?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
kvm tools, setup: Create private directory
[linux-2.6/next.git] / fs / gfs2 / dir.c
blob091ee4779538fc5dd5b7e996007de8b2d18f9f8d
1 /*
2 * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
3 * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
4 *
5 * This copyrighted material is made available to anyone wishing to use,
6 * modify, copy, or redistribute it subject to the terms and conditions
7 * of the GNU General Public License version 2.
8 */
9
10 /*
11 * Implements Extendible Hashing as described in:
12 * "Extendible Hashing" by Fagin, et al in
13 * __ACM Trans. on Database Systems__, Sept 1979.
14 *
15 *
16 * Here's the layout of dirents which is essentially the same as that of ext2
17 * within a single block. The field de_name_len is the number of bytes
18 * actually required for the name (no null terminator). The field de_rec_len
19 * is the number of bytes allocated to the dirent. The offset of the next
20 * dirent in the block is (dirent + dirent->de_rec_len). When a dirent is
21 * deleted, the preceding dirent inherits its allocated space, ie
22 * prev->de_rec_len += deleted->de_rec_len. Since the next dirent is obtained
23 * by adding de_rec_len to the current dirent, this essentially causes the
24 * deleted dirent to get jumped over when iterating through all the dirents.
25 *
26 * When deleting the first dirent in a block, there is no previous dirent so
27 * the field de_ino is set to zero to designate it as deleted. When allocating
28 * a dirent, gfs2_dirent_alloc iterates through the dirents in a block. If the
29 * first dirent has (de_ino == 0) and de_rec_len is large enough, this first
30 * dirent is allocated. Otherwise it must go through all the 'used' dirents
31 * searching for one in which the amount of total space minus the amount of
32 * used space will provide enough space for the new dirent.
33 *
34 * There are two types of blocks in which dirents reside. In a stuffed dinode,
35 * the dirents begin at offset sizeof(struct gfs2_dinode) from the beginning of
36 * the block. In leaves, they begin at offset sizeof(struct gfs2_leaf) from the
37 * beginning of the leaf block. The dirents reside in leaves when
38 *
39 * dip->i_diskflags & GFS2_DIF_EXHASH is true
40 *
41 * Otherwise, the dirents are "linear", within a single stuffed dinode block.
42 *
43 * When the dirents are in leaves, the actual contents of the directory file are
44 * used as an array of 64-bit block pointers pointing to the leaf blocks. The
45 * dirents are NOT in the directory file itself. There can be more than one
46 * block pointer in the array that points to the same leaf. In fact, when a
47 * directory is first converted from linear to exhash, all of the pointers
48 * point to the same leaf.
49 *
50 * When a leaf is completely full, the size of the hash table can be
51 * doubled unless it is already at the maximum size which is hard coded into
52 * GFS2_DIR_MAX_DEPTH. After that, leaves are chained together in a linked list,
53 * but never before the maximum hash table size has been reached.
54 */
55
56 #include <linux/slab.h>
57 #include <linux/spinlock.h>
58 #include <linux/buffer_head.h>
59 #include <linux/sort.h>
60 #include <linux/gfs2_ondisk.h>
61 #include <linux/crc32.h>
62 #include <linux/vmalloc.h>
63
64 #include "gfs2.h"
65 #include "incore.h"
66 #include "dir.h"
67 #include "glock.h"
68 #include "inode.h"
69 #include "meta_io.h"
70 #include "quota.h"
71 #include "rgrp.h"
72 #include "trans.h"
73 #include "bmap.h"
74 #include "util.h"
75
76 #define IS_LEAF 1 /* Hashed (leaf) directory */
77 #define IS_DINODE 2 /* Linear (stuffed dinode block) directory */
78
79 #define gfs2_disk_hash2offset(h) (((u64)(h)) >> 1)
80 #define gfs2_dir_offset2hash(p) ((u32)(((u64)(p)) << 1))
81
82 struct qstr gfs2_qdot __read_mostly;
83 struct qstr gfs2_qdotdot __read_mostly;
84
85 typedef int (*gfs2_dscan_t)(const struct gfs2_dirent *dent,
86 const struct qstr *name, void *opaque);
87
88 int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block,
89 struct buffer_head **bhp)
90 {
91 struct buffer_head *bh;
92
93 bh = gfs2_meta_new(ip->i_gl, block);
94 gfs2_trans_add_bh(ip->i_gl, bh, 1);
95 gfs2_metatype_set(bh, GFS2_METATYPE_JD, GFS2_FORMAT_JD);
96 gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header));
97 *bhp = bh;
98 return 0;
99 }
100
101 static int gfs2_dir_get_existing_buffer(struct gfs2_inode *ip, u64 block,
102 struct buffer_head **bhp)
103 {
104 struct buffer_head *bh;
105 int error;
106
107 error = gfs2_meta_read(ip->i_gl, block, DIO_WAIT, &bh);
108 if (error)
109 return error;
110 if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), bh, GFS2_METATYPE_JD)) {
111 brelse(bh);
112 return -EIO;
113 }
114 *bhp = bh;
115 return 0;
116 }
117
118 static int gfs2_dir_write_stuffed(struct gfs2_inode *ip, const char *buf,
119 unsigned int offset, unsigned int size)
120 {
121 struct buffer_head *dibh;
122 int error;
123
124 error = gfs2_meta_inode_buffer(ip, &dibh);
125 if (error)
126 return error;
127
128 gfs2_trans_add_bh(ip->i_gl, dibh, 1);
129 memcpy(dibh->b_data + offset + sizeof(struct gfs2_dinode), buf, size);
130 if (ip->i_inode.i_size < offset + size)
131 i_size_write(&ip->i_inode, offset + size);
132 ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME;
133 gfs2_dinode_out(ip, dibh->b_data);
134
135 brelse(dibh);
136
137 return size;
138 }
139
140
141
142 /**
143 * gfs2_dir_write_data - Write directory information to the inode
144 * @ip: The GFS2 inode
145 * @buf: The buffer containing information to be written
146 * @offset: The file offset to start writing at
147 * @size: The amount of data to write
148 *
149 * Returns: The number of bytes correctly written or error code
150 */
151 static int gfs2_dir_write_data(struct gfs2_inode *ip, const char *buf,
152 u64 offset, unsigned int size)
153 {
154 struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
155 struct buffer_head *dibh;
156 u64 lblock, dblock;
157 u32 extlen = 0;
158 unsigned int o;
159 int copied = 0;
160 int error = 0;
161 int new = 0;
162
163 if (!size)
164 return 0;
165
166 if (gfs2_is_stuffed(ip) &&
167 offset + size <= sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode))
168 return gfs2_dir_write_stuffed(ip, buf, (unsigned int)offset,
169 size);
170
171 if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip)))
172 return -EINVAL;
173
174 if (gfs2_is_stuffed(ip)) {
175 error = gfs2_unstuff_dinode(ip, NULL);
176 if (error)
177 return error;
178 }
179
180 lblock = offset;
181 o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header);
182
183 while (copied < size) {
184 unsigned int amount;
185 struct buffer_head *bh;
186
187 amount = size - copied;
188 if (amount > sdp->sd_sb.sb_bsize - o)
189 amount = sdp->sd_sb.sb_bsize - o;
190
191 if (!extlen) {
192 new = 1;
193 error = gfs2_extent_map(&ip->i_inode, lblock, &new,
194 &dblock, &extlen);
195 if (error)
196 goto fail;
197 error = -EIO;
198 if (gfs2_assert_withdraw(sdp, dblock))
199 goto fail;
200 }
201
202 if (amount == sdp->sd_jbsize || new)
203 error = gfs2_dir_get_new_buffer(ip, dblock, &bh);
204 else
205 error = gfs2_dir_get_existing_buffer(ip, dblock, &bh);
206
207 if (error)
208 goto fail;
209
210 gfs2_trans_add_bh(ip->i_gl, bh, 1);
211 memcpy(bh->b_data + o, buf, amount);
212 brelse(bh);
213
214 buf += amount;
215 copied += amount;
216 lblock++;
217 dblock++;
218 extlen--;
219
220 o = sizeof(struct gfs2_meta_header);
221 }
222
223 out:
224 error = gfs2_meta_inode_buffer(ip, &dibh);
225 if (error)
226 return error;
227
228 if (ip->i_inode.i_size < offset + copied)
229 i_size_write(&ip->i_inode, offset + copied);
230 ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME;
231
232 gfs2_trans_add_bh(ip->i_gl, dibh, 1);
233 gfs2_dinode_out(ip, dibh->b_data);
234 brelse(dibh);
235
236 return copied;
237 fail:
238 if (copied)
239 goto out;
240 return error;
241 }
242
243 static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf,
244 u64 offset, unsigned int size)
245 {
246 struct buffer_head *dibh;
247 int error;
248
249 error = gfs2_meta_inode_buffer(ip, &dibh);
250 if (!error) {
251 offset += sizeof(struct gfs2_dinode);
252 memcpy(buf, dibh->b_data + offset, size);
253 brelse(dibh);
254 }
255
256 return (error) ? error : size;
257 }
258
259
260 /**
261 * gfs2_dir_read_data - Read a data from a directory inode
262 * @ip: The GFS2 Inode
263 * @buf: The buffer to place result into
264 * @offset: File offset to begin jdata_readng from
265 * @size: Amount of data to transfer
266 *
267 * Returns: The amount of data actually copied or the error
268 */
269 static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
270 unsigned int size, unsigned ra)
271 {
272 struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
273 u64 lblock, dblock;
274 u32 extlen = 0;
275 unsigned int o;
276 int copied = 0;
277 int error = 0;
278 u64 disksize = i_size_read(&ip->i_inode);
279
280 if (offset >= disksize)
281 return 0;
282
283 if (offset + size > disksize)
284 size = disksize - offset;
285
286 if (!size)
287 return 0;
288
289 if (gfs2_is_stuffed(ip))
290 return gfs2_dir_read_stuffed(ip, buf, offset, size);
291
292 if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip)))
293 return -EINVAL;
294
295 lblock = offset;
296 o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header);
297
298 while (copied < size) {
299 unsigned int amount;
300 struct buffer_head *bh;
301 int new;
302
303 amount = size - copied;
304 if (amount > sdp->sd_sb.sb_bsize - o)
305 amount = sdp->sd_sb.sb_bsize - o;
306
307 if (!extlen) {
308 new = 0;
309 error = gfs2_extent_map(&ip->i_inode, lblock, &new,
310 &dblock, &extlen);
311 if (error || !dblock)
312 goto fail;
313 BUG_ON(extlen < 1);
314 if (!ra)
315 extlen = 1;
316 bh = gfs2_meta_ra(ip->i_gl, dblock, extlen);
317 } else {
318 error = gfs2_meta_read(ip->i_gl, dblock, DIO_WAIT, &bh);
319 if (error)
320 goto fail;
321 }
322 error = gfs2_metatype_check(sdp, bh, GFS2_METATYPE_JD);
323 if (error) {
324 brelse(bh);
325 goto fail;
326 }
327 dblock++;
328 extlen--;
329 memcpy(buf, bh->b_data + o, amount);
330 brelse(bh);
331 buf += amount;
332 copied += amount;
333 lblock++;
334 o = sizeof(struct gfs2_meta_header);
335 }
336
337 return copied;
338 fail:
339 return (copied) ? copied : error;
340 }
341
342 static inline int gfs2_dirent_sentinel(const struct gfs2_dirent *dent)
343 {
344 return dent->de_inum.no_addr == 0 || dent->de_inum.no_formal_ino == 0;
345 }
346
347 static inline int __gfs2_dirent_find(const struct gfs2_dirent *dent,
348 const struct qstr *name, int ret)
349 {
350 if (!gfs2_dirent_sentinel(dent) &&
351 be32_to_cpu(dent->de_hash) == name->hash &&
352 be16_to_cpu(dent->de_name_len) == name->len &&
353 memcmp(dent+1, name->name, name->len) == 0)
354 return ret;
355 return 0;
356 }
357
358 static int gfs2_dirent_find(const struct gfs2_dirent *dent,
359 const struct qstr *name,
360 void *opaque)
361 {
362 return __gfs2_dirent_find(dent, name, 1);
363 }
364
365 static int gfs2_dirent_prev(const struct gfs2_dirent *dent,
366 const struct qstr *name,
367 void *opaque)
368 {
369 return __gfs2_dirent_find(dent, name, 2);
370 }
371
372 /*
373 * name->name holds ptr to start of block.
374 * name->len holds size of block.
375 */
376 static int gfs2_dirent_last(const struct gfs2_dirent *dent,
377 const struct qstr *name,
378 void *opaque)
379 {
380 const char *start = name->name;
381 const char *end = (const char *)dent + be16_to_cpu(dent->de_rec_len);
382 if (name->len == (end - start))
383 return 1;
384 return 0;
385 }
386
387 static int gfs2_dirent_find_space(const struct gfs2_dirent *dent,
388 const struct qstr *name,
389 void *opaque)
390 {
391 unsigned required = GFS2_DIRENT_SIZE(name->len);
392 unsigned actual = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len));
393 unsigned totlen = be16_to_cpu(dent->de_rec_len);
394
395 if (gfs2_dirent_sentinel(dent))
396 actual = 0;
397 if (totlen - actual >= required)
398 return 1;
399 return 0;
400 }
401
402 struct dirent_gather {
403 const struct gfs2_dirent **pdent;
404 unsigned offset;
405 };
406
407 static int gfs2_dirent_gather(const struct gfs2_dirent *dent,
408 const struct qstr *name,
409 void *opaque)
410 {
411 struct dirent_gather *g = opaque;
412 if (!gfs2_dirent_sentinel(dent)) {
413 g->pdent[g->offset++] = dent;
414 }
415 return 0;
416 }
417
418 /*
419 * Other possible things to check:
420 * - Inode located within filesystem size (and on valid block)
421 * - Valid directory entry type
422 * Not sure how heavy-weight we want to make this... could also check
423 * hash is correct for example, but that would take a lot of extra time.
424 * For now the most important thing is to check that the various sizes
425 * are correct.
426 */
427 static int gfs2_check_dirent(struct gfs2_dirent *dent, unsigned int offset,
428 unsigned int size, unsigned int len, int first)
429 {
430 const char *msg = "gfs2_dirent too small";
431 if (unlikely(size < sizeof(struct gfs2_dirent)))
432 goto error;
433 msg = "gfs2_dirent misaligned";
434 if (unlikely(offset & 0x7))
435 goto error;
436 msg = "gfs2_dirent points beyond end of block";
437 if (unlikely(offset + size > len))
438 goto error;
439 msg = "zero inode number";
440 if (unlikely(!first && gfs2_dirent_sentinel(dent)))
441 goto error;
442 msg = "name length is greater than space in dirent";
443 if (!gfs2_dirent_sentinel(dent) &&
444 unlikely(sizeof(struct gfs2_dirent)+be16_to_cpu(dent->de_name_len) >
445 size))
446 goto error;
447 return 0;
448 error:
449 printk(KERN_WARNING "gfs2_check_dirent: %s (%s)\n", msg,
450 first ? "first in block" : "not first in block");
451 return -EIO;
452 }
453
454 static int gfs2_dirent_offset(const void *buf)
455 {
456 const struct gfs2_meta_header *h = buf;
457 int offset;
458
459 BUG_ON(buf == NULL);
460
461 switch(be32_to_cpu(h->mh_type)) {
462 case GFS2_METATYPE_LF:
463 offset = sizeof(struct gfs2_leaf);
464 break;
465 case GFS2_METATYPE_DI:
466 offset = sizeof(struct gfs2_dinode);
467 break;
468 default:
469 goto wrong_type;
470 }
471 return offset;
472 wrong_type:
473 printk(KERN_WARNING "gfs2_scan_dirent: wrong block type %u\n",
474 be32_to_cpu(h->mh_type));
475 return -1;
476 }
477
478 static struct gfs2_dirent *gfs2_dirent_scan(struct inode *inode, void *buf,
479 unsigned int len, gfs2_dscan_t scan,
480 const struct qstr *name,
481 void *opaque)
482 {
483 struct gfs2_dirent *dent, *prev;
484 unsigned offset;
485 unsigned size;
486 int ret = 0;
487
488 ret = gfs2_dirent_offset(buf);
489 if (ret < 0)
490 goto consist_inode;
491
492 offset = ret;
493 prev = NULL;
494 dent = buf + offset;
495 size = be16_to_cpu(dent->de_rec_len);
496 if (gfs2_check_dirent(dent, offset, size, len, 1))
497 goto consist_inode;
498 do {
499 ret = scan(dent, name, opaque);
500 if (ret)
501 break;
502 offset += size;
503 if (offset == len)
504 break;
505 prev = dent;
506 dent = buf + offset;
507 size = be16_to_cpu(dent->de_rec_len);
508 if (gfs2_check_dirent(dent, offset, size, len, 0))
509 goto consist_inode;
510 } while(1);
511
512 switch(ret) {
513 case 0:
514 return NULL;
515 case 1:
516 return dent;
517 case 2:
518 return prev ? prev : dent;
519 default:
520 BUG_ON(ret > 0);
521 return ERR_PTR(ret);
522 }
523
524 consist_inode:
525 gfs2_consist_inode(GFS2_I(inode));
526 return ERR_PTR(-EIO);
527 }
528
529 static int dirent_check_reclen(struct gfs2_inode *dip,
530 const struct gfs2_dirent *d, const void *end_p)
531 {
532 const void *ptr = d;
533 u16 rec_len = be16_to_cpu(d->de_rec_len);
534
535 if (unlikely(rec_len < sizeof(struct gfs2_dirent)))
536 goto broken;
537 ptr += rec_len;
538 if (ptr < end_p)
539 return rec_len;
540 if (ptr == end_p)
541 return -ENOENT;
542 broken:
543 gfs2_consist_inode(dip);
544 return -EIO;
545 }
546
547 /**
548 * dirent_next - Next dirent
549 * @dip: the directory
550 * @bh: The buffer
551 * @dent: Pointer to list of dirents
552 *
553 * Returns: 0 on success, error code otherwise
554 */
555
556 static int dirent_next(struct gfs2_inode *dip, struct buffer_head *bh,
557 struct gfs2_dirent **dent)
558 {
559 struct gfs2_dirent *cur = *dent, *tmp;
560 char *bh_end = bh->b_data + bh->b_size;
561 int ret;
562
563 ret = dirent_check_reclen(dip, cur, bh_end);
564 if (ret < 0)
565 return ret;
566
567 tmp = (void *)cur + ret;
568 ret = dirent_check_reclen(dip, tmp, bh_end);
569 if (ret == -EIO)
570 return ret;
571
572 /* Only the first dent could ever have de_inum.no_addr == 0 */
573 if (gfs2_dirent_sentinel(tmp)) {
574 gfs2_consist_inode(dip);
575 return -EIO;
576 }
577
578 *dent = tmp;
579 return 0;
580 }
581
582 /**
583 * dirent_del - Delete a dirent
584 * @dip: The GFS2 inode
585 * @bh: The buffer
586 * @prev: The previous dirent
587 * @cur: The current dirent
588 *
589 */
590
591 static void dirent_del(struct gfs2_inode *dip, struct buffer_head *bh,
592 struct gfs2_dirent *prev, struct gfs2_dirent *cur)
593 {
594 u16 cur_rec_len, prev_rec_len;
595
596 if (gfs2_dirent_sentinel(cur)) {
597 gfs2_consist_inode(dip);
598 return;
599 }
600
601 gfs2_trans_add_bh(dip->i_gl, bh, 1);
602
603 /* If there is no prev entry, this is the first entry in the block.
604 The de_rec_len is already as big as it needs to be. Just zero
605 out the inode number and return. */
606
607 if (!prev) {
608 cur->de_inum.no_addr = 0;
609 cur->de_inum.no_formal_ino = 0;
610 return;
611 }
612
613 /* Combine this dentry with the previous one. */
614
615 prev_rec_len = be16_to_cpu(prev->de_rec_len);
616 cur_rec_len = be16_to_cpu(cur->de_rec_len);
617
618 if ((char *)prev + prev_rec_len != (char *)cur)
619 gfs2_consist_inode(dip);
620 if ((char *)cur + cur_rec_len > bh->b_data + bh->b_size)
621 gfs2_consist_inode(dip);
622
623 prev_rec_len += cur_rec_len;
624 prev->de_rec_len = cpu_to_be16(prev_rec_len);
625 }
626
627 /*
628 * Takes a dent from which to grab space as an argument. Returns the
629 * newly created dent.
630 */
631 static struct gfs2_dirent *gfs2_init_dirent(struct inode *inode,
632 struct gfs2_dirent *dent,
633 const struct qstr *name,
634 struct buffer_head *bh)
635 {
636 struct gfs2_inode *ip = GFS2_I(inode);
637 struct gfs2_dirent *ndent;
638 unsigned offset = 0, totlen;
639
640 if (!gfs2_dirent_sentinel(dent))
641 offset = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len));
642 totlen = be16_to_cpu(dent->de_rec_len);
643 BUG_ON(offset + name->len > totlen);
644 gfs2_trans_add_bh(ip->i_gl, bh, 1);
645 ndent = (struct gfs2_dirent *)((char *)dent + offset);
646 dent->de_rec_len = cpu_to_be16(offset);
647 gfs2_qstr2dirent(name, totlen - offset, ndent);
648 return ndent;
649 }
650
651 static struct gfs2_dirent *gfs2_dirent_alloc(struct inode *inode,
652 struct buffer_head *bh,
653 const struct qstr *name)
654 {
655 struct gfs2_dirent *dent;
656 dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
657 gfs2_dirent_find_space, name, NULL);
658 if (!dent || IS_ERR(dent))
659 return dent;
660 return gfs2_init_dirent(inode, dent, name, bh);
661 }
662
663 static int get_leaf(struct gfs2_inode *dip, u64 leaf_no,
664 struct buffer_head **bhp)
665 {
666 int error;
667
668 error = gfs2_meta_read(dip->i_gl, leaf_no, DIO_WAIT, bhp);
669 if (!error && gfs2_metatype_check(GFS2_SB(&dip->i_inode), *bhp, GFS2_METATYPE_LF)) {
670 /* printk(KERN_INFO "block num=%llu\n", leaf_no); */
671 error = -EIO;
672 }
673
674 return error;
675 }
676
677 /**
678 * get_leaf_nr - Get a leaf number associated with the index
679 * @dip: The GFS2 inode
680 * @index:
681 * @leaf_out:
682 *
683 * Returns: 0 on success, error code otherwise
684 */
685
686 static int get_leaf_nr(struct gfs2_inode *dip, u32 index,
687 u64 *leaf_out)
688 {
689 __be64 leaf_no;
690 int error;
691
692 error = gfs2_dir_read_data(dip, (char *)&leaf_no,
693 index * sizeof(__be64),
694 sizeof(__be64), 0);
695 if (error != sizeof(u64))
696 return (error < 0) ? error : -EIO;
697
698 *leaf_out = be64_to_cpu(leaf_no);
699
700 return 0;
701 }
702
703 static int get_first_leaf(struct gfs2_inode *dip, u32 index,
704 struct buffer_head **bh_out)
705 {
706 u64 leaf_no;
707 int error;
708
709 error = get_leaf_nr(dip, index, &leaf_no);
710 if (!error)
711 error = get_leaf(dip, leaf_no, bh_out);
712
713 return error;
714 }
715
716 static struct gfs2_dirent *gfs2_dirent_search(struct inode *inode,
717 const struct qstr *name,
718 gfs2_dscan_t scan,
719 struct buffer_head **pbh)
720 {
721 struct buffer_head *bh;
722 struct gfs2_dirent *dent;
723 struct gfs2_inode *ip = GFS2_I(inode);
724 int error;
725
726 if (ip->i_diskflags & GFS2_DIF_EXHASH) {
727 struct gfs2_leaf *leaf;
728 unsigned hsize = 1 << ip->i_depth;
729 unsigned index;
730 u64 ln;
731 if (hsize * sizeof(u64) != i_size_read(inode)) {
732 gfs2_consist_inode(ip);
733 return ERR_PTR(-EIO);
734 }
735
736 index = name->hash >> (32 - ip->i_depth);
737 error = get_first_leaf(ip, index, &bh);
738 if (error)
739 return ERR_PTR(error);
740 do {
741 dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
742 scan, name, NULL);
743 if (dent)
744 goto got_dent;
745 leaf = (struct gfs2_leaf *)bh->b_data;
746 ln = be64_to_cpu(leaf->lf_next);
747 brelse(bh);
748 if (!ln)
749 break;
750
751 error = get_leaf(ip, ln, &bh);
752 } while(!error);
753
754 return error ? ERR_PTR(error) : NULL;
755 }
756
757
758 error = gfs2_meta_inode_buffer(ip, &bh);
759 if (error)
760 return ERR_PTR(error);
761 dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size, scan, name, NULL);
762 got_dent:
763 if (unlikely(dent == NULL || IS_ERR(dent))) {
764 brelse(bh);
765 bh = NULL;
766 }
767 *pbh = bh;
768 return dent;
769 }
770
771 static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, u16 depth)
772 {
773 struct gfs2_inode *ip = GFS2_I(inode);
774 unsigned int n = 1;
775 u64 bn;
776 int error;
777 struct buffer_head *bh;
778 struct gfs2_leaf *leaf;
779 struct gfs2_dirent *dent;
780 struct qstr name = { .name = "", .len = 0, .hash = 0 };
781
782 error = gfs2_alloc_block(ip, &bn, &n);
783 if (error)
784 return NULL;
785 bh = gfs2_meta_new(ip->i_gl, bn);
786 if (!bh)
787 return NULL;
788
789 gfs2_trans_add_unrevoke(GFS2_SB(inode), bn, 1);
790 gfs2_trans_add_bh(ip->i_gl, bh, 1);
791 gfs2_metatype_set(bh, GFS2_METATYPE_LF, GFS2_FORMAT_LF);
792 leaf = (struct gfs2_leaf *)bh->b_data;
793 leaf->lf_depth = cpu_to_be16(depth);
794 leaf->lf_entries = 0;
795 leaf->lf_dirent_format = cpu_to_be32(GFS2_FORMAT_DE);
796 leaf->lf_next = 0;
797 memset(leaf->lf_reserved, 0, sizeof(leaf->lf_reserved));
798 dent = (struct gfs2_dirent *)(leaf+1);
799 gfs2_qstr2dirent(&name, bh->b_size - sizeof(struct gfs2_leaf), dent);
800 *pbh = bh;
801 return leaf;
802 }
803
804 /**
805 * dir_make_exhash - Convert a stuffed directory into an ExHash directory
806 * @dip: The GFS2 inode
807 *
808 * Returns: 0 on success, error code otherwise
809 */
810
811 static int dir_make_exhash(struct inode *inode)
812 {
813 struct gfs2_inode *dip = GFS2_I(inode);
814 struct gfs2_sbd *sdp = GFS2_SB(inode);
815 struct gfs2_dirent *dent;
816 struct qstr args;
817 struct buffer_head *bh, *dibh;
818 struct gfs2_leaf *leaf;
819 int y;
820 u32 x;
821 __be64 *lp;
822 u64 bn;
823 int error;
824
825 error = gfs2_meta_inode_buffer(dip, &dibh);
826 if (error)
827 return error;
828
829 /* Turn over a new leaf */
830
831 leaf = new_leaf(inode, &bh, 0);
832 if (!leaf)
833 return -ENOSPC;
834 bn = bh->b_blocknr;
835
836 gfs2_assert(sdp, dip->i_entries < (1 << 16));
837 leaf->lf_entries = cpu_to_be16(dip->i_entries);
838
839 /* Copy dirents */
840
841 gfs2_buffer_copy_tail(bh, sizeof(struct gfs2_leaf), dibh,
842 sizeof(struct gfs2_dinode));
843
844 /* Find last entry */
845
846 x = 0;
847 args.len = bh->b_size - sizeof(struct gfs2_dinode) +
848 sizeof(struct gfs2_leaf);
849 args.name = bh->b_data;
850 dent = gfs2_dirent_scan(&dip->i_inode, bh->b_data, bh->b_size,
851 gfs2_dirent_last, &args, NULL);
852 if (!dent) {
853 brelse(bh);
854 brelse(dibh);
855 return -EIO;
856 }
857 if (IS_ERR(dent)) {
858 brelse(bh);
859 brelse(dibh);
860 return PTR_ERR(dent);
861 }
862
863 /* Adjust the last dirent's record length
864 (Remember that dent still points to the last entry.) */
865
866 dent->de_rec_len = cpu_to_be16(be16_to_cpu(dent->de_rec_len) +
867 sizeof(struct gfs2_dinode) -
868 sizeof(struct gfs2_leaf));
869
870 brelse(bh);
871
872 /* We're done with the new leaf block, now setup the new
873 hash table. */
874
875 gfs2_trans_add_bh(dip->i_gl, dibh, 1);
876 gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
877
878 lp = (__be64 *)(dibh->b_data + sizeof(struct gfs2_dinode));
879
880 for (x = sdp->sd_hash_ptrs; x--; lp++)
881 *lp = cpu_to_be64(bn);
882
883 i_size_write(inode, sdp->sd_sb.sb_bsize / 2);
884 gfs2_add_inode_blocks(&dip->i_inode, 1);
885 dip->i_diskflags |= GFS2_DIF_EXHASH;
886
887 for (x = sdp->sd_hash_ptrs, y = -1; x; x >>= 1, y++) ;
888 dip->i_depth = y;
889
890 gfs2_dinode_out(dip, dibh->b_data);
891
892 brelse(dibh);
893
894 return 0;
895 }
896
897 /**
898 * dir_split_leaf - Split a leaf block into two
899 * @dip: The GFS2 inode
900 * @index:
901 * @leaf_no:
902 *
903 * Returns: 0 on success, error code on failure
904 */
905
906 static int dir_split_leaf(struct inode *inode, const struct qstr *name)
907 {
908 struct gfs2_inode *dip = GFS2_I(inode);
909 struct buffer_head *nbh, *obh, *dibh;
910 struct gfs2_leaf *nleaf, *oleaf;
911 struct gfs2_dirent *dent = NULL, *prev = NULL, *next = NULL, *new;
912 u32 start, len, half_len, divider;
913 u64 bn, leaf_no;
914 __be64 *lp;
915 u32 index;
916 int x, moved = 0;
917 int error;
918
919 index = name->hash >> (32 - dip->i_depth);
920 error = get_leaf_nr(dip, index, &leaf_no);
921 if (error)
922 return error;
923
924 /* Get the old leaf block */
925 error = get_leaf(dip, leaf_no, &obh);
926 if (error)
927 return error;
928
929 oleaf = (struct gfs2_leaf *)obh->b_data;
930 if (dip->i_depth == be16_to_cpu(oleaf->lf_depth)) {
931 brelse(obh);
932 return 1; /* can't split */
933 }
934
935 gfs2_trans_add_bh(dip->i_gl, obh, 1);
936
937 nleaf = new_leaf(inode, &nbh, be16_to_cpu(oleaf->lf_depth) + 1);
938 if (!nleaf) {
939 brelse(obh);
940 return -ENOSPC;
941 }
942 bn = nbh->b_blocknr;
943
944 /* Compute the start and len of leaf pointers in the hash table. */
945 len = 1 << (dip->i_depth - be16_to_cpu(oleaf->lf_depth));
946 half_len = len >> 1;
947 if (!half_len) {
948 printk(KERN_WARNING "i_depth %u lf_depth %u index %u\n", dip->i_depth, be16_to_cpu(oleaf->lf_depth), index);
949 gfs2_consist_inode(dip);
950 error = -EIO;
951 goto fail_brelse;
952 }
953
954 start = (index & ~(len - 1));
955
956 /* Change the pointers.
957 Don't bother distinguishing stuffed from non-stuffed.
958 This code is complicated enough already. */
959 lp = kmalloc(half_len * sizeof(__be64), GFP_NOFS);
960 if (!lp) {
961 error = -ENOMEM;
962 goto fail_brelse;
963 }
964
965 /* Change the pointers */
966 for (x = 0; x < half_len; x++)
967 lp[x] = cpu_to_be64(bn);
968
969 error = gfs2_dir_write_data(dip, (char *)lp, start * sizeof(u64),
970 half_len * sizeof(u64));
971 if (error != half_len * sizeof(u64)) {
972 if (error >= 0)
973 error = -EIO;
974 goto fail_lpfree;
975 }
976
977 kfree(lp);
978
979 /* Compute the divider */
980 divider = (start + half_len) << (32 - dip->i_depth);
981
982 /* Copy the entries */
983 dent = (struct gfs2_dirent *)(obh->b_data + sizeof(struct gfs2_leaf));
984
985 do {
986 next = dent;
987 if (dirent_next(dip, obh, &next))
988 next = NULL;
989
990 if (!gfs2_dirent_sentinel(dent) &&
991 be32_to_cpu(dent->de_hash) < divider) {
992 struct qstr str;
993 str.name = (char*)(dent+1);
994 str.len = be16_to_cpu(dent->de_name_len);
995 str.hash = be32_to_cpu(dent->de_hash);
996 new = gfs2_dirent_alloc(inode, nbh, &str);
997 if (IS_ERR(new)) {
998 error = PTR_ERR(new);
999 break;
1000 }
1001
1002 new->de_inum = dent->de_inum; /* No endian worries */
1003 new->de_type = dent->de_type; /* No endian worries */
1004 be16_add_cpu(&nleaf->lf_entries, 1);
1005
1006 dirent_del(dip, obh, prev, dent);
1007
1008 if (!oleaf->lf_entries)
1009 gfs2_consist_inode(dip);
1010 be16_add_cpu(&oleaf->lf_entries, -1);
1011
1012 if (!prev)
1013 prev = dent;
1014
1015 moved = 1;
1016 } else {
1017 prev = dent;
1018 }
1019 dent = next;
1020 } while (dent);
1021
1022 oleaf->lf_depth = nleaf->lf_depth;
1023
1024 error = gfs2_meta_inode_buffer(dip, &dibh);
1025 if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) {
1026 gfs2_trans_add_bh(dip->i_gl, dibh, 1);
1027 gfs2_add_inode_blocks(&dip->i_inode, 1);
1028 gfs2_dinode_out(dip, dibh->b_data);
1029 brelse(dibh);
1030 }
1031
1032 brelse(obh);
1033 brelse(nbh);
1034
1035 return error;
1036
1037 fail_lpfree:
1038 kfree(lp);
1039
1040 fail_brelse:
1041 brelse(obh);
1042 brelse(nbh);
1043 return error;
1044 }
1045
1046 /**
1047 * dir_double_exhash - Double size of ExHash table
1048 * @dip: The GFS2 dinode
1049 *
1050 * Returns: 0 on success, error code on failure
1051 */
1052
1053 static int dir_double_exhash(struct gfs2_inode *dip)
1054 {
1055 struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1056 struct buffer_head *dibh;
1057 u32 hsize;
1058 u64 *buf;
1059 u64 *from, *to;
1060 u64 block;
1061 u64 disksize = i_size_read(&dip->i_inode);
1062 int x;
1063 int error = 0;
1064
1065 hsize = 1 << dip->i_depth;
1066 if (hsize * sizeof(u64) != disksize) {
1067 gfs2_consist_inode(dip);
1068 return -EIO;
1069 }
1070
1071 /* Allocate both the "from" and "to" buffers in one big chunk */
1072
1073 buf = kcalloc(3, sdp->sd_hash_bsize, GFP_NOFS);
1074 if (!buf)
1075 return -ENOMEM;
1076
1077 for (block = disksize >> sdp->sd_hash_bsize_shift; block--;) {
1078 error = gfs2_dir_read_data(dip, (char *)buf,
1079 block * sdp->sd_hash_bsize,
1080 sdp->sd_hash_bsize, 1);
1081 if (error != sdp->sd_hash_bsize) {
1082 if (error >= 0)
1083 error = -EIO;
1084 goto fail;
1085 }
1086
1087 from = buf;
1088 to = (u64 *)((char *)buf + sdp->sd_hash_bsize);
1089
1090 for (x = sdp->sd_hash_ptrs; x--; from++) {
1091 *to++ = *from; /* No endianess worries */
1092 *to++ = *from;
1093 }
1094
1095 error = gfs2_dir_write_data(dip,
1096 (char *)buf + sdp->sd_hash_bsize,
1097 block * sdp->sd_sb.sb_bsize,
1098 sdp->sd_sb.sb_bsize);
1099 if (error != sdp->sd_sb.sb_bsize) {
1100 if (error >= 0)
1101 error = -EIO;
1102 goto fail;
1103 }
1104 }
1105
1106 kfree(buf);
1107
1108 error = gfs2_meta_inode_buffer(dip, &dibh);
1109 if (!gfs2_assert_withdraw(sdp, !error)) {
1110 dip->i_depth++;
1111 gfs2_dinode_out(dip, dibh->b_data);
1112 brelse(dibh);
1113 }
1114
1115 return error;
1116
1117 fail:
1118 kfree(buf);
1119 return error;
1120 }
1121
1122 /**
1123 * compare_dents - compare directory entries by hash value
1124 * @a: first dent
1125 * @b: second dent
1126 *
1127 * When comparing the hash entries of @a to @b:
1128 * gt: returns 1
1129 * lt: returns -1
1130 * eq: returns 0
1131 */
1132
1133 static int compare_dents(const void *a, const void *b)
1134 {
1135 const struct gfs2_dirent *dent_a, *dent_b;
1136 u32 hash_a, hash_b;
1137 int ret = 0;
1138
1139 dent_a = *(const struct gfs2_dirent **)a;
1140 hash_a = be32_to_cpu(dent_a->de_hash);
1141
1142 dent_b = *(const struct gfs2_dirent **)b;
1143 hash_b = be32_to_cpu(dent_b->de_hash);
1144
1145 if (hash_a > hash_b)
1146 ret = 1;
1147 else if (hash_a < hash_b)
1148 ret = -1;
1149 else {
1150 unsigned int len_a = be16_to_cpu(dent_a->de_name_len);
1151 unsigned int len_b = be16_to_cpu(dent_b->de_name_len);
1152
1153 if (len_a > len_b)
1154 ret = 1;
1155 else if (len_a < len_b)
1156 ret = -1;
1157 else
1158 ret = memcmp(dent_a + 1, dent_b + 1, len_a);
1159 }
1160
1161 return ret;
1162 }
1163
1164 /**
1165 * do_filldir_main - read out directory entries
1166 * @dip: The GFS2 inode
1167 * @offset: The offset in the file to read from
1168 * @opaque: opaque data to pass to filldir
1169 * @filldir: The function to pass entries to
1170 * @darr: an array of struct gfs2_dirent pointers to read
1171 * @entries: the number of entries in darr
1172 * @copied: pointer to int that's non-zero if a entry has been copied out
1173 *
1174 * Jump through some hoops to make sure that if there are hash collsions,
1175 * they are read out at the beginning of a buffer. We want to minimize
1176 * the possibility that they will fall into different readdir buffers or
1177 * that someone will want to seek to that location.
1178 *
1179 * Returns: errno, >0 on exception from filldir
1180 */
1181
1182 static int do_filldir_main(struct gfs2_inode *dip, u64 *offset,
1183 void *opaque, filldir_t filldir,
1184 const struct gfs2_dirent **darr, u32 entries,
1185 int *copied)
1186 {
1187 const struct gfs2_dirent *dent, *dent_next;
1188 u64 off, off_next;
1189 unsigned int x, y;
1190 int run = 0;
1191 int error = 0;
1192
1193 sort(darr, entries, sizeof(struct gfs2_dirent *), compare_dents, NULL);
1194
1195 dent_next = darr[0];
1196 off_next = be32_to_cpu(dent_next->de_hash);
1197 off_next = gfs2_disk_hash2offset(off_next);
1198
1199 for (x = 0, y = 1; x < entries; x++, y++) {
1200 dent = dent_next;
1201 off = off_next;
1202
1203 if (y < entries) {
1204 dent_next = darr[y];
1205 off_next = be32_to_cpu(dent_next->de_hash);
1206 off_next = gfs2_disk_hash2offset(off_next);
1207
1208 if (off < *offset)
1209 continue;
1210 *offset = off;
1211
1212 if (off_next == off) {
1213 if (*copied && !run)
1214 return 1;
1215 run = 1;
1216 } else
1217 run = 0;
1218 } else {
1219 if (off < *offset)
1220 continue;
1221 *offset = off;
1222 }
1223
1224 error = filldir(opaque, (const char *)(dent + 1),
1225 be16_to_cpu(dent->de_name_len),
1226 off, be64_to_cpu(dent->de_inum.no_addr),
1227 be16_to_cpu(dent->de_type));
1228 if (error)
1229 return 1;
1230
1231 *copied = 1;
1232 }
1233
1234 /* Increment the *offset by one, so the next time we come into the
1235 do_filldir fxn, we get the next entry instead of the last one in the
1236 current leaf */
1237
1238 (*offset)++;
1239
1240 return 0;
1241 }
1242
1243 static void *gfs2_alloc_sort_buffer(unsigned size)
1244 {
1245 void *ptr = NULL;
1246
1247 if (size < KMALLOC_MAX_SIZE)
1248 ptr = kmalloc(size, GFP_NOFS | __GFP_NOWARN);
1249 if (!ptr)
1250 ptr = __vmalloc(size, GFP_NOFS, PAGE_KERNEL);
1251 return ptr;
1252 }
1253
1254 static void gfs2_free_sort_buffer(void *ptr)
1255 {
1256 if (is_vmalloc_addr(ptr))
1257 vfree(ptr);
1258 else
1259 kfree(ptr);
1260 }
1261
1262 static int gfs2_dir_read_leaf(struct inode *inode, u64 *offset, void *opaque,
1263 filldir_t filldir, int *copied, unsigned *depth,
1264 u64 leaf_no)
1265 {
1266 struct gfs2_inode *ip = GFS2_I(inode);
1267 struct gfs2_sbd *sdp = GFS2_SB(inode);
1268 struct buffer_head *bh;
1269 struct gfs2_leaf *lf;
1270 unsigned entries = 0, entries2 = 0;
1271 unsigned leaves = 0;
1272 const struct gfs2_dirent **darr, *dent;
1273 struct dirent_gather g;
1274 struct buffer_head **larr;
1275 int leaf = 0;
1276 int error, i;
1277 u64 lfn = leaf_no;
1278
1279 do {
1280 error = get_leaf(ip, lfn, &bh);
1281 if (error)
1282 goto out;
1283 lf = (struct gfs2_leaf *)bh->b_data;
1284 if (leaves == 0)
1285 *depth = be16_to_cpu(lf->lf_depth);
1286 entries += be16_to_cpu(lf->lf_entries);
1287 leaves++;
1288 lfn = be64_to_cpu(lf->lf_next);
1289 brelse(bh);
1290 } while(lfn);
1291
1292 if (!entries)
1293 return 0;
1294
1295 error = -ENOMEM;
1296 /*
1297 * The extra 99 entries are not normally used, but are a buffer
1298 * zone in case the number of entries in the leaf is corrupt.
1299 * 99 is the maximum number of entries that can fit in a single
1300 * leaf block.
1301 */
1302 larr = gfs2_alloc_sort_buffer((leaves + entries + 99) * sizeof(void *));
1303 if (!larr)
1304 goto out;
1305 darr = (const struct gfs2_dirent **)(larr + leaves);
1306 g.pdent = darr;
1307 g.offset = 0;
1308 lfn = leaf_no;
1309
1310 do {
1311 error = get_leaf(ip, lfn, &bh);
1312 if (error)
1313 goto out_free;
1314 lf = (struct gfs2_leaf *)bh->b_data;
1315 lfn = be64_to_cpu(lf->lf_next);
1316 if (lf->lf_entries) {
1317 entries2 += be16_to_cpu(lf->lf_entries);
1318 dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
1319 gfs2_dirent_gather, NULL, &g);
1320 error = PTR_ERR(dent);
1321 if (IS_ERR(dent))
1322 goto out_free;
1323 if (entries2 != g.offset) {
1324 fs_warn(sdp, "Number of entries corrupt in dir "
1325 "leaf %llu, entries2 (%u) != "
1326 "g.offset (%u)\n",
1327 (unsigned long long)bh->b_blocknr,
1328 entries2, g.offset);
1329
1330 error = -EIO;
1331 goto out_free;
1332 }
1333 error = 0;
1334 larr[leaf++] = bh;
1335 } else {
1336 brelse(bh);
1337 }
1338 } while(lfn);
1339
1340 BUG_ON(entries2 != entries);
1341 error = do_filldir_main(ip, offset, opaque, filldir, darr,
1342 entries, copied);
1343 out_free:
1344 for(i = 0; i < leaf; i++)
1345 brelse(larr[i]);
1346 gfs2_free_sort_buffer(larr);
1347 out:
1348 return error;
1349 }
1350
1351 /**
1352 * dir_e_read - Reads the entries from a directory into a filldir buffer
1353 * @dip: dinode pointer
1354 * @offset: the hash of the last entry read shifted to the right once
1355 * @opaque: buffer for the filldir function to fill
1356 * @filldir: points to the filldir function to use
1357 *
1358 * Returns: errno
1359 */
1360
1361 static int dir_e_read(struct inode *inode, u64 *offset, void *opaque,
1362 filldir_t filldir)
1363 {
1364 struct gfs2_inode *dip = GFS2_I(inode);
1365 struct gfs2_sbd *sdp = GFS2_SB(inode);
1366 u32 hsize, len = 0;
1367 u32 ht_offset, lp_offset, ht_offset_cur = -1;
1368 u32 hash, index;
1369 __be64 *lp;
1370 int copied = 0;
1371 int error = 0;
1372 unsigned depth = 0;
1373
1374 hsize = 1 << dip->i_depth;
1375 if (hsize * sizeof(u64) != i_size_read(inode)) {
1376 gfs2_consist_inode(dip);
1377 return -EIO;
1378 }
1379
1380 hash = gfs2_dir_offset2hash(*offset);
1381 index = hash >> (32 - dip->i_depth);
1382
1383 lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS);
1384 if (!lp)
1385 return -ENOMEM;
1386
1387 while (index < hsize) {
1388 lp_offset = index & (sdp->sd_hash_ptrs - 1);
1389 ht_offset = index - lp_offset;
1390
1391 if (ht_offset_cur != ht_offset) {
1392 error = gfs2_dir_read_data(dip, (char *)lp,
1393 ht_offset * sizeof(__be64),
1394 sdp->sd_hash_bsize, 1);
1395 if (error != sdp->sd_hash_bsize) {
1396 if (error >= 0)
1397 error = -EIO;
1398 goto out;
1399 }
1400 ht_offset_cur = ht_offset;
1401 }
1402
1403 error = gfs2_dir_read_leaf(inode, offset, opaque, filldir,
1404 &copied, &depth,
1405 be64_to_cpu(lp[lp_offset]));
1406 if (error)
1407 break;
1408
1409 len = 1 << (dip->i_depth - depth);
1410 index = (index & ~(len - 1)) + len;
1411 }
1412
1413 out:
1414 kfree(lp);
1415 if (error > 0)
1416 error = 0;
1417 return error;
1418 }
1419
1420 int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque,
1421 filldir_t filldir)
1422 {
1423 struct gfs2_inode *dip = GFS2_I(inode);
1424 struct gfs2_sbd *sdp = GFS2_SB(inode);
1425 struct dirent_gather g;
1426 const struct gfs2_dirent **darr, *dent;
1427 struct buffer_head *dibh;
1428 int copied = 0;
1429 int error;
1430
1431 if (!dip->i_entries)
1432 return 0;
1433
1434 if (dip->i_diskflags & GFS2_DIF_EXHASH)
1435 return dir_e_read(inode, offset, opaque, filldir);
1436
1437 if (!gfs2_is_stuffed(dip)) {
1438 gfs2_consist_inode(dip);
1439 return -EIO;
1440 }
1441
1442 error = gfs2_meta_inode_buffer(dip, &dibh);
1443 if (error)
1444 return error;
1445
1446 error = -ENOMEM;
1447 /* 96 is max number of dirents which can be stuffed into an inode */
1448 darr = kmalloc(96 * sizeof(struct gfs2_dirent *), GFP_NOFS);
1449 if (darr) {
1450 g.pdent = darr;
1451 g.offset = 0;
1452 dent = gfs2_dirent_scan(inode, dibh->b_data, dibh->b_size,
1453 gfs2_dirent_gather, NULL, &g);
1454 if (IS_ERR(dent)) {
1455 error = PTR_ERR(dent);
1456 goto out;
1457 }
1458 if (dip->i_entries != g.offset) {
1459 fs_warn(sdp, "Number of entries corrupt in dir %llu, "
1460 "ip->i_entries (%u) != g.offset (%u)\n",
1461 (unsigned long long)dip->i_no_addr,
1462 dip->i_entries,
1463 g.offset);
1464 error = -EIO;
1465 goto out;
1466 }
1467 error = do_filldir_main(dip, offset, opaque, filldir, darr,
1468 dip->i_entries, &copied);
1469 out:
1470 kfree(darr);
1471 }
1472
1473 if (error > 0)
1474 error = 0;
1475
1476 brelse(dibh);
1477
1478 return error;
1479 }
1480
1481 /**
1482 * gfs2_dir_search - Search a directory
1483 * @dip: The GFS2 inode
1484 * @filename:
1485 * @inode:
1486 *
1487 * This routine searches a directory for a file or another directory.
1488 * Assumes a glock is held on dip.
1489 *
1490 * Returns: errno
1491 */
1492
1493 struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name)
1494 {
1495 struct buffer_head *bh;
1496 struct gfs2_dirent *dent;
1497 struct inode *inode;
1498
1499 dent = gfs2_dirent_search(dir, name, gfs2_dirent_find, &bh);
1500 if (dent) {
1501 if (IS_ERR(dent))
1502 return ERR_CAST(dent);
1503 inode = gfs2_inode_lookup(dir->i_sb,
1504 be16_to_cpu(dent->de_type),
1505 be64_to_cpu(dent->de_inum.no_addr),
1506 be64_to_cpu(dent->de_inum.no_formal_ino), 0);
1507 brelse(bh);
1508 return inode;
1509 }
1510 return ERR_PTR(-ENOENT);
1511 }
1512
1513 int gfs2_dir_check(struct inode *dir, const struct qstr *name,
1514 const struct gfs2_inode *ip)
1515 {
1516 struct buffer_head *bh;
1517 struct gfs2_dirent *dent;
1518 int ret = -ENOENT;
1519
1520 dent = gfs2_dirent_search(dir, name, gfs2_dirent_find, &bh);
1521 if (dent) {
1522 if (IS_ERR(dent))
1523 return PTR_ERR(dent);
1524 if (ip) {
1525 if (be64_to_cpu(dent->de_inum.no_addr) != ip->i_no_addr)
1526 goto out;
1527 if (be64_to_cpu(dent->de_inum.no_formal_ino) !=
1528 ip->i_no_formal_ino)
1529 goto out;
1530 if (unlikely(IF2DT(ip->i_inode.i_mode) !=
1531 be16_to_cpu(dent->de_type))) {
1532 gfs2_consist_inode(GFS2_I(dir));
1533 ret = -EIO;
1534 goto out;
1535 }
1536 }
1537 ret = 0;
1538 out:
1539 brelse(bh);
1540 }
1541 return ret;
1542 }
1543
1544 static int dir_new_leaf(struct inode *inode, const struct qstr *name)
1545 {
1546 struct buffer_head *bh, *obh;
1547 struct gfs2_inode *ip = GFS2_I(inode);
1548 struct gfs2_leaf *leaf, *oleaf;
1549 int error;
1550 u32 index;
1551 u64 bn;
1552
1553 index = name->hash >> (32 - ip->i_depth);
1554 error = get_first_leaf(ip, index, &obh);
1555 if (error)
1556 return error;
1557 do {
1558 oleaf = (struct gfs2_leaf *)obh->b_data;
1559 bn = be64_to_cpu(oleaf->lf_next);
1560 if (!bn)
1561 break;
1562 brelse(obh);
1563 error = get_leaf(ip, bn, &obh);
1564 if (error)
1565 return error;
1566 } while(1);
1567
1568 gfs2_trans_add_bh(ip->i_gl, obh, 1);
1569
1570 leaf = new_leaf(inode, &bh, be16_to_cpu(oleaf->lf_depth));
1571 if (!leaf) {
1572 brelse(obh);
1573 return -ENOSPC;
1574 }
1575 oleaf->lf_next = cpu_to_be64(bh->b_blocknr);
1576 brelse(bh);
1577 brelse(obh);
1578
1579 error = gfs2_meta_inode_buffer(ip, &bh);
1580 if (error)
1581 return error;
1582 gfs2_trans_add_bh(ip->i_gl, bh, 1);
1583 gfs2_add_inode_blocks(&ip->i_inode, 1);
1584 gfs2_dinode_out(ip, bh->b_data);
1585 brelse(bh);
1586 return 0;
1587 }
1588
1589 /**
1590 * gfs2_dir_add - Add new filename into directory
1591 * @dip: The GFS2 inode
1592 * @filename: The new name
1593 * @inode: The inode number of the entry
1594 * @type: The type of the entry
1595 *
1596 * Returns: 0 on success, error code on failure
1597 */
1598
1599 int gfs2_dir_add(struct inode *inode, const struct qstr *name,
1600 const struct gfs2_inode *nip)
1601 {
1602 struct gfs2_inode *ip = GFS2_I(inode);
1603 struct buffer_head *bh;
1604 struct gfs2_dirent *dent;
1605 struct gfs2_leaf *leaf;
1606 int error;
1607
1608 while(1) {
1609 dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space,
1610 &bh);
1611 if (dent) {
1612 if (IS_ERR(dent))
1613 return PTR_ERR(dent);
1614 dent = gfs2_init_dirent(inode, dent, name, bh);
1615 gfs2_inum_out(nip, dent);
1616 dent->de_type = cpu_to_be16(IF2DT(nip->i_inode.i_mode));
1617 if (ip->i_diskflags & GFS2_DIF_EXHASH) {
1618 leaf = (struct gfs2_leaf *)bh->b_data;
1619 be16_add_cpu(&leaf->lf_entries, 1);
1620 }
1621 brelse(bh);
1622 error = gfs2_meta_inode_buffer(ip, &bh);
1623 if (error)
1624 break;
1625 gfs2_trans_add_bh(ip->i_gl, bh, 1);
1626 ip->i_entries++;
1627 ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME;
1628 if (S_ISDIR(nip->i_inode.i_mode))
1629 inc_nlink(&ip->i_inode);
1630 gfs2_dinode_out(ip, bh->b_data);
1631 brelse(bh);
1632 error = 0;
1633 break;
1634 }
1635 if (!(ip->i_diskflags & GFS2_DIF_EXHASH)) {
1636 error = dir_make_exhash(inode);
1637 if (error)
1638 break;
1639 continue;
1640 }
1641 error = dir_split_leaf(inode, name);
1642 if (error == 0)
1643 continue;
1644 if (error < 0)
1645 break;
1646 if (ip->i_depth < GFS2_DIR_MAX_DEPTH) {
1647 error = dir_double_exhash(ip);
1648 if (error)
1649 break;
1650 error = dir_split_leaf(inode, name);
1651 if (error < 0)
1652 break;
1653 if (error == 0)
1654 continue;
1655 }
1656 error = dir_new_leaf(inode, name);
1657 if (!error)
1658 continue;
1659 error = -ENOSPC;
1660 break;
1661 }
1662 return error;
1663 }
1664
1665
1666 /**
1667 * gfs2_dir_del - Delete a directory entry
1668 * @dip: The GFS2 inode
1669 * @filename: The filename
1670 *
1671 * Returns: 0 on success, error code on failure
1672 */
1673
1674 int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry)
1675 {
1676 const struct qstr *name = &dentry->d_name;
1677 struct gfs2_dirent *dent, *prev = NULL;
1678 struct buffer_head *bh;
1679 int error;
1680
1681 /* Returns _either_ the entry (if its first in block) or the
1682 previous entry otherwise */
1683 dent = gfs2_dirent_search(&dip->i_inode, name, gfs2_dirent_prev, &bh);
1684 if (!dent) {
1685 gfs2_consist_inode(dip);
1686 return -EIO;
1687 }
1688 if (IS_ERR(dent)) {
1689 gfs2_consist_inode(dip);
1690 return PTR_ERR(dent);
1691 }
1692 /* If not first in block, adjust pointers accordingly */
1693 if (gfs2_dirent_find(dent, name, NULL) == 0) {
1694 prev = dent;
1695 dent = (struct gfs2_dirent *)((char *)dent + be16_to_cpu(prev->de_rec_len));
1696 }
1697
1698 dirent_del(dip, bh, prev, dent);
1699 if (dip->i_diskflags & GFS2_DIF_EXHASH) {
1700 struct gfs2_leaf *leaf = (struct gfs2_leaf *)bh->b_data;
1701 u16 entries = be16_to_cpu(leaf->lf_entries);
1702 if (!entries)
1703 gfs2_consist_inode(dip);
1704 leaf->lf_entries = cpu_to_be16(--entries);
1705 }
1706 brelse(bh);
1707
1708 error = gfs2_meta_inode_buffer(dip, &bh);
1709 if (error)
1710 return error;
1711
1712 if (!dip->i_entries)
1713 gfs2_consist_inode(dip);
1714 gfs2_trans_add_bh(dip->i_gl, bh, 1);
1715 dip->i_entries--;
1716 dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME;
1717 if (S_ISDIR(dentry->d_inode->i_mode))
1718 drop_nlink(&dip->i_inode);
1719 gfs2_dinode_out(dip, bh->b_data);
1720 brelse(bh);
1721 mark_inode_dirty(&dip->i_inode);
1722
1723 return error;
1724 }
1725
1726 /**
1727 * gfs2_dir_mvino - Change inode number of directory entry
1728 * @dip: The GFS2 inode
1729 * @filename:
1730 * @new_inode:
1731 *
1732 * This routine changes the inode number of a directory entry. It's used
1733 * by rename to change ".." when a directory is moved.
1734 * Assumes a glock is held on dvp.
1735 *
1736 * Returns: errno
1737 */
1738
1739 int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename,
1740 const struct gfs2_inode *nip, unsigned int new_type)
1741 {
1742 struct buffer_head *bh;
1743 struct gfs2_dirent *dent;
1744 int error;
1745
1746 dent = gfs2_dirent_search(&dip->i_inode, filename, gfs2_dirent_find, &bh);
1747 if (!dent) {
1748 gfs2_consist_inode(dip);
1749 return -EIO;
1750 }
1751 if (IS_ERR(dent))
1752 return PTR_ERR(dent);
1753
1754 gfs2_trans_add_bh(dip->i_gl, bh, 1);
1755 gfs2_inum_out(nip, dent);
1756 dent->de_type = cpu_to_be16(new_type);
1757
1758 if (dip->i_diskflags & GFS2_DIF_EXHASH) {
1759 brelse(bh);
1760 error = gfs2_meta_inode_buffer(dip, &bh);
1761 if (error)
1762 return error;
1763 gfs2_trans_add_bh(dip->i_gl, bh, 1);
1764 }
1765
1766 dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME;
1767 gfs2_dinode_out(dip, bh->b_data);
1768 brelse(bh);
1769 return 0;
1770 }
1771
1772 /**
1773 * leaf_dealloc - Deallocate a directory leaf
1774 * @dip: the directory
1775 * @index: the hash table offset in the directory
1776 * @len: the number of pointers to this leaf
1777 * @leaf_no: the leaf number
1778 * @leaf_bh: buffer_head for the starting leaf
1779 * last_dealloc: 1 if this is the final dealloc for the leaf, else 0
1780 *
1781 * Returns: errno
1782 */
1783
1784 static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len,
1785 u64 leaf_no, struct buffer_head *leaf_bh,
1786 int last_dealloc)
1787 {
1788 struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1789 struct gfs2_leaf *tmp_leaf;
1790 struct gfs2_rgrp_list rlist;
1791 struct buffer_head *bh, *dibh;
1792 u64 blk, nblk;
1793 unsigned int rg_blocks = 0, l_blocks = 0;
1794 char *ht;
1795 unsigned int x, size = len * sizeof(u64);
1796 int error;
1797
1798 memset(&rlist, 0, sizeof(struct gfs2_rgrp_list));
1799
1800 ht = kzalloc(size, GFP_NOFS);
1801 if (!ht)
1802 return -ENOMEM;
1803
1804 if (!gfs2_alloc_get(dip)) {
1805 error = -ENOMEM;
1806 goto out;
1807 }
1808
1809 error = gfs2_quota_hold(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE);
1810 if (error)
1811 goto out_put;
1812
1813 error = gfs2_rindex_hold(sdp, &dip->i_alloc->al_ri_gh);
1814 if (error)
1815 goto out_qs;
1816
1817 /* Count the number of leaves */
1818 bh = leaf_bh;
1819
1820 for (blk = leaf_no; blk; blk = nblk) {
1821 if (blk != leaf_no) {
1822 error = get_leaf(dip, blk, &bh);
1823 if (error)
1824 goto out_rlist;
1825 }
1826 tmp_leaf = (struct gfs2_leaf *)bh->b_data;
1827 nblk = be64_to_cpu(tmp_leaf->lf_next);
1828 if (blk != leaf_no)
1829 brelse(bh);
1830
1831 gfs2_rlist_add(sdp, &rlist, blk);
1832 l_blocks++;
1833 }
1834
1835 gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE);
1836
1837 for (x = 0; x < rlist.rl_rgrps; x++) {
1838 struct gfs2_rgrpd *rgd;
1839 rgd = rlist.rl_ghs[x].gh_gl->gl_object;
1840 rg_blocks += rgd->rd_length;
1841 }
1842
1843 error = gfs2_glock_nq_m(rlist.rl_rgrps, rlist.rl_ghs);
1844 if (error)
1845 goto out_rlist;
1846
1847 error = gfs2_trans_begin(sdp,
1848 rg_blocks + (DIV_ROUND_UP(size, sdp->sd_jbsize) + 1) +
1849 RES_DINODE + RES_STATFS + RES_QUOTA, l_blocks);
1850 if (error)
1851 goto out_rg_gunlock;
1852
1853 bh = leaf_bh;
1854
1855 for (blk = leaf_no; blk; blk = nblk) {
1856 if (blk != leaf_no) {
1857 error = get_leaf(dip, blk, &bh);
1858 if (error)
1859 goto out_end_trans;
1860 }
1861 tmp_leaf = (struct gfs2_leaf *)bh->b_data;
1862 nblk = be64_to_cpu(tmp_leaf->lf_next);
1863 if (blk != leaf_no)
1864 brelse(bh);
1865
1866 gfs2_free_meta(dip, blk, 1);
1867 gfs2_add_inode_blocks(&dip->i_inode, -1);
1868 }
1869
1870 error = gfs2_dir_write_data(dip, ht, index * sizeof(u64), size);
1871 if (error != size) {
1872 if (error >= 0)
1873 error = -EIO;
1874 goto out_end_trans;
1875 }
1876
1877 error = gfs2_meta_inode_buffer(dip, &dibh);
1878 if (error)
1879 goto out_end_trans;
1880
1881 gfs2_trans_add_bh(dip->i_gl, dibh, 1);
1882 /* On the last dealloc, make this a regular file in case we crash.
1883 (We don't want to free these blocks a second time.) */
1884 if (last_dealloc)
1885 dip->i_inode.i_mode = S_IFREG;
1886 gfs2_dinode_out(dip, dibh->b_data);
1887 brelse(dibh);
1888
1889 out_end_trans:
1890 gfs2_trans_end(sdp);
1891 out_rg_gunlock:
1892 gfs2_glock_dq_m(rlist.rl_rgrps, rlist.rl_ghs);
1893 out_rlist:
1894 gfs2_rlist_free(&rlist);
1895 gfs2_glock_dq_uninit(&dip->i_alloc->al_ri_gh);
1896 out_qs:
1897 gfs2_quota_unhold(dip);
1898 out_put:
1899 gfs2_alloc_put(dip);
1900 out:
1901 kfree(ht);
1902 return error;
1903 }
1904
1905 /**
1906 * gfs2_dir_exhash_dealloc - free all the leaf blocks in a directory
1907 * @dip: the directory
1908 *
1909 * Dealloc all on-disk directory leaves to FREEMETA state
1910 * Change on-disk inode type to "regular file"
1911 *
1912 * Returns: errno
1913 */
1914
1915 int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip)
1916 {
1917 struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1918 struct buffer_head *bh;
1919 struct gfs2_leaf *leaf;
1920 u32 hsize, len;
1921 u32 ht_offset, lp_offset, ht_offset_cur = -1;
1922 u32 index = 0, next_index;
1923 __be64 *lp;
1924 u64 leaf_no;
1925 int error = 0, last;
1926
1927 hsize = 1 << dip->i_depth;
1928 if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) {
1929 gfs2_consist_inode(dip);
1930 return -EIO;
1931 }
1932
1933 lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS);
1934 if (!lp)
1935 return -ENOMEM;
1936
1937 while (index < hsize) {
1938 lp_offset = index & (sdp->sd_hash_ptrs - 1);
1939 ht_offset = index - lp_offset;
1940
1941 if (ht_offset_cur != ht_offset) {
1942 error = gfs2_dir_read_data(dip, (char *)lp,
1943 ht_offset * sizeof(__be64),
1944 sdp->sd_hash_bsize, 1);
1945 if (error != sdp->sd_hash_bsize) {
1946 if (error >= 0)
1947 error = -EIO;
1948 goto out;
1949 }
1950 ht_offset_cur = ht_offset;
1951 }
1952
1953 leaf_no = be64_to_cpu(lp[lp_offset]);
1954 if (leaf_no) {
1955 error = get_leaf(dip, leaf_no, &bh);
1956 if (error)
1957 goto out;
1958 leaf = (struct gfs2_leaf *)bh->b_data;
1959 len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth));
1960
1961 next_index = (index & ~(len - 1)) + len;
1962 last = ((next_index >= hsize) ? 1 : 0);
1963 error = leaf_dealloc(dip, index, len, leaf_no, bh,
1964 last);
1965 brelse(bh);
1966 if (error)
1967 goto out;
1968 index = next_index;
1969 } else
1970 index++;
1971 }
1972
1973 if (index != hsize) {
1974 gfs2_consist_inode(dip);
1975 error = -EIO;
1976 }
1977
1978 out:
1979 kfree(lp);
1980
1981 return error;
1982 }
1983
1984 /**
1985 * gfs2_diradd_alloc_required - find if adding entry will require an allocation
1986 * @ip: the file being written to
1987 * @filname: the filename that's going to be added
1988 *
1989 * Returns: 1 if alloc required, 0 if not, -ve on error
1990 */
1991
1992 int gfs2_diradd_alloc_required(struct inode *inode, const struct qstr *name)
1993 {
1994 struct gfs2_dirent *dent;
1995 struct buffer_head *bh;
1996
1997 dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space, &bh);
1998 if (!dent) {
1999 return 1;
2000 }
2001 if (IS_ERR(dent))
2002 return PTR_ERR(dent);
2003 brelse(bh);
2004 return 0;
2005 }
2006
linux-next