fs/gfs2/dir.c

   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_di.di_flags & 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/sched.h>
  57 #include <linux/slab.h>
  58 #include <linux/spinlock.h>
  59 #include <linux/buffer_head.h>
  60 #include <linux/sort.h>
  61 #include <linux/gfs2_ondisk.h>
  62 #include <linux/crc32.h>
  63 #include <linux/vmalloc.h>
  64 #include <linux/lm_interface.h>
  65
  66 #include "gfs2.h"
  67 #include "incore.h"
  68 #include "dir.h"
  69 #include "glock.h"
  70 #include "inode.h"
  71 #include "meta_io.h"
  72 #include "quota.h"
  73 #include "rgrp.h"
  74 #include "trans.h"
  75 #include "bmap.h"
  76 #include "util.h"
  77
  78 #define IS_LEAF     1 /* Hashed (leaf) directory */
  79 #define IS_DINODE   2 /* Linear (stuffed dinode block) directory */
  80
  81 #define gfs2_disk_hash2offset(h) (((u64)(h)) >> 1)
  82 #define gfs2_dir_offset2hash(p) ((u32)(((u64)(p)) << 1))
  83
  84 typedef int (*leaf_call_t) (struct gfs2_inode *dip, u32 index, u32 len,
  85                             u64 leaf_no, void *data);
  86 typedef int (*gfs2_dscan_t)(const struct gfs2_dirent *dent,
  87                             const struct qstr *name, void *opaque);
  88
  89
  90 int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block,
  91                             struct buffer_head **bhp)
  92 {
  93         struct buffer_head *bh;
  94
  95         bh = gfs2_meta_new(ip->i_gl, block);
  96         gfs2_trans_add_bh(ip->i_gl, bh, 1);
  97         gfs2_metatype_set(bh, GFS2_METATYPE_JD, GFS2_FORMAT_JD);
  98         gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header));
  99         *bhp = bh;
 100         return 0;
 101 }
 102
 103 static int gfs2_dir_get_existing_buffer(struct gfs2_inode *ip, u64 block,
 104                                         struct buffer_head **bhp)
 105 {
 106         struct buffer_head *bh;
 107         int error;
 108
 109         error = gfs2_meta_read(ip->i_gl, block, DIO_WAIT, &bh);
 110         if (error)
 111                 return error;
 112         if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), bh, GFS2_METATYPE_JD)) {
 113                 brelse(bh);
 114                 return -EIO;
 115         }
 116         *bhp = bh;
 117         return 0;
 118 }
 119
 120 static int gfs2_dir_write_stuffed(struct gfs2_inode *ip, const char *buf,
 121                                   unsigned int offset, unsigned int size)
 122 {
 123         struct buffer_head *dibh;
 124         int error;
 125
 126         error = gfs2_meta_inode_buffer(ip, &dibh);
 127         if (error)
 128                 return error;
 129
 130         gfs2_trans_add_bh(ip->i_gl, dibh, 1);
 131         memcpy(dibh->b_data + offset + sizeof(struct gfs2_dinode), buf, size);
 132         if (ip->i_di.di_size < offset + size)
 133                 ip->i_di.di_size = offset + size;
 134         ip->i_inode.i_mtime.tv_sec = ip->i_inode.i_ctime.tv_sec = get_seconds();
 135         gfs2_dinode_out(ip, dibh->b_data);
 136
 137         brelse(dibh);
 138
 139         return size;
 140 }
 141
 142
 143
 144 /**
 145  * gfs2_dir_write_data - Write directory information to the inode
 146  * @ip: The GFS2 inode
 147  * @buf: The buffer containing information to be written
 148  * @offset: The file offset to start writing at
 149  * @size: The amount of data to write
 150  *
 151  * Returns: The number of bytes correctly written or error code
 152  */
 153 static int gfs2_dir_write_data(struct gfs2_inode *ip, const char *buf,
 154                                u64 offset, unsigned int size)
 155 {
 156         struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 157         struct buffer_head *dibh;
 158         u64 lblock, dblock;
 159         u32 extlen = 0;
 160         unsigned int o;
 161         int copied = 0;
 162         int error = 0;
 163
 164         if (!size)
 165                 return 0;
 166
 167         if (gfs2_is_stuffed(ip) &&
 168             offset + size <= sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode))
 169                 return gfs2_dir_write_stuffed(ip, buf, (unsigned int)offset,
 170                                               size);
 171
 172         if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip)))
 173                 return -EINVAL;
 174
 175         if (gfs2_is_stuffed(ip)) {
 176                 error = gfs2_unstuff_dinode(ip, NULL);
 177                 if (error)
 178                         return error;
 179         }
 180
 181         lblock = offset;
 182         o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header);
 183
 184         while (copied < size) {
 185                 unsigned int amount;
 186                 struct buffer_head *bh;
 187                 int new = 0;
 188
 189                 amount = size - copied;
 190                 if (amount > sdp->sd_sb.sb_bsize - o)
 191                         amount = sdp->sd_sb.sb_bsize - o;
 192
 193                 if (!extlen) {
 194                         new = 1;
 195                         error = gfs2_extent_map(&ip->i_inode, lblock, &new,
 196                                                 &dblock, &extlen);
 197                         if (error)
 198                                 goto fail;
 199                         error = -EIO;
 200                         if (gfs2_assert_withdraw(sdp, dblock))
 201                                 goto fail;
 202                 }
 203
 204                 if (amount == sdp->sd_jbsize || new)
 205                         error = gfs2_dir_get_new_buffer(ip, dblock, &bh);
 206                 else
 207                         error = gfs2_dir_get_existing_buffer(ip, dblock, &bh);
 208
 209                 if (error)
 210                         goto fail;
 211
 212                 gfs2_trans_add_bh(ip->i_gl, bh, 1);
 213                 memcpy(bh->b_data + o, buf, amount);
 214                 brelse(bh);
 215
 216                 buf += amount;
 217                 copied += amount;
 218                 lblock++;
 219                 dblock++;
 220                 extlen--;
 221
 222                 o = sizeof(struct gfs2_meta_header);
 223         }
 224
 225 out:
 226         error = gfs2_meta_inode_buffer(ip, &dibh);
 227         if (error)
 228                 return error;
 229
 230         if (ip->i_di.di_size < offset + copied)
 231                 ip->i_di.di_size = offset + copied;
 232         ip->i_inode.i_mtime.tv_sec = ip->i_inode.i_ctime.tv_sec = get_seconds();
 233
 234         gfs2_trans_add_bh(ip->i_gl, dibh, 1);
 235         gfs2_dinode_out(ip, dibh->b_data);
 236         brelse(dibh);
 237
 238         return copied;
 239 fail:
 240         if (copied)
 241                 goto out;
 242         return error;
 243 }
 244
 245 static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf,
 246                                  u64 offset, unsigned int size)
 247 {
 248         struct buffer_head *dibh;
 249         int error;
 250
 251         error = gfs2_meta_inode_buffer(ip, &dibh);
 252         if (!error) {
 253                 offset += sizeof(struct gfs2_dinode);
 254                 memcpy(buf, dibh->b_data + offset, size);
 255                 brelse(dibh);
 256         }
 257
 258         return (error) ? error : size;
 259 }
 260
 261
 262 /**
 263  * gfs2_dir_read_data - Read a data from a directory inode
 264  * @ip: The GFS2 Inode
 265  * @buf: The buffer to place result into
 266  * @offset: File offset to begin jdata_readng from
 267  * @size: Amount of data to transfer
 268  *
 269  * Returns: The amount of data actually copied or the error
 270  */
 271 static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
 272                               unsigned int size, unsigned ra)
 273 {
 274         struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 275         u64 lblock, dblock;
 276         u32 extlen = 0;
 277         unsigned int o;
 278         int copied = 0;
 279         int error = 0;
 280
 281         if (offset >= ip->i_di.di_size)
 282                 return 0;
 283
 284         if (offset + size > ip->i_di.di_size)
 285                 size = ip->i_di.di_size - offset;
 286
 287         if (!size)
 288                 return 0;
 289
 290         if (gfs2_is_stuffed(ip))
 291                 return gfs2_dir_read_stuffed(ip, buf, offset, size);
 292
 293         if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip)))
 294                 return -EINVAL;
 295
 296         lblock = offset;
 297         o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header);
 298
 299         while (copied < size) {
 300                 unsigned int amount;
 301                 struct buffer_head *bh;
 302                 int new;
 303
 304                 amount = size - copied;
 305                 if (amount > sdp->sd_sb.sb_bsize - o)
 306                         amount = sdp->sd_sb.sb_bsize - o;
 307
 308                 if (!extlen) {
 309                         new = 0;
 310                         error = gfs2_extent_map(&ip->i_inode, lblock, &new,
 311                                                 &dblock, &extlen);
 312                         if (error || !dblock)
 313                                 goto fail;
 314                         BUG_ON(extlen < 1);
 315                         if (!ra)
 316                                 extlen = 1;
 317                         bh = gfs2_meta_ra(ip->i_gl, dblock, extlen);
 318                 } else {
 319                         error = gfs2_meta_read(ip->i_gl, dblock, DIO_WAIT, &bh);
 320                         if (error)
 321                                 goto fail;
 322                 }
 323                 error = gfs2_metatype_check(sdp, bh, GFS2_METATYPE_JD);
 324                 if (error) {
 325                         brelse(bh);
 326                         goto fail;
 327                 }
 328                 dblock++;
 329                 extlen--;
 330                 memcpy(buf, bh->b_data + o, amount);
 331                 brelse(bh);
 332                 buf += amount;
 333                 copied += amount;
 334                 lblock++;
 335                 o = sizeof(struct gfs2_meta_header);
 336         }
 337
 338         return copied;
 339 fail:
 340         return (copied) ? copied : error;
 341 }
 342
 343 static inline int gfs2_dirent_sentinel(const struct gfs2_dirent *dent)
 344 {
 345         return dent->de_inum.no_addr == 0 || dent->de_inum.no_formal_ino == 0;
 346 }
 347
 348 static inline int __gfs2_dirent_find(const struct gfs2_dirent *dent,
 349                                      const struct qstr *name, int ret)
 350 {
 351         if (!gfs2_dirent_sentinel(dent) &&
 352             be32_to_cpu(dent->de_hash) == name->hash &&
 353             be16_to_cpu(dent->de_name_len) == name->len &&
 354             memcmp(dent+1, name->name, name->len) == 0)
 355                 return ret;
 356         return 0;
 357 }
 358
 359 static int gfs2_dirent_find(const struct gfs2_dirent *dent,
 360                             const struct qstr *name,
 361                             void *opaque)
 362 {
 363         return __gfs2_dirent_find(dent, name, 1);
 364 }
 365
 366 static int gfs2_dirent_prev(const struct gfs2_dirent *dent,
 367                             const struct qstr *name,
 368                             void *opaque)
 369 {
 370         return __gfs2_dirent_find(dent, name, 2);
 371 }
 372
 373 /*
 374  * name->name holds ptr to start of block.
 375  * name->len holds size of block.
 376  */
 377 static int gfs2_dirent_last(const struct gfs2_dirent *dent,
 378                             const struct qstr *name,
 379                             void *opaque)
 380 {
 381         const char *start = name->name;
 382         const char *end = (const char *)dent + be16_to_cpu(dent->de_rec_len);
 383         if (name->len == (end - start))
 384                 return 1;
 385         return 0;
 386 }
 387
 388 static int gfs2_dirent_find_space(const struct gfs2_dirent *dent,
 389                                   const struct qstr *name,
 390                                   void *opaque)
 391 {
 392         unsigned required = GFS2_DIRENT_SIZE(name->len);
 393         unsigned actual = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len));
 394         unsigned totlen = be16_to_cpu(dent->de_rec_len);
 395
 396         if (gfs2_dirent_sentinel(dent))
 397                 actual = GFS2_DIRENT_SIZE(0);
 398         if (totlen - actual >= required)
 399                 return 1;
 400         return 0;
 401 }
 402
 403 struct dirent_gather {
 404         const struct gfs2_dirent **pdent;
 405         unsigned offset;
 406 };
 407
 408 static int gfs2_dirent_gather(const struct gfs2_dirent *dent,
 409                               const struct qstr *name,
 410                               void *opaque)
 411 {
 412         struct dirent_gather *g = opaque;
 413         if (!gfs2_dirent_sentinel(dent)) {
 414                 g->pdent[g->offset++] = dent;
 415         }
 416         return 0;
 417 }
 418
 419 /*
 420  * Other possible things to check:
 421  * - Inode located within filesystem size (and on valid block)
 422  * - Valid directory entry type
 423  * Not sure how heavy-weight we want to make this... could also check
 424  * hash is correct for example, but that would take a lot of extra time.
 425  * For now the most important thing is to check that the various sizes
 426  * are correct.
 427  */
 428 static int gfs2_check_dirent(struct gfs2_dirent *dent, unsigned int offset,
 429                              unsigned int size, unsigned int len, int first)
 430 {
 431         const char *msg = "gfs2_dirent too small";
 432         if (unlikely(size < sizeof(struct gfs2_dirent)))
 433                 goto error;
 434         msg = "gfs2_dirent misaligned";
 435         if (unlikely(offset & 0x7))
 436                 goto error;
 437         msg = "gfs2_dirent points beyond end of block";
 438         if (unlikely(offset + size > len))
 439                 goto error;
 440         msg = "zero inode number";
 441         if (unlikely(!first && gfs2_dirent_sentinel(dent)))
 442                 goto error;
 443         msg = "name length is greater than space in dirent";
 444         if (!gfs2_dirent_sentinel(dent) &&
 445             unlikely(sizeof(struct gfs2_dirent)+be16_to_cpu(dent->de_name_len) >
 446                      size))
 447                 goto error;
 448         return 0;
 449 error:
 450         printk(KERN_WARNING "gfs2_check_dirent: %s (%s)\n", msg,
 451                first ? "first in block" : "not first in block");
 452         return -EIO;
 453 }
 454
 455 static int gfs2_dirent_offset(const void *buf)
 456 {
 457         const struct gfs2_meta_header *h = buf;
 458         int offset;
 459
 460         BUG_ON(buf == NULL);
 461
 462         switch(be32_to_cpu(h->mh_type)) {
 463         case GFS2_METATYPE_LF:
 464                 offset = sizeof(struct gfs2_leaf);
 465                 break;
 466         case GFS2_METATYPE_DI:
 467                 offset = sizeof(struct gfs2_dinode);
 468                 break;
 469         default:
 470                 goto wrong_type;
 471         }
 472         return offset;
 473 wrong_type:
 474         printk(KERN_WARNING "gfs2_scan_dirent: wrong block type %u\n",
 475                be32_to_cpu(h->mh_type));
 476         return -1;
 477 }
 478
 479 static struct gfs2_dirent *gfs2_dirent_scan(struct inode *inode, void *buf,
 480                                             unsigned int len, gfs2_dscan_t scan,
 481                                             const struct qstr *name,
 482                                             void *opaque)
 483 {
 484         struct gfs2_dirent *dent, *prev;
 485         unsigned offset;
 486         unsigned size;
 487         int ret = 0;
 488
 489         ret = gfs2_dirent_offset(buf);
 490         if (ret < 0)
 491                 goto consist_inode;
 492
 493         offset = ret;
 494         prev = NULL;
 495         dent = buf + offset;
 496         size = be16_to_cpu(dent->de_rec_len);
 497         if (gfs2_check_dirent(dent, offset, size, len, 1))
 498                 goto consist_inode;
 499         do {
 500                 ret = scan(dent, name, opaque);
 501                 if (ret)
 502                         break;
 503                 offset += size;
 504                 if (offset == len)
 505                         break;
 506                 prev = dent;
 507                 dent = buf + offset;
 508                 size = be16_to_cpu(dent->de_rec_len);
 509                 if (gfs2_check_dirent(dent, offset, size, len, 0))
 510                         goto consist_inode;
 511         } while(1);
 512
 513         switch(ret) {
 514         case 0:
 515                 return NULL;
 516         case 1:
 517                 return dent;
 518         case 2:
 519                 return prev ? prev : dent;
 520         default:
 521                 BUG_ON(ret > 0);
 522                 return ERR_PTR(ret);
 523         }
 524
 525 consist_inode:
 526         gfs2_consist_inode(GFS2_I(inode));
 527         return ERR_PTR(-EIO);
 528 }
 529
 530
 531 /**
 532  * dirent_first - Return the first dirent
 533  * @dip: the directory
 534  * @bh: The buffer
 535  * @dent: Pointer to list of dirents
 536  *
 537  * return first dirent whether bh points to leaf or stuffed dinode
 538  *
 539  * Returns: IS_LEAF, IS_DINODE, or -errno
 540  */
 541
 542 static int dirent_first(struct gfs2_inode *dip, struct buffer_head *bh,
 543                         struct gfs2_dirent **dent)
 544 {
 545         struct gfs2_meta_header *h = (struct gfs2_meta_header *)bh->b_data;
 546
 547         if (be32_to_cpu(h->mh_type) == GFS2_METATYPE_LF) {
 548                 if (gfs2_meta_check(GFS2_SB(&dip->i_inode), bh))
 549                         return -EIO;
 550                 *dent = (struct gfs2_dirent *)(bh->b_data +
 551                                                sizeof(struct gfs2_leaf));
 552                 return IS_LEAF;
 553         } else {
 554                 if (gfs2_metatype_check(GFS2_SB(&dip->i_inode), bh, GFS2_METATYPE_DI))
 555                         return -EIO;
 556                 *dent = (struct gfs2_dirent *)(bh->b_data +
 557                                                sizeof(struct gfs2_dinode));
 558                 return IS_DINODE;
 559         }
 560 }
 561
 562 static int dirent_check_reclen(struct gfs2_inode *dip,
 563                                const struct gfs2_dirent *d, const void *end_p)
 564 {
 565         const void *ptr = d;
 566         u16 rec_len = be16_to_cpu(d->de_rec_len);
 567
 568         if (unlikely(rec_len < sizeof(struct gfs2_dirent)))
 569                 goto broken;
 570         ptr += rec_len;
 571         if (ptr < end_p)
 572                 return rec_len;
 573         if (ptr == end_p)
 574                 return -ENOENT;
 575 broken:
 576         gfs2_consist_inode(dip);
 577         return -EIO;
 578 }
 579
 580 /**
 581  * dirent_next - Next dirent
 582  * @dip: the directory
 583  * @bh: The buffer
 584  * @dent: Pointer to list of dirents
 585  *
 586  * Returns: 0 on success, error code otherwise
 587  */
 588
 589 static int dirent_next(struct gfs2_inode *dip, struct buffer_head *bh,
 590                        struct gfs2_dirent **dent)
 591 {
 592         struct gfs2_dirent *cur = *dent, *tmp;
 593         char *bh_end = bh->b_data + bh->b_size;
 594         int ret;
 595
 596         ret = dirent_check_reclen(dip, cur, bh_end);
 597         if (ret < 0)
 598                 return ret;
 599
 600         tmp = (void *)cur + ret;
 601         ret = dirent_check_reclen(dip, tmp, bh_end);
 602         if (ret == -EIO)
 603                 return ret;
 604
 605         /* Only the first dent could ever have de_inum.no_addr == 0 */
 606         if (gfs2_dirent_sentinel(tmp)) {
 607                 gfs2_consist_inode(dip);
 608                 return -EIO;
 609         }
 610
 611         *dent = tmp;
 612         return 0;
 613 }
 614
 615 /**
 616  * dirent_del - Delete a dirent
 617  * @dip: The GFS2 inode
 618  * @bh: The buffer
 619  * @prev: The previous dirent
 620  * @cur: The current dirent
 621  *
 622  */
 623
 624 static void dirent_del(struct gfs2_inode *dip, struct buffer_head *bh,
 625                        struct gfs2_dirent *prev, struct gfs2_dirent *cur)
 626 {
 627         u16 cur_rec_len, prev_rec_len;
 628
 629         if (gfs2_dirent_sentinel(cur)) {
 630                 gfs2_consist_inode(dip);
 631                 return;
 632         }
 633
 634         gfs2_trans_add_bh(dip->i_gl, bh, 1);
 635
 636         /* If there is no prev entry, this is the first entry in the block.
 637            The de_rec_len is already as big as it needs to be.  Just zero
 638            out the inode number and return.  */
 639
 640         if (!prev) {
 641                 cur->de_inum.no_addr = 0;
 642                 cur->de_inum.no_formal_ino = 0;
 643                 return;
 644         }
 645
 646         /*  Combine this dentry with the previous one.  */
 647
 648         prev_rec_len = be16_to_cpu(prev->de_rec_len);
 649         cur_rec_len = be16_to_cpu(cur->de_rec_len);
 650
 651         if ((char *)prev + prev_rec_len != (char *)cur)
 652                 gfs2_consist_inode(dip);
 653         if ((char *)cur + cur_rec_len > bh->b_data + bh->b_size)
 654                 gfs2_consist_inode(dip);
 655
 656         prev_rec_len += cur_rec_len;
 657         prev->de_rec_len = cpu_to_be16(prev_rec_len);
 658 }
 659
 660 /*
 661  * Takes a dent from which to grab space as an argument. Returns the
 662  * newly created dent.
 663  */
 664 static struct gfs2_dirent *gfs2_init_dirent(struct inode *inode,
 665                                             struct gfs2_dirent *dent,
 666                                             const struct qstr *name,
 667                                             struct buffer_head *bh)
 668 {
 669         struct gfs2_inode *ip = GFS2_I(inode);
 670         struct gfs2_dirent *ndent;
 671         unsigned offset = 0, totlen;
 672
 673         if (!gfs2_dirent_sentinel(dent))
 674                 offset = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len));
 675         totlen = be16_to_cpu(dent->de_rec_len);
 676         BUG_ON(offset + name->len > totlen);
 677         gfs2_trans_add_bh(ip->i_gl, bh, 1);
 678         ndent = (struct gfs2_dirent *)((char *)dent + offset);
 679         dent->de_rec_len = cpu_to_be16(offset);
 680         gfs2_qstr2dirent(name, totlen - offset, ndent);
 681         return ndent;
 682 }
 683
 684 static struct gfs2_dirent *gfs2_dirent_alloc(struct inode *inode,
 685                                              struct buffer_head *bh,
 686                                              const struct qstr *name)
 687 {
 688         struct gfs2_dirent *dent;
 689         dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
 690                                 gfs2_dirent_find_space, name, NULL);
 691         if (!dent || IS_ERR(dent))
 692                 return dent;
 693         return gfs2_init_dirent(inode, dent, name, bh);
 694 }
 695
 696 static int get_leaf(struct gfs2_inode *dip, u64 leaf_no,
 697                     struct buffer_head **bhp)
 698 {
 699         int error;
 700
 701         error = gfs2_meta_read(dip->i_gl, leaf_no, DIO_WAIT, bhp);
 702         if (!error && gfs2_metatype_check(GFS2_SB(&dip->i_inode), *bhp, GFS2_METATYPE_LF)) {
 703                 /* printk(KERN_INFO "block num=%llu\n", leaf_no); */
 704                 error = -EIO;
 705         }
 706
 707         return error;
 708 }
 709
 710 /**
 711  * get_leaf_nr - Get a leaf number associated with the index
 712  * @dip: The GFS2 inode
 713  * @index:
 714  * @leaf_out:
 715  *
 716  * Returns: 0 on success, error code otherwise
 717  */
 718
 719 static int get_leaf_nr(struct gfs2_inode *dip, u32 index,
 720                        u64 *leaf_out)
 721 {
 722         __be64 leaf_no;
 723         int error;
 724
 725         error = gfs2_dir_read_data(dip, (char *)&leaf_no,
 726                                     index * sizeof(__be64),
 727                                     sizeof(__be64), 0);
 728         if (error != sizeof(u64))
 729                 return (error < 0) ? error : -EIO;
 730
 731         *leaf_out = be64_to_cpu(leaf_no);
 732
 733         return 0;
 734 }
 735
 736 static int get_first_leaf(struct gfs2_inode *dip, u32 index,
 737                           struct buffer_head **bh_out)
 738 {
 739         u64 leaf_no;
 740         int error;
 741
 742         error = get_leaf_nr(dip, index, &leaf_no);
 743         if (!error)
 744                 error = get_leaf(dip, leaf_no, bh_out);
 745
 746         return error;
 747 }
 748
 749 static struct gfs2_dirent *gfs2_dirent_search(struct inode *inode,
 750                                               const struct qstr *name,
 751                                               gfs2_dscan_t scan,
 752                                               struct buffer_head **pbh)
 753 {
 754         struct buffer_head *bh;
 755         struct gfs2_dirent *dent;
 756         struct gfs2_inode *ip = GFS2_I(inode);
 757         int error;
 758
 759         if (ip->i_di.di_flags & GFS2_DIF_EXHASH) {
 760                 struct gfs2_leaf *leaf;
 761                 unsigned hsize = 1 << ip->i_di.di_depth;
 762                 unsigned index;
 763                 u64 ln;
 764                 if (hsize * sizeof(u64) != ip->i_di.di_size) {
 765                         gfs2_consist_inode(ip);
 766                         return ERR_PTR(-EIO);
 767                 }
 768
 769                 index = name->hash >> (32 - ip->i_di.di_depth);
 770                 error = get_first_leaf(ip, index, &bh);
 771                 if (error)
 772                         return ERR_PTR(error);
 773                 do {
 774                         dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
 775                                                 scan, name, NULL);
 776                         if (dent)
 777                                 goto got_dent;
 778                         leaf = (struct gfs2_leaf *)bh->b_data;
 779                         ln = be64_to_cpu(leaf->lf_next);
 780                         brelse(bh);
 781                         if (!ln)
 782                                 break;
 783
 784                         error = get_leaf(ip, ln, &bh);
 785                 } while(!error);
 786
 787                 return error ? ERR_PTR(error) : NULL;
 788         }
 789
 790
 791         error = gfs2_meta_inode_buffer(ip, &bh);
 792         if (error)
 793                 return ERR_PTR(error);
 794         dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size, scan, name, NULL);
 795 got_dent:
 796         if (unlikely(dent == NULL || IS_ERR(dent))) {
 797                 brelse(bh);
 798                 bh = NULL;
 799         }
 800         *pbh = bh;
 801         return dent;
 802 }
 803
 804 static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, u16 depth)
 805 {
 806         struct gfs2_inode *ip = GFS2_I(inode);
 807         u64 bn = gfs2_alloc_meta(ip);
 808         struct buffer_head *bh = gfs2_meta_new(ip->i_gl, bn);
 809         struct gfs2_leaf *leaf;
 810         struct gfs2_dirent *dent;
 811         struct qstr name = { .name = "", .len = 0, .hash = 0 };
 812         if (!bh)
 813                 return NULL;
 814
 815         gfs2_trans_add_bh(ip->i_gl, bh, 1);
 816         gfs2_metatype_set(bh, GFS2_METATYPE_LF, GFS2_FORMAT_LF);
 817         leaf = (struct gfs2_leaf *)bh->b_data;
 818         leaf->lf_depth = cpu_to_be16(depth);
 819         leaf->lf_entries = 0;
 820         leaf->lf_dirent_format = cpu_to_be32(GFS2_FORMAT_DE);
 821         leaf->lf_next = 0;
 822         memset(leaf->lf_reserved, 0, sizeof(leaf->lf_reserved));
 823         dent = (struct gfs2_dirent *)(leaf+1);
 824         gfs2_qstr2dirent(&name, bh->b_size - sizeof(struct gfs2_leaf), dent);
 825         *pbh = bh;
 826         return leaf;
 827 }
 828
 829 /**
 830  * dir_make_exhash - Convert a stuffed directory into an ExHash directory
 831  * @dip: The GFS2 inode
 832  *
 833  * Returns: 0 on success, error code otherwise
 834  */
 835
 836 static int dir_make_exhash(struct inode *inode)
 837 {
 838         struct gfs2_inode *dip = GFS2_I(inode);
 839         struct gfs2_sbd *sdp = GFS2_SB(inode);
 840         struct gfs2_dirent *dent;
 841         struct qstr args;
 842         struct buffer_head *bh, *dibh;
 843         struct gfs2_leaf *leaf;
 844         int y;
 845         u32 x;
 846         __be64 *lp;
 847         u64 bn;
 848         int error;
 849
 850         error = gfs2_meta_inode_buffer(dip, &dibh);
 851         if (error)
 852                 return error;
 853
 854         /*  Turn over a new leaf  */
 855
 856         leaf = new_leaf(inode, &bh, 0);
 857         if (!leaf)
 858                 return -ENOSPC;
 859         bn = bh->b_blocknr;
 860
 861         gfs2_assert(sdp, dip->i_di.di_entries < (1 << 16));
 862         leaf->lf_entries = cpu_to_be16(dip->i_di.di_entries);
 863
 864         /*  Copy dirents  */
 865
 866         gfs2_buffer_copy_tail(bh, sizeof(struct gfs2_leaf), dibh,
 867                              sizeof(struct gfs2_dinode));
 868
 869         /*  Find last entry  */
 870
 871         x = 0;
 872         args.len = bh->b_size - sizeof(struct gfs2_dinode) +
 873                    sizeof(struct gfs2_leaf);
 874         args.name = bh->b_data;
 875         dent = gfs2_dirent_scan(&dip->i_inode, bh->b_data, bh->b_size,
 876                                 gfs2_dirent_last, &args, NULL);
 877         if (!dent) {
 878                 brelse(bh);
 879                 brelse(dibh);
 880                 return -EIO;
 881         }
 882         if (IS_ERR(dent)) {
 883                 brelse(bh);
 884                 brelse(dibh);
 885                 return PTR_ERR(dent);
 886         }
 887
 888         /*  Adjust the last dirent's record length
 889            (Remember that dent still points to the last entry.)  */
 890
 891         dent->de_rec_len = cpu_to_be16(be16_to_cpu(dent->de_rec_len) +
 892                 sizeof(struct gfs2_dinode) -
 893                 sizeof(struct gfs2_leaf));
 894
 895         brelse(bh);
 896
 897         /*  We're done with the new leaf block, now setup the new
 898             hash table.  */
 899
 900         gfs2_trans_add_bh(dip->i_gl, dibh, 1);
 901         gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode));
 902
 903         lp = (__be64 *)(dibh->b_data + sizeof(struct gfs2_dinode));
 904
 905         for (x = sdp->sd_hash_ptrs; x--; lp++)
 906                 *lp = cpu_to_be64(bn);
 907
 908         dip->i_di.di_size = sdp->sd_sb.sb_bsize / 2;
 909         dip->i_di.di_blocks++;
 910         gfs2_set_inode_blocks(&dip->i_inode);
 911         dip->i_di.di_flags |= GFS2_DIF_EXHASH;
 912
 913         for (x = sdp->sd_hash_ptrs, y = -1; x; x >>= 1, y++) ;
 914         dip->i_di.di_depth = y;
 915
 916         gfs2_dinode_out(dip, dibh->b_data);
 917
 918         brelse(dibh);
 919
 920         return 0;
 921 }
 922
 923 /**
 924  * dir_split_leaf - Split a leaf block into two
 925  * @dip: The GFS2 inode
 926  * @index:
 927  * @leaf_no:
 928  *
 929  * Returns: 0 on success, error code on failure
 930  */
 931
 932 static int dir_split_leaf(struct inode *inode, const struct qstr *name)
 933 {
 934         struct gfs2_inode *dip = GFS2_I(inode);
 935         struct buffer_head *nbh, *obh, *dibh;
 936         struct gfs2_leaf *nleaf, *oleaf;
 937         struct gfs2_dirent *dent = NULL, *prev = NULL, *next = NULL, *new;
 938         u32 start, len, half_len, divider;
 939         u64 bn, leaf_no;
 940         __be64 *lp;
 941         u32 index;
 942         int x, moved = 0;
 943         int error;
 944
 945         index = name->hash >> (32 - dip->i_di.di_depth);
 946         error = get_leaf_nr(dip, index, &leaf_no);
 947         if (error)
 948                 return error;
 949
 950         /*  Get the old leaf block  */
 951         error = get_leaf(dip, leaf_no, &obh);
 952         if (error)
 953                 return error;
 954
 955         oleaf = (struct gfs2_leaf *)obh->b_data;
 956         if (dip->i_di.di_depth == be16_to_cpu(oleaf->lf_depth)) {
 957                 brelse(obh);
 958                 return 1; /* can't split */
 959         }
 960
 961         gfs2_trans_add_bh(dip->i_gl, obh, 1);
 962
 963         nleaf = new_leaf(inode, &nbh, be16_to_cpu(oleaf->lf_depth) + 1);
 964         if (!nleaf) {
 965                 brelse(obh);
 966                 return -ENOSPC;
 967         }
 968         bn = nbh->b_blocknr;
 969
 970         /*  Compute the start and len of leaf pointers in the hash table.  */
 971         len = 1 << (dip->i_di.di_depth - be16_to_cpu(oleaf->lf_depth));
 972         half_len = len >> 1;
 973         if (!half_len) {
 974                 printk(KERN_WARNING "di_depth %u lf_depth %u index %u\n", dip->i_di.di_depth, be16_to_cpu(oleaf->lf_depth), index);
 975                 gfs2_consist_inode(dip);
 976                 error = -EIO;
 977                 goto fail_brelse;
 978         }
 979
 980         start = (index & ~(len - 1));
 981
 982         /* Change the pointers.
 983            Don't bother distinguishing stuffed from non-stuffed.
 984            This code is complicated enough already. */
 985         lp = kmalloc(half_len * sizeof(__be64), GFP_NOFS | __GFP_NOFAIL);
 986         /*  Change the pointers  */
 987         for (x = 0; x < half_len; x++)
 988                 lp[x] = cpu_to_be64(bn);
 989
 990         error = gfs2_dir_write_data(dip, (char *)lp, start * sizeof(u64),
 991                                     half_len * sizeof(u64));
 992         if (error != half_len * sizeof(u64)) {
 993                 if (error >= 0)
 994                         error = -EIO;
 995                 goto fail_lpfree;
 996         }
 997
 998         kfree(lp);
 999
1000         /*  Compute the divider  */
1001         divider = (start + half_len) << (32 - dip->i_di.di_depth);
1002
1003         /*  Copy the entries  */
1004         dirent_first(dip, obh, &dent);
1005
1006         do {
1007                 next = dent;
1008                 if (dirent_next(dip, obh, &next))
1009                         next = NULL;
1010
1011                 if (!gfs2_dirent_sentinel(dent) &&
1012                     be32_to_cpu(dent->de_hash) < divider) {
1013                         struct qstr str;
1014                         str.name = (char*)(dent+1);
1015                         str.len = be16_to_cpu(dent->de_name_len);
1016                         str.hash = be32_to_cpu(dent->de_hash);
1017                         new = gfs2_dirent_alloc(inode, nbh, &str);
1018                         if (IS_ERR(new)) {
1019                                 error = PTR_ERR(new);
1020                                 break;
1021                         }
1022
1023                         new->de_inum = dent->de_inum; /* No endian worries */
1024                         new->de_type = dent->de_type; /* No endian worries */
1025                         nleaf->lf_entries = cpu_to_be16(be16_to_cpu(nleaf->lf_entries)+1);
1026
1027                         dirent_del(dip, obh, prev, dent);
1028
1029                         if (!oleaf->lf_entries)
1030                                 gfs2_consist_inode(dip);
1031                         oleaf->lf_entries = cpu_to_be16(be16_to_cpu(oleaf->lf_entries)-1);
1032
1033                         if (!prev)
1034                                 prev = dent;
1035
1036                         moved = 1;
1037                 } else {
1038                         prev = dent;
1039                 }
1040                 dent = next;
1041         } while (dent);
1042
1043         oleaf->lf_depth = nleaf->lf_depth;
1044
1045         error = gfs2_meta_inode_buffer(dip, &dibh);
1046         if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) {
1047                 dip->i_di.di_blocks++;
1048                 gfs2_set_inode_blocks(&dip->i_inode);
1049                 gfs2_dinode_out(dip, dibh->b_data);
1050                 brelse(dibh);
1051         }
1052
1053         brelse(obh);
1054         brelse(nbh);
1055
1056         return error;
1057
1058 fail_lpfree:
1059         kfree(lp);
1060
1061 fail_brelse:
1062         brelse(obh);
1063         brelse(nbh);
1064         return error;
1065 }
1066
1067 /**
1068  * dir_double_exhash - Double size of ExHash table
1069  * @dip: The GFS2 dinode
1070  *
1071  * Returns: 0 on success, error code on failure
1072  */
1073
1074 static int dir_double_exhash(struct gfs2_inode *dip)
1075 {
1076         struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1077         struct buffer_head *dibh;
1078         u32 hsize;
1079         u64 *buf;
1080         u64 *from, *to;
1081         u64 block;
1082         int x;
1083         int error = 0;
1084
1085         hsize = 1 << dip->i_di.di_depth;
1086         if (hsize * sizeof(u64) != dip->i_di.di_size) {
1087                 gfs2_consist_inode(dip);
1088                 return -EIO;
1089         }
1090
1091         /*  Allocate both the "from" and "to" buffers in one big chunk  */
1092
1093         buf = kcalloc(3, sdp->sd_hash_bsize, GFP_KERNEL | __GFP_NOFAIL);
1094
1095         for (block = dip->i_di.di_size >> sdp->sd_hash_bsize_shift; block--;) {
1096                 error = gfs2_dir_read_data(dip, (char *)buf,
1097                                             block * sdp->sd_hash_bsize,
1098                                             sdp->sd_hash_bsize, 1);
1099                 if (error != sdp->sd_hash_bsize) {
1100                         if (error >= 0)
1101                                 error = -EIO;
1102                         goto fail;
1103                 }
1104
1105                 from = buf;
1106                 to = (u64 *)((char *)buf + sdp->sd_hash_bsize);
1107
1108                 for (x = sdp->sd_hash_ptrs; x--; from++) {
1109                         *to++ = *from;  /*  No endianess worries  */
1110                         *to++ = *from;
1111                 }
1112
1113                 error = gfs2_dir_write_data(dip,
1114                                              (char *)buf + sdp->sd_hash_bsize,
1115                                              block * sdp->sd_sb.sb_bsize,
1116                                              sdp->sd_sb.sb_bsize);
1117                 if (error != sdp->sd_sb.sb_bsize) {
1118                         if (error >= 0)
1119                                 error = -EIO;
1120                         goto fail;
1121                 }
1122         }
1123
1124         kfree(buf);
1125
1126         error = gfs2_meta_inode_buffer(dip, &dibh);
1127         if (!gfs2_assert_withdraw(sdp, !error)) {
1128                 dip->i_di.di_depth++;
1129                 gfs2_dinode_out(dip, dibh->b_data);
1130                 brelse(dibh);
1131         }
1132
1133         return error;
1134
1135 fail:
1136         kfree(buf);
1137         return error;
1138 }
1139
1140 /**
1141  * compare_dents - compare directory entries by hash value
1142  * @a: first dent
1143  * @b: second dent
1144  *
1145  * When comparing the hash entries of @a to @b:
1146  *   gt: returns 1
1147  *   lt: returns -1
1148  *   eq: returns 0
1149  */
1150
1151 static int compare_dents(const void *a, const void *b)
1152 {
1153         const struct gfs2_dirent *dent_a, *dent_b;
1154         u32 hash_a, hash_b;
1155         int ret = 0;
1156
1157         dent_a = *(const struct gfs2_dirent **)a;
1158         hash_a = be32_to_cpu(dent_a->de_hash);
1159
1160         dent_b = *(const struct gfs2_dirent **)b;
1161         hash_b = be32_to_cpu(dent_b->de_hash);
1162
1163         if (hash_a > hash_b)
1164                 ret = 1;
1165         else if (hash_a < hash_b)
1166                 ret = -1;
1167         else {
1168                 unsigned int len_a = be16_to_cpu(dent_a->de_name_len);
1169                 unsigned int len_b = be16_to_cpu(dent_b->de_name_len);
1170
1171                 if (len_a > len_b)
1172                         ret = 1;
1173                 else if (len_a < len_b)
1174                         ret = -1;
1175                 else
1176                         ret = memcmp(dent_a + 1, dent_b + 1, len_a);
1177         }
1178
1179         return ret;
1180 }
1181
1182 /**
1183  * do_filldir_main - read out directory entries
1184  * @dip: The GFS2 inode
1185  * @offset: The offset in the file to read from
1186  * @opaque: opaque data to pass to filldir
1187  * @filldir: The function to pass entries to
1188  * @darr: an array of struct gfs2_dirent pointers to read
1189  * @entries: the number of entries in darr
1190  * @copied: pointer to int that's non-zero if a entry has been copied out
1191  *
1192  * Jump through some hoops to make sure that if there are hash collsions,
1193  * they are read out at the beginning of a buffer.  We want to minimize
1194  * the possibility that they will fall into different readdir buffers or
1195  * that someone will want to seek to that location.
1196  *
1197  * Returns: errno, >0 on exception from filldir
1198  */
1199
1200 static int do_filldir_main(struct gfs2_inode *dip, u64 *offset,
1201                            void *opaque, gfs2_filldir_t filldir,
1202                            const struct gfs2_dirent **darr, u32 entries,
1203                            int *copied)
1204 {
1205         const struct gfs2_dirent *dent, *dent_next;
1206         struct gfs2_inum_host inum;
1207         u64 off, off_next;
1208         unsigned int x, y;
1209         int run = 0;
1210         int error = 0;
1211
1212         sort(darr, entries, sizeof(struct gfs2_dirent *), compare_dents, NULL);
1213
1214         dent_next = darr[0];
1215         off_next = be32_to_cpu(dent_next->de_hash);
1216         off_next = gfs2_disk_hash2offset(off_next);
1217
1218         for (x = 0, y = 1; x < entries; x++, y++) {
1219                 dent = dent_next;
1220                 off = off_next;
1221
1222                 if (y < entries) {
1223                         dent_next = darr[y];
1224                         off_next = be32_to_cpu(dent_next->de_hash);
1225                         off_next = gfs2_disk_hash2offset(off_next);
1226
1227                         if (off < *offset)
1228                                 continue;
1229                         *offset = off;
1230
1231                         if (off_next == off) {
1232                                 if (*copied && !run)
1233                                         return 1;
1234                                 run = 1;
1235                         } else
1236                                 run = 0;
1237                 } else {
1238                         if (off < *offset)
1239                                 continue;
1240                         *offset = off;
1241                 }
1242
1243                 gfs2_inum_in(&inum, (char *)&dent->de_inum);
1244
1245                 error = filldir(opaque, (const char *)(dent + 1),
1246                                 be16_to_cpu(dent->de_name_len),
1247                                 off, &inum,
1248                                 be16_to_cpu(dent->de_type));
1249                 if (error)
1250                         return 1;
1251
1252                 *copied = 1;
1253         }
1254
1255         /* Increment the *offset by one, so the next time we come into the
1256            do_filldir fxn, we get the next entry instead of the last one in the
1257            current leaf */
1258
1259         (*offset)++;
1260
1261         return 0;
1262 }
1263
1264 static int gfs2_dir_read_leaf(struct inode *inode, u64 *offset, void *opaque,
1265                               gfs2_filldir_t filldir, int *copied,
1266                               unsigned *depth, u64 leaf_no)
1267 {
1268         struct gfs2_inode *ip = GFS2_I(inode);
1269         struct buffer_head *bh;
1270         struct gfs2_leaf *lf;
1271         unsigned entries = 0;
1272         unsigned leaves = 0;
1273         const struct gfs2_dirent **darr, *dent;
1274         struct dirent_gather g;
1275         struct buffer_head **larr;
1276         int leaf = 0;
1277         int error, i;
1278         u64 lfn = leaf_no;
1279
1280         do {
1281                 error = get_leaf(ip, lfn, &bh);
1282                 if (error)
1283                         goto out;
1284                 lf = (struct gfs2_leaf *)bh->b_data;
1285                 if (leaves == 0)
1286                         *depth = be16_to_cpu(lf->lf_depth);
1287                 entries += be16_to_cpu(lf->lf_entries);
1288                 leaves++;
1289                 lfn = be64_to_cpu(lf->lf_next);
1290                 brelse(bh);
1291         } while(lfn);
1292
1293         if (!entries)
1294                 return 0;
1295
1296         error = -ENOMEM;
1297         larr = vmalloc((leaves + entries) * sizeof(void *));
1298         if (!larr)
1299                 goto out;
1300         darr = (const struct gfs2_dirent **)(larr + leaves);
1301         g.pdent = darr;
1302         g.offset = 0;
1303         lfn = leaf_no;
1304
1305         do {
1306                 error = get_leaf(ip, lfn, &bh);
1307                 if (error)
1308                         goto out_kfree;
1309                 lf = (struct gfs2_leaf *)bh->b_data;
1310                 lfn = be64_to_cpu(lf->lf_next);
1311                 if (lf->lf_entries) {
1312                         dent = gfs2_dirent_scan(inode, bh->b_data, bh->b_size,
1313                                                 gfs2_dirent_gather, NULL, &g);
1314                         error = PTR_ERR(dent);
1315                         if (IS_ERR(dent)) {
1316                                 goto out_kfree;
1317                         }
1318                         error = 0;
1319                         larr[leaf++] = bh;
1320                 } else {
1321                         brelse(bh);
1322                 }
1323         } while(lfn);
1324
1325         error = do_filldir_main(ip, offset, opaque, filldir, darr,
1326                                 entries, copied);
1327 out_kfree:
1328         for(i = 0; i < leaf; i++)
1329                 brelse(larr[i]);
1330         vfree(larr);
1331 out:
1332         return error;
1333 }
1334
1335 /**
1336  * dir_e_read - Reads the entries from a directory into a filldir buffer
1337  * @dip: dinode pointer
1338  * @offset: the hash of the last entry read shifted to the right once
1339  * @opaque: buffer for the filldir function to fill
1340  * @filldir: points to the filldir function to use
1341  *
1342  * Returns: errno
1343  */
1344
1345 static int dir_e_read(struct inode *inode, u64 *offset, void *opaque,
1346                       gfs2_filldir_t filldir)
1347 {
1348         struct gfs2_inode *dip = GFS2_I(inode);
1349         struct gfs2_sbd *sdp = GFS2_SB(inode);
1350         u32 hsize, len = 0;
1351         u32 ht_offset, lp_offset, ht_offset_cur = -1;
1352         u32 hash, index;
1353         __be64 *lp;
1354         int copied = 0;
1355         int error = 0;
1356         unsigned depth = 0;
1357
1358         hsize = 1 << dip->i_di.di_depth;
1359         if (hsize * sizeof(u64) != dip->i_di.di_size) {
1360                 gfs2_consist_inode(dip);
1361                 return -EIO;
1362         }
1363
1364         hash = gfs2_dir_offset2hash(*offset);
1365         index = hash >> (32 - dip->i_di.di_depth);
1366
1367         lp = kmalloc(sdp->sd_hash_bsize, GFP_KERNEL);
1368         if (!lp)
1369                 return -ENOMEM;
1370
1371         while (index < hsize) {
1372                 lp_offset = index & (sdp->sd_hash_ptrs - 1);
1373                 ht_offset = index - lp_offset;
1374
1375                 if (ht_offset_cur != ht_offset) {
1376                         error = gfs2_dir_read_data(dip, (char *)lp,
1377                                                 ht_offset * sizeof(__be64),
1378                                                 sdp->sd_hash_bsize, 1);
1379                         if (error != sdp->sd_hash_bsize) {
1380                                 if (error >= 0)
1381                                         error = -EIO;
1382                                 goto out;
1383                         }
1384                         ht_offset_cur = ht_offset;
1385                 }
1386
1387                 error = gfs2_dir_read_leaf(inode, offset, opaque, filldir,
1388                                            &copied, &depth,
1389                                            be64_to_cpu(lp[lp_offset]));
1390                 if (error)
1391                         break;
1392
1393                 len = 1 << (dip->i_di.di_depth - depth);
1394                 index = (index & ~(len - 1)) + len;
1395         }
1396
1397 out:
1398         kfree(lp);
1399         if (error > 0)
1400                 error = 0;
1401         return error;
1402 }
1403
1404 int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque,
1405                   gfs2_filldir_t filldir)
1406 {
1407         struct gfs2_inode *dip = GFS2_I(inode);
1408         struct dirent_gather g;
1409         const struct gfs2_dirent **darr, *dent;
1410         struct buffer_head *dibh;
1411         int copied = 0;
1412         int error;
1413
1414         if (!dip->i_di.di_entries)
1415                 return 0;
1416
1417         if (dip->i_di.di_flags & GFS2_DIF_EXHASH)
1418                 return dir_e_read(inode, offset, opaque, filldir);
1419
1420         if (!gfs2_is_stuffed(dip)) {
1421                 gfs2_consist_inode(dip);
1422                 return -EIO;
1423         }
1424
1425         error = gfs2_meta_inode_buffer(dip, &dibh);
1426         if (error)
1427                 return error;
1428
1429         error = -ENOMEM;
1430         darr = kmalloc(dip->i_di.di_entries * sizeof(struct gfs2_dirent *),
1431                        GFP_KERNEL);
1432         if (darr) {
1433                 g.pdent = darr;
1434                 g.offset = 0;
1435                 dent = gfs2_dirent_scan(inode, dibh->b_data, dibh->b_size,
1436                                         gfs2_dirent_gather, NULL, &g);
1437                 if (IS_ERR(dent)) {
1438                         error = PTR_ERR(dent);
1439                         goto out;
1440                 }
1441                 error = do_filldir_main(dip, offset, opaque, filldir, darr,
1442                                         dip->i_di.di_entries, &copied);
1443 out:
1444                 kfree(darr);
1445         }
1446
1447         if (error > 0)
1448                 error = 0;
1449
1450         brelse(dibh);
1451
1452         return error;
1453 }
1454
1455 /**
1456  * gfs2_dir_search - Search a directory
1457  * @dip: The GFS2 inode
1458  * @filename:
1459  * @inode:
1460  *
1461  * This routine searches a directory for a file or another directory.
1462  * Assumes a glock is held on dip.
1463  *
1464  * Returns: errno
1465  */
1466
1467 int gfs2_dir_search(struct inode *dir, const struct qstr *name,
1468                     struct gfs2_inum_host *inum, unsigned int *type)
1469 {
1470         struct buffer_head *bh;
1471         struct gfs2_dirent *dent;
1472
1473         dent = gfs2_dirent_search(dir, name, gfs2_dirent_find, &bh);
1474         if (dent) {
1475                 if (IS_ERR(dent))
1476                         return PTR_ERR(dent);
1477                 if (inum)
1478                         gfs2_inum_in(inum, (char *)&dent->de_inum);
1479                 if (type)
1480                         *type = be16_to_cpu(dent->de_type);
1481                 brelse(bh);
1482                 return 0;
1483         }
1484         return -ENOENT;
1485 }
1486
1487 static int dir_new_leaf(struct inode *inode, const struct qstr *name)
1488 {
1489         struct buffer_head *bh, *obh;
1490         struct gfs2_inode *ip = GFS2_I(inode);
1491         struct gfs2_leaf *leaf, *oleaf;
1492         int error;
1493         u32 index;
1494         u64 bn;
1495
1496         index = name->hash >> (32 - ip->i_di.di_depth);
1497         error = get_first_leaf(ip, index, &obh);
1498         if (error)
1499                 return error;
1500         do {
1501                 oleaf = (struct gfs2_leaf *)obh->b_data;
1502                 bn = be64_to_cpu(oleaf->lf_next);
1503                 if (!bn)
1504                         break;
1505                 brelse(obh);
1506                 error = get_leaf(ip, bn, &obh);
1507                 if (error)
1508                         return error;
1509         } while(1);
1510
1511         gfs2_trans_add_bh(ip->i_gl, obh, 1);
1512
1513         leaf = new_leaf(inode, &bh, be16_to_cpu(oleaf->lf_depth));
1514         if (!leaf) {
1515                 brelse(obh);
1516                 return -ENOSPC;
1517         }
1518         oleaf->lf_next = cpu_to_be64(bh->b_blocknr);
1519         brelse(bh);
1520         brelse(obh);
1521
1522         error = gfs2_meta_inode_buffer(ip, &bh);
1523         if (error)
1524                 return error;
1525         gfs2_trans_add_bh(ip->i_gl, bh, 1);
1526         ip->i_di.di_blocks++;
1527         gfs2_set_inode_blocks(&ip->i_inode);
1528         gfs2_dinode_out(ip, bh->b_data);
1529         brelse(bh);
1530         return 0;
1531 }
1532
1533 /**
1534  * gfs2_dir_add - Add new filename into directory
1535  * @dip: The GFS2 inode
1536  * @filename: The new name
1537  * @inode: The inode number of the entry
1538  * @type: The type of the entry
1539  *
1540  * Returns: 0 on success, error code on failure
1541  */
1542
1543 int gfs2_dir_add(struct inode *inode, const struct qstr *name,
1544                  const struct gfs2_inum_host *inum, unsigned type)
1545 {
1546         struct gfs2_inode *ip = GFS2_I(inode);
1547         struct buffer_head *bh;
1548         struct gfs2_dirent *dent;
1549         struct gfs2_leaf *leaf;
1550         int error;
1551
1552         while(1) {
1553                 dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space,
1554                                           &bh);
1555                 if (dent) {
1556                         if (IS_ERR(dent))
1557                                 return PTR_ERR(dent);
1558                         dent = gfs2_init_dirent(inode, dent, name, bh);
1559                         gfs2_inum_out(inum, (char *)&dent->de_inum);
1560                         dent->de_type = cpu_to_be16(type);
1561                         if (ip->i_di.di_flags & GFS2_DIF_EXHASH) {
1562                                 leaf = (struct gfs2_leaf *)bh->b_data;
1563                                 leaf->lf_entries = cpu_to_be16(be16_to_cpu(leaf->lf_entries) + 1);
1564                         }
1565                         brelse(bh);
1566                         error = gfs2_meta_inode_buffer(ip, &bh);
1567                         if (error)
1568                                 break;
1569                         gfs2_trans_add_bh(ip->i_gl, bh, 1);
1570                         ip->i_di.di_entries++;
1571                         ip->i_inode.i_mtime.tv_sec = ip->i_inode.i_ctime.tv_sec = get_seconds();
1572                         gfs2_dinode_out(ip, bh->b_data);
1573                         brelse(bh);
1574                         error = 0;
1575                         break;
1576                 }
1577                 if (!(ip->i_di.di_flags & GFS2_DIF_EXHASH)) {
1578                         error = dir_make_exhash(inode);
1579                         if (error)
1580                                 break;
1581                         continue;
1582                 }
1583                 error = dir_split_leaf(inode, name);
1584                 if (error == 0)
1585                         continue;
1586                 if (error < 0)
1587                         break;
1588                 if (ip->i_di.di_depth < GFS2_DIR_MAX_DEPTH) {
1589                         error = dir_double_exhash(ip);
1590                         if (error)
1591                                 break;
1592                         error = dir_split_leaf(inode, name);
1593                         if (error < 0)
1594                                 break;
1595                         if (error == 0)
1596                                 continue;
1597                 }
1598                 error = dir_new_leaf(inode, name);
1599                 if (!error)
1600                         continue;
1601                 error = -ENOSPC;
1602                 break;
1603         }
1604         return error;
1605 }
1606
1607
1608 /**
1609  * gfs2_dir_del - Delete a directory entry
1610  * @dip: The GFS2 inode
1611  * @filename: The filename
1612  *
1613  * Returns: 0 on success, error code on failure
1614  */
1615
1616 int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *name)
1617 {
1618         struct gfs2_dirent *dent, *prev = NULL;
1619         struct buffer_head *bh;
1620         int error;
1621
1622         /* Returns _either_ the entry (if its first in block) or the
1623            previous entry otherwise */
1624         dent = gfs2_dirent_search(&dip->i_inode, name, gfs2_dirent_prev, &bh);
1625         if (!dent) {
1626                 gfs2_consist_inode(dip);
1627                 return -EIO;
1628         }
1629         if (IS_ERR(dent)) {
1630                 gfs2_consist_inode(dip);
1631                 return PTR_ERR(dent);
1632         }
1633         /* If not first in block, adjust pointers accordingly */
1634         if (gfs2_dirent_find(dent, name, NULL) == 0) {
1635                 prev = dent;
1636                 dent = (struct gfs2_dirent *)((char *)dent + be16_to_cpu(prev->de_rec_len));
1637         }
1638
1639         dirent_del(dip, bh, prev, dent);
1640         if (dip->i_di.di_flags & GFS2_DIF_EXHASH) {
1641                 struct gfs2_leaf *leaf = (struct gfs2_leaf *)bh->b_data;
1642                 u16 entries = be16_to_cpu(leaf->lf_entries);
1643                 if (!entries)
1644                         gfs2_consist_inode(dip);
1645                 leaf->lf_entries = cpu_to_be16(--entries);
1646         }
1647         brelse(bh);
1648
1649         error = gfs2_meta_inode_buffer(dip, &bh);
1650         if (error)
1651                 return error;
1652
1653         if (!dip->i_di.di_entries)
1654                 gfs2_consist_inode(dip);
1655         gfs2_trans_add_bh(dip->i_gl, bh, 1);
1656         dip->i_di.di_entries--;
1657         dip->i_inode.i_mtime.tv_sec = dip->i_inode.i_ctime.tv_sec = get_seconds();
1658         gfs2_dinode_out(dip, bh->b_data);
1659         brelse(bh);
1660         mark_inode_dirty(&dip->i_inode);
1661
1662         return error;
1663 }
1664
1665 /**
1666  * gfs2_dir_mvino - Change inode number of directory entry
1667  * @dip: The GFS2 inode
1668  * @filename:
1669  * @new_inode:
1670  *
1671  * This routine changes the inode number of a directory entry.  It's used
1672  * by rename to change ".." when a directory is moved.
1673  * Assumes a glock is held on dvp.
1674  *
1675  * Returns: errno
1676  */
1677
1678 int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename,
1679                    struct gfs2_inum_host *inum, unsigned int new_type)
1680 {
1681         struct buffer_head *bh;
1682         struct gfs2_dirent *dent;
1683         int error;
1684
1685         dent = gfs2_dirent_search(&dip->i_inode, filename, gfs2_dirent_find, &bh);
1686         if (!dent) {
1687                 gfs2_consist_inode(dip);
1688                 return -EIO;
1689         }
1690         if (IS_ERR(dent))
1691                 return PTR_ERR(dent);
1692
1693         gfs2_trans_add_bh(dip->i_gl, bh, 1);
1694         gfs2_inum_out(inum, (char *)&dent->de_inum);
1695         dent->de_type = cpu_to_be16(new_type);
1696
1697         if (dip->i_di.di_flags & GFS2_DIF_EXHASH) {
1698                 brelse(bh);
1699                 error = gfs2_meta_inode_buffer(dip, &bh);
1700                 if (error)
1701                         return error;
1702                 gfs2_trans_add_bh(dip->i_gl, bh, 1);
1703         }
1704
1705         dip->i_inode.i_mtime.tv_sec = dip->i_inode.i_ctime.tv_sec = get_seconds();
1706         gfs2_dinode_out(dip, bh->b_data);
1707         brelse(bh);
1708         return 0;
1709 }
1710
1711 /**
1712  * foreach_leaf - call a function for each leaf in a directory
1713  * @dip: the directory
1714  * @lc: the function to call for each each
1715  * @data: private data to pass to it
1716  *
1717  * Returns: errno
1718  */
1719
1720 static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data)
1721 {
1722         struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1723         struct buffer_head *bh;
1724         struct gfs2_leaf *leaf;
1725         u32 hsize, len;
1726         u32 ht_offset, lp_offset, ht_offset_cur = -1;
1727         u32 index = 0;
1728         __be64 *lp;
1729         u64 leaf_no;
1730         int error = 0;
1731
1732         hsize = 1 << dip->i_di.di_depth;
1733         if (hsize * sizeof(u64) != dip->i_di.di_size) {
1734                 gfs2_consist_inode(dip);
1735                 return -EIO;
1736         }
1737
1738         lp = kmalloc(sdp->sd_hash_bsize, GFP_KERNEL);
1739         if (!lp)
1740                 return -ENOMEM;
1741
1742         while (index < hsize) {
1743                 lp_offset = index & (sdp->sd_hash_ptrs - 1);
1744                 ht_offset = index - lp_offset;
1745
1746                 if (ht_offset_cur != ht_offset) {
1747                         error = gfs2_dir_read_data(dip, (char *)lp,
1748                                                 ht_offset * sizeof(__be64),
1749                                                 sdp->sd_hash_bsize, 1);
1750                         if (error != sdp->sd_hash_bsize) {
1751                                 if (error >= 0)
1752                                         error = -EIO;
1753                                 goto out;
1754                         }
1755                         ht_offset_cur = ht_offset;
1756                 }
1757
1758                 leaf_no = be64_to_cpu(lp[lp_offset]);
1759                 if (leaf_no) {
1760                         error = get_leaf(dip, leaf_no, &bh);
1761                         if (error)
1762                                 goto out;
1763                         leaf = (struct gfs2_leaf *)bh->b_data;
1764                         len = 1 << (dip->i_di.di_depth - be16_to_cpu(leaf->lf_depth));
1765                         brelse(bh);
1766
1767                         error = lc(dip, index, len, leaf_no, data);
1768                         if (error)
1769                                 goto out;
1770
1771                         index = (index & ~(len - 1)) + len;
1772                 } else
1773                         index++;
1774         }
1775
1776         if (index != hsize) {
1777                 gfs2_consist_inode(dip);
1778                 error = -EIO;
1779         }
1780
1781 out:
1782         kfree(lp);
1783
1784         return error;
1785 }
1786
1787 /**
1788  * leaf_dealloc - Deallocate a directory leaf
1789  * @dip: the directory
1790  * @index: the hash table offset in the directory
1791  * @len: the number of pointers to this leaf
1792  * @leaf_no: the leaf number
1793  * @data: not used
1794  *
1795  * Returns: errno
1796  */
1797
1798 static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len,
1799                         u64 leaf_no, void *data)
1800 {
1801         struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
1802         struct gfs2_leaf *tmp_leaf;
1803         struct gfs2_rgrp_list rlist;
1804         struct buffer_head *bh, *dibh;
1805         u64 blk, nblk;
1806         unsigned int rg_blocks = 0, l_blocks = 0;
1807         char *ht;
1808         unsigned int x, size = len * sizeof(u64);
1809         int error;
1810
1811         memset(&rlist, 0, sizeof(struct gfs2_rgrp_list));
1812
1813         ht = kzalloc(size, GFP_KERNEL);
1814         if (!ht)
1815                 return -ENOMEM;
1816
1817         gfs2_alloc_get(dip);
1818
1819         error = gfs2_quota_hold(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE);
1820         if (error)
1821                 goto out;
1822
1823         error = gfs2_rindex_hold(sdp, &dip->i_alloc.al_ri_gh);
1824         if (error)
1825                 goto out_qs;
1826
1827         /*  Count the number of leaves  */
1828
1829         for (blk = leaf_no; blk; blk = nblk) {
1830                 error = get_leaf(dip, blk, &bh);
1831                 if (error)
1832                         goto out_rlist;
1833                 tmp_leaf = (struct gfs2_leaf *)bh->b_data;
1834                 nblk = be64_to_cpu(tmp_leaf->lf_next);
1835                 brelse(bh);
1836
1837                 gfs2_rlist_add(sdp, &rlist, blk);
1838                 l_blocks++;
1839         }
1840
1841         gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE, 0);
1842
1843         for (x = 0; x < rlist.rl_rgrps; x++) {
1844                 struct gfs2_rgrpd *rgd;
1845                 rgd = rlist.rl_ghs[x].gh_gl->gl_object;
1846                 rg_blocks += rgd->rd_ri.ri_length;
1847         }
1848
1849         error = gfs2_glock_nq_m(rlist.rl_rgrps, rlist.rl_ghs);
1850         if (error)
1851                 goto out_rlist;
1852
1853         error = gfs2_trans_begin(sdp,
1854                         rg_blocks + (DIV_ROUND_UP(size, sdp->sd_jbsize) + 1) +
1855                         RES_DINODE + RES_STATFS + RES_QUOTA, l_blocks);
1856         if (error)
1857                 goto out_rg_gunlock;
1858
1859         for (blk = leaf_no; blk; blk = nblk) {
1860                 error = get_leaf(dip, blk, &bh);
1861                 if (error)
1862                         goto out_end_trans;
1863                 tmp_leaf = (struct gfs2_leaf *)bh->b_data;
1864                 nblk = be64_to_cpu(tmp_leaf->lf_next);
1865                 brelse(bh);
1866
1867                 gfs2_free_meta(dip, blk, 1);
1868
1869                 if (!dip->i_di.di_blocks)
1870                         gfs2_consist_inode(dip);
1871                 dip->i_di.di_blocks--;
1872                 gfs2_set_inode_blocks(&dip->i_inode);
1873         }
1874
1875         error = gfs2_dir_write_data(dip, ht, index * sizeof(u64), size);
1876         if (error != size) {
1877                 if (error >= 0)
1878                         error = -EIO;
1879                 goto out_end_trans;
1880         }
1881
1882         error = gfs2_meta_inode_buffer(dip, &dibh);
1883         if (error)
1884                 goto out_end_trans;
1885
1886         gfs2_trans_add_bh(dip->i_gl, dibh, 1);
1887         gfs2_dinode_out(dip, dibh->b_data);
1888         brelse(dibh);
1889
1890 out_end_trans:
1891         gfs2_trans_end(sdp);
1892 out_rg_gunlock:
1893         gfs2_glock_dq_m(rlist.rl_rgrps, rlist.rl_ghs);
1894 out_rlist:
1895         gfs2_rlist_free(&rlist);
1896         gfs2_glock_dq_uninit(&dip->i_alloc.al_ri_gh);
1897 out_qs:
1898         gfs2_quota_unhold(dip);
1899 out:
1900         gfs2_alloc_put(dip);
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         int error;
1920
1921         /* Dealloc on-disk leaves to FREEMETA state */
1922         error = foreach_leaf(dip, leaf_dealloc, NULL);
1923         if (error)
1924                 return error;
1925
1926         /* Make this a regular file in case we crash.
1927            (We don't want to free these blocks a second time.)  */
1928
1929         error = gfs2_trans_begin(sdp, RES_DINODE, 0);
1930         if (error)
1931                 return error;
1932
1933         error = gfs2_meta_inode_buffer(dip, &bh);
1934         if (!error) {
1935                 gfs2_trans_add_bh(dip->i_gl, bh, 1);
1936                 ((struct gfs2_dinode *)bh->b_data)->di_mode =
1937                                                 cpu_to_be32(S_IFREG);
1938                 brelse(bh);
1939         }
1940
1941         gfs2_trans_end(sdp);
1942
1943         return error;
1944 }
1945
1946 /**
1947  * gfs2_diradd_alloc_required - find if adding entry will require an allocation
1948  * @ip: the file being written to
1949  * @filname: the filename that's going to be added
1950  *
1951  * Returns: 1 if alloc required, 0 if not, -ve on error
1952  */
1953
1954 int gfs2_diradd_alloc_required(struct inode *inode, const struct qstr *name)
1955 {
1956         struct gfs2_dirent *dent;
1957         struct buffer_head *bh;
1958
1959         dent = gfs2_dirent_search(inode, name, gfs2_dirent_find_space, &bh);
1960         if (!dent) {
1961                 return 1;
1962         }
1963         if (IS_ERR(dent))
1964                 return PTR_ERR(dent);
1965         brelse(bh);
1966         return 0;
1967 }
1968