (new_file_length < le_ih_k_offset(le_ih)) {
941 		*cut_size = -(IH_SIZE + ih_item_len(le_ih));
942 		return M_DELETE;	/* Delete this item. */
943 	}
944 	/* Calculate first position and size for cutting from item. */
945 	*cut_size = -(ih_item_len(le_ih) -
946 		      (pos_in_item(path) =
947 		       new_file_length + 1 - le_ih_k_offset(le_ih)));
948 	return M_CUT;		/* Cut from this item. */
949 }
950 
prepare_for_direntry_item(struct treepath * path,struct item_head * le_ih,struct inode * inode,loff_t new_file_length,int * cut_size)951 static inline int prepare_for_direntry_item(struct treepath *path,
952 					    struct item_head *le_ih,
953 					    struct inode *inode,
954 					    loff_t new_file_length,
955 					    int *cut_size)
956 {
957 	if (le_ih_k_offset(le_ih) == DOT_OFFSET &&
958 	    new_file_length == max_reiserfs_offset(inode)) {
959 		RFALSE(ih_entry_count(le_ih) != 2,
960 		       "PAP-5220: incorrect empty directory item (%h)", le_ih);
961 		*cut_size = -(IH_SIZE + ih_item_len(le_ih));
962 		return M_DELETE;	/* Delete the directory item containing "." and ".." entry. */
963 	}
964 
965 	if (ih_entry_count(le_ih) == 1) {
966 		/* Delete the directory item such as there is one record only
967 		   in this item */
968 		*cut_size = -(IH_SIZE + ih_item_len(le_ih));
969 		return M_DELETE;
970 	}
971 
972 	/* Cut one record from the directory item. */
973 	*cut_size =
974 	    -(DEH_SIZE +
975 	      entry_length(get_last_bh(path), le_ih, pos_in_item(path)));
976 	return M_CUT;
977 }
978 
979 #define JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD (2 * JOURNAL_PER_BALANCE_CNT + 1)
980 
981 /*  If the path points to a directory or direct item, calculate mode and the size cut, for balance.
982     If the path points to an indirect item, remove some number of its unformatted nodes.
983     In case of file truncate calculate whether this item must be deleted/truncated or last
984     unformatted node of this item will be converted to a direct item.
985     This function returns a determination of what balance mode the calling function should employ. */
prepare_for_delete_or_cut(struct reiserfs_transaction_handle * th,struct inode * inode,struct treepath * path,const struct cpu_key * item_key,int * removed,int * cut_size,unsigned long long new_file_length)986 static char prepare_for_delete_or_cut(struct reiserfs_transaction_handle *th, struct inode *inode, struct treepath *path, const struct cpu_key *item_key, int *removed,	/* Number of unformatted nodes which were removed
987 																						   from end of the file. */
988 				      int *cut_size, unsigned long long new_file_length	/* MAX_KEY_OFFSET in case of delete. */
989     )
990 {
991 	struct super_block *sb = inode->i_sb;
992 	struct item_head *p_le_ih = PATH_PITEM_HEAD(path);
993 	struct buffer_head *bh = PATH_PLAST_BUFFER(path);
994 
995 	BUG_ON(!th->t_trans_id);
996 
997 	/* Stat_data item. */
998 	if (is_statdata_le_ih(p_le_ih)) {
999 
1000 		RFALSE(new_file_length != max_reiserfs_offset(inode),
1001 		       "PAP-5210: mode must be M_DELETE");
1002 
1003 		*cut_size = -(IH_SIZE + ih_item_len(p_le_ih));
1004 		return M_DELETE;
1005 	}
1006 
1007 	/* Directory item. */
1008 	if (is_direntry_le_ih(p_le_ih))
1009 		return prepare_for_direntry_item(path, p_le_ih, inode,
1010 						 new_file_length,
1011 						 cut_size);
1012 
1013 	/* Direct item. */
1014 	if (is_direct_le_ih(p_le_ih))
1015 		return prepare_for_direct_item(path, p_le_ih, inode,
1016 					       new_file_length, cut_size);
1017 
1018 	/* Case of an indirect item. */
1019 	{
1020 	    int blk_size = sb->s_blocksize;
1021 	    struct item_head s_ih;
1022 	    int need_re_search;
1023 	    int delete = 0;
1024 	    int result = M_CUT;
1025 	    int pos = 0;
1026 
1027 	    if ( new_file_length == max_reiserfs_offset (inode) ) {
1028 		/* prepare_for_delete_or_cut() is called by
1029 		 * reiserfs_delete_item() */
1030 		new_file_length = 0;
1031 		delete = 1;
1032 	    }
1033 
1034 	    do {
1035 		need_re_search = 0;
1036 		*cut_size = 0;
1037 		bh = PATH_PLAST_BUFFER(path);
1038 		copy_item_head(&s_ih, PATH_PITEM_HEAD(path));
1039 		pos = I_UNFM_NUM(&s_ih);
1040 
1041 		while (le_ih_k_offset (&s_ih) + (pos - 1) * blk_size > new_file_length) {
1042 		    __le32 *unfm;
1043 		    __u32 block;
1044 
1045 		    /* Each unformatted block deletion may involve one additional
1046 		     * bitmap block into the transaction, thereby the initial
1047 		     * journal space reservation might not be enough. */
1048 		    if (!delete && (*cut_size) != 0 &&
1049 			reiserfs_transaction_free_space(th) < JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD)
1050 			break;
1051 
1052 		    unfm = (__le32 *)B_I_PITEM(bh, &s_ih) + pos - 1;
1053 		    block = get_block_num(unfm, 0);
1054 
1055 		    if (block != 0) {
1056 			reiserfs_prepare_for_journal(sb, bh, 1);
1057 			put_block_num(unfm, 0, 0);
1058 			journal_mark_dirty(th, sb, bh);
1059 			reiserfs_free_block(th, inode, block, 1);
1060 		    }
1061 
1062 		    reiserfs_write_unlock(sb);
1063 		    cond_resched();
1064 		    reiserfs_write_lock(sb);
1065 
1066 		    if (item_moved (&s_ih, path))  {
1067 			need_re_search = 1;
1068 			break;
1069 		    }
1070 
1071 		    pos --;
1072 		    (*removed)++;
1073 		    (*cut_size) -= UNFM_P_SIZE;
1074 
1075 		    if (pos == 0) {
1076 			(*cut_size) -= IH_SIZE;
1077 			result = M_DELETE;
1078 			break;
1079 		    }
1080 		}
1081 		/* a trick.  If the buffer has been logged, this will do nothing.  If
1082 		** we've broken the loop without logging it, it will restore the
1083 		** buffer */
1084 		reiserfs_restore_prepared_buffer(sb, bh);
1085 	    } while (need_re_search &&
1086 		     search_for_position_by_key(sb, item_key, path) == POSITION_FOUND);
1087 	    pos_in_item(path) = pos * UNFM_P_SIZE;
1088 
1089 	    if (*cut_size == 0) {
1090 		/* Nothing were cut. maybe convert last unformatted node to the
1091 		 * direct item? */
1092 		result = M_CONVERT;
1093 	    }
1094 	    return result;
1095 	}
1096 }
1097 
1098 /* Calculate number of bytes which will be deleted or cut during balance */
calc_deleted_bytes_number(struct tree_balance * tb,char mode)1099 static int calc_deleted_bytes_number(struct tree_balance *tb, char mode)
1100 {
1101 	int del_size;
1102 	struct item_head *p_le_ih = PATH_PITEM_HEAD(tb->tb_path);
1103 
1104 	if (is_statdata_le_ih(p_le_ih))
1105 		return 0;
1106 
1107 	del_size =
1108 	    (mode ==
1109 	     M_DELETE) ? ih_item_len(p_le_ih) : -tb->insert_size[0];
1110 	if (is_direntry_le_ih(p_le_ih)) {
1111 		/* return EMPTY_DIR_SIZE; We delete emty directoris only.
1112 		 * we can't use EMPTY_DIR_SIZE, as old format dirs have a different
1113 		 * empty size.  ick. FIXME, is this right? */
1114 		return del_size;
1115 	}
1116 
1117 	if (is_indirect_le_ih(p_le_ih))
1118 		del_size = (del_size / UNFM_P_SIZE) *
1119 				(PATH_PLAST_BUFFER(tb->tb_path)->b_size);
1120 	return del_size;
1121 }
1122 
init_tb_struct(struct reiserfs_transaction_handle * th,struct tree_balance * tb,struct super_block * sb,struct treepath * path,int size)1123 static void init_tb_struct(struct reiserfs_transaction_handle *th,
1124 			   struct tree_balance *tb,
1125 			   struct super_block *sb,
1126 			   struct treepath *path, int size)
1127 {
1128 
1129 	BUG_ON(!th->t_trans_id);
1130 
1131 	memset(tb, '\0', sizeof(struct tree_balance));
1132 	tb->transaction_handle = th;
1133 	tb->tb_sb = sb;
1134 	tb->tb_path = path;
1135 	PATH_OFFSET_PBUFFER(path, ILLEGAL_PATH_ELEMENT_OFFSET) = NULL;
1136 	PATH_OFFSET_POSITION(path, ILLEGAL_PATH_ELEMENT_OFFSET) = 0;
1137 	tb->insert_size[0] = size;
1138 }
1139 
padd_item(char * item,int total_length,int length)1140 void padd_item(char *item, int total_length, int length)
1141 {
1142 	int i;
1143 
1144 	for (i = total_length; i > length;)
1145 		item[--i] = 0;
1146 }
1147 
1148 #ifdef REISERQUOTA_DEBUG
key2type(struct reiserfs_key * ih)1149 char key2type(struct reiserfs_key *ih)
1150 {
1151 	if (is_direntry_le_key(2, ih))
1152 		return 'd';
1153 	if (is_direct_le_key(2, ih))
1154 		return 'D';
1155 	if (is_indirect_le_key(2, ih))
1156 		return 'i';
1157 	if (is_statdata_le_key(2, ih))
1158 		return 's';
1159 	return 'u';
1160 }
1161 
head2type(struct item_head * ih)1162 char head2type(struct item_head *ih)
1163 {
1164 	if (is_direntry_le_ih(ih))
1165 		return 'd';
1166 	if (is_direct_le_ih(ih))
1167 		return 'D';
1168 	if (is_indirect_le_ih(ih))
1169 		return 'i';
1170 	if (is_statdata_le_ih(ih))
1171 		return 's';
1172 	return 'u';
1173 }
1174 #endif
1175 
1176 /* Delete object item.
1177  * th       - active transaction handle
1178  * path     - path to the deleted item
1179  * item_key - key to search for the deleted item
1180  * indode   - used for updating i_blocks and quotas
1181  * un_bh    - NULL or unformatted node pointer
1182  */
reiserfs_delete_item(struct reiserfs_transaction_handle * th,struct treepath * path,const struct cpu_key * item_key,struct inode * inode,struct buffer_head * un_bh)1183 int reiserfs_delete_item(struct reiserfs_transaction_handle *th,
1184 			 struct treepath *path, const struct cpu_key *item_key,
1185 			 struct inode *inode, struct buffer_head *un_bh)
1186 {
1187 	struct super_block *sb = inode->i_sb;
1188 	struct tree_balance s_del_balance;
1189 	struct item_head s_ih;
1190 	struct item_head *q_ih;
1191 	int quota_cut_bytes;
1192 	int ret_value, del_size, removed;
1193 
1194 #ifdef CONFIG_REISERFS_CHECK
1195 	char mode;
1196 	int iter = 0;
1197 #endif
1198 
1199 	BUG_ON(!th->t_trans_id);
1200 
1201 	init_tb_struct(th, &s_del_balance, sb, path,
1202 		       0 /*size is unknown */ );
1203 
1204 	while (1) {
1205 		removed = 0;
1206 
1207 #ifdef CONFIG_REISERFS_CHECK
1208 		iter++;
1209 		mode =
1210 #endif
1211 		    prepare_for_delete_or_cut(th, inode, path,
1212 					      item_key, &removed,
1213 					      &del_size,
1214 					      max_reiserfs_offset(inode));
1215 
1216 		RFALSE(mode != M_DELETE, "PAP-5320: mode must be M_DELETE");
1217 
1218 		copy_item_head(&s_ih, PATH_PITEM_HEAD(path));
1219 		s_del_balance.insert_size[0] = del_size;
1220 
1221 		ret_value = fix_nodes(M_DELETE, &s_del_balance, NULL, NULL);
1222 		if (ret_value != REPEAT_SEARCH)
1223 			break;
1224 
1225 		PROC_INFO_INC(sb, delete_item_restarted);
1226 
1227 		// file system changed, repeat search
1228 		ret_value =
1229 		    search_for_position_by_key(sb, item_key, path);
1230 		if (ret_value == IO_ERROR)
1231 			break;
1232 		if (ret_value == FILE_NOT_FOUND) {
1233 			reiserfs_warning(sb, "vs-5340",
1234 					 "no items of the file %K found",
1235 					 item_key);
1236 			break;
1237 		}
1238 	}			/* while (1) */
1239 
1240 	if (ret_value != CARRY_ON) {
1241 		unfix_nodes(&s_del_balance);
1242 		return 0;
1243 	}
1244 	// reiserfs_delete_item returns item length when success
1245 	ret_value = calc_deleted_bytes_number(&s_del_balance, M_DELETE);
1246 	q_ih = get_ih(path);
1247 	quota_cut_bytes = ih_item_len(q_ih);
1248 
1249 	/* hack so the quota code doesn't have to guess if the file
1250 	 ** has a tail.  On tail insert, we allocate quota for 1 unformatted node.
1251 	 ** We test the offset because the tail might have been
1252 	 ** split into multiple items, and we only want to decrement for
1253 	 ** the unfm node once
1254 	 */
1255 	if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(q_ih)) {
1256 		if ((le_ih_k_offset(q_ih) & (sb->s_blocksize - 1)) == 1) {
1257 			quota_cut_bytes = sb->s_blocksize + UNFM_P_SIZE;
1258 		} else {
1259 			quota_cut_bytes = 0;
1260 		}
1261 	}
1262 
1263 	if (un_bh) {
1264 		int off;
1265 		char *data;
1266 
1267 		/* We are in direct2indirect conversion, so move tail contents
1268 		   to the unformatted node */
1269 		/* note, we do the copy before preparing the buffer because we
1270 		 ** don't care about the contents of the unformatted node yet.
1271 		 ** the only thing we really care about is the direct item's data
1272 		 ** is in the unformatted node.
1273 		 **
1274 		 ** Otherwise, we would have to call reiserfs_prepare_for_journal on
1275 		 ** the unformatted node, which might schedule, meaning we'd have to
1276 		 ** loop all the way back up to the start of the while loop.
1277 		 **
1278 		 ** The unformatted node must be dirtied later on.  We can't be
1279 		 ** sure here if the entire tail has been deleted yet.
1280 		 **
1281 		 ** un_bh is from the page cache (all unformatted nodes are
1282 		 ** from the page cache) and might be a highmem page.  So, we
1283 		 ** can't use un_bh->b_data.
1284 		 ** -clm
1285 		 */
1286 
1287 		data = kmap_atomic(un_bh->b_page);
1288 		off = ((le_ih_k_offset(&s_ih) - 1) & (PAGE_CACHE_SIZE - 1));
1289 		memcpy(data + off,
1290 		       B_I_PITEM(PATH_PLAST_BUFFER(path), &s_ih),
1291 		       ret_value);
1292 		kunmap_atomic(data);
1293 	}
1294 	/* Perform balancing after all resources have been collected at once. */
1295 	do_balance(&s_del_balance, NULL, NULL, M_DELETE);
1296 
1297 #ifdef REISERQUOTA_DEBUG
1298 	reiserfs_debug(sb, REISERFS_DEBUG_CODE,
1299 		       "reiserquota delete_item(): freeing %u, id=%u type=%c",
1300 		       quota_cut_bytes, inode->i_uid, head2type(&s_ih));
1301 #endif
1302 	dquot_free_space_nodirty(inode, quota_cut_bytes);
1303 
1304 	/* Return deleted body length */
1305 	return ret_value;
1306 }
1307 
1308 /* Summary Of Mechanisms For Handling Collisions Between Processes:
1309 
1310  deletion of the body of the object is performed by iput(), with the
1311  result that if multiple processes are operating on a file, the
1312  deletion of the body of the file is deferred until the last process
1313  that has an open inode performs its iput().
1314 
1315  writes and truncates are protected from collisions by use of
1316  semaphores.
1317 
1318  creates, linking, and mknod are protected from collisions with other
1319  processes by making the reiserfs_add_entry() the last step in the
1320  creation, and then rolling back all changes if there was a collision.
1321  - Hans
1322 */
1323 
1324 /* this deletes item which never gets split */
reiserfs_delete_solid_item(struct reiserfs_transaction_handle * th,struct inode * inode,struct reiserfs_key * key)1325 void reiserfs_delete_solid_item(struct reiserfs_transaction_handle *th,
1326 				struct inode *inode, struct reiserfs_key *key)
1327 {
1328 	struct tree_balance tb;
1329 	INITIALIZE_PATH(path);
1330 	int item_len = 0;
1331 	int tb_init = 0;
1332 	struct cpu_key cpu_key;
1333 	int retval;
1334 	int quota_cut_bytes = 0;
1335 
1336 	BUG_ON(!th->t_trans_id);
1337 
1338 	le_key2cpu_key(&cpu_key, key);
1339 
1340 	while (1) {
1341 		retval = search_item(th->t_super, &cpu_key, &path);
1342 		if (retval == IO_ERROR) {
1343 			reiserfs_error(th->t_super, "vs-5350",
1344 				       "i/o failure occurred trying "
1345 				       "to delete %K", &cpu_key);
1346 			break;
1347 		}
1348 		if (retval != ITEM_FOUND) {
1349 			pathrelse(&path);
1350 			// No need for a warning, if there is just no free space to insert '..' item into the newly-created subdir
1351 			if (!
1352 			    ((unsigned long long)
1353 			     GET_HASH_VALUE(le_key_k_offset
1354 					    (le_key_version(key), key)) == 0
1355 			     && (unsigned long long)
1356 			     GET_GENERATION_NUMBER(le_key_k_offset
1357 						   (le_key_version(key),
1358 						    key)) == 1))
1359 				reiserfs_warning(th->t_super, "vs-5355",
1360 						 "%k not found", key);
1361 			break;
1362 		}
1363 		if (!tb_init) {
1364 			tb_init = 1;
1365 			item_len = ih_item_len(PATH_PITEM_HEAD(&path));
1366 			init_tb_struct(th, &tb, th->t_super, &path,
1367 				       -(IH_SIZE + item_len));
1368 		}
1369 		quota_cut_bytes = ih_item_len(PATH_PITEM_HEAD(&path));
1370 
1371 		retval = fix_nodes(M_DELETE, &tb, NULL, NULL);
1372 		if (retval == REPEAT_SEARCH) {
1373 			PROC_INFO_INC(th->t_super, delete_solid_item_restarted);
1374 			continue;
1375 		}
1376 
1377 		if (retval == CARRY_ON) {
1378 			do_balance(&tb, NULL, NULL, M_DELETE);
1379 			if (inode) {	/* Should we count quota for item? (we don't count quotas for save-links) */
1380 #ifdef REISERQUOTA_DEBUG
1381 				reiserfs_debug(th->t_super, REISERFS_DEBUG_CODE,
1382 					       "reiserquota delete_solid_item(): freeing %u id=%u type=%c",
1383 					       quota_cut_bytes, inode->i_uid,
1384 					       key2type(key));
1385 #endif
1386 				dquot_free_space_nodirty(inode,
1387 							 quota_cut_bytes);
1388 			}
1389 			break;
1390 		}
1391 		// IO_ERROR, NO_DISK_SPACE, etc
1392 		reiserfs_warning(th->t_super, "vs-5360",
1393 				 "could not delete %K due to fix_nodes failure",
1394 				 &cpu_key);
1395 		unfix_nodes(&tb);
1396 		break;
1397 	}
1398 
1399 	reiserfs_check_path(&path);
1400 }
1401 
reiserfs_delete_object(struct reiserfs_transaction_handle * th,struct inode * inode)1402 int reiserfs_delete_object(struct reiserfs_transaction_handle *th,
1403 			   struct inode *inode)
1404 {
1405 	int err;
1406 	inode->i_size = 0;
1407 	BUG_ON(!th->t_trans_id);
1408 
1409 	/* for directory this deletes item containing "." and ".." */
1410 	err =
1411 	    reiserfs_do_truncate(th, inode, NULL, 0 /*no timestamp updates */ );
1412 	if (err)
1413 		return err;
1414 
1415 #if defined( USE_INODE_GENERATION_COUNTER )
1416 	if (!old_format_only(th->t_super)) {
1417 		__le32 *inode_generation;
1418 
1419 		inode_generation =
1420 		    &REISERFS_SB(th->t_super)->s_rs->s_inode_generation;
1421 		le32_add_cpu(inode_generation, 1);
1422 	}
1423 /* USE_INODE_GENERATION_COUNTER */
1424 #endif
1425 	reiserfs_delete_solid_item(th, inode, INODE_PKEY(inode));
1426 
1427 	return err;
1428 }
1429 
unmap_buffers(struct page * page,loff_t pos)1430 static void unmap_buffers(struct page *page, loff_t pos)
1431 {
1432 	struct buffer_head *bh;
1433 	struct buffer_head *head;
1434 	struct buffer_head *next;
1435 	unsigned long tail_index;
1436 	unsigned long cur_index;
1437 
1438 	if (page) {
1439 		if (page_has_buffers(page)) {
1440 			tail_index = pos & (PAGE_CACHE_SIZE - 1);
1441 			cur_index = 0;
1442 			head = page_buffers(page);
1443 			bh = head;
1444 			do {
1445 				next = bh->b_this_page;
1446 
1447 				/* we want to unmap the buffers that contain the tail, and
1448 				 ** all the buffers after it (since the tail must be at the
1449 				 ** end of the file).  We don't want to unmap file data
1450 				 ** before the tail, since it might be dirty and waiting to
1451 				 ** reach disk
1452 				 */
1453 				cur_index += bh->b_size;
1454 				if (cur_index > tail_index) {
1455 					reiserfs_unmap_buffer(bh);
1456 				}
1457 				bh = next;
1458 			} while (bh != head);
1459 		}
1460 	}
1461 }
1462 
maybe_indirect_to_direct(struct reiserfs_transaction_handle * th,struct inode * inode,struct page * page,struct treepath * path,const struct cpu_key * item_key,loff_t new_file_size,char * mode)1463 static int maybe_indirect_to_direct(struct reiserfs_transaction_handle *th,
1464 				    struct inode *inode,
1465 				    struct page *page,
1466 				    struct treepath *path,
1467 				    const struct cpu_key *item_key,
1468 				    loff_t new_file_size, char *mode)
1469 {
1470 	struct super_block *sb = inode->i_sb;
1471 	int block_size = sb->s_blocksize;
1472 	int cut_bytes;
1473 	BUG_ON(!th->t_trans_id);
1474 	BUG_ON(new_file_size != inode->i_size);
1475 
1476 	/* the page being sent in could be NULL if there was an i/o error
1477 	 ** reading in the last block.  The user will hit problems trying to
1478 	 ** read the file, but for now we just skip the indirect2direct
1479 	 */
1480 	if (atomic_read(&inode->i_count) > 1 ||
1481 	    !tail_has_to_be_packed(inode) ||
1482 	    !page || (REISERFS_I(inode)->i_flags & i_nopack_mask)) {
1483 		/* leave tail in an unformatted node */
1484 		*mode = M_SKIP_BALANCING;
1485 		cut_bytes =
1486 		    block_size - (new_file_size & (block_size - 1));
1487 		pathrelse(path);
1488 		return cut_bytes;
1489 	}
1490 	/* Perform the conversion to a direct_item. */
1491 	/* return indirect_to_direct(inode, path, item_key,
1492 				  new_file_size, mode); */
1493 	return indirect2direct(th, inode, page, path, item_key,
1494 			       new_file_size, mode);
1495 }
1496 
1497 /* we did indirect_to_direct conversion. And we have inserted direct
1498    item successesfully, but there were no disk space to cut unfm
1499    pointer being converted. Therefore we have to delete inserted
1500    direct item(s) */
indirect_to_direct_roll_back(struct reiserfs_transaction_handle * th,struct inode * inode,struct treepath * path)1501 static void indirect_to_direct_roll_back(struct reiserfs_transaction_handle *th,
1502 					 struct inode *inode, struct treepath *path)
1503 {
1504 	struct cpu_key tail_key;
1505 	int tail_len;
1506 	int removed;
1507 	BUG_ON(!th->t_trans_id);
1508 
1509 	make_cpu_key(&tail_key, inode, inode->i_size + 1, TYPE_DIRECT, 4);	// !!!!
1510 	tail_key.key_length = 4;
1511 
1512 	tail_len =
1513 	    (cpu_key_k_offset(&tail_key) & (inode->i_sb->s_blocksize - 1)) - 1;
1514 	while (tail_len) {
1515 		/* look for the last byte of the tail */
1516 		if (search_for_position_by_key(inode->i_sb, &tail_key, path) ==
1517 		    POSITION_NOT_FOUND)
1518 			reiserfs_panic(inode->i_sb, "vs-5615",
1519 				       "found invalid item");
1520 		RFALSE(path->pos_in_item !=
1521 		       ih_item_len(PATH_PITEM_HEAD(path)) - 1,
1522 		       "vs-5616: appended bytes found");
1523 		PATH_LAST_POSITION(path)--;
1524 
1525 		removed =
1526 		    reiserfs_delete_item(th, path, &tail_key, inode,
1527 					 NULL /*unbh not needed */ );
1528 		RFALSE(removed <= 0
1529 		       || removed > tail_len,
1530 		       "vs-5617: there was tail %d bytes, removed item length %d bytes",
1531 		       tail_len, removed);
1532 		tail_len -= removed;
1533 		set_cpu_key_k_offset(&tail_key,
1534 				     cpu_key_k_offset(&tail_key) - removed);
1535 	}
1536 	reiserfs_warning(inode->i_sb, "reiserfs-5091", "indirect_to_direct "
1537 			 "conversion has been rolled back due to "
1538 			 "lack of disk space");
1539 	//mark_file_without_tail (inode);
1540 	mark_inode_dirty(inode);
1541 }
1542 
1543 /* (Truncate or cut entry) or delete object item. Returns < 0 on failure */
reiserfs_cut_from_item(struct reiserfs_transaction_handle * th,struct treepath * path,struct cpu_key * item_key,struct inode * inode,struct page * page,loff_t new_file_size)1544 int reiserfs_cut_from_item(struct reiserfs_transaction_handle *th,
1545 			   struct treepath *path,
1546 			   struct cpu_key *item_key,
1547 			   struct inode *inode,
1548 			   struct page *page, loff_t new_file_size)
1549 {
1550 	struct super_block *sb = inode->i_sb;
1551 	/* Every function which is going to call do_balance must first
1552 	   create a tree_balance structure.  Then it must fill up this
1553 	   structure by using the init_tb_struct and fix_nodes functions.
1554 	   After that we can make tree balancing. */
1555 	struct tree_balance s_cut_balance;
1556 	struct item_head *p_le_ih;
1557 	int cut_size = 0,	/* Amount to be cut. */
1558 	    ret_value = CARRY_ON, removed = 0,	/* Number of the removed unformatted nodes. */
1559 	    is_inode_locked = 0;
1560 	char mode;		/* Mode of the balance. */
1561 	int retval2 = -1;
1562 	int quota_cut_bytes;
1563 	loff_t tail_pos = 0;
1564 
1565 	BUG_ON(!th->t_trans_id);
1566 
1567 	init_tb_struct(th, &s_cut_balance, inode->i_sb, path,
1568 		       cut_size);
1569 
1570 	/* Repeat this loop until we either cut the item without needing
1571 	   to balance, or we fix_nodes without schedule occurring */
1572 	while (1) {
1573 		/* Determine the balance mode, position of the first byte to
1574 		   be cut, and size to be cut.  In case of the indirect item
1575 		   free unformatted nodes which are pointed to by the cut
1576 		   pointers. */
1577 
1578 		mode =
1579 		    prepare_for_delete_or_cut(th, inode, path,
1580 					      item_key, &removed,
1581 					      &cut_size, new_file_size);
1582 		if (mode == M_CONVERT) {
1583 			/* convert last unformatted node to direct item or leave
1584 			   tail in the unformatted node */
1585 			RFALSE(ret_value != CARRY_ON,
1586 			       "PAP-5570: can not convert twice");
1587 
1588 			ret_value =
1589 			    maybe_indirect_to_direct(th, inode, page,
1590 						     path, item_key,
1591 						     new_file_size, &mode);
1592 			if (mode == M_SKIP_BALANCING)
1593 				/* tail has been left in the unformatted node */
1594 				return ret_value;
1595 
1596 			is_inode_locked = 1;
1597 
1598 			/* removing of last unformatted node will change value we
1599 			   have to return to truncate. Save it */
1600 			retval2 = ret_value;
1601 			/*retval2 = sb->s_blocksize - (new_file_size & (sb->s_blocksize - 1)); */
1602 
1603 			/* So, we have performed the first part of the conversion:
1604 			   inserting the new direct item.  Now we are removing the
1605 			   last unformatted node pointer. Set key to search for
1606 			   it. */
1607 			set_cpu_key_k_type(item_key, TYPE_INDIRECT);
1608 			item_key->key_length = 4;
1609 			new_file_size -=
1610 			    (new_file_size & (sb->s_blocksize - 1));
1611 			tail_pos = new_file_size;
1612 			set_cpu_key_k_offset(item_key, new_file_size + 1);
1613 			if (search_for_position_by_key
1614 			    (sb, item_key,
1615 			     path) == POSITION_NOT_FOUND) {
1616 				print_block(PATH_PLAST_BUFFER(path), 3,
1617 					    PATH_LAST_POSITION(path) - 1,
1618 					    PATH_LAST_POSITION(path) + 1);
1619 				reiserfs_panic(sb, "PAP-5580", "item to "
1620 					       "convert does not exist (%K)",
1621 					       item_key);
1622 			}
1623 			continue;
1624 		}
1625 		if (cut_size == 0) {
1626 			pathrelse(path);
1627 			return 0;
1628 		}
1629 
1630 		s_cut_balance.insert_size[0] = cut_size;
1631 
1632 		ret_value = fix_nodes(mode, &s_cut_balance, NULL, NULL);
1633 		if (ret_value != REPEAT_SEARCH)
1634 			break;
1635 
1636 		PROC_INFO_INC(sb, cut_from_item_restarted);
1637 
1638 		ret_value =
1639 		    search_for_position_by_key(sb, item_key, path);
1640 		if (ret_value == POSITION_FOUND)
1641 			continue;
1642 
1643 		reiserfs_warning(sb, "PAP-5610", "item %K not found",
1644 				 item_key);
1645 		unfix_nodes(&s_cut_balance);
1646 		return (ret_value == IO_ERROR) ? -EIO : -ENOENT;
1647 	}			/* while */
1648 
1649 	// check fix_nodes results (IO_ERROR or NO_DISK_SPACE)
1650 	if (ret_value != CARRY_ON) {
1651 		if (is_inode_locked) {
1652 			// FIXME: this seems to be not needed: we are always able
1653 			// to cut item
1654 			indirect_to_direct_roll_back(th, inode, path);
1655 		}
1656 		if (ret_value == NO_DISK_SPACE)
1657 			reiserfs_warning(sb, "reiserfs-5092",
1658 					 "NO_DISK_SPACE");
1659 		unfix_nodes(&s_cut_balance);
1660 		return -EIO;
1661 	}
1662 
1663 	/* go ahead and perform balancing */
1664 
1665 	RFALSE(mode == M_PASTE || mode == M_INSERT, "invalid mode");
1666 
1667 	/* Calculate number of bytes that need to be cut from the item. */
1668 	quota_cut_bytes =
1669 	    (mode ==
1670 	     M_DELETE) ? ih_item_len(get_ih(path)) : -s_cut_balance.
1671 	    insert_size[0];
1672 	if (retval2 == -1)
1673 		ret_value = calc_deleted_bytes_number(&s_cut_balance, mode);
1674 	else
1675 		ret_value = retval2;
1676 
1677 	/* For direct items, we only change the quota when deleting the last
1678 	 ** item.
1679 	 */
1680 	p_le_ih = PATH_PITEM_HEAD(s_cut_balance.tb_path);
1681 	if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(p_le_ih)) {
1682 		if (mode == M_DELETE &&
1683 		    (le_ih_k_offset(p_le_ih) & (sb->s_blocksize - 1)) ==
1684 		    1) {
1685 			// FIXME: this is to keep 3.5 happy
1686 			REISERFS_I(inode)->i_first_direct_byte = U32_MAX;
1687 			quota_cut_bytes = sb->s_blocksize + UNFM_P_SIZE;
1688 		} else {
1689 			quota_cut_bytes = 0;
1690 		}
1691 	}
1692 #ifdef CONFIG_REISERFS_CHECK
1693 	if (is_inode_locked) {
1694 		struct item_head *le_ih =
1695 		    PATH_PITEM_HEAD(s_cut_balance.tb_path);
1696 		/* we are going to complete indirect2direct conversion. Make
1697 		   sure, that we exactly remove last unformatted node pointer
1698 		   of the item */
1699 		if (!is_indirect_le_ih(le_ih))
1700 			reiserfs_panic(sb, "vs-5652",
1701 				       "item must be indirect %h", le_ih);
1702 
1703 		if (mode == M_DELETE && ih_item_len(le_ih) != UNFM_P_SIZE)
1704 			reiserfs_panic(sb, "vs-5653", "completing "
1705 				       "indirect2direct conversion indirect "
1706 				       "item %h being deleted must be of "
1707 				       "4 byte long", le_ih);
1708 
1709 		if (mode == M_CUT
1710 		    && s_cut_balance.insert_size[0] != -UNFM_P_SIZE) {
1711 			reiserfs_panic(sb, "vs-5654", "can not complete "
1712 				       "indirect2direct conversion of %h "
1713 				       "(CUT, insert_size==%d)",
1714 				       le_ih, s_cut_balance.insert_size[0]);
1715 		}
1716 		/* it would be useful to make sure, that right neighboring
1717 		   item is direct item of this file */
1718 	}
1719 #endif
1720 
1721 	do_balance(&s_cut_balance, NULL, NULL, mode);
1722 	if (is_inode_locked) {
1723 		/* we've done an indirect->direct conversion.  when the data block
1724 		 ** was freed, it was removed from the list of blocks that must
1725 		 ** be flushed before the transaction commits, make sure to
1726 		 ** unmap and invalidate it
1727 		 */
1728 		unmap_buffers(page, tail_pos);
1729 		REISERFS_I(inode)->i_flags &= ~i_pack_on_close_mask;
1730 	}
1731 #ifdef REISERQUOTA_DEBUG
1732 	reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE,
1733 		       "reiserquota cut_from_item(): freeing %u id=%u type=%c",
1734 		       quota_cut_bytes, inode->i_uid, '?');
1735 #endif
1736 	dquot_free_space_nodirty(inode, quota_cut_bytes);
1737 	return ret_value;
1738 }
1739 
truncate_directory(struct reiserfs_transaction_handle * th,struct inode * inode)1740 static void truncate_directory(struct reiserfs_transaction_handle *th,
1741 			       struct inode *inode)
1742 {
1743 	BUG_ON(!th->t_trans_id);
1744 	if (inode->i_nlink)
1745 		reiserfs_error(inode->i_sb, "vs-5655", "link count != 0");
1746 
1747 	set_le_key_k_offset(KEY_FORMAT_3_5, INODE_PKEY(inode), DOT_OFFSET);
1748 	set_le_key_k_type(KEY_FORMAT_3_5, INODE_PKEY(inode), TYPE_DIRENTRY);
1749 	reiserfs_delete_solid_item(th, inode, INODE_PKEY(inode));
1750 	reiserfs_update_sd(th, inode);
1751 	set_le_key_k_offset(KEY_FORMAT_3_5, INODE_PKEY(inode), SD_OFFSET);
1752 	set_le_key_k_type(KEY_FORMAT_3_5, INODE_PKEY(inode), TYPE_STAT_DATA);
1753 }
1754 
1755 /* Truncate file to the new size. Note, this must be called with a transaction
1756    already started */
reiserfs_do_truncate(struct reiserfs_transaction_handle * th,struct inode * inode,struct page * page,int update_timestamps)1757 int reiserfs_do_truncate(struct reiserfs_transaction_handle *th,
1758 			  struct inode *inode,	/* ->i_size contains new size */
1759 			 struct page *page,	/* up to date for last block */
1760 			 int update_timestamps	/* when it is called by
1761 						   file_release to convert
1762 						   the tail - no timestamps
1763 						   should be updated */
1764     )
1765 {
1766 	INITIALIZE_PATH(s_search_path);	/* Path to the current object item. */
1767 	struct item_head *p_le_ih;	/* Pointer to an item header. */
1768 	struct cpu_key s_item_key;	/* Key to search for a previous file item. */
1769 	loff_t file_size,	/* Old file size. */
1770 	 new_file_size;	/* New file size. */
1771 	int deleted;		/* Number of deleted or truncated bytes. */
1772 	int retval;
1773 	int err = 0;
1774 
1775 	BUG_ON(!th->t_trans_id);
1776 	if (!
1777 	    (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)
1778 	     || S_ISLNK(inode->i_mode)))
1779 		return 0;
1780 
1781 	if (S_ISDIR(inode->i_mode)) {
1782 		// deletion of directory - no need to update timestamps
1783 		truncate_directory(th, inode);
1784 		return 0;
1785 	}
1786 
1787 	/* Get new file size. */
1788 	new_file_size = inode->i_size;
1789 
1790 	// FIXME: note, that key type is unimportant here
1791 	make_cpu_key(&s_item_key, inode, max_reiserfs_offset(inode),
1792 		     TYPE_DIRECT, 3);
1793 
1794 	retval =
1795 	    search_for_position_by_key(inode->i_sb, &s_item_key,
1796 				       &s_search_path);
1797 	if (retval == IO_ERROR) {
1798 		reiserfs_error(inode->i_sb, "vs-5657",
1799 			       "i/o failure occurred trying to truncate %K",
1800 			       &s_item_key);
1801 		err = -EIO;
1802 		goto out;
1803 	}
1804 	if (retval == POSITION_FOUND || retval == FILE_NOT_FOUND) {
1805 		reiserfs_error(inode->i_sb, "PAP-5660",
1806 			       "wrong result %d of search for %K", retval,
1807 			       &s_item_key);
1808 
1809 		err = -EIO;
1810 		goto out;
1811 	}
1812 
1813 	s_search_path.pos_in_item--;
1814 
1815 	/* Get real file size (total length of all file items) */
1816 	p_le_ih = PATH_PITEM_HEAD(&s_search_path);
1817 	if (is_statdata_le_ih(p_le_ih))
1818 		file_size = 0;
1819 	else {
1820 		loff_t offset = le_ih_k_offset(p_le_ih);
1821 		int bytes =
1822 		    op_bytes_number(p_le_ih, inode->i_sb->s_blocksize);
1823 
1824 		/* this may mismatch with real file size: if last direct item
1825 		   had no padding zeros and last unformatted node had no free
1826 		   space, this file would have this file size */
1827 		file_size = offset + bytes - 1;
1828 	}
1829 	/*
1830 	 * are we doing a full truncate or delete, if so
1831 	 * kick in the reada code
1832 	 */
1833 	if (new_file_size == 0)
1834 		s_search_path.reada = PATH_READA | PATH_READA_BACK;
1835 
1836 	if (file_size == 0 || file_size < new_file_size) {
1837 		goto update_and_out;
1838 	}
1839 
1840 	/* Update key to search for the last file item. */
1841 	set_cpu_key_k_offset(&s_item_key, file_size);
1842 
1843 	do {
1844 		/* Cut or delete file item. */
1845 		deleted =
1846 		    reiserfs_cut_from_item(th, &s_search_path, &s_item_key,
1847 					   inode, page, new_file_size);
1848 		if (deleted < 0) {
1849 			reiserfs_warning(inode->i_sb, "vs-5665",
1850 					 "reiserfs_cut_from_item failed");
1851 			reiserfs_check_path(&s_search_path);
1852 			return 0;
1853 		}
1854 
1855 		RFALSE(deleted > file_size,
1856 		       "PAP-5670: reiserfs_cut_from_item: too many bytes deleted: deleted %d, file_size %lu, item_key %K",
1857 		       deleted, file_size, &s_item_key);
1858 
1859 		/* Change key to search the last file item. */
1860 		file_size -= deleted;
1861 
1862 		set_cpu_key_k_offset(&s_item_key, file_size);
1863 
1864 		/* While there are bytes to truncate and previous file item is presented in the tree. */
1865 
1866 		/*
1867 		 ** This loop could take a really long time, and could log
1868 		 ** many more blocks than a transaction can hold.  So, we do a polite
1869 		 ** journal end here, and if the transaction needs ending, we make
1870 		 ** sure the file is consistent before ending the current trans
1871 		 ** and starting a new one
1872 		 */
1873 		if (journal_transaction_should_end(th, 0) ||
1874 		    reiserfs_transaction_free_space(th) <= JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD) {
1875 			int orig_len_alloc = th->t_blocks_allocated;
1876 			pathrelse(&s_search_path);
1877 
1878 			if (update_timestamps) {
1879 				inode->i_mtime = CURRENT_TIME_SEC;
1880 				inode->i_ctime = CURRENT_TIME_SEC;
1881 			}
1882 			reiserfs_update_sd(th, inode);
1883 
1884 			err = journal_end(th, inode->i_sb, orig_len_alloc);
1885 			if (err)
1886 				goto out;
1887 			err = journal_begin(th, inode->i_sb,
1888 					    JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD + JOURNAL_PER_BALANCE_CNT * 4) ;
1889 			if (err)
1890 				goto out;
1891 			reiserfs_update_inode_transaction(inode);
1892 		}
1893 	} while (file_size > ROUND_UP(new_file_size) &&
1894 		 search_for_position_by_key(inode->i_sb, &s_item_key,
1895 					    &s_search_path) == POSITION_FOUND);
1896 
1897 	RFALSE(file_size > ROUND_UP(new_file_size),
1898 	       "PAP-5680: truncate did not finish: new_file_size %Ld, current %Ld, oid %d",
1899 	       new_file_size, file_size, s_item_key.on_disk_key.k_objectid);
1900 
1901       update_and_out:
1902 	if (update_timestamps) {
1903 		// this is truncate, not file closing
1904 		inode->i_mtime = CURRENT_TIME_SEC;
1905 		inode->i_ctime = CURRENT_TIME_SEC;
1906 	}
1907 	reiserfs_update_sd(th, inode);
1908 
1909       out:
1910 	pathrelse(&s_search_path);
1911 	return err;
1912 }
1913 
1914 #ifdef CONFIG_REISERFS_CHECK
1915 // this makes sure, that we __append__, not overwrite or add holes
check_research_for_paste(struct treepath * path,const struct cpu_key * key)1916 static void check_research_for_paste(struct treepath *path,
1917 				     const struct cpu_key *key)
1918 {
1919 	struct item_head *found_ih = get_ih(path);
1920 
1921 	if (is_direct_le_ih(found_ih)) {
1922 		if (le_ih_k_offset(found_ih) +
1923 		    op_bytes_number(found_ih,
1924 				    get_last_bh(path)->b_size) !=
1925 		    cpu_key_k_offset(key)
1926 		    || op_bytes_number(found_ih,
1927 				       get_last_bh(path)->b_size) !=
1928 		    pos_in_item(path))
1929 			reiserfs_panic(NULL, "PAP-5720", "found direct item "
1930 				       "%h or position (%d) does not match "
1931 				       "to key %K", found_ih,
1932 				       pos_in_item(path), key);
1933 	}
1934 	if (is_indirect_le_ih(found_ih)) {
1935 		if (le_ih_k_offset(found_ih) +
1936 		    op_bytes_number(found_ih,
1937 				    get_last_bh(path)->b_size) !=
1938 		    cpu_key_k_offset(key)
1939 		    || I_UNFM_NUM(found_ih) != pos_in_item(path)
1940 		    || get_ih_free_space(found_ih) != 0)
1941 			reiserfs_panic(NULL, "PAP-5730", "found indirect "
1942 				       "item (%h) or position (%d) does not "
1943 				       "match to key (%K)",
1944 				       found_ih, pos_in_item(path), key);
1945 	}
1946 }
1947 #endif				/* config reiserfs check */
1948 
1949 /* Paste bytes to the existing item. Returns bytes number pasted into the item. */
reiserfs_paste_into_item(struct reiserfs_transaction_handle * th,struct treepath * search_path,const struct cpu_key * key,struct inode * inode,const char * body,int pasted_size)1950 int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, struct treepath *search_path,	/* Path to the pasted item.	  */
1951 			     const struct cpu_key *key,	/* Key to search for the needed item. */
1952 			     struct inode *inode,	/* Inode item belongs to */
1953 			     const char *body,	/* Pointer to the bytes to paste.    */
1954 			     int pasted_size)
1955 {				/* Size of pasted bytes.             */
1956 	struct tree_balance s_paste_balance;
1957 	int retval;
1958 	int fs_gen;
1959 
1960 	BUG_ON(!th->t_trans_id);
1961 
1962 	fs_gen = get_generation(inode->i_sb);
1963 
1964 #ifdef REISERQUOTA_DEBUG
1965 	reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE,
1966 		       "reiserquota paste_into_item(): allocating %u id=%u type=%c",
1967 		       pasted_size, inode->i_uid,
1968 		       key2type(&(key->on_disk_key)));
1969 #endif
1970 
1971 	reiserfs_write_unlock(inode->i_sb);
1972 	retval = dquot_alloc_space_nodirty(inode, pasted_size);
1973 	reiserfs_write_lock(inode->i_sb);
1974 	if (retval) {
1975 		pathrelse(search_path);
1976 		return retval;
1977 	}
1978 	init_tb_struct(th, &s_paste_balance, th->t_super, search_path,
1979 		       pasted_size);
1980 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
1981 	s_paste_balance.key = key->on_disk_key;
1982 #endif
1983 
1984 	/* DQUOT_* can schedule, must check before the fix_nodes */
1985 	if (fs_changed(fs_gen, inode->i_sb)) {
1986 		goto search_again;
1987 	}
1988 
1989 	while ((retval =
1990 		fix_nodes(M_PASTE, &s_paste_balance, NULL,
1991 			  body)) == REPEAT_SEARCH) {
1992 	      search_again:
1993 		/* file system changed while we were in the fix_nodes */
1994 		PROC_INFO_INC(th->t_super, paste_into_item_restarted);
1995 		retval =
1996 		    search_for_position_by_key(th->t_super, key,
1997 					       search_path);
1998 		if (retval == IO_ERROR) {
1999 			retval = -EIO;
2000 			goto error_out;
2001 		}
2002 		if (retval == POSITION_FOUND) {
2003 			reiserfs_warning(inode->i_sb, "PAP-5710",
2004 					 "entry or pasted byte (%K) exists",
2005 					 key);
2006 			retval = -EEXIST;
2007 			goto error_out;
2008 		}
2009 #ifdef CONFIG_REISERFS_CHECK
2010 		check_research_for_paste(search_path, key);
2011 #endif
2012 	}
2013 
2014 	/* Perform balancing after all resources are collected by fix_nodes, and
2015 	   accessing them will not risk triggering schedule. */
2016 	if (retval == CARRY_ON) {
2017 		do_balance(&s_paste_balance, NULL /*ih */ , body, M_PASTE);
2018 		return 0;
2019 	}
2020 	retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO;
2021       error_out:
2022 	/* this also releases the path */
2023 	unfix_nodes(&s_paste_balance);
2024 #ifdef REISERQUOTA_DEBUG
2025 	reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE,
2026 		       "reiserquota paste_into_item(): freeing %u id=%u type=%c",
2027 		       pasted_size, inode->i_uid,
2028 		       key2type(&(key->on_disk_key)));
2029 #endif
2030 	dquot_free_space_nodirty(inode, pasted_size);
2031 	return retval;
2032 }
2033 
2034 /* Insert new item into the buffer at the path.
2035  * th   - active transaction handle
2036  * path - path to the inserted item
2037  * ih   - pointer to the item header to insert
2038  * body - pointer to the bytes to insert
2039  */
reiserfs_insert_item(struct reiserfs_transaction_handle * th,struct treepath * path,const struct cpu_key * key,struct item_head * ih,struct inode * inode,const char * body)2040 int reiserfs_insert_item(struct reiserfs_transaction_handle *th,
2041 			 struct treepath *path, const struct cpu_key *key,
2042 			 struct item_head *ih, struct inode *inode,
2043 			 const char *body)
2044 {
2045 	struct tree_balance s_ins_balance;
2046 	int retval;
2047 	int fs_gen = 0;
2048 	int quota_bytes = 0;
2049 
2050 	BUG_ON(!th->t_trans_id);
2051 
2052 	if (inode) {		/* Do we count quotas for item? */
2053 		fs_gen = get_generation(inode->i_sb);
2054 		quota_bytes = ih_item_len(ih);
2055 
2056 		/* hack so the quota code doesn't have to guess if the file has
2057 		 ** a tail, links are always tails, so there's no guessing needed
2058 		 */
2059 		if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(ih))
2060 			quota_bytes = inode->i_sb->s_blocksize + UNFM_P_SIZE;
2061 #ifdef REISERQUOTA_DEBUG
2062 		reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE,
2063 			       "reiserquota insert_item(): allocating %u id=%u type=%c",
2064 			       quota_bytes, inode->i_uid, head2type(ih));
2065 #endif
2066 		reiserfs_write_unlock(inode->i_sb);
2067 		/* We can't dirty inode here. It would be immediately written but
2068 		 * appropriate stat item isn't inserted yet... */
2069 		retval = dquot_alloc_space_nodirty(inode, quota_bytes);
2070 		reiserfs_write_lock(inode->i_sb);
2071 		if (retval) {
2072 			pathrelse(path);
2073 			return retval;
2074 		}
2075 	}
2076 	init_tb_struct(th, &s_ins_balance, th->t_super, path,
2077 		       IH_SIZE + ih_item_len(ih));
2078 #ifdef DISPLACE_NEW_PACKING_LOCALITIES
2079 	s_ins_balance.key = key->on_disk_key;
2080 #endif
2081 	/* DQUOT_* can schedule, must check to be sure calling fix_nodes is safe */
2082 	if (inode && fs_changed(fs_gen, inode->i_sb)) {
2083 		goto search_again;
2084 	}
2085 
2086 	while ((retval =
2087 		fix_nodes(M_INSERT, &s_ins_balance, ih,
2088 			  body)) == REPEAT_SEARCH) {
2089 	      search_again:
2090 		/* file system changed while we were in the fix_nodes */
2091 		PROC_INFO_INC(th->t_super, insert_item_restarted);
2092 		retval = search_item(th->t_super, key, path);
2093 		if (retval == IO_ERROR) {
2094 			retval = -EIO;
2095 			goto error_out;
2096 		}
2097 		if (retval == ITEM_FOUND) {
2098 			reiserfs_warning(th->t_super, "PAP-5760",
2099 					 "key %K already exists in the tree",
2100 					 key);
2101 			retval = -EEXIST;
2102 			goto error_out;
2103 		}
2104 	}
2105 
2106 	/* make balancing after all resources will be collected at a time */
2107 	if (retval == CARRY_ON) {
2108 		do_balance(&s_ins_balance, ih, body, M_INSERT);
2109 		return 0;
2110 	}
2111 
2112 	retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO;
2113       error_out:
2114 	/* also releases the path */
2115 	unfix_nodes(&s_ins_balance);
2116 #ifdef REISERQUOTA_DEBUG
2117 	reiserfs_debug(th->t_super, REISERFS_DEBUG_CODE,
2118 		       "reiserquota insert_item(): freeing %u id=%u type=%c",
2119 		       quota_bytes, inode->i_uid, head2type(ih));
2120 #endif
2121 	if (inode)
2122 		dquot_free_space_nodirty(inode, quota_bytes);
2123 	return retval;
2124 }
2125