fs/ext4/extents.c

   1 /*
   2  * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
   3  * Written by Alex Tomas <alex@clusterfs.com>
   4  *
   5  * Architecture independence:
   6  *   Copyright (c) 2005, Bull S.A.
   7  *   Written by Pierre Peiffer <pierre.peiffer@bull.net>
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License version 2 as
  11  * published by the Free Software Foundation.
  12  *
  13  * This program is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public Licens
  19  * along with this program; if not, write to the Free Software
  20  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-
  21  */
  22
  23 /*
  24  * Extents support for EXT4
  25  *
  26  * TODO:
  27  *   - ext4*_error() should be used in some situations
  28  *   - analyze all BUG()/BUG_ON(), use -EIO where appropriate
  29  *   - smart tree reduction
  30  */
  31
  32 #include <linux/module.h>
  33 #include <linux/fs.h>
  34 #include <linux/time.h>
  35 #include <linux/jbd2.h>
  36 #include <linux/highuid.h>
  37 #include <linux/pagemap.h>
  38 #include <linux/quotaops.h>
  39 #include <linux/string.h>
  40 #include <linux/slab.h>
  41 #include <linux/falloc.h>
  42 #include <asm/uaccess.h>
  43 #include <linux/fiemap.h>
  44 #include "ext4_jbd2.h"
  45 #include "ext4_extents.h"
  46
  47 #include <trace/events/ext4.h>
  48
  49 static int ext4_split_extent(handle_t *handle,
  50                                 struct inode *inode,
  51                                 struct ext4_ext_path *path,
  52                                 struct ext4_map_blocks *map,
  53                                 int split_flag,
  54                                 int flags);
  55
  56 static int ext4_ext_truncate_extend_restart(handle_t *handle,
  57                                             struct inode *inode,
  58                                             int needed)
  59 {
  60         int err;
  61
  62         if (!ext4_handle_valid(handle))
  63                 return 0;
  64         if (handle->h_buffer_credits > needed)
  65                 return 0;
  66         err = ext4_journal_extend(handle, needed);
  67         if (err <= 0)
  68                 return err;
  69         err = ext4_truncate_restart_trans(handle, inode, needed);
  70         if (err == 0)
  71                 err = -EAGAIN;
  72
  73         return err;
  74 }
  75
  76 /*
  77  * could return:
  78  *  - EROFS
  79  *  - ENOMEM
  80  */
  81 static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
  82                                 struct ext4_ext_path *path)
  83 {
  84         if (path->p_bh) {
  85                 /* path points to block */
  86                 return ext4_journal_get_write_access(handle, path->p_bh);
  87         }
  88         /* path points to leaf/index in inode body */
  89         /* we use in-core data, no need to protect them */
  90         return 0;
  91 }
  92
  93 /*
  94  * could return:
  95  *  - EROFS
  96  *  - ENOMEM
  97  *  - EIO
  98  */
  99 #define ext4_ext_dirty(handle, inode, path) \
 100                 __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path))
 101 static int __ext4_ext_dirty(const char *where, unsigned int line,
 102                             handle_t *handle, struct inode *inode,
 103                             struct ext4_ext_path *path)
 104 {
 105         int err;
 106         if (path->p_bh) {
 107                 /* path points to block */
 108                 err = __ext4_handle_dirty_metadata(where, line, handle,
 109                                                    inode, path->p_bh);
 110         } else {
 111                 /* path points to leaf/index in inode body */
 112                 err = ext4_mark_inode_dirty(handle, inode);
 113         }
 114         return err;
 115 }
 116
 117 static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
 118                               struct ext4_ext_path *path,
 119                               ext4_lblk_t block)
 120 {
 121         int depth;
 122
 123         if (path) {
 124                 struct ext4_extent *ex;
 125                 depth = path->p_depth;
 126
 127                 /*
 128                  * Try to predict block placement assuming that we are
 129                  * filling in a file which will eventually be
 130                  * non-sparse --- i.e., in the case of libbfd writing
 131                  * an ELF object sections out-of-order but in a way
 132                  * the eventually results in a contiguous object or
 133                  * executable file, or some database extending a table
 134                  * space file.  However, this is actually somewhat
 135                  * non-ideal if we are writing a sparse file such as
 136                  * qemu or KVM writing a raw image file that is going
 137                  * to stay fairly sparse, since it will end up
 138                  * fragmenting the file system's free space.  Maybe we
 139                  * should have some hueristics or some way to allow
 140                  * userspace to pass a hint to file system,
 141                  * especially if the latter case turns out to be
 142                  * common.
 143                  */
 144                 ex = path[depth].p_ext;
 145                 if (ex) {
 146                         ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
 147                         ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block);
 148
 149                         if (block > ext_block)
 150                                 return ext_pblk + (block - ext_block);
 151                         else
 152                                 return ext_pblk - (ext_block - block);
 153                 }
 154
 155                 /* it looks like index is empty;
 156                  * try to find starting block from index itself */
 157                 if (path[depth].p_bh)
 158                         return path[depth].p_bh->b_blocknr;
 159         }
 160
 161         /* OK. use inode's group */
 162         return ext4_inode_to_goal_block(inode);
 163 }
 164
 165 /*
 166  * Allocation for a meta data block
 167  */
 168 static ext4_fsblk_t
 169 ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
 170                         struct ext4_ext_path *path,
 171                         struct ext4_extent *ex, int *err, unsigned int flags)
 172 {
 173         ext4_fsblk_t goal, newblock;
 174
 175         goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
 176         newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
 177                                         NULL, err);
 178         return newblock;
 179 }
 180
 181 static inline int ext4_ext_space_block(struct inode *inode, int check)
 182 {
 183         int size;
 184
 185         size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 186                         / sizeof(struct ext4_extent);
 187         if (!check) {
 188 #ifdef AGGRESSIVE_TEST
 189                 if (size > 6)
 190                         size = 6;
 191 #endif
 192         }
 193         return size;
 194 }
 195
 196 static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
 197 {
 198         int size;
 199
 200         size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 201                         / sizeof(struct ext4_extent_idx);
 202         if (!check) {
 203 #ifdef AGGRESSIVE_TEST
 204                 if (size > 5)
 205                         size = 5;
 206 #endif
 207         }
 208         return size;
 209 }
 210
 211 static inline int ext4_ext_space_root(struct inode *inode, int check)
 212 {
 213         int size;
 214
 215         size = sizeof(EXT4_I(inode)->i_data);
 216         size -= sizeof(struct ext4_extent_header);
 217         size /= sizeof(struct ext4_extent);
 218         if (!check) {
 219 #ifdef AGGRESSIVE_TEST
 220                 if (size > 3)
 221                         size = 3;
 222 #endif
 223         }
 224         return size;
 225 }
 226
 227 static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
 228 {
 229         int size;
 230
 231         size = sizeof(EXT4_I(inode)->i_data);
 232         size -= sizeof(struct ext4_extent_header);
 233         size /= sizeof(struct ext4_extent_idx);
 234         if (!check) {
 235 #ifdef AGGRESSIVE_TEST
 236                 if (size > 4)
 237                         size = 4;
 238 #endif
 239         }
 240         return size;
 241 }
 242
 243 /*
 244  * Calculate the number of metadata blocks needed
 245  * to allocate @blocks
 246  * Worse case is one block per extent
 247  */
 248 int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
 249 {
 250         struct ext4_inode_info *ei = EXT4_I(inode);
 251         int idxs, num = 0;
 252
 253         idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 254                 / sizeof(struct ext4_extent_idx));
 255
 256         /*
 257          * If the new delayed allocation block is contiguous with the
 258          * previous da block, it can share index blocks with the
 259          * previous block, so we only need to allocate a new index
 260          * block every idxs leaf blocks.  At ldxs**2 blocks, we need
 261          * an additional index block, and at ldxs**3 blocks, yet
 262          * another index blocks.
 263          */
 264         if (ei->i_da_metadata_calc_len &&
 265             ei->i_da_metadata_calc_last_lblock+1 == lblock) {
 266                 if ((ei->i_da_metadata_calc_len % idxs) == 0)
 267                         num++;
 268                 if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
 269                         num++;
 270                 if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
 271                         num++;
 272                         ei->i_da_metadata_calc_len = 0;
 273                 } else
 274                         ei->i_da_metadata_calc_len++;
 275                 ei->i_da_metadata_calc_last_lblock++;
 276                 return num;
 277         }
 278
 279         /*
 280          * In the worst case we need a new set of index blocks at
 281          * every level of the inode's extent tree.
 282          */
 283         ei->i_da_metadata_calc_len = 1;
 284         ei->i_da_metadata_calc_last_lblock = lblock;
 285         return ext_depth(inode) + 1;
 286 }
 287
 288 static int
 289 ext4_ext_max_entries(struct inode *inode, int depth)
 290 {
 291         int max;
 292
 293         if (depth == ext_depth(inode)) {
 294                 if (depth == 0)
 295                         max = ext4_ext_space_root(inode, 1);
 296                 else
 297                         max = ext4_ext_space_root_idx(inode, 1);
 298         } else {
 299                 if (depth == 0)
 300                         max = ext4_ext_space_block(inode, 1);
 301                 else
 302                         max = ext4_ext_space_block_idx(inode, 1);
 303         }
 304
 305         return max;
 306 }
 307
 308 static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
 309 {
 310         ext4_fsblk_t block = ext4_ext_pblock(ext);
 311         int len = ext4_ext_get_actual_len(ext);
 312
 313         return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
 314 }
 315
 316 static int ext4_valid_extent_idx(struct inode *inode,
 317                                 struct ext4_extent_idx *ext_idx)
 318 {
 319         ext4_fsblk_t block = ext4_idx_pblock(ext_idx);
 320
 321         return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
 322 }
 323
 324 static int ext4_valid_extent_entries(struct inode *inode,
 325                                 struct ext4_extent_header *eh,
 326                                 int depth)
 327 {
 328         struct ext4_extent *ext;
 329         struct ext4_extent_idx *ext_idx;
 330         unsigned short entries;
 331         if (eh->eh_entries == 0)
 332                 return 1;
 333
 334         entries = le16_to_cpu(eh->eh_entries);
 335
 336         if (depth == 0) {
 337                 /* leaf entries */
 338                 ext = EXT_FIRST_EXTENT(eh);
 339                 while (entries) {
 340                         if (!ext4_valid_extent(inode, ext))
 341                                 return 0;
 342                         ext++;
 343                         entries--;
 344                 }
 345         } else {
 346                 ext_idx = EXT_FIRST_INDEX(eh);
 347                 while (entries) {
 348                         if (!ext4_valid_extent_idx(inode, ext_idx))
 349                                 return 0;
 350                         ext_idx++;
 351                         entries--;
 352                 }
 353         }
 354         return 1;
 355 }
 356
 357 static int __ext4_ext_check(const char *function, unsigned int line,
 358                             struct inode *inode, struct ext4_extent_header *eh,
 359                             int depth)
 360 {
 361         const char *error_msg;
 362         int max = 0;
 363
 364         if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
 365                 error_msg = "invalid magic";
 366                 goto corrupted;
 367         }
 368         if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
 369                 error_msg = "unexpected eh_depth";
 370                 goto corrupted;
 371         }
 372         if (unlikely(eh->eh_max == 0)) {
 373                 error_msg = "invalid eh_max";
 374                 goto corrupted;
 375         }
 376         max = ext4_ext_max_entries(inode, depth);
 377         if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
 378                 error_msg = "too large eh_max";
 379                 goto corrupted;
 380         }
 381         if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
 382                 error_msg = "invalid eh_entries";
 383                 goto corrupted;
 384         }
 385         if (!ext4_valid_extent_entries(inode, eh, depth)) {
 386                 error_msg = "invalid extent entries";
 387                 goto corrupted;
 388         }
 389         return 0;
 390
 391 corrupted:
 392         ext4_error_inode(inode, function, line, 0,
 393                         "bad header/extent: %s - magic %x, "
 394                         "entries %u, max %u(%u), depth %u(%u)",
 395                         error_msg, le16_to_cpu(eh->eh_magic),
 396                         le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
 397                         max, le16_to_cpu(eh->eh_depth), depth);
 398
 399         return -EIO;
 400 }
 401
 402 #define ext4_ext_check(inode, eh, depth)        \
 403         __ext4_ext_check(__func__, __LINE__, inode, eh, depth)
 404
 405 int ext4_ext_check_inode(struct inode *inode)
 406 {
 407         return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
 408 }
 409
 410 #ifdef EXT_DEBUG
 411 static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
 412 {
 413         int k, l = path->p_depth;
 414
 415         ext_debug("path:");
 416         for (k = 0; k <= l; k++, path++) {
 417                 if (path->p_idx) {
 418                   ext_debug("  %d->%llu", le32_to_cpu(path->p_idx->ei_block),
 419                             ext4_idx_pblock(path->p_idx));
 420                 } else if (path->p_ext) {
 421                         ext_debug("  %d:[%d]%d:%llu ",
 422                                   le32_to_cpu(path->p_ext->ee_block),
 423                                   ext4_ext_is_uninitialized(path->p_ext),
 424                                   ext4_ext_get_actual_len(path->p_ext),
 425                                   ext4_ext_pblock(path->p_ext));
 426                 } else
 427                         ext_debug("  []");
 428         }
 429         ext_debug("\n");
 430 }
 431
 432 static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
 433 {
 434         int depth = ext_depth(inode);
 435         struct ext4_extent_header *eh;
 436         struct ext4_extent *ex;
 437         int i;
 438
 439         if (!path)
 440                 return;
 441
 442         eh = path[depth].p_hdr;
 443         ex = EXT_FIRST_EXTENT(eh);
 444
 445         ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
 446
 447         for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
 448                 ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
 449                           ext4_ext_is_uninitialized(ex),
 450                           ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
 451         }
 452         ext_debug("\n");
 453 }
 454
 455 static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path,
 456                         ext4_fsblk_t newblock, int level)
 457 {
 458         int depth = ext_depth(inode);
 459         struct ext4_extent *ex;
 460
 461         if (depth != level) {
 462                 struct ext4_extent_idx *idx;
 463                 idx = path[level].p_idx;
 464                 while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
 465                         ext_debug("%d: move %d:%llu in new index %llu\n", level,
 466                                         le32_to_cpu(idx->ei_block),
 467                                         ext4_idx_pblock(idx),
 468                                         newblock);
 469                         idx++;
 470                 }
 471
 472                 return;
 473         }
 474
 475         ex = path[depth].p_ext;
 476         while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
 477                 ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
 478                                 le32_to_cpu(ex->ee_block),
 479                                 ext4_ext_pblock(ex),
 480                                 ext4_ext_is_uninitialized(ex),
 481                                 ext4_ext_get_actual_len(ex),
 482                                 newblock);
 483                 ex++;
 484         }
 485 }
 486
 487 #else
 488 #define ext4_ext_show_path(inode, path)
 489 #define ext4_ext_show_leaf(inode, path)
 490 #define ext4_ext_show_move(inode, path, newblock, level)
 491 #endif
 492
 493 void ext4_ext_drop_refs(struct ext4_ext_path *path)
 494 {
 495         int depth = path->p_depth;
 496         int i;
 497
 498         for (i = 0; i <= depth; i++, path++)
 499                 if (path->p_bh) {
 500                         brelse(path->p_bh);
 501                         path->p_bh = NULL;
 502                 }
 503 }
 504
 505 /*
 506  * ext4_ext_binsearch_idx:
 507  * binary search for the closest index of the given block
 508  * the header must be checked before calling this
 509  */
 510 static void
 511 ext4_ext_binsearch_idx(struct inode *inode,
 512                         struct ext4_ext_path *path, ext4_lblk_t block)
 513 {
 514         struct ext4_extent_header *eh = path->p_hdr;
 515         struct ext4_extent_idx *r, *l, *m;
 516
 517
 518         ext_debug("binsearch for %u(idx):  ", block);
 519
 520         l = EXT_FIRST_INDEX(eh) + 1;
 521         r = EXT_LAST_INDEX(eh);
 522         while (l <= r) {
 523                 m = l + (r - l) / 2;
 524                 if (block < le32_to_cpu(m->ei_block))
 525                         r = m - 1;
 526                 else
 527                         l = m + 1;
 528                 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
 529                                 m, le32_to_cpu(m->ei_block),
 530                                 r, le32_to_cpu(r->ei_block));
 531         }
 532
 533         path->p_idx = l - 1;
 534         ext_debug("  -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
 535                   ext4_idx_pblock(path->p_idx));
 536
 537 #ifdef CHECK_BINSEARCH
 538         {
 539                 struct ext4_extent_idx *chix, *ix;
 540                 int k;
 541
 542                 chix = ix = EXT_FIRST_INDEX(eh);
 543                 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
 544                   if (k != 0 &&
 545                       le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
 546                                 printk(KERN_DEBUG "k=%d, ix=0x%p, "
 547                                        "first=0x%p\n", k,
 548                                        ix, EXT_FIRST_INDEX(eh));
 549                                 printk(KERN_DEBUG "%u <= %u\n",
 550                                        le32_to_cpu(ix->ei_block),
 551                                        le32_to_cpu(ix[-1].ei_block));
 552                         }
 553                         BUG_ON(k && le32_to_cpu(ix->ei_block)
 554                                            <= le32_to_cpu(ix[-1].ei_block));
 555                         if (block < le32_to_cpu(ix->ei_block))
 556                                 break;
 557                         chix = ix;
 558                 }
 559                 BUG_ON(chix != path->p_idx);
 560         }
 561 #endif
 562
 563 }
 564
 565 /*
 566  * ext4_ext_binsearch:
 567  * binary search for closest extent of the given block
 568  * the header must be checked before calling this
 569  */
 570 static void
 571 ext4_ext_binsearch(struct inode *inode,
 572                 struct ext4_ext_path *path, ext4_lblk_t block)
 573 {
 574         struct ext4_extent_header *eh = path->p_hdr;
 575         struct ext4_extent *r, *l, *m;
 576
 577         if (eh->eh_entries == 0) {
 578                 /*
 579                  * this leaf is empty:
 580                  * we get such a leaf in split/add case
 581                  */
 582                 return;
 583         }
 584
 585         ext_debug("binsearch for %u:  ", block);
 586
 587         l = EXT_FIRST_EXTENT(eh) + 1;
 588         r = EXT_LAST_EXTENT(eh);
 589
 590         while (l <= r) {
 591                 m = l + (r - l) / 2;
 592                 if (block < le32_to_cpu(m->ee_block))
 593                         r = m - 1;
 594                 else
 595                         l = m + 1;
 596                 ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
 597                                 m, le32_to_cpu(m->ee_block),
 598                                 r, le32_to_cpu(r->ee_block));
 599         }
 600
 601         path->p_ext = l - 1;
 602         ext_debug("  -> %d:%llu:[%d]%d ",
 603                         le32_to_cpu(path->p_ext->ee_block),
 604                         ext4_ext_pblock(path->p_ext),
 605                         ext4_ext_is_uninitialized(path->p_ext),
 606                         ext4_ext_get_actual_len(path->p_ext));
 607
 608 #ifdef CHECK_BINSEARCH
 609         {
 610                 struct ext4_extent *chex, *ex;
 611                 int k;
 612
 613                 chex = ex = EXT_FIRST_EXTENT(eh);
 614                 for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
 615                         BUG_ON(k && le32_to_cpu(ex->ee_block)
 616                                           <= le32_to_cpu(ex[-1].ee_block));
 617                         if (block < le32_to_cpu(ex->ee_block))
 618                                 break;
 619                         chex = ex;
 620                 }
 621                 BUG_ON(chex != path->p_ext);
 622         }
 623 #endif
 624
 625 }
 626
 627 int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
 628 {
 629         struct ext4_extent_header *eh;
 630
 631         eh = ext_inode_hdr(inode);
 632         eh->eh_depth = 0;
 633         eh->eh_entries = 0;
 634         eh->eh_magic = EXT4_EXT_MAGIC;
 635         eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
 636         ext4_mark_inode_dirty(handle, inode);
 637         ext4_ext_invalidate_cache(inode);
 638         return 0;
 639 }
 640
 641 struct ext4_ext_path *
 642 ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
 643                                         struct ext4_ext_path *path)
 644 {
 645         struct ext4_extent_header *eh;
 646         struct buffer_head *bh;
 647         short int depth, i, ppos = 0, alloc = 0;
 648
 649         eh = ext_inode_hdr(inode);
 650         depth = ext_depth(inode);
 651
 652         /* account possible depth increase */
 653         if (!path) {
 654                 path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
 655                                 GFP_NOFS);
 656                 if (!path)
 657                         return ERR_PTR(-ENOMEM);
 658                 alloc = 1;
 659         }
 660         path[0].p_hdr = eh;
 661         path[0].p_bh = NULL;
 662
 663         i = depth;
 664         /* walk through the tree */
 665         while (i) {
 666                 int need_to_validate = 0;
 667
 668                 ext_debug("depth %d: num %d, max %d\n",
 669                           ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
 670
 671                 ext4_ext_binsearch_idx(inode, path + ppos, block);
 672                 path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
 673                 path[ppos].p_depth = i;
 674                 path[ppos].p_ext = NULL;
 675
 676                 bh = sb_getblk(inode->i_sb, path[ppos].p_block);
 677                 if (unlikely(!bh))
 678                         goto err;
 679                 if (!bh_uptodate_or_lock(bh)) {
 680                         trace_ext4_ext_load_extent(inode, block,
 681                                                 path[ppos].p_block);
 682                         if (bh_submit_read(bh) < 0) {
 683                                 put_bh(bh);
 684                                 goto err;
 685                         }
 686                         /* validate the extent entries */
 687                         need_to_validate = 1;
 688                 }
 689                 eh = ext_block_hdr(bh);
 690                 ppos++;
 691                 if (unlikely(ppos > depth)) {
 692                         put_bh(bh);
 693                         EXT4_ERROR_INODE(inode,
 694                                          "ppos %d > depth %d", ppos, depth);
 695                         goto err;
 696                 }
 697                 path[ppos].p_bh = bh;
 698                 path[ppos].p_hdr = eh;
 699                 i--;
 700
 701                 if (need_to_validate && ext4_ext_check(inode, eh, i))
 702                         goto err;
 703         }
 704
 705         path[ppos].p_depth = i;
 706         path[ppos].p_ext = NULL;
 707         path[ppos].p_idx = NULL;
 708
 709         /* find extent */
 710         ext4_ext_binsearch(inode, path + ppos, block);
 711         /* if not an empty leaf */
 712         if (path[ppos].p_ext)
 713                 path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);
 714
 715         ext4_ext_show_path(inode, path);
 716
 717         return path;
 718
 719 err:
 720         ext4_ext_drop_refs(path);
 721         if (alloc)
 722                 kfree(path);
 723         return ERR_PTR(-EIO);
 724 }
 725
 726 /*
 727  * ext4_ext_insert_index:
 728  * insert new index [@logical;@ptr] into the block at @curp;
 729  * check where to insert: before @curp or after @curp
 730  */
 731 static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
 732                                  struct ext4_ext_path *curp,
 733                                  int logical, ext4_fsblk_t ptr)
 734 {
 735         struct ext4_extent_idx *ix;
 736         int len, err;
 737
 738         err = ext4_ext_get_access(handle, inode, curp);
 739         if (err)
 740                 return err;
 741
 742         if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
 743                 EXT4_ERROR_INODE(inode,
 744                                  "logical %d == ei_block %d!",
 745                                  logical, le32_to_cpu(curp->p_idx->ei_block));
 746                 return -EIO;
 747         }
 748
 749         if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
 750                              >= le16_to_cpu(curp->p_hdr->eh_max))) {
 751                 EXT4_ERROR_INODE(inode,
 752                                  "eh_entries %d >= eh_max %d!",
 753                                  le16_to_cpu(curp->p_hdr->eh_entries),
 754                                  le16_to_cpu(curp->p_hdr->eh_max));
 755                 return -EIO;
 756         }
 757
 758         len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx;
 759         if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
 760                 /* insert after */
 761                 if (curp->p_idx != EXT_LAST_INDEX(curp->p_hdr)) {
 762                         len = (len - 1) * sizeof(struct ext4_extent_idx);
 763                         len = len < 0 ? 0 : len;
 764                         ext_debug("insert new index %d after: %llu. "
 765                                         "move %d from 0x%p to 0x%p\n",
 766                                         logical, ptr, len,
 767                                         (curp->p_idx + 1), (curp->p_idx + 2));
 768                         memmove(curp->p_idx + 2, curp->p_idx + 1, len);
 769                 }
 770                 ix = curp->p_idx + 1;
 771         } else {
 772                 /* insert before */
 773                 len = len * sizeof(struct ext4_extent_idx);
 774                 len = len < 0 ? 0 : len;
 775                 ext_debug("insert new index %d before: %llu. "
 776                                 "move %d from 0x%p to 0x%p\n",
 777                                 logical, ptr, len,
 778                                 curp->p_idx, (curp->p_idx + 1));
 779                 memmove(curp->p_idx + 1, curp->p_idx, len);
 780                 ix = curp->p_idx;
 781         }
 782
 783         ix->ei_block = cpu_to_le32(logical);
 784         ext4_idx_store_pblock(ix, ptr);
 785         le16_add_cpu(&curp->p_hdr->eh_entries, 1);
 786
 787         if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
 788                 EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
 789                 return -EIO;
 790         }
 791
 792         err = ext4_ext_dirty(handle, inode, curp);
 793         ext4_std_error(inode->i_sb, err);
 794
 795         return err;
 796 }
 797
 798 /*
 799  * ext4_ext_split:
 800  * inserts new subtree into the path, using free index entry
 801  * at depth @at:
 802  * - allocates all needed blocks (new leaf and all intermediate index blocks)
 803  * - makes decision where to split
 804  * - moves remaining extents and index entries (right to the split point)
 805  *   into the newly allocated blocks
 806  * - initializes subtree
 807  */
 808 static int ext4_ext_split(handle_t *handle, struct inode *inode,
 809                           unsigned int flags,
 810                           struct ext4_ext_path *path,
 811                           struct ext4_extent *newext, int at)
 812 {
 813         struct buffer_head *bh = NULL;
 814         int depth = ext_depth(inode);
 815         struct ext4_extent_header *neh;
 816         struct ext4_extent_idx *fidx;
 817         int i = at, k, m, a;
 818         ext4_fsblk_t newblock, oldblock;
 819         __le32 border;
 820         ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
 821         int err = 0;
 822
 823         /* make decision: where to split? */
 824         /* FIXME: now decision is simplest: at current extent */
 825
 826         /* if current leaf will be split, then we should use
 827          * border from split point */
 828         if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
 829                 EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
 830                 return -EIO;
 831         }
 832         if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
 833                 border = path[depth].p_ext[1].ee_block;
 834                 ext_debug("leaf will be split."
 835                                 " next leaf starts at %d\n",
 836                                   le32_to_cpu(border));
 837         } else {
 838                 border = newext->ee_block;
 839                 ext_debug("leaf will be added."
 840                                 " next leaf starts at %d\n",
 841                                 le32_to_cpu(border));
 842         }
 843
 844         /*
 845          * If error occurs, then we break processing
 846          * and mark filesystem read-only. index won't
 847          * be inserted and tree will be in consistent
 848          * state. Next mount will repair buffers too.
 849          */
 850
 851         /*
 852          * Get array to track all allocated blocks.
 853          * We need this to handle errors and free blocks
 854          * upon them.
 855          */
 856         ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
 857         if (!ablocks)
 858                 return -ENOMEM;
 859
 860         /* allocate all needed blocks */
 861         ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
 862         for (a = 0; a < depth - at; a++) {
 863                 newblock = ext4_ext_new_meta_block(handle, inode, path,
 864                                                    newext, &err, flags);
 865                 if (newblock == 0)
 866                         goto cleanup;
 867                 ablocks[a] = newblock;
 868         }
 869
 870         /* initialize new leaf */
 871         newblock = ablocks[--a];
 872         if (unlikely(newblock == 0)) {
 873                 EXT4_ERROR_INODE(inode, "newblock == 0!");
 874                 err = -EIO;
 875                 goto cleanup;
 876         }
 877         bh = sb_getblk(inode->i_sb, newblock);
 878         if (!bh) {
 879                 err = -EIO;
 880                 goto cleanup;
 881         }
 882         lock_buffer(bh);
 883
 884         err = ext4_journal_get_create_access(handle, bh);
 885         if (err)
 886                 goto cleanup;
 887
 888         neh = ext_block_hdr(bh);
 889         neh->eh_entries = 0;
 890         neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
 891         neh->eh_magic = EXT4_EXT_MAGIC;
 892         neh->eh_depth = 0;
 893
 894         /* move remainder of path[depth] to the new leaf */
 895         if (unlikely(path[depth].p_hdr->eh_entries !=
 896                      path[depth].p_hdr->eh_max)) {
 897                 EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
 898                                  path[depth].p_hdr->eh_entries,
 899                                  path[depth].p_hdr->eh_max);
 900                 err = -EIO;
 901                 goto cleanup;
 902         }
 903         /* start copy from next extent */
 904         m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++;
 905         ext4_ext_show_move(inode, path, newblock, depth);
 906         if (m) {
 907                 struct ext4_extent *ex;
 908                 ex = EXT_FIRST_EXTENT(neh);
 909                 memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m);
 910                 le16_add_cpu(&neh->eh_entries, m);
 911         }
 912
 913         set_buffer_uptodate(bh);
 914         unlock_buffer(bh);
 915
 916         err = ext4_handle_dirty_metadata(handle, inode, bh);
 917         if (err)
 918                 goto cleanup;
 919         brelse(bh);
 920         bh = NULL;
 921
 922         /* correct old leaf */
 923         if (m) {
 924                 err = ext4_ext_get_access(handle, inode, path + depth);
 925                 if (err)
 926                         goto cleanup;
 927                 le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
 928                 err = ext4_ext_dirty(handle, inode, path + depth);
 929                 if (err)
 930                         goto cleanup;
 931
 932         }
 933
 934         /* create intermediate indexes */
 935         k = depth - at - 1;
 936         if (unlikely(k < 0)) {
 937                 EXT4_ERROR_INODE(inode, "k %d < 0!", k);
 938                 err = -EIO;
 939                 goto cleanup;
 940         }
 941         if (k)
 942                 ext_debug("create %d intermediate indices\n", k);
 943         /* insert new index into current index block */
 944         /* current depth stored in i var */
 945         i = depth - 1;
 946         while (k--) {
 947                 oldblock = newblock;
 948                 newblock = ablocks[--a];
 949                 bh = sb_getblk(inode->i_sb, newblock);
 950                 if (!bh) {
 951                         err = -EIO;
 952                         goto cleanup;
 953                 }
 954                 lock_buffer(bh);
 955
 956                 err = ext4_journal_get_create_access(handle, bh);
 957                 if (err)
 958                         goto cleanup;
 959
 960                 neh = ext_block_hdr(bh);
 961                 neh->eh_entries = cpu_to_le16(1);
 962                 neh->eh_magic = EXT4_EXT_MAGIC;
 963                 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
 964                 neh->eh_depth = cpu_to_le16(depth - i);
 965                 fidx = EXT_FIRST_INDEX(neh);
 966                 fidx->ei_block = border;
 967                 ext4_idx_store_pblock(fidx, oldblock);
 968
 969                 ext_debug("int.index at %d (block %llu): %u -> %llu\n",
 970                                 i, newblock, le32_to_cpu(border), oldblock);
 971
 972                 /* move remainder of path[i] to the new index block */
 973                 if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
 974                                         EXT_LAST_INDEX(path[i].p_hdr))) {
 975                         EXT4_ERROR_INODE(inode,
 976                                          "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
 977                                          le32_to_cpu(path[i].p_ext->ee_block));
 978                         err = -EIO;
 979                         goto cleanup;
 980                 }
 981                 /* start copy indexes */
 982                 m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
 983                 ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
 984                                 EXT_MAX_INDEX(path[i].p_hdr));
 985                 ext4_ext_show_move(inode, path, newblock, i);
 986                 if (m) {
 987                         memmove(++fidx, path[i].p_idx,
 988                                 sizeof(struct ext4_extent_idx) * m);
 989                         le16_add_cpu(&neh->eh_entries, m);
 990                 }
 991                 set_buffer_uptodate(bh);
 992                 unlock_buffer(bh);
 993
 994                 err = ext4_handle_dirty_metadata(handle, inode, bh);
 995                 if (err)
 996                         goto cleanup;
 997                 brelse(bh);
 998                 bh = NULL;
 999
1000                 /* correct old index */
1001                 if (m) {
1002                         err = ext4_ext_get_access(handle, inode, path + i);
1003                         if (err)
1004                                 goto cleanup;
1005                         le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
1006                         err = ext4_ext_dirty(handle, inode, path + i);
1007                         if (err)
1008                                 goto cleanup;
1009                 }
1010
1011                 i--;
1012         }
1013
1014         /* insert new index */
1015         err = ext4_ext_insert_index(handle, inode, path + at,
1016                                     le32_to_cpu(border), newblock);
1017
1018 cleanup:
1019         if (bh) {
1020                 if (buffer_locked(bh))
1021                         unlock_buffer(bh);
1022                 brelse(bh);
1023         }
1024
1025         if (err) {
1026                 /* free all allocated blocks in error case */
1027                 for (i = 0; i < depth; i++) {
1028                         if (!ablocks[i])
1029                                 continue;
1030                         ext4_free_blocks(handle, inode, NULL, ablocks[i], 1,
1031                                          EXT4_FREE_BLOCKS_METADATA);
1032                 }
1033         }
1034         kfree(ablocks);
1035
1036         return err;
1037 }
1038
1039 /*
1040  * ext4_ext_grow_indepth:
1041  * implements tree growing procedure:
1042  * - allocates new block
1043  * - moves top-level data (index block or leaf) into the new block
1044  * - initializes new top-level, creating index that points to the
1045  *   just created block
1046  */
1047 static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
1048                                  unsigned int flags,
1049                                  struct ext4_ext_path *path,
1050                                  struct ext4_extent *newext)
1051 {
1052         struct ext4_ext_path *curp = path;
1053         struct ext4_extent_header *neh;
1054         struct buffer_head *bh;
1055         ext4_fsblk_t newblock;
1056         int err = 0;
1057
1058         newblock = ext4_ext_new_meta_block(handle, inode, path,
1059                 newext, &err, flags);
1060         if (newblock == 0)
1061                 return err;
1062
1063         bh = sb_getblk(inode->i_sb, newblock);
1064         if (!bh) {
1065                 err = -EIO;
1066                 ext4_std_error(inode->i_sb, err);
1067                 return err;
1068         }
1069         lock_buffer(bh);
1070
1071         err = ext4_journal_get_create_access(handle, bh);
1072         if (err) {
1073                 unlock_buffer(bh);
1074                 goto out;
1075         }
1076
1077         /* move top-level index/leaf into new block */
1078         memmove(bh->b_data, curp->p_hdr, sizeof(EXT4_I(inode)->i_data));
1079
1080         /* set size of new block */
1081         neh = ext_block_hdr(bh);
1082         /* old root could have indexes or leaves
1083          * so calculate e_max right way */
1084         if (ext_depth(inode))
1085                 neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
1086         else
1087                 neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
1088         neh->eh_magic = EXT4_EXT_MAGIC;
1089         set_buffer_uptodate(bh);
1090         unlock_buffer(bh);
1091
1092         err = ext4_handle_dirty_metadata(handle, inode, bh);
1093         if (err)
1094                 goto out;
1095
1096         /* create index in new top-level index: num,max,pointer */
1097         err = ext4_ext_get_access(handle, inode, curp);
1098         if (err)
1099                 goto out;
1100
1101         curp->p_hdr->eh_magic = EXT4_EXT_MAGIC;
1102         curp->p_hdr->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
1103         curp->p_hdr->eh_entries = cpu_to_le16(1);
1104         curp->p_idx = EXT_FIRST_INDEX(curp->p_hdr);
1105
1106         if (path[0].p_hdr->eh_depth)
1107                 curp->p_idx->ei_block =
1108                         EXT_FIRST_INDEX(path[0].p_hdr)->ei_block;
1109         else
1110                 curp->p_idx->ei_block =
1111                         EXT_FIRST_EXTENT(path[0].p_hdr)->ee_block;
1112         ext4_idx_store_pblock(curp->p_idx, newblock);
1113
1114         neh = ext_inode_hdr(inode);
1115         ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
1116                   le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
1117                   le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
1118                   ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
1119
1120         neh->eh_depth = cpu_to_le16(path->p_depth + 1);
1121         err = ext4_ext_dirty(handle, inode, curp);
1122 out:
1123         brelse(bh);
1124
1125         return err;
1126 }
1127
1128 /*
1129  * ext4_ext_create_new_leaf:
1130  * finds empty index and adds new leaf.
1131  * if no free index is found, then it requests in-depth growing.
1132  */
1133 static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
1134                                     unsigned int flags,
1135                                     struct ext4_ext_path *path,
1136                                     struct ext4_extent *newext)
1137 {
1138         struct ext4_ext_path *curp;
1139         int depth, i, err = 0;
1140
1141 repeat:
1142         i = depth = ext_depth(inode);
1143
1144         /* walk up to the tree and look for free index entry */
1145         curp = path + depth;
1146         while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
1147                 i--;
1148                 curp--;
1149         }
1150
1151         /* we use already allocated block for index block,
1152          * so subsequent data blocks should be contiguous */
1153         if (EXT_HAS_FREE_INDEX(curp)) {
1154                 /* if we found index with free entry, then use that
1155                  * entry: create all needed subtree and add new leaf */
1156                 err = ext4_ext_split(handle, inode, flags, path, newext, i);
1157                 if (err)
1158                         goto out;
1159
1160                 /* refill path */
1161                 ext4_ext_drop_refs(path);
1162                 path = ext4_ext_find_extent(inode,
1163                                     (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1164                                     path);
1165                 if (IS_ERR(path))
1166                         err = PTR_ERR(path);
1167         } else {
1168                 /* tree is full, time to grow in depth */
1169                 err = ext4_ext_grow_indepth(handle, inode, flags,
1170                                             path, newext);
1171                 if (err)
1172                         goto out;
1173
1174                 /* refill path */
1175                 ext4_ext_drop_refs(path);
1176                 path = ext4_ext_find_extent(inode,
1177                                    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1178                                     path);
1179                 if (IS_ERR(path)) {
1180                         err = PTR_ERR(path);
1181                         goto out;
1182                 }
1183
1184                 /*
1185                  * only first (depth 0 -> 1) produces free space;
1186                  * in all other cases we have to split the grown tree
1187                  */
1188                 depth = ext_depth(inode);
1189                 if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
1190                         /* now we need to split */
1191                         goto repeat;
1192                 }
1193         }
1194
1195 out:
1196         return err;
1197 }
1198
1199 /*
1200  * search the closest allocated block to the left for *logical
1201  * and returns it at @logical + it's physical address at @phys
1202  * if *logical is the smallest allocated block, the function
1203  * returns 0 at @phys
1204  * return value contains 0 (success) or error code
1205  */
1206 static int ext4_ext_search_left(struct inode *inode,
1207                                 struct ext4_ext_path *path,
1208                                 ext4_lblk_t *logical, ext4_fsblk_t *phys)
1209 {
1210         struct ext4_extent_idx *ix;
1211         struct ext4_extent *ex;
1212         int depth, ee_len;
1213
1214         if (unlikely(path == NULL)) {
1215                 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1216                 return -EIO;
1217         }
1218         depth = path->p_depth;
1219         *phys = 0;
1220
1221         if (depth == 0 && path->p_ext == NULL)
1222                 return 0;
1223
1224         /* usually extent in the path covers blocks smaller
1225          * then *logical, but it can be that extent is the
1226          * first one in the file */
1227
1228         ex = path[depth].p_ext;
1229         ee_len = ext4_ext_get_actual_len(ex);
1230         if (*logical < le32_to_cpu(ex->ee_block)) {
1231                 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1232                         EXT4_ERROR_INODE(inode,
1233                                          "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
1234                                          *logical, le32_to_cpu(ex->ee_block));
1235                         return -EIO;
1236                 }
1237                 while (--depth >= 0) {
1238                         ix = path[depth].p_idx;
1239                         if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1240                                 EXT4_ERROR_INODE(inode,
1241                                   "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
1242                                   ix != NULL ? ix->ei_block : 0,
1243                                   EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
1244                                     EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block : 0,
1245                                   depth);
1246                                 return -EIO;
1247                         }
1248                 }
1249                 return 0;
1250         }
1251
1252         if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1253                 EXT4_ERROR_INODE(inode,
1254                                  "logical %d < ee_block %d + ee_len %d!",
1255                                  *logical, le32_to_cpu(ex->ee_block), ee_len);
1256                 return -EIO;
1257         }
1258
1259         *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
1260         *phys = ext4_ext_pblock(ex) + ee_len - 1;
1261         return 0;
1262 }
1263
1264 /*
1265  * search the closest allocated block to the right for *logical
1266  * and returns it at @logical + it's physical address at @phys
1267  * if *logical is the smallest allocated block, the function
1268  * returns 0 at @phys
1269  * return value contains 0 (success) or error code
1270  */
1271 static int ext4_ext_search_right(struct inode *inode,
1272                                  struct ext4_ext_path *path,
1273                                  ext4_lblk_t *logical, ext4_fsblk_t *phys)
1274 {
1275         struct buffer_head *bh = NULL;
1276         struct ext4_extent_header *eh;
1277         struct ext4_extent_idx *ix;
1278         struct ext4_extent *ex;
1279         ext4_fsblk_t block;
1280         int depth;      /* Note, NOT eh_depth; depth from top of tree */
1281         int ee_len;
1282
1283         if (unlikely(path == NULL)) {
1284                 EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1285                 return -EIO;
1286         }
1287         depth = path->p_depth;
1288         *phys = 0;
1289
1290         if (depth == 0 && path->p_ext == NULL)
1291                 return 0;
1292
1293         /* usually extent in the path covers blocks smaller
1294          * then *logical, but it can be that extent is the
1295          * first one in the file */
1296
1297         ex = path[depth].p_ext;
1298         ee_len = ext4_ext_get_actual_len(ex);
1299         if (*logical < le32_to_cpu(ex->ee_block)) {
1300                 if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1301                         EXT4_ERROR_INODE(inode,
1302                                          "first_extent(path[%d].p_hdr) != ex",
1303                                          depth);
1304                         return -EIO;
1305                 }
1306                 while (--depth >= 0) {
1307                         ix = path[depth].p_idx;
1308                         if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1309                                 EXT4_ERROR_INODE(inode,
1310                                                  "ix != EXT_FIRST_INDEX *logical %d!",
1311                                                  *logical);
1312                                 return -EIO;
1313                         }
1314                 }
1315                 *logical = le32_to_cpu(ex->ee_block);
1316                 *phys = ext4_ext_pblock(ex);
1317                 return 0;
1318         }
1319
1320         if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1321                 EXT4_ERROR_INODE(inode,
1322                                  "logical %d < ee_block %d + ee_len %d!",
1323                                  *logical, le32_to_cpu(ex->ee_block), ee_len);
1324                 return -EIO;
1325         }
1326
1327         if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
1328                 /* next allocated block in this leaf */
1329                 ex++;
1330                 *logical = le32_to_cpu(ex->ee_block);
1331                 *phys = ext4_ext_pblock(ex);
1332                 return 0;
1333         }
1334
1335         /* go up and search for index to the right */
1336         while (--depth >= 0) {
1337                 ix = path[depth].p_idx;
1338                 if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
1339                         goto got_index;
1340         }
1341
1342         /* we've gone up to the root and found no index to the right */
1343         return 0;
1344
1345 got_index:
1346         /* we've found index to the right, let's
1347          * follow it and find the closest allocated
1348          * block to the right */
1349         ix++;
1350         block = ext4_idx_pblock(ix);
1351         while (++depth < path->p_depth) {
1352                 bh = sb_bread(inode->i_sb, block);
1353                 if (bh == NULL)
1354                         return -EIO;
1355                 eh = ext_block_hdr(bh);
1356                 /* subtract from p_depth to get proper eh_depth */
1357                 if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
1358                         put_bh(bh);
1359                         return -EIO;
1360                 }
1361                 ix = EXT_FIRST_INDEX(eh);
1362                 block = ext4_idx_pblock(ix);
1363                 put_bh(bh);
1364         }
1365
1366         bh = sb_bread(inode->i_sb, block);
1367         if (bh == NULL)
1368                 return -EIO;
1369         eh = ext_block_hdr(bh);
1370         if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
1371                 put_bh(bh);
1372                 return -EIO;
1373         }
1374         ex = EXT_FIRST_EXTENT(eh);
1375         *logical = le32_to_cpu(ex->ee_block);
1376         *phys = ext4_ext_pblock(ex);
1377         put_bh(bh);
1378         return 0;
1379 }
1380
1381 /*
1382  * ext4_ext_next_allocated_block:
1383  * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
1384  * NOTE: it considers block number from index entry as
1385  * allocated block. Thus, index entries have to be consistent
1386  * with leaves.
1387  */
1388 static ext4_lblk_t
1389 ext4_ext_next_allocated_block(struct ext4_ext_path *path)
1390 {
1391         int depth;
1392
1393         BUG_ON(path == NULL);
1394         depth = path->p_depth;
1395
1396         if (depth == 0 && path->p_ext == NULL)
1397                 return EXT_MAX_BLOCKS;
1398
1399         while (depth >= 0) {
1400                 if (depth == path->p_depth) {
1401                         /* leaf */
1402                         if (path[depth].p_ext !=
1403                                         EXT_LAST_EXTENT(path[depth].p_hdr))
1404                           return le32_to_cpu(path[depth].p_ext[1].ee_block);
1405                 } else {
1406                         /* index */
1407                         if (path[depth].p_idx !=
1408                                         EXT_LAST_INDEX(path[depth].p_hdr))
1409                           return le32_to_cpu(path[depth].p_idx[1].ei_block);
1410                 }
1411                 depth--;
1412         }
1413
1414         return EXT_MAX_BLOCKS;
1415 }
1416
1417 /*
1418  * ext4_ext_next_leaf_block:
1419  * returns first allocated block from next leaf or EXT_MAX_BLOCKS
1420  */
1421 static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path)
1422 {
1423         int depth;
1424
1425         BUG_ON(path == NULL);
1426         depth = path->p_depth;
1427
1428         /* zero-tree has no leaf blocks at all */
1429         if (depth == 0)
1430                 return EXT_MAX_BLOCKS;
1431
1432         /* go to index block */
1433         depth--;
1434
1435         while (depth >= 0) {
1436                 if (path[depth].p_idx !=
1437                                 EXT_LAST_INDEX(path[depth].p_hdr))
1438                         return (ext4_lblk_t)
1439                                 le32_to_cpu(path[depth].p_idx[1].ei_block);
1440                 depth--;
1441         }
1442
1443         return EXT_MAX_BLOCKS;
1444 }
1445
1446 /*
1447  * ext4_ext_correct_indexes:
1448  * if leaf gets modified and modified extent is first in the leaf,
1449  * then we have to correct all indexes above.
1450  * TODO: do we need to correct tree in all cases?
1451  */
1452 static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
1453                                 struct ext4_ext_path *path)
1454 {
1455         struct ext4_extent_header *eh;
1456         int depth = ext_depth(inode);
1457         struct ext4_extent *ex;
1458         __le32 border;
1459         int k, err = 0;
1460
1461         eh = path[depth].p_hdr;
1462         ex = path[depth].p_ext;
1463
1464         if (unlikely(ex == NULL || eh == NULL)) {
1465                 EXT4_ERROR_INODE(inode,
1466                                  "ex %p == NULL or eh %p == NULL", ex, eh);
1467                 return -EIO;
1468         }
1469
1470         if (depth == 0) {
1471                 /* there is no tree at all */
1472                 return 0;
1473         }
1474
1475         if (ex != EXT_FIRST_EXTENT(eh)) {
1476                 /* we correct tree if first leaf got modified only */
1477                 return 0;
1478         }
1479
1480         /*
1481          * TODO: we need correction if border is smaller than current one
1482          */
1483         k = depth - 1;
1484         border = path[depth].p_ext->ee_block;
1485         err = ext4_ext_get_access(handle, inode, path + k);
1486         if (err)
1487                 return err;
1488         path[k].p_idx->ei_block = border;
1489         err = ext4_ext_dirty(handle, inode, path + k);
1490         if (err)
1491                 return err;
1492
1493         while (k--) {
1494                 /* change all left-side indexes */
1495                 if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
1496                         break;
1497                 err = ext4_ext_get_access(handle, inode, path + k);
1498                 if (err)
1499                         break;
1500                 path[k].p_idx->ei_block = border;
1501                 err = ext4_ext_dirty(handle, inode, path + k);
1502                 if (err)
1503                         break;
1504         }
1505
1506         return err;
1507 }
1508
1509 int
1510 ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
1511                                 struct ext4_extent *ex2)
1512 {
1513         unsigned short ext1_ee_len, ext2_ee_len, max_len;
1514
1515         /*
1516          * Make sure that either both extents are uninitialized, or
1517          * both are _not_.
1518          */
1519         if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
1520                 return 0;
1521
1522         if (ext4_ext_is_uninitialized(ex1))
1523                 max_len = EXT_UNINIT_MAX_LEN;
1524         else
1525                 max_len = EXT_INIT_MAX_LEN;
1526
1527         ext1_ee_len = ext4_ext_get_actual_len(ex1);
1528         ext2_ee_len = ext4_ext_get_actual_len(ex2);
1529
1530         if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
1531                         le32_to_cpu(ex2->ee_block))
1532                 return 0;
1533
1534         /*
1535          * To allow future support for preallocated extents to be added
1536          * as an RO_COMPAT feature, refuse to merge to extents if
1537          * this can result in the top bit of ee_len being set.
1538          */
1539         if (ext1_ee_len + ext2_ee_len > max_len)
1540                 return 0;
1541 #ifdef AGGRESSIVE_TEST
1542         if (ext1_ee_len >= 4)
1543                 return 0;
1544 #endif
1545
1546         if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
1547                 return 1;
1548         return 0;
1549 }
1550
1551 /*
1552  * This function tries to merge the "ex" extent to the next extent in the tree.
1553  * It always tries to merge towards right. If you want to merge towards
1554  * left, pass "ex - 1" as argument instead of "ex".
1555  * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
1556  * 1 if they got merged.
1557  */
1558 static int ext4_ext_try_to_merge_right(struct inode *inode,
1559                                  struct ext4_ext_path *path,
1560                                  struct ext4_extent *ex)
1561 {
1562         struct ext4_extent_header *eh;
1563         unsigned int depth, len;
1564         int merge_done = 0;
1565         int uninitialized = 0;
1566
1567         depth = ext_depth(inode);
1568         BUG_ON(path[depth].p_hdr == NULL);
1569         eh = path[depth].p_hdr;
1570
1571         while (ex < EXT_LAST_EXTENT(eh)) {
1572                 if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
1573                         break;
1574                 /* merge with next extent! */
1575                 if (ext4_ext_is_uninitialized(ex))
1576                         uninitialized = 1;
1577                 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1578                                 + ext4_ext_get_actual_len(ex + 1));
1579                 if (uninitialized)
1580                         ext4_ext_mark_uninitialized(ex);
1581
1582                 if (ex + 1 < EXT_LAST_EXTENT(eh)) {
1583                         len = (EXT_LAST_EXTENT(eh) - ex - 1)
1584                                 * sizeof(struct ext4_extent);
1585                         memmove(ex + 1, ex + 2, len);
1586                 }
1587                 le16_add_cpu(&eh->eh_entries, -1);
1588                 merge_done = 1;
1589                 WARN_ON(eh->eh_entries == 0);
1590                 if (!eh->eh_entries)
1591                         EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
1592         }
1593
1594         return merge_done;
1595 }
1596
1597 /*
1598  * This function tries to merge the @ex extent to neighbours in the tree.
1599  * return 1 if merge left else 0.
1600  */
1601 static int ext4_ext_try_to_merge(struct inode *inode,
1602                                   struct ext4_ext_path *path,
1603                                   struct ext4_extent *ex) {
1604         struct ext4_extent_header *eh;
1605         unsigned int depth;
1606         int merge_done = 0;
1607         int ret = 0;
1608
1609         depth = ext_depth(inode);
1610         BUG_ON(path[depth].p_hdr == NULL);
1611         eh = path[depth].p_hdr;
1612
1613         if (ex > EXT_FIRST_EXTENT(eh))
1614                 merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1);
1615
1616         if (!merge_done)
1617                 ret = ext4_ext_try_to_merge_right(inode, path, ex);
1618
1619         return ret;
1620 }
1621
1622 /*
1623  * check if a portion of the "newext" extent overlaps with an
1624  * existing extent.
1625  *
1626  * If there is an overlap discovered, it updates the length of the newext
1627  * such that there will be no overlap, and then returns 1.
1628  * If there is no overlap found, it returns 0.
1629  */
1630 static unsigned int ext4_ext_check_overlap(struct inode *inode,
1631                                            struct ext4_extent *newext,
1632                                            struct ext4_ext_path *path)
1633 {
1634         ext4_lblk_t b1, b2;
1635         unsigned int depth, len1;
1636         unsigned int ret = 0;
1637
1638         b1 = le32_to_cpu(newext->ee_block);
1639         len1 = ext4_ext_get_actual_len(newext);
1640         depth = ext_depth(inode);
1641         if (!path[depth].p_ext)
1642                 goto out;
1643         b2 = le32_to_cpu(path[depth].p_ext->ee_block);
1644
1645         /*
1646          * get the next allocated block if the extent in the path
1647          * is before the requested block(s)
1648          */
1649         if (b2 < b1) {
1650                 b2 = ext4_ext_next_allocated_block(path);
1651                 if (b2 == EXT_MAX_BLOCKS)
1652                         goto out;
1653         }
1654
1655         /* check for wrap through zero on extent logical start block*/
1656         if (b1 + len1 < b1) {
1657                 len1 = EXT_MAX_BLOCKS - b1;
1658                 newext->ee_len = cpu_to_le16(len1);
1659                 ret = 1;
1660         }
1661
1662         /* check for overlap */
1663         if (b1 + len1 > b2) {
1664                 newext->ee_len = cpu_to_le16(b2 - b1);
1665                 ret = 1;
1666         }
1667 out:
1668         return ret;
1669 }
1670
1671 /*
1672  * ext4_ext_insert_extent:
1673  * tries to merge requsted extent into the existing extent or
1674  * inserts requested extent as new one into the tree,
1675  * creating new leaf in the no-space case.
1676  */
1677 int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
1678                                 struct ext4_ext_path *path,
1679                                 struct ext4_extent *newext, int flag)
1680 {
1681         struct ext4_extent_header *eh;
1682         struct ext4_extent *ex, *fex;
1683         struct ext4_extent *nearex; /* nearest extent */
1684         struct ext4_ext_path *npath = NULL;
1685         int depth, len, err;
1686         ext4_lblk_t next;
1687         unsigned uninitialized = 0;
1688         int flags = 0;
1689
1690         if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
1691                 EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
1692                 return -EIO;
1693         }
1694         depth = ext_depth(inode);
1695         ex = path[depth].p_ext;
1696         if (unlikely(path[depth].p_hdr == NULL)) {
1697                 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1698                 return -EIO;
1699         }
1700
1701         /* try to insert block into found extent and return */
1702         if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
1703                 && ext4_can_extents_be_merged(inode, ex, newext)) {
1704                 ext_debug("append [%d]%d block to %d:[%d]%d (from %llu)\n",
1705                           ext4_ext_is_uninitialized(newext),
1706                           ext4_ext_get_actual_len(newext),
1707                           le32_to_cpu(ex->ee_block),
1708                           ext4_ext_is_uninitialized(ex),
1709                           ext4_ext_get_actual_len(ex),
1710                           ext4_ext_pblock(ex));
1711                 err = ext4_ext_get_access(handle, inode, path + depth);
1712                 if (err)
1713                         return err;
1714
1715                 /*
1716                  * ext4_can_extents_be_merged should have checked that either
1717                  * both extents are uninitialized, or both aren't. Thus we
1718                  * need to check only one of them here.
1719                  */
1720                 if (ext4_ext_is_uninitialized(ex))
1721                         uninitialized = 1;
1722                 ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1723                                         + ext4_ext_get_actual_len(newext));
1724                 if (uninitialized)
1725                         ext4_ext_mark_uninitialized(ex);
1726                 eh = path[depth].p_hdr;
1727                 nearex = ex;
1728                 goto merge;
1729         }
1730
1731         depth = ext_depth(inode);
1732         eh = path[depth].p_hdr;
1733         if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
1734                 goto has_space;
1735
1736         /* probably next leaf has space for us? */
1737         fex = EXT_LAST_EXTENT(eh);
1738         next = EXT_MAX_BLOCKS;
1739         if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block))
1740                 next = ext4_ext_next_leaf_block(path);
1741         if (next != EXT_MAX_BLOCKS) {
1742                 ext_debug("next leaf block - %d\n", next);
1743                 BUG_ON(npath != NULL);
1744                 npath = ext4_ext_find_extent(inode, next, NULL);
1745                 if (IS_ERR(npath))
1746                         return PTR_ERR(npath);
1747                 BUG_ON(npath->p_depth != path->p_depth);
1748                 eh = npath[depth].p_hdr;
1749                 if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
1750                         ext_debug("next leaf isn't full(%d)\n",
1751                                   le16_to_cpu(eh->eh_entries));
1752                         path = npath;
1753                         goto has_space;
1754                 }
1755                 ext_debug("next leaf has no free space(%d,%d)\n",
1756                           le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
1757         }
1758
1759         /*
1760          * There is no free space in the found leaf.
1761          * We're gonna add a new leaf in the tree.
1762          */
1763         if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT)
1764                 flags = EXT4_MB_USE_ROOT_BLOCKS;
1765         err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
1766         if (err)
1767                 goto cleanup;
1768         depth = ext_depth(inode);
1769         eh = path[depth].p_hdr;
1770
1771 has_space:
1772         nearex = path[depth].p_ext;
1773
1774         err = ext4_ext_get_access(handle, inode, path + depth);
1775         if (err)
1776                 goto cleanup;
1777
1778         if (!nearex) {
1779                 /* there is no extent in this leaf, create first one */
1780                 ext_debug("first extent in the leaf: %d:%llu:[%d]%d\n",
1781                                 le32_to_cpu(newext->ee_block),
1782                                 ext4_ext_pblock(newext),
1783                                 ext4_ext_is_uninitialized(newext),
1784                                 ext4_ext_get_actual_len(newext));
1785                 path[depth].p_ext = EXT_FIRST_EXTENT(eh);
1786         } else if (le32_to_cpu(newext->ee_block)
1787                            > le32_to_cpu(nearex->ee_block)) {
1788 /*              BUG_ON(newext->ee_block == nearex->ee_block); */
1789                 if (nearex != EXT_LAST_EXTENT(eh)) {
1790                         len = EXT_MAX_EXTENT(eh) - nearex;
1791                         len = (len - 1) * sizeof(struct ext4_extent);
1792                         len = len < 0 ? 0 : len;
1793                         ext_debug("insert %d:%llu:[%d]%d after: nearest 0x%p, "
1794                                         "move %d from 0x%p to 0x%p\n",
1795                                         le32_to_cpu(newext->ee_block),
1796                                         ext4_ext_pblock(newext),
1797                                         ext4_ext_is_uninitialized(newext),
1798                                         ext4_ext_get_actual_len(newext),
1799                                         nearex, len, nearex + 1, nearex + 2);
1800                         memmove(nearex + 2, nearex + 1, len);
1801                 }
1802                 path[depth].p_ext = nearex + 1;
1803         } else {
1804                 BUG_ON(newext->ee_block == nearex->ee_block);
1805                 len = (EXT_MAX_EXTENT(eh) - nearex) * sizeof(struct ext4_extent);
1806                 len = len < 0 ? 0 : len;
1807                 ext_debug("insert %d:%llu:[%d]%d before: nearest 0x%p, "
1808                                 "move %d from 0x%p to 0x%p\n",
1809                                 le32_to_cpu(newext->ee_block),
1810                                 ext4_ext_pblock(newext),
1811                                 ext4_ext_is_uninitialized(newext),
1812                                 ext4_ext_get_actual_len(newext),
1813                                 nearex, len, nearex, nearex + 1);
1814                 memmove(nearex + 1, nearex, len);
1815                 path[depth].p_ext = nearex;
1816         }
1817
1818         le16_add_cpu(&eh->eh_entries, 1);
1819         nearex = path[depth].p_ext;
1820         nearex->ee_block = newext->ee_block;
1821         ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
1822         nearex->ee_len = newext->ee_len;
1823
1824 merge:
1825         /* try to merge extents to the right */
1826         if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
1827                 ext4_ext_try_to_merge(inode, path, nearex);
1828
1829         /* try to merge extents to the left */
1830
1831         /* time to correct all indexes above */
1832         err = ext4_ext_correct_indexes(handle, inode, path);
1833         if (err)
1834                 goto cleanup;
1835
1836         err = ext4_ext_dirty(handle, inode, path + depth);
1837
1838 cleanup:
1839         if (npath) {
1840                 ext4_ext_drop_refs(npath);
1841                 kfree(npath);
1842         }
1843         ext4_ext_invalidate_cache(inode);
1844         return err;
1845 }
1846
1847 static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
1848                                ext4_lblk_t num, ext_prepare_callback func,
1849                                void *cbdata)
1850 {
1851         struct ext4_ext_path *path = NULL;
1852         struct ext4_ext_cache cbex;
1853         struct ext4_extent *ex;
1854         ext4_lblk_t next, start = 0, end = 0;
1855         ext4_lblk_t last = block + num;
1856         int depth, exists, err = 0;
1857
1858         BUG_ON(func == NULL);
1859         BUG_ON(inode == NULL);
1860
1861         while (block < last && block != EXT_MAX_BLOCKS) {
1862                 num = last - block;
1863                 /* find extent for this block */
1864                 down_read(&EXT4_I(inode)->i_data_sem);
1865                 path = ext4_ext_find_extent(inode, block, path);
1866                 up_read(&EXT4_I(inode)->i_data_sem);
1867                 if (IS_ERR(path)) {
1868                         err = PTR_ERR(path);
1869                         path = NULL;
1870                         break;
1871                 }
1872
1873                 depth = ext_depth(inode);
1874                 if (unlikely(path[depth].p_hdr == NULL)) {
1875                         EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1876                         err = -EIO;
1877                         break;
1878                 }
1879                 ex = path[depth].p_ext;
1880                 next = ext4_ext_next_allocated_block(path);
1881
1882                 exists = 0;
1883                 if (!ex) {
1884                         /* there is no extent yet, so try to allocate
1885                          * all requested space */
1886                         start = block;
1887                         end = block + num;
1888                 } else if (le32_to_cpu(ex->ee_block) > block) {
1889                         /* need to allocate space before found extent */
1890                         start = block;
1891                         end = le32_to_cpu(ex->ee_block);
1892                         if (block + num < end)
1893                                 end = block + num;
1894                 } else if (block >= le32_to_cpu(ex->ee_block)
1895                                         + ext4_ext_get_actual_len(ex)) {
1896                         /* need to allocate space after found extent */
1897                         start = block;
1898                         end = block + num;
1899                         if (end >= next)
1900                                 end = next;
1901                 } else if (block >= le32_to_cpu(ex->ee_block)) {
1902                         /*
1903                          * some part of requested space is covered
1904                          * by found extent
1905                          */
1906                         start = block;
1907                         end = le32_to_cpu(ex->ee_block)
1908                                 + ext4_ext_get_actual_len(ex);
1909                         if (block + num < end)
1910                                 end = block + num;
1911                         exists = 1;
1912                 } else {
1913                         BUG();
1914                 }
1915                 BUG_ON(end <= start);
1916
1917                 if (!exists) {
1918                         cbex.ec_block = start;
1919                         cbex.ec_len = end - start;
1920                         cbex.ec_start = 0;
1921                 } else {
1922                         cbex.ec_block = le32_to_cpu(ex->ee_block);
1923                         cbex.ec_len = ext4_ext_get_actual_len(ex);
1924                         cbex.ec_start = ext4_ext_pblock(ex);
1925                 }
1926
1927                 if (unlikely(cbex.ec_len == 0)) {
1928                         EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
1929                         err = -EIO;
1930                         break;
1931                 }
1932                 err = func(inode, next, &cbex, ex, cbdata);
1933                 ext4_ext_drop_refs(path);
1934
1935                 if (err < 0)
1936                         break;
1937
1938                 if (err == EXT_REPEAT)
1939                         continue;
1940                 else if (err == EXT_BREAK) {
1941                         err = 0;
1942                         break;
1943                 }
1944
1945                 if (ext_depth(inode) != depth) {
1946                         /* depth was changed. we have to realloc path */
1947                         kfree(path);
1948                         path = NULL;
1949                 }
1950
1951                 block = cbex.ec_block + cbex.ec_len;
1952         }
1953
1954         if (path) {
1955                 ext4_ext_drop_refs(path);
1956                 kfree(path);
1957         }
1958
1959         return err;
1960 }
1961
1962 static void
1963 ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
1964                         __u32 len, ext4_fsblk_t start)
1965 {
1966         struct ext4_ext_cache *cex;
1967         BUG_ON(len == 0);
1968         spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
1969         cex = &EXT4_I(inode)->i_cached_extent;
1970         cex->ec_block = block;
1971         cex->ec_len = len;
1972         cex->ec_start = start;
1973         spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
1974 }
1975
1976 /*
1977  * ext4_ext_put_gap_in_cache:
1978  * calculate boundaries of the gap that the requested block fits into
1979  * and cache this gap
1980  */
1981 static void
1982 ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
1983                                 ext4_lblk_t block)
1984 {
1985         int depth = ext_depth(inode);
1986         unsigned long len;
1987         ext4_lblk_t lblock;
1988         struct ext4_extent *ex;
1989
1990         ex = path[depth].p_ext;
1991         if (ex == NULL) {
1992                 /* there is no extent yet, so gap is [0;-] */
1993                 lblock = 0;
1994                 len = EXT_MAX_BLOCKS;
1995                 ext_debug("cache gap(whole file):");
1996         } else if (block < le32_to_cpu(ex->ee_block)) {
1997                 lblock = block;
1998                 len = le32_to_cpu(ex->ee_block) - block;
1999                 ext_debug("cache gap(before): %u [%u:%u]",
2000                                 block,
2001                                 le32_to_cpu(ex->ee_block),
2002                                  ext4_ext_get_actual_len(ex));
2003         } else if (block >= le32_to_cpu(ex->ee_block)
2004                         + ext4_ext_get_actual_len(ex)) {
2005                 ext4_lblk_t next;
2006                 lblock = le32_to_cpu(ex->ee_block)
2007                         + ext4_ext_get_actual_len(ex);
2008
2009                 next = ext4_ext_next_allocated_block(path);
2010                 ext_debug("cache gap(after): [%u:%u] %u",
2011                                 le32_to_cpu(ex->ee_block),
2012                                 ext4_ext_get_actual_len(ex),
2013                                 block);
2014                 BUG_ON(next == lblock);
2015                 len = next - lblock;
2016         } else {
2017                 lblock = len = 0;
2018                 BUG();
2019         }
2020
2021         ext_debug(" -> %u:%lu\n", lblock, len);
2022         ext4_ext_put_in_cache(inode, lblock, len, 0);
2023 }
2024
2025 /*
2026  * ext4_ext_check_cache()
2027  * Checks to see if the given block is in the cache.
2028  * If it is, the cached extent is stored in the given
2029  * cache extent pointer.  If the cached extent is a hole,
2030  * this routine should be used instead of
2031  * ext4_ext_in_cache if the calling function needs to
2032  * know the size of the hole.
2033  *
2034  * @inode: The files inode
2035  * @block: The block to look for in the cache
2036  * @ex:    Pointer where the cached extent will be stored
2037  *         if it contains block
2038  *
2039  * Return 0 if cache is invalid; 1 if the cache is valid
2040  */
2041 static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
2042         struct ext4_ext_cache *ex){
2043         struct ext4_ext_cache *cex;
2044         struct ext4_sb_info *sbi;
2045         int ret = 0;
2046
2047         /*
2048          * We borrow i_block_reservation_lock to protect i_cached_extent
2049          */
2050         spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
2051         cex = &EXT4_I(inode)->i_cached_extent;
2052         sbi = EXT4_SB(inode->i_sb);
2053
2054         /* has cache valid data? */
2055         if (cex->ec_len == 0)
2056                 goto errout;
2057
2058         if (in_range(block, cex->ec_block, cex->ec_len)) {
2059                 memcpy(ex, cex, sizeof(struct ext4_ext_cache));
2060                 ext_debug("%u cached by %u:%u:%llu\n",
2061                                 block,
2062                                 cex->ec_block, cex->ec_len, cex->ec_start);
2063                 ret = 1;
2064         }
2065 errout:
2066         if (!ret)
2067                 sbi->extent_cache_misses++;
2068         else
2069                 sbi->extent_cache_hits++;
2070         spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
2071         return ret;
2072 }
2073
2074 /*
2075  * ext4_ext_in_cache()
2076  * Checks to see if the given block is in the cache.
2077  * If it is, the cached extent is stored in the given
2078  * extent pointer.
2079  *
2080  * @inode: The files inode
2081  * @block: The block to look for in the cache
2082  * @ex:    Pointer where the cached extent will be stored
2083  *         if it contains block
2084  *
2085  * Return 0 if cache is invalid; 1 if the cache is valid
2086  */
2087 static int
2088 ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
2089                         struct ext4_extent *ex)
2090 {
2091         struct ext4_ext_cache cex;
2092         int ret = 0;
2093
2094         if (ext4_ext_check_cache(inode, block, &cex)) {
2095                 ex->ee_block = cpu_to_le32(cex.ec_block);
2096                 ext4_ext_store_pblock(ex, cex.ec_start);
2097                 ex->ee_len = cpu_to_le16(cex.ec_len);
2098                 ret = 1;
2099         }
2100
2101         return ret;
2102 }
2103
2104
2105 /*
2106  * ext4_ext_rm_idx:
2107  * removes index from the index block.
2108  */
2109 static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
2110                         struct ext4_ext_path *path)
2111 {
2112         int err;
2113         ext4_fsblk_t leaf;
2114
2115         /* free index block */
2116         path--;
2117         leaf = ext4_idx_pblock(path->p_idx);
2118         if (unlikely(path->p_hdr->eh_entries == 0)) {
2119                 EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
2120                 return -EIO;
2121         }
2122         err = ext4_ext_get_access(handle, inode, path);
2123         if (err)
2124                 return err;
2125
2126         if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) {
2127                 int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx;
2128                 len *= sizeof(struct ext4_extent_idx);
2129                 memmove(path->p_idx, path->p_idx + 1, len);
2130         }
2131
2132         le16_add_cpu(&path->p_hdr->eh_entries, -1);
2133         err = ext4_ext_dirty(handle, inode, path);
2134         if (err)
2135                 return err;
2136         ext_debug("index is empty, remove it, free block %llu\n", leaf);
2137         ext4_free_blocks(handle, inode, NULL, leaf, 1,
2138                          EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
2139         return err;
2140 }
2141
2142 /*
2143  * ext4_ext_calc_credits_for_single_extent:
2144  * This routine returns max. credits that needed to insert an extent
2145  * to the extent tree.
2146  * When pass the actual path, the caller should calculate credits
2147  * under i_data_sem.
2148  */
2149 int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
2150                                                 struct ext4_ext_path *path)
2151 {
2152         if (path) {
2153                 int depth = ext_depth(inode);
2154                 int ret = 0;
2155
2156                 /* probably there is space in leaf? */
2157                 if (le16_to_cpu(path[depth].p_hdr->eh_entries)
2158                                 < le16_to_cpu(path[depth].p_hdr->eh_max)) {
2159
2160                         /*
2161                          *  There are some space in the leaf tree, no
2162                          *  need to account for leaf block credit
2163                          *
2164                          *  bitmaps and block group descriptor blocks
2165                          *  and other metadat blocks still need to be
2166                          *  accounted.
2167                          */
2168                         /* 1 bitmap, 1 block group descriptor */
2169                         ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
2170                         return ret;
2171                 }
2172         }
2173
2174         return ext4_chunk_trans_blocks(inode, nrblocks);
2175 }
2176
2177 /*
2178  * How many index/leaf blocks need to change/allocate to modify nrblocks?
2179  *
2180  * if nrblocks are fit in a single extent (chunk flag is 1), then
2181  * in the worse case, each tree level index/leaf need to be changed
2182  * if the tree split due to insert a new extent, then the old tree
2183  * index/leaf need to be updated too
2184  *
2185  * If the nrblocks are discontiguous, they could cause
2186  * the whole tree split more than once, but this is really rare.
2187  */
2188 int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
2189 {
2190         int index;
2191         int depth = ext_depth(inode);
2192
2193         if (chunk)
2194                 index = depth * 2;
2195         else
2196                 index = depth * 3;
2197
2198         return index;
2199 }
2200
2201 static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
2202                                 struct ext4_extent *ex,
2203                                 ext4_lblk_t from, ext4_lblk_t to)
2204 {
2205         unsigned short ee_len =  ext4_ext_get_actual_len(ex);
2206         int flags = EXT4_FREE_BLOCKS_FORGET;
2207
2208         if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2209                 flags |= EXT4_FREE_BLOCKS_METADATA;
2210 #ifdef EXTENTS_STATS
2211         {
2212                 struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2213                 spin_lock(&sbi->s_ext_stats_lock);
2214                 sbi->s_ext_blocks += ee_len;
2215                 sbi->s_ext_extents++;
2216                 if (ee_len < sbi->s_ext_min)
2217                         sbi->s_ext_min = ee_len;
2218                 if (ee_len > sbi->s_ext_max)
2219                         sbi->s_ext_max = ee_len;
2220                 if (ext_depth(inode) > sbi->s_depth_max)
2221                         sbi->s_depth_max = ext_depth(inode);
2222                 spin_unlock(&sbi->s_ext_stats_lock);
2223         }
2224 #endif
2225         if (from >= le32_to_cpu(ex->ee_block)
2226             && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
2227                 /* tail removal */
2228                 ext4_lblk_t num;
2229                 ext4_fsblk_t start;
2230
2231                 num = le32_to_cpu(ex->ee_block) + ee_len - from;
2232                 start = ext4_ext_pblock(ex) + ee_len - num;
2233                 ext_debug("free last %u blocks starting %llu\n", num, start);
2234                 ext4_free_blocks(handle, inode, NULL, start, num, flags);
2235         } else if (from == le32_to_cpu(ex->ee_block)
2236                    && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
2237                 /* head removal */
2238                 ext4_lblk_t num;
2239                 ext4_fsblk_t start;
2240
2241                 num = to - from;
2242                 start = ext4_ext_pblock(ex);
2243
2244                 ext_debug("free first %u blocks starting %llu\n", num, start);
2245                 ext4_free_blocks(handle, inode, 0, start, num, flags);
2246
2247         } else {
2248                 printk(KERN_INFO "strange request: removal(2) "
2249                                 "%u-%u from %u:%u\n",
2250                                 from, to, le32_to_cpu(ex->ee_block), ee_len);
2251         }
2252         return 0;
2253 }
2254
2255
2256 /*
2257  * ext4_ext_rm_leaf() Removes the extents associated with the
2258  * blocks appearing between "start" and "end", and splits the extents
2259  * if "start" and "end" appear in the same extent
2260  *
2261  * @handle: The journal handle
2262  * @inode:  The files inode
2263  * @path:   The path to the leaf
2264  * @start:  The first block to remove
2265  * @end:   The last block to remove
2266  */
2267 static int
2268 ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
2269                 struct ext4_ext_path *path, ext4_lblk_t start,
2270                 ext4_lblk_t end)
2271 {
2272         int err = 0, correct_index = 0;
2273         int depth = ext_depth(inode), credits;
2274         struct ext4_extent_header *eh;
2275         ext4_lblk_t a, b, block;
2276         unsigned num;
2277         ext4_lblk_t ex_ee_block;
2278         unsigned short ex_ee_len;
2279         unsigned uninitialized = 0;
2280         struct ext4_extent *ex;
2281         struct ext4_map_blocks map;
2282
2283         /* the header must be checked already in ext4_ext_remove_space() */
2284         ext_debug("truncate since %u in leaf\n", start);
2285         if (!path[depth].p_hdr)
2286                 path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
2287         eh = path[depth].p_hdr;
2288         if (unlikely(path[depth].p_hdr == NULL)) {
2289                 EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
2290                 return -EIO;
2291         }
2292         /* find where to start removing */
2293         ex = EXT_LAST_EXTENT(eh);
2294
2295         ex_ee_block = le32_to_cpu(ex->ee_block);
2296         ex_ee_len = ext4_ext_get_actual_len(ex);
2297
2298         while (ex >= EXT_FIRST_EXTENT(eh) &&
2299                         ex_ee_block + ex_ee_len > start) {
2300
2301                 if (ext4_ext_is_uninitialized(ex))
2302                         uninitialized = 1;
2303                 else
2304                         uninitialized = 0;
2305
2306                 ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
2307                          uninitialized, ex_ee_len);
2308                 path[depth].p_ext = ex;
2309
2310                 a = ex_ee_block > start ? ex_ee_block : start;
2311                 b = ex_ee_block+ex_ee_len - 1 < end ?
2312                         ex_ee_block+ex_ee_len - 1 : end;
2313
2314                 ext_debug("  border %u:%u\n", a, b);
2315
2316                 /* If this extent is beyond the end of the hole, skip it */
2317                 if (end <= ex_ee_block) {
2318                         ex--;
2319                         ex_ee_block = le32_to_cpu(ex->ee_block);
2320                         ex_ee_len = ext4_ext_get_actual_len(ex);
2321                         continue;
2322                 } else if (a != ex_ee_block &&
2323                         b != ex_ee_block + ex_ee_len - 1) {
2324                         /*
2325                          * If this is a truncate, then this condition should
2326                          * never happen because at least one of the end points
2327                          * needs to be on the edge of the extent.
2328                          */
2329                         if (end == EXT_MAX_BLOCKS - 1) {
2330                                 ext_debug("  bad truncate %u:%u\n",
2331                                                 start, end);
2332                                 block = 0;
2333                                 num = 0;
2334                                 err = -EIO;
2335                                 goto out;
2336                         }
2337                         /*
2338                          * else this is a hole punch, so the extent needs to
2339                          * be split since neither edge of the hole is on the
2340                          * extent edge
2341                          */
2342                         else{
2343                                 map.m_pblk = ext4_ext_pblock(ex);
2344                                 map.m_lblk = ex_ee_block;
2345                                 map.m_len = b - ex_ee_block;
2346
2347                                 err = ext4_split_extent(handle,
2348                                         inode, path, &map, 0,
2349                                         EXT4_GET_BLOCKS_PUNCH_OUT_EXT |
2350                                         EXT4_GET_BLOCKS_PRE_IO);
2351
2352                                 if (err < 0)
2353                                         goto out;
2354
2355                                 ex_ee_len = ext4_ext_get_actual_len(ex);
2356
2357                                 b = ex_ee_block+ex_ee_len - 1 < end ?
2358                                         ex_ee_block+ex_ee_len - 1 : end;
2359
2360                                 /* Then remove tail of this extent */
2361                                 block = ex_ee_block;
2362                                 num = a - block;
2363                         }
2364                 } else if (a != ex_ee_block) {
2365                         /* remove tail of the extent */
2366                         block = ex_ee_block;
2367                         num = a - block;
2368                 } else if (b != ex_ee_block + ex_ee_len - 1) {
2369                         /* remove head of the extent */
2370                         block = b;
2371                         num =  ex_ee_block + ex_ee_len - b;
2372
2373                         /*
2374                          * If this is a truncate, this condition
2375                          * should never happen
2376                          */
2377                         if (end == EXT_MAX_BLOCKS - 1) {
2378                                 ext_debug("  bad truncate %u:%u\n",
2379                                         start, end);
2380                                 err = -EIO;
2381                                 goto out;
2382                         }
2383                 } else {
2384                         /* remove whole extent: excellent! */
2385                         block = ex_ee_block;
2386                         num = 0;
2387                         if (a != ex_ee_block) {
2388                                 ext_debug("  bad truncate %u:%u\n",
2389                                         start, end);
2390                                 err = -EIO;
2391                                 goto out;
2392                         }
2393
2394                         if (b != ex_ee_block + ex_ee_len - 1) {
2395                                 ext_debug("  bad truncate %u:%u\n",
2396                                         start, end);
2397                                 err = -EIO;
2398                                 goto out;
2399                         }
2400                 }
2401
2402                 /*
2403                  * 3 for leaf, sb, and inode plus 2 (bmap and group
2404                  * descriptor) for each block group; assume two block
2405                  * groups plus ex_ee_len/blocks_per_block_group for
2406                  * the worst case
2407                  */
2408                 credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb));
2409                 if (ex == EXT_FIRST_EXTENT(eh)) {
2410                         correct_index = 1;
2411                         credits += (ext_depth(inode)) + 1;
2412                 }
2413                 credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
2414
2415                 err = ext4_ext_truncate_extend_restart(handle, inode, credits);
2416                 if (err)
2417                         goto out;
2418
2419                 err = ext4_ext_get_access(handle, inode, path + depth);
2420                 if (err)
2421                         goto out;
2422
2423                 err = ext4_remove_blocks(handle, inode, ex, a, b);
2424                 if (err)
2425                         goto out;
2426
2427                 if (num == 0) {
2428                         /* this extent is removed; mark slot entirely unused */
2429                         ext4_ext_store_pblock(ex, 0);
2430                 } else if (block != ex_ee_block) {
2431                         /*
2432                          * If this was a head removal, then we need to update
2433                          * the physical block since it is now at a different
2434                          * location
2435                          */
2436                         ext4_ext_store_pblock(ex, ext4_ext_pblock(ex) + (b-a));
2437                 }
2438
2439                 ex->ee_block = cpu_to_le32(block);
2440                 ex->ee_len = cpu_to_le16(num);
2441                 /*
2442                  * Do not mark uninitialized if all the blocks in the
2443                  * extent have been removed.
2444                  */
2445                 if (uninitialized && num)
2446                         ext4_ext_mark_uninitialized(ex);
2447
2448                 err = ext4_ext_dirty(handle, inode, path + depth);
2449                 if (err)
2450                         goto out;
2451
2452                 /*
2453                  * If the extent was completely released,
2454                  * we need to remove it from the leaf
2455                  */
2456                 if (num == 0) {
2457                         if (end != EXT_MAX_BLOCKS - 1) {
2458                                 /*
2459                                  * For hole punching, we need to scoot all the
2460                                  * extents up when an extent is removed so that
2461                                  * we dont have blank extents in the middle
2462                                  */
2463                                 memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) *
2464                                         sizeof(struct ext4_extent));
2465
2466                                 /* Now get rid of the one at the end */
2467                                 memset(EXT_LAST_EXTENT(eh), 0,
2468                                         sizeof(struct ext4_extent));
2469                         }
2470                         le16_add_cpu(&eh->eh_entries, -1);
2471                 }
2472
2473                 ext_debug("new extent: %u:%u:%llu\n", block, num,
2474                                 ext4_ext_pblock(ex));
2475                 ex--;
2476                 ex_ee_block = le32_to_cpu(ex->ee_block);
2477                 ex_ee_len = ext4_ext_get_actual_len(ex);
2478         }
2479
2480         if (correct_index && eh->eh_entries)
2481                 err = ext4_ext_correct_indexes(handle, inode, path);
2482
2483         /* if this leaf is free, then we should
2484          * remove it from index block above */
2485         if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
2486                 err = ext4_ext_rm_idx(handle, inode, path + depth);
2487
2488 out:
2489         return err;
2490 }
2491
2492 /*
2493  * ext4_ext_more_to_rm:
2494  * returns 1 if current index has to be freed (even partial)
2495  */
2496 static int
2497 ext4_ext_more_to_rm(struct ext4_ext_path *path)
2498 {
2499         BUG_ON(path->p_idx == NULL);
2500
2501         if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
2502                 return 0;
2503
2504         /*
2505          * if truncate on deeper level happened, it wasn't partial,
2506          * so we have to consider current index for truncation
2507          */
2508         if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
2509                 return 0;
2510         return 1;
2511 }
2512
2513 static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
2514 {
2515         struct super_block *sb = inode->i_sb;
2516         int depth = ext_depth(inode);
2517         struct ext4_ext_path *path;
2518         handle_t *handle;
2519         int i, err;
2520
2521         ext_debug("truncate since %u\n", start);
2522
2523         /* probably first extent we're gonna free will be last in block */
2524         handle = ext4_journal_start(inode, depth + 1);
2525         if (IS_ERR(handle))
2526                 return PTR_ERR(handle);
2527
2528 again:
2529         ext4_ext_invalidate_cache(inode);
2530
2531         /*
2532          * We start scanning from right side, freeing all the blocks
2533          * after i_size and walking into the tree depth-wise.
2534          */
2535         depth = ext_depth(inode);
2536         path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
2537         if (path == NULL) {
2538                 ext4_journal_stop(handle);
2539                 return -ENOMEM;
2540         }
2541         path[0].p_depth = depth;
2542         path[0].p_hdr = ext_inode_hdr(inode);
2543         if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
2544                 err = -EIO;
2545                 goto out;
2546         }
2547         i = err = 0;
2548
2549         while (i >= 0 && err == 0) {
2550                 if (i == depth) {
2551                         /* this is leaf block */
2552                         err = ext4_ext_rm_leaf(handle, inode, path,
2553                                         start, EXT_MAX_BLOCKS - 1);
2554                         /* root level has p_bh == NULL, brelse() eats this */
2555                         brelse(path[i].p_bh);
2556                         path[i].p_bh = NULL;
2557                         i--;
2558                         continue;
2559                 }
2560
2561                 /* this is index block */
2562                 if (!path[i].p_hdr) {
2563                         ext_debug("initialize header\n");
2564                         path[i].p_hdr = ext_block_hdr(path[i].p_bh);
2565                 }
2566
2567                 if (!path[i].p_idx) {
2568                         /* this level hasn't been touched yet */
2569                         path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2570                         path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
2571                         ext_debug("init index ptr: hdr 0x%p, num %d\n",
2572                                   path[i].p_hdr,
2573                                   le16_to_cpu(path[i].p_hdr->eh_entries));
2574                 } else {
2575                         /* we were already here, see at next index */
2576                         path[i].p_idx--;
2577                 }
2578
2579                 ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
2580                                 i, EXT_FIRST_INDEX(path[i].p_hdr),
2581                                 path[i].p_idx);
2582                 if (ext4_ext_more_to_rm(path + i)) {
2583                         struct buffer_head *bh;
2584                         /* go to the next level */
2585                         ext_debug("move to level %d (block %llu)\n",
2586                                   i + 1, ext4_idx_pblock(path[i].p_idx));
2587                         memset(path + i + 1, 0, sizeof(*path));
2588                         bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
2589                         if (!bh) {
2590                                 /* should we reset i_size? */
2591                                 err = -EIO;
2592                                 break;
2593                         }
2594                         if (WARN_ON(i + 1 > depth)) {
2595                                 err = -EIO;
2596                                 break;
2597                         }
2598                         if (ext4_ext_check(inode, ext_block_hdr(bh),
2599                                                         depth - i - 1)) {
2600                                 err = -EIO;
2601                                 break;
2602                         }
2603                         path[i + 1].p_bh = bh;
2604
2605                         /* save actual number of indexes since this
2606                          * number is changed at the next iteration */
2607                         path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
2608                         i++;
2609                 } else {
2610                         /* we finished processing this index, go up */
2611                         if (path[i].p_hdr->eh_entries == 0 && i > 0) {
2612                                 /* index is empty, remove it;
2613                                  * handle must be already prepared by the
2614                                  * truncatei_leaf() */
2615                                 err = ext4_ext_rm_idx(handle, inode, path + i);
2616                         }
2617                         /* root level has p_bh == NULL, brelse() eats this */
2618                         brelse(path[i].p_bh);
2619                         path[i].p_bh = NULL;
2620                         i--;
2621                         ext_debug("return to level %d\n", i);
2622                 }
2623         }
2624
2625         /* TODO: flexible tree reduction should be here */
2626         if (path->p_hdr->eh_entries == 0) {
2627                 /*
2628                  * truncate to zero freed all the tree,
2629                  * so we need to correct eh_depth
2630                  */
2631                 err = ext4_ext_get_access(handle, inode, path);
2632                 if (err == 0) {
2633                         ext_inode_hdr(inode)->eh_depth = 0;
2634                         ext_inode_hdr(inode)->eh_max =
2635                                 cpu_to_le16(ext4_ext_space_root(inode, 0));
2636                         err = ext4_ext_dirty(handle, inode, path);
2637                 }
2638         }
2639 out:
2640         ext4_ext_drop_refs(path);
2641         kfree(path);
2642         if (err == -EAGAIN)
2643                 goto again;
2644         ext4_journal_stop(handle);
2645
2646         return err;
2647 }
2648
2649 /*
2650  * called at mount time
2651  */
2652 void ext4_ext_init(struct super_block *sb)
2653 {
2654         /*
2655          * possible initialization would be here
2656          */
2657
2658         if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
2659 #if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
2660                 printk(KERN_INFO "EXT4-fs: file extents enabled");
2661 #ifdef AGGRESSIVE_TEST
2662                 printk(", aggressive tests");
2663 #endif
2664 #ifdef CHECK_BINSEARCH
2665                 printk(", check binsearch");
2666 #endif
2667 #ifdef EXTENTS_STATS
2668                 printk(", stats");
2669 #endif
2670                 printk("\n");
2671 #endif
2672 #ifdef EXTENTS_STATS
2673                 spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
2674                 EXT4_SB(sb)->s_ext_min = 1 << 30;
2675                 EXT4_SB(sb)->s_ext_max = 0;
2676 #endif
2677         }
2678 }
2679
2680 /*
2681  * called at umount time
2682  */
2683 void ext4_ext_release(struct super_block *sb)
2684 {
2685         if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
2686                 return;
2687
2688 #ifdef EXTENTS_STATS
2689         if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
2690                 struct ext4_sb_info *sbi = EXT4_SB(sb);
2691                 printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
2692                         sbi->s_ext_blocks, sbi->s_ext_extents,
2693                         sbi->s_ext_blocks / sbi->s_ext_extents);
2694                 printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
2695                         sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
2696         }
2697 #endif
2698 }
2699
2700 /* FIXME!! we need to try to merge to left or right after zero-out  */
2701 static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
2702 {
2703         ext4_fsblk_t ee_pblock;
2704         unsigned int ee_len;
2705         int ret;
2706
2707         ee_len    = ext4_ext_get_actual_len(ex);
2708         ee_pblock = ext4_ext_pblock(ex);
2709
2710         ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
2711         if (ret > 0)
2712                 ret = 0;
2713
2714         return ret;
2715 }
2716
2717 /*
2718  * used by extent splitting.
2719  */
2720 #define EXT4_EXT_MAY_ZEROOUT    0x1  /* safe to zeroout if split fails \
2721                                         due to ENOSPC */
2722 #define EXT4_EXT_MARK_UNINIT1   0x2  /* mark first half uninitialized */
2723 #define EXT4_EXT_MARK_UNINIT2   0x4  /* mark second half uninitialized */
2724
2725 /*
2726  * ext4_split_extent_at() splits an extent at given block.
2727  *
2728  * @handle: the journal handle
2729  * @inode: the file inode
2730  * @path: the path to the extent
2731  * @split: the logical block where the extent is splitted.
2732  * @split_flags: indicates if the extent could be zeroout if split fails, and
2733  *               the states(init or uninit) of new extents.
2734  * @flags: flags used to insert new extent to extent tree.
2735  *
2736  *
2737  * Splits extent [a, b] into two extents [a, @split) and [@split, b], states
2738  * of which are deterimined by split_flag.
2739  *
2740  * There are two cases:
2741  *  a> the extent are splitted into two extent.
2742  *  b> split is not needed, and just mark the extent.
2743  *
2744  * return 0 on success.
2745  */
2746 static int ext4_split_extent_at(handle_t *handle,
2747                              struct inode *inode,
2748                              struct ext4_ext_path *path,
2749                              ext4_lblk_t split,
2750                              int split_flag,
2751                              int flags)
2752 {
2753         ext4_fsblk_t newblock;
2754         ext4_lblk_t ee_block;
2755         struct ext4_extent *ex, newex, orig_ex;
2756         struct ext4_extent *ex2 = NULL;
2757         unsigned int ee_len, depth;
2758         int err = 0;
2759
2760         ext_debug("ext4_split_extents_at: inode %lu, logical"
2761                 "block %llu\n", inode->i_ino, (unsigned long long)split);
2762
2763         ext4_ext_show_leaf(inode, path);
2764
2765         depth = ext_depth(inode);
2766         ex = path[depth].p_ext;
2767         ee_block = le32_to_cpu(ex->ee_block);
2768         ee_len = ext4_ext_get_actual_len(ex);
2769         newblock = split - ee_block + ext4_ext_pblock(ex);
2770
2771         BUG_ON(split < ee_block || split >= (ee_block + ee_len));
2772
2773         err = ext4_ext_get_access(handle, inode, path + depth);
2774         if (err)
2775                 goto out;
2776
2777         if (split == ee_block) {
2778                 /*
2779                  * case b: block @split is the block that the extent begins with
2780                  * then we just change the state of the extent, and splitting
2781                  * is not needed.
2782                  */
2783                 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2784                         ext4_ext_mark_uninitialized(ex);
2785                 else
2786                         ext4_ext_mark_initialized(ex);
2787
2788                 if (!(flags & EXT4_GET_BLOCKS_PRE_IO))
2789                         ext4_ext_try_to_merge(inode, path, ex);
2790
2791                 err = ext4_ext_dirty(handle, inode, path + depth);
2792                 goto out;
2793         }
2794
2795         /* case a */
2796         memcpy(&orig_ex, ex, sizeof(orig_ex));
2797         ex->ee_len = cpu_to_le16(split - ee_block);
2798         if (split_flag & EXT4_EXT_MARK_UNINIT1)
2799                 ext4_ext_mark_uninitialized(ex);
2800
2801         /*
2802          * path may lead to new leaf, not to original leaf any more
2803          * after ext4_ext_insert_extent() returns,
2804          */
2805         err = ext4_ext_dirty(handle, inode, path + depth);
2806         if (err)
2807                 goto fix_extent_len;
2808
2809         ex2 = &newex;
2810         ex2->ee_block = cpu_to_le32(split);
2811         ex2->ee_len   = cpu_to_le16(ee_len - (split - ee_block));
2812         ext4_ext_store_pblock(ex2, newblock);
2813         if (split_flag & EXT4_EXT_MARK_UNINIT2)
2814                 ext4_ext_mark_uninitialized(ex2);
2815
2816         err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
2817         if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2818                 err = ext4_ext_zeroout(inode, &orig_ex);
2819                 if (err)
2820                         goto fix_extent_len;
2821                 /* update the extent length and mark as initialized */
2822                 ex->ee_len = cpu_to_le32(ee_len);
2823                 ext4_ext_try_to_merge(inode, path, ex);
2824                 err = ext4_ext_dirty(handle, inode, path + depth);
2825                 goto out;
2826         } else if (err)
2827                 goto fix_extent_len;
2828
2829 out:
2830         ext4_ext_show_leaf(inode, path);
2831         return err;
2832
2833 fix_extent_len:
2834         ex->ee_len = orig_ex.ee_len;
2835         ext4_ext_dirty(handle, inode, path + depth);
2836         return err;
2837 }
2838
2839 /*
2840  * ext4_split_extents() splits an extent and mark extent which is covered
2841  * by @map as split_flags indicates
2842  *
2843  * It may result in splitting the extent into multiple extents (upto three)
2844  * There are three possibilities:
2845  *   a> There is no split required
2846  *   b> Splits in two extents: Split is happening at either end of the extent
2847  *   c> Splits in three extents: Somone is splitting in middle of the extent
2848  *
2849  */
2850 static int ext4_split_extent(handle_t *handle,
2851                               struct inode *inode,
2852                               struct ext4_ext_path *path,
2853                               struct ext4_map_blocks *map,
2854                               int split_flag,
2855                               int flags)
2856 {
2857         ext4_lblk_t ee_block;
2858         struct ext4_extent *ex;
2859         unsigned int ee_len, depth;
2860         int err = 0;
2861         int uninitialized;
2862         int split_flag1, flags1;
2863
2864         depth = ext_depth(inode);
2865         ex = path[depth].p_ext;
2866         ee_block = le32_to_cpu(ex->ee_block);
2867         ee_len = ext4_ext_get_actual_len(ex);
2868         uninitialized = ext4_ext_is_uninitialized(ex);
2869
2870         if (map->m_lblk + map->m_len < ee_block + ee_len) {
2871                 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2872                               EXT4_EXT_MAY_ZEROOUT : 0;
2873                 flags1 = flags | EXT4_GET_BLOCKS_PRE_IO;
2874                 if (uninitialized)
2875                         split_flag1 |= EXT4_EXT_MARK_UNINIT1 |
2876                                        EXT4_EXT_MARK_UNINIT2;
2877                 err = ext4_split_extent_at(handle, inode, path,
2878                                 map->m_lblk + map->m_len, split_flag1, flags1);
2879                 if (err)
2880                         goto out;
2881         }
2882
2883         ext4_ext_drop_refs(path);
2884         path = ext4_ext_find_extent(inode, map->m_lblk, path);
2885         if (IS_ERR(path))
2886                 return PTR_ERR(path);
2887
2888         if (map->m_lblk >= ee_block) {
2889                 split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2890                               EXT4_EXT_MAY_ZEROOUT : 0;
2891                 if (uninitialized)
2892                         split_flag1 |= EXT4_EXT_MARK_UNINIT1;
2893                 if (split_flag & EXT4_EXT_MARK_UNINIT2)
2894                         split_flag1 |= EXT4_EXT_MARK_UNINIT2;
2895                 err = ext4_split_extent_at(handle, inode, path,
2896                                 map->m_lblk, split_flag1, flags);
2897                 if (err)
2898                         goto out;
2899         }
2900
2901         ext4_ext_show_leaf(inode, path);
2902 out:
2903         return err ? err : map->m_len;
2904 }
2905
2906 #define EXT4_EXT_ZERO_LEN 7
2907 /*
2908  * This function is called by ext4_ext_map_blocks() if someone tries to write
2909  * to an uninitialized extent. It may result in splitting the uninitialized
2910  * extent into multiple extents (up to three - one initialized and two
2911  * uninitialized).
2912  * There are three possibilities:
2913  *   a> There is no split required: Entire extent should be initialized
2914  *   b> Splits in two extents: Write is happening at either end of the extent
2915  *   c> Splits in three extents: Somone is writing in middle of the extent
2916  */
2917 static int ext4_ext_convert_to_initialized(handle_t *handle,
2918                                            struct inode *inode,
2919                                            struct ext4_map_blocks *map,
2920                                            struct ext4_ext_path *path)
2921 {
2922         struct ext4_map_blocks split_map;
2923         struct ext4_extent zero_ex;
2924         struct ext4_extent *ex;
2925         ext4_lblk_t ee_block, eof_block;
2926         unsigned int allocated, ee_len, depth;
2927         int err = 0;
2928         int split_flag = 0;
2929
2930         ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
2931                 "block %llu, max_blocks %u\n", inode->i_ino,
2932                 (unsigned long long)map->m_lblk, map->m_len);
2933
2934         eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
2935                 inode->i_sb->s_blocksize_bits;
2936         if (eof_block < map->m_lblk + map->m_len)
2937                 eof_block = map->m_lblk + map->m_len;
2938
2939         depth = ext_depth(inode);
2940         ex = path[depth].p_ext;
2941         ee_block = le32_to_cpu(ex->ee_block);
2942         ee_len = ext4_ext_get_actual_len(ex);
2943         allocated = ee_len - (map->m_lblk - ee_block);
2944
2945         WARN_ON(map->m_lblk < ee_block);
2946         /*
2947          * It is safe to convert extent to initialized via explicit
2948          * zeroout only if extent is fully insde i_size or new_size.
2949          */
2950         split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
2951
2952         /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
2953         if (ee_len <= 2*EXT4_EXT_ZERO_LEN &&
2954             (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2955                 err = ext4_ext_zeroout(inode, ex);
2956                 if (err)
2957                         goto out;
2958
2959                 err = ext4_ext_get_access(handle, inode, path + depth);
2960                 if (err)
2961                         goto out;
2962                 ext4_ext_mark_initialized(ex);
2963                 ext4_ext_try_to_merge(inode, path, ex);
2964                 err = ext4_ext_dirty(handle, inode, path + depth);
2965                 goto out;
2966         }
2967
2968         /*
2969          * four cases:
2970          * 1. split the extent into three extents.
2971          * 2. split the extent into two extents, zeroout the first half.
2972          * 3. split the extent into two extents, zeroout the second half.
2973          * 4. split the extent into two extents with out zeroout.
2974          */
2975         split_map.m_lblk = map->m_lblk;
2976         split_map.m_len = map->m_len;
2977
2978         if (allocated > map->m_len) {
2979                 if (allocated <= EXT4_EXT_ZERO_LEN &&
2980                     (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2981                         /* case 3 */
2982                         zero_ex.ee_block =
2983                                          cpu_to_le32(map->m_lblk);
2984                         zero_ex.ee_len = cpu_to_le16(allocated);
2985                         ext4_ext_store_pblock(&zero_ex,
2986                                 ext4_ext_pblock(ex) + map->m_lblk - ee_block);
2987                         err = ext4_ext_zeroout(inode, &zero_ex);
2988                         if (err)
2989                                 goto out;
2990                         split_map.m_lblk = map->m_lblk;
2991                         split_map.m_len = allocated;
2992                 } else if ((map->m_lblk - ee_block + map->m_len <
2993                            EXT4_EXT_ZERO_LEN) &&
2994                            (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2995                         /* case 2 */
2996                         if (map->m_lblk != ee_block) {
2997                                 zero_ex.ee_block = ex->ee_block;
2998                                 zero_ex.ee_len = cpu_to_le16(map->m_lblk -
2999                                                         ee_block);
3000                                 ext4_ext_store_pblock(&zero_ex,
3001                                                       ext4_ext_pblock(ex));
3002                                 err = ext4_ext_zeroout(inode, &zero_ex);
3003                                 if (err)
3004                                         goto out;
3005                         }
3006
3007                         split_map.m_lblk = ee_block;
3008                         split_map.m_len = map->m_lblk - ee_block + map->m_len;
3009                         allocated = map->m_len;
3010                 }
3011         }
3012
3013         allocated = ext4_split_extent(handle, inode, path,
3014                                        &split_map, split_flag, 0);
3015         if (allocated < 0)
3016                 err = allocated;
3017
3018 out:
3019         return err ? err : allocated;
3020 }
3021
3022 /*
3023  * This function is called by ext4_ext_map_blocks() from
3024  * ext4_get_blocks_dio_write() when DIO to write
3025  * to an uninitialized extent.
3026  *
3027  * Writing to an uninitialized extent may result in splitting the uninitialized
3028  * extent into multiple /initialized uninitialized extents (up to three)
3029  * There are three possibilities:
3030  *   a> There is no split required: Entire extent should be uninitialized
3031  *   b> Splits in two extents: Write is happening at either end of the extent
3032  *   c> Splits in three extents: Somone is writing in middle of the extent
3033  *
3034  * One of more index blocks maybe needed if the extent tree grow after
3035  * the uninitialized extent split. To prevent ENOSPC occur at the IO
3036  * complete, we need to split the uninitialized extent before DIO submit
3037  * the IO. The uninitialized extent called at this time will be split
3038  * into three uninitialized extent(at most). After IO complete, the part
3039  * being filled will be convert to initialized by the end_io callback function
3040  * via ext4_convert_unwritten_extents().
3041  *
3042  * Returns the size of uninitialized extent to be written on success.
3043  */
3044 static int ext4_split_unwritten_extents(handle_t *handle,
3045                                         struct inode *inode,
3046                                         struct ext4_map_blocks *map,
3047                                         struct ext4_ext_path *path,
3048                                         int flags)
3049 {
3050         ext4_lblk_t eof_block;
3051         ext4_lblk_t ee_block;
3052         struct ext4_extent *ex;
3053         unsigned int ee_len;
3054         int split_flag = 0, depth;
3055
3056         ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
3057                 "block %llu, max_blocks %u\n", inode->i_ino,
3058                 (unsigned long long)map->m_lblk, map->m_len);
3059
3060         eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
3061                 inode->i_sb->s_blocksize_bits;
3062         if (eof_block < map->m_lblk + map->m_len)
3063                 eof_block = map->m_lblk + map->m_len;
3064         /*
3065          * It is safe to convert extent to initialized via explicit
3066          * zeroout only if extent is fully insde i_size or new_size.
3067          */
3068         depth = ext_depth(inode);
3069         ex = path[depth].p_ext;
3070         ee_block = le32_to_cpu(ex->ee_block);
3071         ee_len = ext4_ext_get_actual_len(ex);
3072
3073         split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3074         split_flag |= EXT4_EXT_MARK_UNINIT2;
3075
3076         flags |= EXT4_GET_BLOCKS_PRE_IO;
3077         return ext4_split_extent(handle, inode, path, map, split_flag, flags);
3078 }
3079
3080 static int ext4_convert_unwritten_extents_endio(handle_t *handle,
3081                                               struct inode *inode,
3082                                               struct ext4_ext_path *path)
3083 {
3084         struct ext4_extent *ex;
3085         int depth;
3086         int err = 0;
3087
3088         depth = ext_depth(inode);
3089         ex = path[depth].p_ext;
3090
3091         ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
3092                 "block %llu, max_blocks %u\n", inode->i_ino,
3093                 (unsigned long long)le32_to_cpu(ex->ee_block),
3094                 ext4_ext_get_actual_len(ex));
3095
3096         err = ext4_ext_get_access(handle, inode, path + depth);
3097         if (err)
3098                 goto out;
3099         /* first mark the extent as initialized */
3100         ext4_ext_mark_initialized(ex);
3101
3102         /* note: ext4_ext_correct_indexes() isn't needed here because
3103          * borders are not changed
3104          */
3105         ext4_ext_try_to_merge(inode, path, ex);
3106
3107         /* Mark modified extent as dirty */
3108         err = ext4_ext_dirty(handle, inode, path + depth);
3109 out:
3110         ext4_ext_show_leaf(inode, path);
3111         return err;
3112 }
3113
3114 static void unmap_underlying_metadata_blocks(struct block_device *bdev,
3115                         sector_t block, int count)
3116 {
3117         int i;
3118         for (i = 0; i < count; i++)
3119                 unmap_underlying_metadata(bdev, block + i);
3120 }
3121
3122 /*
3123  * Handle EOFBLOCKS_FL flag, clearing it if necessary
3124  */
3125 static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
3126                               ext4_lblk_t lblk,
3127                               struct ext4_ext_path *path,
3128                               unsigned int len)
3129 {
3130         int i, depth;
3131         struct ext4_extent_header *eh;
3132         struct ext4_extent *last_ex;
3133
3134         if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
3135                 return 0;
3136
3137         depth = ext_depth(inode);
3138         eh = path[depth].p_hdr;
3139
3140         if (unlikely(!eh->eh_entries)) {
3141                 EXT4_ERROR_INODE(inode, "eh->eh_entries == 0 and "
3142                                  "EOFBLOCKS_FL set");
3143                 return -EIO;
3144         }
3145         last_ex = EXT_LAST_EXTENT(eh);
3146         /*
3147          * We should clear the EOFBLOCKS_FL flag if we are writing the
3148          * last block in the last extent in the file.  We test this by
3149          * first checking to see if the caller to
3150          * ext4_ext_get_blocks() was interested in the last block (or
3151          * a block beyond the last block) in the current extent.  If
3152          * this turns out to be false, we can bail out from this
3153          * function immediately.
3154          */
3155         if (lblk + len < le32_to_cpu(last_ex->ee_block) +
3156             ext4_ext_get_actual_len(last_ex))
3157                 return 0;
3158         /*
3159          * If the caller does appear to be planning to write at or
3160          * beyond the end of the current extent, we then test to see
3161          * if the current extent is the last extent in the file, by
3162          * checking to make sure it was reached via the rightmost node
3163          * at each level of the tree.
3164          */
3165         for (i = depth-1; i >= 0; i--)
3166                 if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
3167                         return 0;
3168         ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3169         return ext4_mark_inode_dirty(handle, inode);
3170 }
3171
3172 static int
3173 ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
3174                         struct ext4_map_blocks *map,
3175                         struct ext4_ext_path *path, int flags,
3176                         unsigned int allocated, ext4_fsblk_t newblock)
3177 {
3178         int ret = 0;
3179         int err = 0;
3180         ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3181
3182         ext_debug("ext4_ext_handle_uninitialized_extents: inode %lu, logical"
3183                   "block %llu, max_blocks %u, flags %d, allocated %u",
3184                   inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
3185                   flags, allocated);
3186         ext4_ext_show_leaf(inode, path);
3187
3188         /* get_block() before submit the IO, split the extent */
3189         if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
3190                 ret = ext4_split_unwritten_extents(handle, inode, map,
3191                                                    path, flags);
3192                 /*
3193                  * Flag the inode(non aio case) or end_io struct (aio case)
3194                  * that this IO needs to conversion to written when IO is
3195                  * completed
3196                  */
3197                 if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) {
3198                         io->flag = EXT4_IO_END_UNWRITTEN;
3199                         atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten);
3200                 } else
3201                         ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
3202                 if (ext4_should_dioread_nolock(inode))
3203                         map->m_flags |= EXT4_MAP_UNINIT;
3204                 goto out;
3205         }
3206         /* IO end_io complete, convert the filled extent to written */
3207         if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
3208                 ret = ext4_convert_unwritten_extents_endio(handle, inode,
3209                                                         path);
3210                 if (ret >= 0) {
3211                         ext4_update_inode_fsync_trans(handle, inode, 1);
3212                         err = check_eofblocks_fl(handle, inode, map->m_lblk,
3213                                                  path, map->m_len);
3214                 } else
3215                         err = ret;
3216                 goto out2;
3217         }
3218         /* buffered IO case */
3219         /*
3220          * repeat fallocate creation request
3221          * we already have an unwritten extent
3222          */
3223         if (flags & EXT4_GET_BLOCKS_UNINIT_EXT)
3224                 goto map_out;
3225
3226         /* buffered READ or buffered write_begin() lookup */
3227         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3228                 /*
3229                  * We have blocks reserved already.  We
3230                  * return allocated blocks so that delalloc
3231                  * won't do block reservation for us.  But
3232                  * the buffer head will be unmapped so that
3233                  * a read from the block returns 0s.
3234                  */
3235                 map->m_flags |= EXT4_MAP_UNWRITTEN;
3236                 goto out1;
3237         }
3238
3239         /* buffered write, writepage time, convert*/
3240         ret = ext4_ext_convert_to_initialized(handle, inode, map, path);
3241         if (ret >= 0) {
3242                 ext4_update_inode_fsync_trans(handle, inode, 1);
3243                 err = check_eofblocks_fl(handle, inode, map->m_lblk, path,
3244                                          map->m_len);
3245                 if (err < 0)
3246                         goto out2;
3247         }
3248
3249 out:
3250         if (ret <= 0) {
3251                 err = ret;
3252                 goto out2;
3253         } else
3254                 allocated = ret;
3255         map->m_flags |= EXT4_MAP_NEW;
3256         /*
3257          * if we allocated more blocks than requested
3258          * we need to make sure we unmap the extra block
3259          * allocated. The actual needed block will get
3260          * unmapped later when we find the buffer_head marked
3261          * new.
3262          */
3263         if (allocated > map->m_len) {
3264                 unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
3265                                         newblock + map->m_len,
3266                                         allocated - map->m_len);
3267                 allocated = map->m_len;
3268         }
3269
3270         /*
3271          * If we have done fallocate with the offset that is already
3272          * delayed allocated, we would have block reservation
3273          * and quota reservation done in the delayed write path.
3274          * But fallocate would have already updated quota and block
3275          * count for this offset. So cancel these reservation
3276          */
3277         if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
3278                 ext4_da_update_reserve_space(inode, allocated, 0);
3279
3280 map_out:
3281         map->m_flags |= EXT4_MAP_MAPPED;
3282 out1:
3283         if (allocated > map->m_len)
3284                 allocated = map->m_len;
3285         ext4_ext_show_leaf(inode, path);
3286         map->m_pblk = newblock;
3287         map->m_len = allocated;
3288 out2:
3289         if (path) {
3290                 ext4_ext_drop_refs(path);
3291                 kfree(path);
3292         }
3293         return err ? err : allocated;
3294 }
3295
3296 /*
3297  * Block allocation/map/preallocation routine for extents based files
3298  *
3299  *
3300  * Need to be called with
3301  * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
3302  * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
3303  *
3304  * return > 0, number of of blocks already mapped/allocated
3305  *          if create == 0 and these are pre-allocated blocks
3306  *              buffer head is unmapped
3307  *          otherwise blocks are mapped
3308  *
3309  * return = 0, if plain look up failed (blocks have not been allocated)
3310  *          buffer head is unmapped
3311  *
3312  * return < 0, error case.
3313  */
3314 int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
3315                         struct ext4_map_blocks *map, int flags)
3316 {
3317         struct ext4_ext_path *path = NULL;
3318         struct ext4_extent newex, *ex;
3319         ext4_fsblk_t newblock = 0;
3320         int err = 0, depth, ret;
3321         unsigned int allocated = 0;
3322         unsigned int punched_out = 0;
3323         unsigned int result = 0;
3324         struct ext4_allocation_request ar;
3325         ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3326         struct ext4_map_blocks punch_map;
3327
3328         ext_debug("blocks %u/%u requested for inode %lu\n",
3329                   map->m_lblk, map->m_len, inode->i_ino);
3330         trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
3331
3332         /* check in cache */
3333         if (!(flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) &&
3334                 ext4_ext_in_cache(inode, map->m_lblk, &newex)) {
3335                 if (!newex.ee_start_lo && !newex.ee_start_hi) {
3336                         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3337                                 /*
3338                                  * block isn't allocated yet and
3339                                  * user doesn't want to allocate it
3340                                  */
3341                                 goto out2;
3342                         }
3343                         /* we should allocate requested block */
3344                 } else {
3345                         /* block is already allocated */
3346                         newblock = map->m_lblk
3347                                    - le32_to_cpu(newex.ee_block)
3348                                    + ext4_ext_pblock(&newex);
3349                         /* number of remaining blocks in the extent */
3350                         allocated = ext4_ext_get_actual_len(&newex) -
3351                                 (map->m_lblk - le32_to_cpu(newex.ee_block));
3352                         goto out;
3353                 }
3354         }
3355
3356         /* find extent for this block */
3357         path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
3358         if (IS_ERR(path)) {
3359                 err = PTR_ERR(path);
3360                 path = NULL;
3361                 goto out2;
3362         }
3363
3364         depth = ext_depth(inode);
3365
3366         /*
3367          * consistent leaf must not be empty;
3368          * this situation is possible, though, _during_ tree modification;
3369          * this is why assert can't be put in ext4_ext_find_extent()
3370          */
3371         if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
3372                 EXT4_ERROR_INODE(inode, "bad extent address "
3373                                  "lblock: %lu, depth: %d pblock %lld",
3374                                  (unsigned long) map->m_lblk, depth,
3375                                  path[depth].p_block);
3376                 err = -EIO;
3377                 goto out2;
3378         }
3379
3380         ex = path[depth].p_ext;
3381         if (ex) {
3382                 ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
3383                 ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
3384                 unsigned short ee_len;
3385
3386                 /*
3387                  * Uninitialized extents are treated as holes, except that
3388                  * we split out initialized portions during a write.
3389                  */
3390                 ee_len = ext4_ext_get_actual_len(ex);
3391                 /* if found extent covers block, simply return it */
3392                 if (in_range(map->m_lblk, ee_block, ee_len)) {
3393                         newblock = map->m_lblk - ee_block + ee_start;
3394                         /* number of remaining blocks in the extent */
3395                         allocated = ee_len - (map->m_lblk - ee_block);
3396                         ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
3397                                   ee_block, ee_len, newblock);
3398
3399                         if ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0) {
3400                                 /*
3401                                  * Do not put uninitialized extent
3402                                  * in the cache
3403                                  */
3404                                 if (!ext4_ext_is_uninitialized(ex)) {
3405                                         ext4_ext_put_in_cache(inode, ee_block,
3406                                                 ee_len, ee_start);
3407                                         goto out;
3408                                 }
3409                                 ret = ext4_ext_handle_uninitialized_extents(
3410                                         handle, inode, map, path, flags,
3411                                         allocated, newblock);
3412                                 return ret;
3413                         }
3414
3415                         /*
3416                          * Punch out the map length, but only to the
3417                          * end of the extent
3418                          */
3419                         punched_out = allocated < map->m_len ?
3420                                 allocated : map->m_len;
3421
3422                         /*
3423                          * Sense extents need to be converted to
3424                          * uninitialized, they must fit in an
3425                          * uninitialized extent
3426                          */
3427                         if (punched_out > EXT_UNINIT_MAX_LEN)
3428                                 punched_out = EXT_UNINIT_MAX_LEN;
3429
3430                         punch_map.m_lblk = map->m_lblk;
3431                         punch_map.m_pblk = newblock;
3432                         punch_map.m_len = punched_out;
3433                         punch_map.m_flags = 0;
3434
3435                         /* Check to see if the extent needs to be split */
3436                         if (punch_map.m_len != ee_len ||
3437                                 punch_map.m_lblk != ee_block) {
3438
3439                                 ret = ext4_split_extent(handle, inode,
3440                                 path, &punch_map, 0,
3441                                 EXT4_GET_BLOCKS_PUNCH_OUT_EXT |
3442                                 EXT4_GET_BLOCKS_PRE_IO);
3443
3444                                 if (ret < 0) {
3445                                         err = ret;
3446                                         goto out2;
3447                                 }
3448                                 /*
3449                                  * find extent for the block at
3450                                  * the start of the hole
3451                                  */
3452                                 ext4_ext_drop_refs(path);
3453                                 kfree(path);
3454
3455                                 path = ext4_ext_find_extent(inode,
3456                                 map->m_lblk, NULL);
3457                                 if (IS_ERR(path)) {
3458                                         err = PTR_ERR(path);
3459                                         path = NULL;
3460                                         goto out2;
3461                                 }
3462
3463                                 depth = ext_depth(inode);
3464                                 ex = path[depth].p_ext;
3465                                 ee_len = ext4_ext_get_actual_len(ex);
3466                                 ee_block = le32_to_cpu(ex->ee_block);
3467                                 ee_start = ext4_ext_pblock(ex);
3468
3469                         }
3470
3471                         ext4_ext_mark_uninitialized(ex);
3472
3473                         ext4_ext_invalidate_cache(inode);
3474
3475                         err = ext4_ext_rm_leaf(handle, inode, path,
3476                                 map->m_lblk, map->m_lblk + punched_out);
3477
3478                         if (!err && path->p_hdr->eh_entries == 0) {
3479                                 /*
3480                                  * Punch hole freed all of this sub tree,
3481                                  * so we need to correct eh_depth
3482                                  */
3483                                 err = ext4_ext_get_access(handle, inode, path);
3484                                 if (err == 0) {
3485                                         ext_inode_hdr(inode)->eh_depth = 0;
3486                                         ext_inode_hdr(inode)->eh_max =
3487                                         cpu_to_le16(ext4_ext_space_root(
3488                                                 inode, 0));
3489
3490                                         err = ext4_ext_dirty(
3491                                                 handle, inode, path);
3492                                 }
3493                         }
3494
3495                         goto out2;
3496                 }
3497         }
3498
3499         /*
3500          * requested block isn't allocated yet;
3501          * we couldn't try to create block if create flag is zero
3502          */
3503         if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3504                 /*
3505                  * put just found gap into cache to speed up
3506                  * subsequent requests
3507                  */
3508                 ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
3509                 goto out2;
3510         }
3511         /*
3512          * Okay, we need to do block allocation.
3513          */
3514
3515         /* find neighbour allocated blocks */
3516         ar.lleft = map->m_lblk;
3517         err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
3518         if (err)
3519                 goto out2;
3520         ar.lright = map->m_lblk;
3521         err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright);
3522         if (err)
3523                 goto out2;
3524
3525         /*
3526          * See if request is beyond maximum number of blocks we can have in
3527          * a single extent. For an initialized extent this limit is
3528          * EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
3529          * EXT_UNINIT_MAX_LEN.
3530          */
3531         if (map->m_len > EXT_INIT_MAX_LEN &&
3532             !(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
3533                 map->m_len = EXT_INIT_MAX_LEN;
3534         else if (map->m_len > EXT_UNINIT_MAX_LEN &&
3535                  (flags & EXT4_GET_BLOCKS_UNINIT_EXT))
3536                 map->m_len = EXT_UNINIT_MAX_LEN;
3537
3538         /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */
3539         newex.ee_block = cpu_to_le32(map->m_lblk);
3540         newex.ee_len = cpu_to_le16(map->m_len);
3541         err = ext4_ext_check_overlap(inode, &newex, path);
3542         if (err)
3543                 allocated = ext4_ext_get_actual_len(&newex);
3544         else
3545                 allocated = map->m_len;
3546
3547         /* allocate new block */
3548         ar.inode = inode;
3549         ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk);
3550         ar.logical = map->m_lblk;
3551         ar.len = allocated;
3552         if (S_ISREG(inode->i_mode))
3553                 ar.flags = EXT4_MB_HINT_DATA;
3554         else
3555                 /* disable in-core preallocation for non-regular files */
3556                 ar.flags = 0;
3557         if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE)
3558                 ar.flags |= EXT4_MB_HINT_NOPREALLOC;
3559         newblock = ext4_mb_new_blocks(handle, &ar, &err);
3560         if (!newblock)
3561                 goto out2;
3562         ext_debug("allocate new block: goal %llu, found %llu/%u\n",
3563                   ar.goal, newblock, allocated);
3564
3565         /* try to insert new extent into found leaf and return */
3566         ext4_ext_store_pblock(&newex, newblock);
3567         newex.ee_len = cpu_to_le16(ar.len);
3568         /* Mark uninitialized */
3569         if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
3570                 ext4_ext_mark_uninitialized(&newex);
3571                 /*
3572                  * io_end structure was created for every IO write to an
3573                  * uninitialized extent. To avoid unnecessary conversion,
3574                  * here we flag the IO that really needs the conversion.
3575                  * For non asycn direct IO case, flag the inode state
3576                  * that we need to perform conversion when IO is done.
3577                  */
3578                 if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
3579                         if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) {
3580                                 io->flag = EXT4_IO_END_UNWRITTEN;
3581                                 atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten);
3582                         } else
3583                                 ext4_set_inode_state(inode,
3584                                                      EXT4_STATE_DIO_UNWRITTEN);
3585                 }
3586                 if (ext4_should_dioread_nolock(inode))
3587                         map->m_flags |= EXT4_MAP_UNINIT;
3588         }
3589
3590         err = check_eofblocks_fl(handle, inode, map->m_lblk, path, ar.len);
3591         if (!err)
3592                 err = ext4_ext_insert_extent(handle, inode, path,
3593                                              &newex, flags);
3594         if (err) {
3595                 int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ?
3596                         EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0;
3597                 /* free data blocks we just allocated */
3598                 /* not a good idea to call discard here directly,
3599                  * but otherwise we'd need to call it every free() */
3600                 ext4_discard_preallocations(inode);
3601                 ext4_free_blocks(handle, inode, NULL, ext4_ext_pblock(&newex),
3602                                  ext4_ext_get_actual_len(&newex), fb_flags);
3603                 goto out2;
3604         }
3605
3606         /* previous routine could use block we allocated */
3607         newblock = ext4_ext_pblock(&newex);
3608         allocated = ext4_ext_get_actual_len(&newex);
3609         if (allocated > map->m_len)
3610                 allocated = map->m_len;
3611         map->m_flags |= EXT4_MAP_NEW;
3612
3613         /*
3614          * Update reserved blocks/metadata blocks after successful
3615          * block allocation which had been deferred till now.
3616          */
3617         if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
3618                 ext4_da_update_reserve_space(inode, allocated, 1);
3619
3620         /*
3621          * Cache the extent and update transaction to commit on fdatasync only
3622          * when it is _not_ an uninitialized extent.
3623          */
3624         if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) {
3625                 ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock);
3626                 ext4_update_inode_fsync_trans(handle, inode, 1);
3627         } else
3628                 ext4_update_inode_fsync_trans(handle, inode, 0);
3629 out:
3630         if (allocated > map->m_len)
3631                 allocated = map->m_len;
3632         ext4_ext_show_leaf(inode, path);
3633         map->m_flags |= EXT4_MAP_MAPPED;
3634         map->m_pblk = newblock;
3635         map->m_len = allocated;
3636 out2:
3637         if (path) {
3638                 ext4_ext_drop_refs(path);
3639                 kfree(path);
3640         }
3641         trace_ext4_ext_map_blocks_exit(inode, map->m_lblk,
3642                 newblock, map->m_len, err ? err : allocated);
3643
3644         result = (flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) ?
3645                         punched_out : allocated;
3646
3647         return err ? err : result;
3648 }
3649
3650 void ext4_ext_truncate(struct inode *inode)
3651 {
3652         struct address_space *mapping = inode->i_mapping;
3653         struct super_block *sb = inode->i_sb;
3654         ext4_lblk_t last_block;
3655         handle_t *handle;
3656         int err = 0;
3657
3658         /*
3659          * finish any pending end_io work so we won't run the risk of
3660          * converting any truncated blocks to initialized later
3661          */
3662         ext4_flush_completed_IO(inode);
3663
3664         /*
3665          * probably first extent we're gonna free will be last in block
3666          */
3667         err = ext4_writepage_trans_blocks(inode);
3668         handle = ext4_journal_start(inode, err);
3669         if (IS_ERR(handle))
3670                 return;
3671
3672         if (inode->i_size & (sb->s_blocksize - 1))
3673                 ext4_block_truncate_page(handle, mapping, inode->i_size);
3674
3675         if (ext4_orphan_add(handle, inode))
3676                 goto out_stop;
3677
3678         down_write(&EXT4_I(inode)->i_data_sem);
3679         ext4_ext_invalidate_cache(inode);
3680
3681         ext4_discard_preallocations(inode);
3682
3683         /*
3684          * TODO: optimization is possible here.
3685          * Probably we need not scan at all,
3686          * because page truncation is enough.
3687          */
3688
3689         /* we have to know where to truncate from in crash case */
3690         EXT4_I(inode)->i_disksize = inode->i_size;
3691         ext4_mark_inode_dirty(handle, inode);
3692
3693         last_block = (inode->i_size + sb->s_blocksize - 1)
3694                         >> EXT4_BLOCK_SIZE_BITS(sb);
3695         err = ext4_ext_remove_space(inode, last_block);
3696
3697         /* In a multi-transaction truncate, we only make the final
3698          * transaction synchronous.
3699          */
3700         if (IS_SYNC(inode))
3701                 ext4_handle_sync(handle);
3702
3703         up_write(&EXT4_I(inode)->i_data_sem);
3704
3705 out_stop:
3706         /*
3707          * If this was a simple ftruncate() and the file will remain alive,
3708          * then we need to clear up the orphan record which we created above.
3709          * However, if this was a real unlink then we were called by
3710          * ext4_delete_inode(), and we allow that function to clean up the
3711          * orphan info for us.
3712          */
3713         if (inode->i_nlink)
3714                 ext4_orphan_del(handle, inode);
3715
3716         inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
3717         ext4_mark_inode_dirty(handle, inode);
3718         ext4_journal_stop(handle);
3719 }
3720
3721 static void ext4_falloc_update_inode(struct inode *inode,
3722                                 int mode, loff_t new_size, int update_ctime)
3723 {
3724         struct timespec now;
3725
3726         if (update_ctime) {
3727                 now = current_fs_time(inode->i_sb);
3728                 if (!timespec_equal(&inode->i_ctime, &now))
3729                         inode->i_ctime = now;
3730         }
3731         /*
3732          * Update only when preallocation was requested beyond
3733          * the file size.
3734          */
3735         if (!(mode & FALLOC_FL_KEEP_SIZE)) {
3736                 if (new_size > i_size_read(inode))
3737                         i_size_write(inode, new_size);
3738                 if (new_size > EXT4_I(inode)->i_disksize)
3739                         ext4_update_i_disksize(inode, new_size);
3740         } else {
3741                 /*
3742                  * Mark that we allocate beyond EOF so the subsequent truncate
3743                  * can proceed even if the new size is the same as i_size.
3744                  */
3745                 if (new_size > i_size_read(inode))
3746                         ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3747         }
3748
3749 }
3750
3751 /*
3752  * preallocate space for a file. This implements ext4's fallocate file
3753  * operation, which gets called from sys_fallocate system call.
3754  * For block-mapped files, posix_fallocate should fall back to the method
3755  * of writing zeroes to the required new blocks (the same behavior which is
3756  * expected for file systems which do not support fallocate() system call).
3757  */
3758 long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
3759 {
3760         struct inode *inode = file->f_path.dentry->d_inode;
3761         handle_t *handle;
3762         loff_t new_size;
3763         unsigned int max_blocks;
3764         int ret = 0;
3765         int ret2 = 0;
3766         int retries = 0;
3767         struct ext4_map_blocks map;
3768         unsigned int credits, blkbits = inode->i_blkbits;
3769
3770         /*
3771          * currently supporting (pre)allocate mode for extent-based
3772          * files _only_
3773          */
3774         if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
3775                 return -EOPNOTSUPP;
3776
3777         /* Return error if mode is not supported */
3778         if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
3779                 return -EOPNOTSUPP;
3780
3781         if (mode & FALLOC_FL_PUNCH_HOLE)
3782                 return ext4_punch_hole(file, offset, len);
3783
3784         trace_ext4_fallocate_enter(inode, offset, len, mode);
3785         map.m_lblk = offset >> blkbits;
3786         /*
3787          * We can't just convert len to max_blocks because
3788          * If blocksize = 4096 offset = 3072 and len = 2048
3789          */
3790         max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
3791                 - map.m_lblk;
3792         /*
3793          * credits to insert 1 extent into extent tree
3794          */
3795         credits = ext4_chunk_trans_blocks(inode, max_blocks);
3796         mutex_lock(&inode->i_mutex);
3797         ret = inode_newsize_ok(inode, (len + offset));
3798         if (ret) {
3799                 mutex_unlock(&inode->i_mutex);
3800                 trace_ext4_fallocate_exit(inode, offset, max_blocks, ret);
3801                 return ret;
3802         }
3803 retry:
3804         while (ret >= 0 && ret < max_blocks) {
3805                 map.m_lblk = map.m_lblk + ret;
3806                 map.m_len = max_blocks = max_blocks - ret;
3807                 handle = ext4_journal_start(inode, credits);
3808                 if (IS_ERR(handle)) {
3809                         ret = PTR_ERR(handle);
3810                         break;
3811                 }
3812                 ret = ext4_map_blocks(handle, inode, &map,
3813                                       EXT4_GET_BLOCKS_CREATE_UNINIT_EXT |
3814                                       EXT4_GET_BLOCKS_NO_NORMALIZE);
3815                 if (ret <= 0) {
3816 #ifdef EXT4FS_DEBUG
3817                         WARN_ON(ret <= 0);
3818                         printk(KERN_ERR "%s: ext4_ext_map_blocks "
3819                                     "returned error inode#%lu, block=%u, "
3820                                     "max_blocks=%u", __func__,
3821                                     inode->i_ino, map.m_lblk, max_blocks);
3822 #endif
3823                         ext4_mark_inode_dirty(handle, inode);
3824                         ret2 = ext4_journal_stop(handle);
3825                         break;
3826                 }
3827                 if ((map.m_lblk + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
3828                                                 blkbits) >> blkbits))
3829                         new_size = offset + len;
3830                 else
3831                         new_size = ((loff_t) map.m_lblk + ret) << blkbits;
3832
3833                 ext4_falloc_update_inode(inode, mode, new_size,
3834                                          (map.m_flags & EXT4_MAP_NEW));
3835                 ext4_mark_inode_dirty(handle, inode);
3836                 ret2 = ext4_journal_stop(handle);
3837                 if (ret2)
3838                         break;
3839         }
3840         if (ret == -ENOSPC &&
3841                         ext4_should_retry_alloc(inode->i_sb, &retries)) {
3842                 ret = 0;
3843                 goto retry;
3844         }
3845         mutex_unlock(&inode->i_mutex);
3846         trace_ext4_fallocate_exit(inode, offset, max_blocks,
3847                                 ret > 0 ? ret2 : ret);
3848         return ret > 0 ? ret2 : ret;
3849 }
3850
3851 /*
3852  * This function convert a range of blocks to written extents
3853  * The caller of this function will pass the start offset and the size.
3854  * all unwritten extents within this range will be converted to
3855  * written extents.
3856  *
3857  * This function is called from the direct IO end io call back
3858  * function, to convert the fallocated extents after IO is completed.
3859  * Returns 0 on success.
3860  */
3861 int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
3862                                     ssize_t len)
3863 {
3864         handle_t *handle;
3865         unsigned int max_blocks;
3866         int ret = 0;
3867         int ret2 = 0;
3868         struct ext4_map_blocks map;
3869         unsigned int credits, blkbits = inode->i_blkbits;
3870
3871         map.m_lblk = offset >> blkbits;
3872         /*
3873          * We can't just convert len to max_blocks because
3874          * If blocksize = 4096 offset = 3072 and len = 2048
3875          */
3876         max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) -
3877                       map.m_lblk);
3878         /*
3879          * credits to insert 1 extent into extent tree
3880          */
3881         credits = ext4_chunk_trans_blocks(inode, max_blocks);
3882         while (ret >= 0 && ret < max_blocks) {
3883                 map.m_lblk += ret;
3884                 map.m_len = (max_blocks -= ret);
3885                 handle = ext4_journal_start(inode, credits);
3886                 if (IS_ERR(handle)) {
3887                         ret = PTR_ERR(handle);
3888                         break;
3889                 }
3890                 ret = ext4_map_blocks(handle, inode, &map,
3891                                       EXT4_GET_BLOCKS_IO_CONVERT_EXT);
3892                 if (ret <= 0) {
3893                         WARN_ON(ret <= 0);
3894                         printk(KERN_ERR "%s: ext4_ext_map_blocks "
3895                                     "returned error inode#%lu, block=%u, "
3896                                     "max_blocks=%u", __func__,
3897                                     inode->i_ino, map.m_lblk, map.m_len);
3898                 }
3899                 ext4_mark_inode_dirty(handle, inode);
3900                 ret2 = ext4_journal_stop(handle);
3901                 if (ret <= 0 || ret2 )
3902                         break;
3903         }
3904         return ret > 0 ? ret2 : ret;
3905 }
3906
3907 /*
3908  * Callback function called for each extent to gather FIEMAP information.
3909  */
3910 static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next,
3911                        struct ext4_ext_cache *newex, struct ext4_extent *ex,
3912                        void *data)
3913 {
3914         __u64   logical;
3915         __u64   physical;
3916         __u64   length;
3917         __u32   flags = 0;
3918         int             ret = 0;
3919         struct fiemap_extent_info *fieinfo = data;
3920         unsigned char blksize_bits;
3921
3922         blksize_bits = inode->i_sb->s_blocksize_bits;
3923         logical = (__u64)newex->ec_block << blksize_bits;
3924
3925         if (newex->ec_start == 0) {
3926                 /*
3927                  * No extent in extent-tree contains block @newex->ec_start,
3928                  * then the block may stay in 1)a hole or 2)delayed-extent.
3929                  *
3930                  * Holes or delayed-extents are processed as follows.
3931                  * 1. lookup dirty pages with specified range in pagecache.
3932                  *    If no page is got, then there is no delayed-extent and
3933                  *    return with EXT_CONTINUE.
3934                  * 2. find the 1st mapped buffer,
3935                  * 3. check if the mapped buffer is both in the request range
3936                  *    and a delayed buffer. If not, there is no delayed-extent,
3937                  *    then return.
3938                  * 4. a delayed-extent is found, the extent will be collected.
3939                  */
3940                 ext4_lblk_t     end = 0;
3941                 pgoff_t         last_offset;
3942                 pgoff_t         offset;
3943                 pgoff_t         index;
3944                 pgoff_t         start_index = 0;
3945                 struct page     **pages = NULL;
3946                 struct buffer_head *bh = NULL;
3947                 struct buffer_head *head = NULL;
3948                 unsigned int nr_pages = PAGE_SIZE / sizeof(struct page *);
3949
3950                 pages = kmalloc(PAGE_SIZE, GFP_KERNEL);
3951                 if (pages == NULL)
3952                         return -ENOMEM;
3953
3954                 offset = logical >> PAGE_SHIFT;
3955 repeat:
3956                 last_offset = offset;
3957                 head = NULL;
3958                 ret = find_get_pages_tag(inode->i_mapping, &offset,
3959                                         PAGECACHE_TAG_DIRTY, nr_pages, pages);
3960
3961                 if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
3962                         /* First time, try to find a mapped buffer. */
3963                         if (ret == 0) {
3964 out:
3965                                 for (index = 0; index < ret; index++)
3966                                         page_cache_release(pages[index]);
3967                                 /* just a hole. */
3968                                 kfree(pages);
3969                                 return EXT_CONTINUE;
3970                         }
3971                         index = 0;
3972
3973 next_page:
3974                         /* Try to find the 1st mapped buffer. */
3975                         end = ((__u64)pages[index]->index << PAGE_SHIFT) >>
3976                                   blksize_bits;
3977                         if (!page_has_buffers(pages[index]))
3978                                 goto out;
3979                         head = page_buffers(pages[index]);
3980                         if (!head)
3981                                 goto out;
3982
3983                         index++;
3984                         bh = head;
3985                         do {
3986                                 if (end >= newex->ec_block +
3987                                         newex->ec_len)
3988                                         /* The buffer is out of
3989                                          * the request range.
3990                                          */
3991                                         goto out;
3992
3993                                 if (buffer_mapped(bh) &&
3994                                     end >= newex->ec_block) {
3995                                         start_index = index - 1;
3996                                         /* get the 1st mapped buffer. */
3997                                         goto found_mapped_buffer;
3998                                 }
3999
4000                                 bh = bh->b_this_page;
4001                                 end++;
4002                         } while (bh != head);
4003
4004                         /* No mapped buffer in the range found in this page,
4005                          * We need to look up next page.
4006                          */
4007                         if (index >= ret) {
4008                                 /* There is no page left, but we need to limit
4009                                  * newex->ec_len.
4010                                  */
4011                                 newex->ec_len = end - newex->ec_block;
4012                                 goto out;
4013                         }
4014                         goto next_page;
4015                 } else {
4016                         /*Find contiguous delayed buffers. */
4017                         if (ret > 0 && pages[0]->index == last_offset)
4018                                 head = page_buffers(pages[0]);
4019                         bh = head;
4020                         index = 1;
4021                         start_index = 0;
4022                 }
4023
4024 found_mapped_buffer:
4025                 if (bh != NULL && buffer_delay(bh)) {
4026                         /* 1st or contiguous delayed buffer found. */
4027                         if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4028                                 /*
4029                                  * 1st delayed buffer found, record
4030                                  * the start of extent.
4031                                  */
4032                                 flags |= FIEMAP_EXTENT_DELALLOC;
4033                                 newex->ec_block = end;
4034                                 logical = (__u64)end << blksize_bits;
4035                         }
4036                         /* Find contiguous delayed buffers. */
4037                         do {
4038                                 if (!buffer_delay(bh))
4039                                         goto found_delayed_extent;
4040                                 bh = bh->b_this_page;
4041                                 end++;
4042                         } while (bh != head);
4043
4044                         for (; index < ret; index++) {
4045                                 if (!page_has_buffers(pages[index])) {
4046                                         bh = NULL;
4047                                         break;
4048                                 }
4049                                 head = page_buffers(pages[index]);
4050                                 if (!head) {
4051                                         bh = NULL;
4052                                         break;
4053                                 }
4054
4055                                 if (pages[index]->index !=
4056                                     pages[start_index]->index + index
4057                                     - start_index) {
4058                                         /* Blocks are not contiguous. */
4059                                         bh = NULL;
4060                                         break;
4061                                 }
4062                                 bh = head;
4063                                 do {
4064                                         if (!buffer_delay(bh))
4065                                                 /* Delayed-extent ends. */
4066                                                 goto found_delayed_extent;
4067                                         bh = bh->b_this_page;
4068                                         end++;
4069                                 } while (bh != head);
4070                         }
4071                 } else if (!(flags & FIEMAP_EXTENT_DELALLOC))
4072                         /* a hole found. */
4073                         goto out;
4074
4075 found_delayed_extent:
4076                 newex->ec_len = min(end - newex->ec_block,
4077                                                 (ext4_lblk_t)EXT_INIT_MAX_LEN);
4078                 if (ret == nr_pages && bh != NULL &&
4079                         newex->ec_len < EXT_INIT_MAX_LEN &&
4080                         buffer_delay(bh)) {
4081                         /* Have not collected an extent and continue. */
4082                         for (index = 0; index < ret; index++)
4083                                 page_cache_release(pages[index]);
4084                         goto repeat;
4085                 }
4086
4087                 for (index = 0; index < ret; index++)
4088                         page_cache_release(pages[index]);
4089                 kfree(pages);
4090         }
4091
4092         physical = (__u64)newex->ec_start << blksize_bits;
4093         length =   (__u64)newex->ec_len << blksize_bits;
4094
4095         if (ex && ext4_ext_is_uninitialized(ex))
4096                 flags |= FIEMAP_EXTENT_UNWRITTEN;
4097
4098         if (next == EXT_MAX_BLOCKS)
4099                 flags |= FIEMAP_EXTENT_LAST;
4100
4101         ret = fiemap_fill_next_extent(fieinfo, logical, physical,
4102                                         length, flags);
4103         if (ret < 0)
4104                 return ret;
4105         if (ret == 1)
4106                 return EXT_BREAK;
4107         return EXT_CONTINUE;
4108 }
4109
4110 /* fiemap flags we can handle specified here */
4111 #define EXT4_FIEMAP_FLAGS       (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
4112
4113 static int ext4_xattr_fiemap(struct inode *inode,
4114                                 struct fiemap_extent_info *fieinfo)
4115 {
4116         __u64 physical = 0;
4117         __u64 length;
4118         __u32 flags = FIEMAP_EXTENT_LAST;
4119         int blockbits = inode->i_sb->s_blocksize_bits;
4120         int error = 0;
4121
4122         /* in-inode? */
4123         if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
4124                 struct ext4_iloc iloc;
4125                 int offset;     /* offset of xattr in inode */
4126
4127                 error = ext4_get_inode_loc(inode, &iloc);
4128                 if (error)
4129                         return error;
4130                 physical = iloc.bh->b_blocknr << blockbits;
4131                 offset = EXT4_GOOD_OLD_INODE_SIZE +
4132                                 EXT4_I(inode)->i_extra_isize;
4133                 physical += offset;
4134                 length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
4135                 flags |= FIEMAP_EXTENT_DATA_INLINE;
4136                 brelse(iloc.bh);
4137         } else { /* external block */
4138                 physical = EXT4_I(inode)->i_file_acl << blockbits;
4139                 length = inode->i_sb->s_blocksize;
4140         }
4141
4142         if (physical)
4143                 error = fiemap_fill_next_extent(fieinfo, 0, physical,
4144                                                 length, flags);
4145         return (error < 0 ? error : 0);
4146 }
4147
4148 /*
4149  * ext4_ext_punch_hole
4150  *
4151  * Punches a hole of "length" bytes in a file starting
4152  * at byte "offset"
4153  *
4154  * @inode:  The inode of the file to punch a hole in
4155  * @offset: The starting byte offset of the hole
4156  * @length: The length of the hole
4157  *
4158  * Returns the number of blocks removed or negative on err
4159  */
4160 int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length)
4161 {
4162         struct inode *inode = file->f_path.dentry->d_inode;
4163         struct super_block *sb = inode->i_sb;
4164         struct ext4_ext_cache cache_ex;
4165         ext4_lblk_t first_block, last_block, num_blocks, iblock, max_blocks;
4166         struct address_space *mapping = inode->i_mapping;
4167         struct ext4_map_blocks map;
4168         handle_t *handle;
4169         loff_t first_page, last_page, page_len;
4170         loff_t first_page_offset, last_page_offset;
4171         int ret, credits, blocks_released, err = 0;
4172
4173         /* No need to punch hole beyond i_size */
4174         if (offset >= inode->i_size)
4175                 return 0;
4176
4177         /*
4178          * If the hole extends beyond i_size, set the hole
4179          * to end after the page that contains i_size
4180          */
4181         if (offset + length > inode->i_size) {
4182                 length = inode->i_size +
4183                    PAGE_CACHE_SIZE - (inode->i_size & (PAGE_CACHE_SIZE - 1)) -
4184                    offset;
4185         }
4186
4187         first_block = (offset + sb->s_blocksize - 1) >>
4188                 EXT4_BLOCK_SIZE_BITS(sb);
4189         last_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb);
4190
4191         first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
4192         last_page = (offset + length) >> PAGE_CACHE_SHIFT;
4193
4194         first_page_offset = first_page << PAGE_CACHE_SHIFT;
4195         last_page_offset = last_page << PAGE_CACHE_SHIFT;
4196
4197         /*
4198          * Write out all dirty pages to avoid race conditions
4199          * Then release them.
4200          */
4201         if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
4202                 err = filemap_write_and_wait_range(mapping,
4203                         offset, offset + length - 1);
4204
4205                 if (err)
4206                         return err;
4207         }
4208
4209         /* Now release the pages */
4210         if (last_page_offset > first_page_offset) {
4211                 truncate_inode_pages_range(mapping, first_page_offset,
4212                                            last_page_offset-1);
4213         }
4214
4215         /* finish any pending end_io work */
4216         ext4_flush_completed_IO(inode);
4217
4218         credits = ext4_writepage_trans_blocks(inode);
4219         handle = ext4_journal_start(inode, credits);
4220         if (IS_ERR(handle))
4221                 return PTR_ERR(handle);
4222
4223         err = ext4_orphan_add(handle, inode);
4224         if (err)
4225                 goto out;
4226
4227         /*
4228          * Now we need to zero out the non-page-aligned data in the
4229          * pages at the start and tail of the hole, and unmap the buffer
4230          * heads for the block aligned regions of the page that were
4231          * completely zeroed.
4232          */
4233         if (first_page > last_page) {
4234                 /*
4235                  * If the file space being truncated is contained within a page
4236                  * just zero out and unmap the middle of that page
4237                  */
4238                 err = ext4_discard_partial_page_buffers(handle,
4239                         mapping, offset, length, 0);
4240
4241                 if (err)
4242                         goto out;
4243         } else {
4244                 /*
4245                  * zero out and unmap the partial page that contains
4246                  * the start of the hole
4247                  */
4248                 page_len  = first_page_offset - offset;
4249                 if (page_len > 0) {
4250                         err = ext4_discard_partial_page_buffers(handle, mapping,
4251                                                    offset, page_len, 0);
4252                         if (err)
4253                                 goto out;
4254                 }
4255
4256                 /*
4257                  * zero out and unmap the partial page that contains
4258                  * the end of the hole
4259                  */
4260                 page_len = offset + length - last_page_offset;
4261                 if (page_len > 0) {
4262                         err = ext4_discard_partial_page_buffers(handle, mapping,
4263                                         last_page_offset, page_len, 0);
4264                         if (err)
4265                                 goto out;
4266                 }
4267         }
4268
4269
4270         /*
4271          * If i_size is contained in the last page, we need to
4272          * unmap and zero the partial page after i_size
4273          */
4274         if (inode->i_size >> PAGE_CACHE_SHIFT == last_page &&
4275            inode->i_size % PAGE_CACHE_SIZE != 0) {
4276
4277                 page_len = PAGE_CACHE_SIZE -
4278                         (inode->i_size & (PAGE_CACHE_SIZE - 1));
4279
4280                 if (page_len > 0) {
4281                         err = ext4_discard_partial_page_buffers(handle,
4282                           mapping, inode->i_size, page_len, 0);
4283
4284                         if (err)
4285                                 goto out;
4286                 }
4287         }
4288
4289         /* If there are no blocks to remove, return now */
4290         if (first_block >= last_block)
4291                 goto out;
4292
4293         down_write(&EXT4_I(inode)->i_data_sem);
4294         ext4_ext_invalidate_cache(inode);
4295         ext4_discard_preallocations(inode);
4296
4297         /*
4298          * Loop over all the blocks and identify blocks
4299          * that need to be punched out
4300          */
4301         iblock = first_block;
4302         blocks_released = 0;
4303         while (iblock < last_block) {
4304                 max_blocks = last_block - iblock;
4305                 num_blocks = 1;
4306                 memset(&map, 0, sizeof(map));
4307                 map.m_lblk = iblock;
4308                 map.m_len = max_blocks;
4309                 ret = ext4_ext_map_blocks(handle, inode, &map,
4310                         EXT4_GET_BLOCKS_PUNCH_OUT_EXT);
4311
4312                 if (ret > 0) {
4313                         blocks_released += ret;
4314                         num_blocks = ret;
4315                 } else if (ret == 0) {
4316                         /*
4317                          * If map blocks could not find the block,
4318                          * then it is in a hole.  If the hole was
4319                          * not already cached, then map blocks should
4320                          * put it in the cache.  So we can get the hole
4321                          * out of the cache
4322                          */
4323                         memset(&cache_ex, 0, sizeof(cache_ex));
4324                         if ((ext4_ext_check_cache(inode, iblock, &cache_ex)) &&
4325                                 !cache_ex.ec_start) {
4326
4327                                 /* The hole is cached */
4328                                 num_blocks = cache_ex.ec_block +
4329                                 cache_ex.ec_len - iblock;
4330
4331                         } else {
4332                                 /* The block could not be identified */
4333                                 err = -EIO;
4334                                 break;
4335                         }
4336                 } else {
4337                         /* Map blocks error */
4338                         err = ret;
4339                         break;
4340                 }
4341
4342                 if (num_blocks == 0) {
4343                         /* This condition should never happen */
4344                         ext_debug("Block lookup failed");
4345                         err = -EIO;
4346                         break;
4347                 }
4348
4349                 iblock += num_blocks;
4350         }
4351
4352         if (blocks_released > 0) {
4353                 ext4_ext_invalidate_cache(inode);
4354                 ext4_discard_preallocations(inode);
4355         }
4356
4357         if (IS_SYNC(inode))
4358                 ext4_handle_sync(handle);
4359
4360         up_write(&EXT4_I(inode)->i_data_sem);
4361
4362 out:
4363         ext4_orphan_del(handle, inode);
4364         inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4365         ext4_mark_inode_dirty(handle, inode);
4366         ext4_journal_stop(handle);
4367         return err;
4368 }
4369 int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
4370                 __u64 start, __u64 len)
4371 {
4372         ext4_lblk_t start_blk;
4373         int error = 0;
4374
4375         /* fallback to generic here if not in extents fmt */
4376         if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
4377                 return generic_block_fiemap(inode, fieinfo, start, len,
4378                         ext4_get_block);
4379
4380         if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
4381                 return -EBADR;
4382
4383         if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
4384                 error = ext4_xattr_fiemap(inode, fieinfo);
4385         } else {
4386                 ext4_lblk_t len_blks;
4387                 __u64 last_blk;
4388
4389                 start_blk = start >> inode->i_sb->s_blocksize_bits;
4390                 last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
4391                 if (last_blk >= EXT_MAX_BLOCKS)
4392                         last_blk = EXT_MAX_BLOCKS-1;
4393                 len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
4394
4395                 /*
4396                  * Walk the extent tree gathering extent information.
4397                  * ext4_ext_fiemap_cb will push extents back to user.
4398                  */
4399                 error = ext4_ext_walk_space(inode, start_blk, len_blks,
4400                                           ext4_ext_fiemap_cb, fieinfo);
4401         }
4402
4403         return error;
4404 }