linux/fs/ext4/extents.c
<<
>>
Prefs
   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
  46#include <trace/events/ext4.h>
  47
  48static int ext4_split_extent(handle_t *handle,
  49                                struct inode *inode,
  50                                struct ext4_ext_path *path,
  51                                struct ext4_map_blocks *map,
  52                                int split_flag,
  53                                int flags);
  54
  55static int ext4_ext_truncate_extend_restart(handle_t *handle,
  56                                            struct inode *inode,
  57                                            int needed)
  58{
  59        int err;
  60
  61        if (!ext4_handle_valid(handle))
  62                return 0;
  63        if (handle->h_buffer_credits > needed)
  64                return 0;
  65        err = ext4_journal_extend(handle, needed);
  66        if (err <= 0)
  67                return err;
  68        err = ext4_truncate_restart_trans(handle, inode, needed);
  69        if (err == 0)
  70                err = -EAGAIN;
  71
  72        return err;
  73}
  74
  75/*
  76 * could return:
  77 *  - EROFS
  78 *  - ENOMEM
  79 */
  80static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
  81                                struct ext4_ext_path *path)
  82{
  83        if (path->p_bh) {
  84                /* path points to block */
  85                return ext4_journal_get_write_access(handle, path->p_bh);
  86        }
  87        /* path points to leaf/index in inode body */
  88        /* we use in-core data, no need to protect them */
  89        return 0;
  90}
  91
  92/*
  93 * could return:
  94 *  - EROFS
  95 *  - ENOMEM
  96 *  - EIO
  97 */
  98#define ext4_ext_dirty(handle, inode, path) \
  99                __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path))
 100static int __ext4_ext_dirty(const char *where, unsigned int line,
 101                            handle_t *handle, struct inode *inode,
 102                            struct ext4_ext_path *path)
 103{
 104        int err;
 105        if (path->p_bh) {
 106                /* path points to block */
 107                err = __ext4_handle_dirty_metadata(where, line, handle,
 108                                                   inode, path->p_bh);
 109        } else {
 110                /* path points to leaf/index in inode body */
 111                err = ext4_mark_inode_dirty(handle, inode);
 112        }
 113        return err;
 114}
 115
 116static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
 117                              struct ext4_ext_path *path,
 118                              ext4_lblk_t block)
 119{
 120        if (path) {
 121                int depth = path->p_depth;
 122                struct ext4_extent *ex;
 123
 124                /*
 125                 * Try to predict block placement assuming that we are
 126                 * filling in a file which will eventually be
 127                 * non-sparse --- i.e., in the case of libbfd writing
 128                 * an ELF object sections out-of-order but in a way
 129                 * the eventually results in a contiguous object or
 130                 * executable file, or some database extending a table
 131                 * space file.  However, this is actually somewhat
 132                 * non-ideal if we are writing a sparse file such as
 133                 * qemu or KVM writing a raw image file that is going
 134                 * to stay fairly sparse, since it will end up
 135                 * fragmenting the file system's free space.  Maybe we
 136                 * should have some hueristics or some way to allow
 137                 * userspace to pass a hint to file system,
 138                 * especially if the latter case turns out to be
 139                 * common.
 140                 */
 141                ex = path[depth].p_ext;
 142                if (ex) {
 143                        ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
 144                        ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block);
 145
 146                        if (block > ext_block)
 147                                return ext_pblk + (block - ext_block);
 148                        else
 149                                return ext_pblk - (ext_block - block);
 150                }
 151
 152                /* it looks like index is empty;
 153                 * try to find starting block from index itself */
 154                if (path[depth].p_bh)
 155                        return path[depth].p_bh->b_blocknr;
 156        }
 157
 158        /* OK. use inode's group */
 159        return ext4_inode_to_goal_block(inode);
 160}
 161
 162/*
 163 * Allocation for a meta data block
 164 */
 165static ext4_fsblk_t
 166ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
 167                        struct ext4_ext_path *path,
 168                        struct ext4_extent *ex, int *err, unsigned int flags)
 169{
 170        ext4_fsblk_t goal, newblock;
 171
 172        goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
 173        newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
 174                                        NULL, err);
 175        return newblock;
 176}
 177
 178static inline int ext4_ext_space_block(struct inode *inode, int check)
 179{
 180        int size;
 181
 182        size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 183                        / sizeof(struct ext4_extent);
 184#ifdef AGGRESSIVE_TEST
 185        if (!check && size > 6)
 186                size = 6;
 187#endif
 188        return size;
 189}
 190
 191static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
 192{
 193        int size;
 194
 195        size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 196                        / sizeof(struct ext4_extent_idx);
 197#ifdef AGGRESSIVE_TEST
 198        if (!check && size > 5)
 199                size = 5;
 200#endif
 201        return size;
 202}
 203
 204static inline int ext4_ext_space_root(struct inode *inode, int check)
 205{
 206        int size;
 207
 208        size = sizeof(EXT4_I(inode)->i_data);
 209        size -= sizeof(struct ext4_extent_header);
 210        size /= sizeof(struct ext4_extent);
 211#ifdef AGGRESSIVE_TEST
 212        if (!check && size > 3)
 213                size = 3;
 214#endif
 215        return size;
 216}
 217
 218static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
 219{
 220        int size;
 221
 222        size = sizeof(EXT4_I(inode)->i_data);
 223        size -= sizeof(struct ext4_extent_header);
 224        size /= sizeof(struct ext4_extent_idx);
 225#ifdef AGGRESSIVE_TEST
 226        if (!check && size > 4)
 227                size = 4;
 228#endif
 229        return size;
 230}
 231
 232/*
 233 * Calculate the number of metadata blocks needed
 234 * to allocate @blocks
 235 * Worse case is one block per extent
 236 */
 237int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
 238{
 239        struct ext4_inode_info *ei = EXT4_I(inode);
 240        int idxs;
 241
 242        idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
 243                / sizeof(struct ext4_extent_idx));
 244
 245        /*
 246         * If the new delayed allocation block is contiguous with the
 247         * previous da block, it can share index blocks with the
 248         * previous block, so we only need to allocate a new index
 249         * block every idxs leaf blocks.  At ldxs**2 blocks, we need
 250         * an additional index block, and at ldxs**3 blocks, yet
 251         * another index blocks.
 252         */
 253        if (ei->i_da_metadata_calc_len &&
 254            ei->i_da_metadata_calc_last_lblock+1 == lblock) {
 255                int num = 0;
 256
 257                if ((ei->i_da_metadata_calc_len % idxs) == 0)
 258                        num++;
 259                if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
 260                        num++;
 261                if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
 262                        num++;
 263                        ei->i_da_metadata_calc_len = 0;
 264                } else
 265                        ei->i_da_metadata_calc_len++;
 266                ei->i_da_metadata_calc_last_lblock++;
 267                return num;
 268        }
 269
 270        /*
 271         * In the worst case we need a new set of index blocks at
 272         * every level of the inode's extent tree.
 273         */
 274        ei->i_da_metadata_calc_len = 1;
 275        ei->i_da_metadata_calc_last_lblock = lblock;
 276        return ext_depth(inode) + 1;
 277}
 278
 279static int
 280ext4_ext_max_entries(struct inode *inode, int depth)
 281{
 282        int max;
 283
 284        if (depth == ext_depth(inode)) {
 285                if (depth == 0)
 286                        max = ext4_ext_space_root(inode, 1);
 287                else
 288                        max = ext4_ext_space_root_idx(inode, 1);
 289        } else {
 290                if (depth == 0)
 291                        max = ext4_ext_space_block(inode, 1);
 292                else
 293                        max = ext4_ext_space_block_idx(inode, 1);
 294        }
 295
 296        return max;
 297}
 298
 299static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
 300{
 301        ext4_fsblk_t block = ext4_ext_pblock(ext);
 302        int len = ext4_ext_get_actual_len(ext);
 303
 304        return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
 305}
 306
 307static int ext4_valid_extent_idx(struct inode *inode,
 308                                struct ext4_extent_idx *ext_idx)
 309{
 310        ext4_fsblk_t block = ext4_idx_pblock(ext_idx);
 311
 312        return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
 313}
 314
 315static int ext4_valid_extent_entries(struct inode *inode,
 316                                struct ext4_extent_header *eh,
 317                                int depth)
 318{
 319        unsigned short entries;
 320        if (eh->eh_entries == 0)
 321                return 1;
 322
 323        entries = le16_to_cpu(eh->eh_entries);
 324
 325        if (depth == 0) {
 326                /* leaf entries */
 327                struct ext4_extent *ext = EXT_FIRST_EXTENT(eh);
 328                while (entries) {
 329                        if (!ext4_valid_extent(inode, ext))
 330                                return 0;
 331                        ext++;
 332                        entries--;
 333                }
 334        } else {
 335                struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh);
 336                while (entries) {
 337                        if (!ext4_valid_extent_idx(inode, ext_idx))
 338                                return 0;
 339                        ext_idx++;
 340                        entries--;
 341                }
 342        }
 343        return 1;
 344}
 345
 346static int __ext4_ext_check(const char *function, unsigned int line,
 347                            struct inode *inode, struct ext4_extent_header *eh,
 348                            int depth)
 349{
 350        const char *error_msg;
 351        int max = 0;
 352
 353        if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
 354                error_msg = "invalid magic";
 355                goto corrupted;
 356        }
 357        if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
 358                error_msg = "unexpected eh_depth";
 359                goto corrupted;
 360        }
 361        if (unlikely(eh->eh_max == 0)) {
 362                error_msg = "invalid eh_max";
 363                goto corrupted;
 364        }
 365        max = ext4_ext_max_entries(inode, depth);
 366        if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
 367                error_msg = "too large eh_max";
 368                goto corrupted;
 369        }
 370        if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
 371                error_msg = "invalid eh_entries";
 372                goto corrupted;
 373        }
 374        if (!ext4_valid_extent_entries(inode, eh, depth)) {
 375                error_msg = "invalid extent entries";
 376                goto corrupted;
 377        }
 378        return 0;
 379
 380corrupted:
 381        ext4_error_inode(inode, function, line, 0,
 382                        "bad header/extent: %s - magic %x, "
 383                        "entries %u, max %u(%u), depth %u(%u)",
 384                        error_msg, le16_to_cpu(eh->eh_magic),
 385                        le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
 386                        max, le16_to_cpu(eh->eh_depth), depth);
 387
 388        return -EIO;
 389}
 390
 391#define ext4_ext_check(inode, eh, depth)        \
 392        __ext4_ext_check(__func__, __LINE__, inode, eh, depth)
 393
 394int ext4_ext_check_inode(struct inode *inode)
 395{
 396        return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
 397}
 398
 399#ifdef EXT_DEBUG
 400static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
 401{
 402        int k, l = path->p_depth;
 403
 404        ext_debug("path:");
 405        for (k = 0; k <= l; k++, path++) {
 406                if (path->p_idx) {
 407                  ext_debug("  %d->%llu", le32_to_cpu(path->p_idx->ei_block),
 408                            ext4_idx_pblock(path->p_idx));
 409                } else if (path->p_ext) {
 410                        ext_debug("  %d:[%d]%d:%llu ",
 411                                  le32_to_cpu(path->p_ext->ee_block),
 412                                  ext4_ext_is_uninitialized(path->p_ext),
 413                                  ext4_ext_get_actual_len(path->p_ext),
 414                                  ext4_ext_pblock(path->p_ext));
 415                } else
 416                        ext_debug("  []");
 417        }
 418        ext_debug("\n");
 419}
 420
 421static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
 422{
 423        int depth = ext_depth(inode);
 424        struct ext4_extent_header *eh;
 425        struct ext4_extent *ex;
 426        int i;
 427
 428        if (!path)
 429                return;
 430
 431        eh = path[depth].p_hdr;
 432        ex = EXT_FIRST_EXTENT(eh);
 433
 434        ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
 435
 436        for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
 437                ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
 438                          ext4_ext_is_uninitialized(ex),
 439                          ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
 440        }
 441        ext_debug("\n");
 442}
 443
 444static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path,
 445                        ext4_fsblk_t newblock, int level)
 446{
 447        int depth = ext_depth(inode);
 448        struct ext4_extent *ex;
 449
 450        if (depth != level) {
 451                struct ext4_extent_idx *idx;
 452                idx = path[level].p_idx;
 453                while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
 454                        ext_debug("%d: move %d:%llu in new index %llu\n", level,
 455                                        le32_to_cpu(idx->ei_block),
 456                                        ext4_idx_pblock(idx),
 457                                        newblock);
 458                        idx++;
 459                }
 460
 461                return;
 462        }
 463
 464        ex = path[depth].p_ext;
 465        while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
 466                ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
 467                                le32_to_cpu(ex->ee_block),
 468                                ext4_ext_pblock(ex),
 469                                ext4_ext_is_uninitialized(ex),
 470                                ext4_ext_get_actual_len(ex),
 471                                newblock);
 472                ex++;
 473        }
 474}
 475
 476#else
 477#define ext4_ext_show_path(inode, path)
 478#define ext4_ext_show_leaf(inode, path)
 479#define ext4_ext_show_move(inode, path, newblock, level)
 480#endif
 481
 482void ext4_ext_drop_refs(struct ext4_ext_path *path)
 483{
 484        int depth = path->p_depth;
 485        int i;
 486
 487        for (i = 0; i <= depth; i++, path++)
 488                if (path->p_bh) {
 489                        brelse(path->p_bh);
 490                        path->p_bh = NULL;
 491                }
 492}
 493
 494/*
 495 * ext4_ext_binsearch_idx:
 496 * binary search for the closest index of the given block
 497 * the header must be checked before calling this
 498 */
 499static void
 500ext4_ext_binsearch_idx(struct inode *inode,
 501                        struct ext4_ext_path *path, ext4_lblk_t block)
 502{
 503        struct ext4_extent_header *eh = path->p_hdr;
 504        struct ext4_extent_idx *r, *l, *m;
 505
 506
 507        ext_debug("binsearch for %u(idx):  ", block);
 508
 509        l = EXT_FIRST_INDEX(eh) + 1;
 510        r = EXT_LAST_INDEX(eh);
 511        while (l <= r) {
 512                m = l + (r - l) / 2;
 513                if (block < le32_to_cpu(m->ei_block))
 514                        r = m - 1;
 515                else
 516                        l = m + 1;
 517                ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
 518                                m, le32_to_cpu(m->ei_block),
 519                                r, le32_to_cpu(r->ei_block));
 520        }
 521
 522        path->p_idx = l - 1;
 523        ext_debug("  -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
 524                  ext4_idx_pblock(path->p_idx));
 525
 526#ifdef CHECK_BINSEARCH
 527        {
 528                struct ext4_extent_idx *chix, *ix;
 529                int k;
 530
 531                chix = ix = EXT_FIRST_INDEX(eh);
 532                for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
 533                  if (k != 0 &&
 534                      le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
 535                                printk(KERN_DEBUG "k=%d, ix=0x%p, "
 536                                       "first=0x%p\n", k,
 537                                       ix, EXT_FIRST_INDEX(eh));
 538                                printk(KERN_DEBUG "%u <= %u\n",
 539                                       le32_to_cpu(ix->ei_block),
 540                                       le32_to_cpu(ix[-1].ei_block));
 541                        }
 542                        BUG_ON(k && le32_to_cpu(ix->ei_block)
 543                                           <= le32_to_cpu(ix[-1].ei_block));
 544                        if (block < le32_to_cpu(ix->ei_block))
 545                                break;
 546                        chix = ix;
 547                }
 548                BUG_ON(chix != path->p_idx);
 549        }
 550#endif
 551
 552}
 553
 554/*
 555 * ext4_ext_binsearch:
 556 * binary search for closest extent of the given block
 557 * the header must be checked before calling this
 558 */
 559static void
 560ext4_ext_binsearch(struct inode *inode,
 561                struct ext4_ext_path *path, ext4_lblk_t block)
 562{
 563        struct ext4_extent_header *eh = path->p_hdr;
 564        struct ext4_extent *r, *l, *m;
 565
 566        if (eh->eh_entries == 0) {
 567                /*
 568                 * this leaf is empty:
 569                 * we get such a leaf in split/add case
 570                 */
 571                return;
 572        }
 573
 574        ext_debug("binsearch for %u:  ", block);
 575
 576        l = EXT_FIRST_EXTENT(eh) + 1;
 577        r = EXT_LAST_EXTENT(eh);
 578
 579        while (l <= r) {
 580                m = l + (r - l) / 2;
 581                if (block < le32_to_cpu(m->ee_block))
 582                        r = m - 1;
 583                else
 584                        l = m + 1;
 585                ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
 586                                m, le32_to_cpu(m->ee_block),
 587                                r, le32_to_cpu(r->ee_block));
 588        }
 589
 590        path->p_ext = l - 1;
 591        ext_debug("  -> %d:%llu:[%d]%d ",
 592                        le32_to_cpu(path->p_ext->ee_block),
 593                        ext4_ext_pblock(path->p_ext),
 594                        ext4_ext_is_uninitialized(path->p_ext),
 595                        ext4_ext_get_actual_len(path->p_ext));
 596
 597#ifdef CHECK_BINSEARCH
 598        {
 599                struct ext4_extent *chex, *ex;
 600                int k;
 601
 602                chex = ex = EXT_FIRST_EXTENT(eh);
 603                for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
 604                        BUG_ON(k && le32_to_cpu(ex->ee_block)
 605                                          <= le32_to_cpu(ex[-1].ee_block));
 606                        if (block < le32_to_cpu(ex->ee_block))
 607                                break;
 608                        chex = ex;
 609                }
 610                BUG_ON(chex != path->p_ext);
 611        }
 612#endif
 613
 614}
 615
 616int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
 617{
 618        struct ext4_extent_header *eh;
 619
 620        eh = ext_inode_hdr(inode);
 621        eh->eh_depth = 0;
 622        eh->eh_entries = 0;
 623        eh->eh_magic = EXT4_EXT_MAGIC;
 624        eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
 625        ext4_mark_inode_dirty(handle, inode);
 626        ext4_ext_invalidate_cache(inode);
 627        return 0;
 628}
 629
 630struct ext4_ext_path *
 631ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
 632                                        struct ext4_ext_path *path)
 633{
 634        struct ext4_extent_header *eh;
 635        struct buffer_head *bh;
 636        short int depth, i, ppos = 0, alloc = 0;
 637
 638        eh = ext_inode_hdr(inode);
 639        depth = ext_depth(inode);
 640
 641        /* account possible depth increase */
 642        if (!path) {
 643                path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
 644                                GFP_NOFS);
 645                if (!path)
 646                        return ERR_PTR(-ENOMEM);
 647                alloc = 1;
 648        }
 649        path[0].p_hdr = eh;
 650        path[0].p_bh = NULL;
 651
 652        i = depth;
 653        /* walk through the tree */
 654        while (i) {
 655                int need_to_validate = 0;
 656
 657                ext_debug("depth %d: num %d, max %d\n",
 658                          ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
 659
 660                ext4_ext_binsearch_idx(inode, path + ppos, block);
 661                path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
 662                path[ppos].p_depth = i;
 663                path[ppos].p_ext = NULL;
 664
 665                bh = sb_getblk(inode->i_sb, path[ppos].p_block);
 666                if (unlikely(!bh))
 667                        goto err;
 668                if (!bh_uptodate_or_lock(bh)) {
 669                        trace_ext4_ext_load_extent(inode, block,
 670                                                path[ppos].p_block);
 671                        if (bh_submit_read(bh) < 0) {
 672                                put_bh(bh);
 673                                goto err;
 674                        }
 675                        /* validate the extent entries */
 676                        need_to_validate = 1;
 677                }
 678                eh = ext_block_hdr(bh);
 679                ppos++;
 680                if (unlikely(ppos > depth)) {
 681                        put_bh(bh);
 682                        EXT4_ERROR_INODE(inode,
 683                                         "ppos %d > depth %d", ppos, depth);
 684                        goto err;
 685                }
 686                path[ppos].p_bh = bh;
 687                path[ppos].p_hdr = eh;
 688                i--;
 689
 690                if (need_to_validate && ext4_ext_check(inode, eh, i))
 691                        goto err;
 692        }
 693
 694        path[ppos].p_depth = i;
 695        path[ppos].p_ext = NULL;
 696        path[ppos].p_idx = NULL;
 697
 698        /* find extent */
 699        ext4_ext_binsearch(inode, path + ppos, block);
 700        /* if not an empty leaf */
 701        if (path[ppos].p_ext)
 702                path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);
 703
 704        ext4_ext_show_path(inode, path);
 705
 706        return path;
 707
 708err:
 709        ext4_ext_drop_refs(path);
 710        if (alloc)
 711                kfree(path);
 712        return ERR_PTR(-EIO);
 713}
 714
 715/*
 716 * ext4_ext_insert_index:
 717 * insert new index [@logical;@ptr] into the block at @curp;
 718 * check where to insert: before @curp or after @curp
 719 */
 720static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
 721                                 struct ext4_ext_path *curp,
 722                                 int logical, ext4_fsblk_t ptr)
 723{
 724        struct ext4_extent_idx *ix;
 725        int len, err;
 726
 727        err = ext4_ext_get_access(handle, inode, curp);
 728        if (err)
 729                return err;
 730
 731        if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
 732                EXT4_ERROR_INODE(inode,
 733                                 "logical %d == ei_block %d!",
 734                                 logical, le32_to_cpu(curp->p_idx->ei_block));
 735                return -EIO;
 736        }
 737
 738        if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
 739                             >= le16_to_cpu(curp->p_hdr->eh_max))) {
 740                EXT4_ERROR_INODE(inode,
 741                                 "eh_entries %d >= eh_max %d!",
 742                                 le16_to_cpu(curp->p_hdr->eh_entries),
 743                                 le16_to_cpu(curp->p_hdr->eh_max));
 744                return -EIO;
 745        }
 746
 747        if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
 748                /* insert after */
 749                ext_debug("insert new index %d after: %llu\n", logical, ptr);
 750                ix = curp->p_idx + 1;
 751        } else {
 752                /* insert before */
 753                ext_debug("insert new index %d before: %llu\n", logical, ptr);
 754                ix = curp->p_idx;
 755        }
 756
 757        len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1;
 758        BUG_ON(len < 0);
 759        if (len > 0) {
 760                ext_debug("insert new index %d: "
 761                                "move %d indices from 0x%p to 0x%p\n",
 762                                logical, len, ix, ix + 1);
 763                memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx));
 764        }
 765
 766        if (unlikely(ix > EXT_MAX_INDEX(curp->p_hdr))) {
 767                EXT4_ERROR_INODE(inode, "ix > EXT_MAX_INDEX!");
 768                return -EIO;
 769        }
 770
 771        ix->ei_block = cpu_to_le32(logical);
 772        ext4_idx_store_pblock(ix, ptr);
 773        le16_add_cpu(&curp->p_hdr->eh_entries, 1);
 774
 775        if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
 776                EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
 777                return -EIO;
 778        }
 779
 780        err = ext4_ext_dirty(handle, inode, curp);
 781        ext4_std_error(inode->i_sb, err);
 782
 783        return err;
 784}
 785
 786/*
 787 * ext4_ext_split:
 788 * inserts new subtree into the path, using free index entry
 789 * at depth @at:
 790 * - allocates all needed blocks (new leaf and all intermediate index blocks)
 791 * - makes decision where to split
 792 * - moves remaining extents and index entries (right to the split point)
 793 *   into the newly allocated blocks
 794 * - initializes subtree
 795 */
 796static int ext4_ext_split(handle_t *handle, struct inode *inode,
 797                          unsigned int flags,
 798                          struct ext4_ext_path *path,
 799                          struct ext4_extent *newext, int at)
 800{
 801        struct buffer_head *bh = NULL;
 802        int depth = ext_depth(inode);
 803        struct ext4_extent_header *neh;
 804        struct ext4_extent_idx *fidx;
 805        int i = at, k, m, a;
 806        ext4_fsblk_t newblock, oldblock;
 807        __le32 border;
 808        ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
 809        int err = 0;
 810
 811        /* make decision: where to split? */
 812        /* FIXME: now decision is simplest: at current extent */
 813
 814        /* if current leaf will be split, then we should use
 815         * border from split point */
 816        if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
 817                EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
 818                return -EIO;
 819        }
 820        if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
 821                border = path[depth].p_ext[1].ee_block;
 822                ext_debug("leaf will be split."
 823                                " next leaf starts at %d\n",
 824                                  le32_to_cpu(border));
 825        } else {
 826                border = newext->ee_block;
 827                ext_debug("leaf will be added."
 828                                " next leaf starts at %d\n",
 829                                le32_to_cpu(border));
 830        }
 831
 832        /*
 833         * If error occurs, then we break processing
 834         * and mark filesystem read-only. index won't
 835         * be inserted and tree will be in consistent
 836         * state. Next mount will repair buffers too.
 837         */
 838
 839        /*
 840         * Get array to track all allocated blocks.
 841         * We need this to handle errors and free blocks
 842         * upon them.
 843         */
 844        ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
 845        if (!ablocks)
 846                return -ENOMEM;
 847
 848        /* allocate all needed blocks */
 849        ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
 850        for (a = 0; a < depth - at; a++) {
 851                newblock = ext4_ext_new_meta_block(handle, inode, path,
 852                                                   newext, &err, flags);
 853                if (newblock == 0)
 854                        goto cleanup;
 855                ablocks[a] = newblock;
 856        }
 857
 858        /* initialize new leaf */
 859        newblock = ablocks[--a];
 860        if (unlikely(newblock == 0)) {
 861                EXT4_ERROR_INODE(inode, "newblock == 0!");
 862                err = -EIO;
 863                goto cleanup;
 864        }
 865        bh = sb_getblk(inode->i_sb, newblock);
 866        if (!bh) {
 867                err = -EIO;
 868                goto cleanup;
 869        }
 870        lock_buffer(bh);
 871
 872        err = ext4_journal_get_create_access(handle, bh);
 873        if (err)
 874                goto cleanup;
 875
 876        neh = ext_block_hdr(bh);
 877        neh->eh_entries = 0;
 878        neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
 879        neh->eh_magic = EXT4_EXT_MAGIC;
 880        neh->eh_depth = 0;
 881
 882        /* move remainder of path[depth] to the new leaf */
 883        if (unlikely(path[depth].p_hdr->eh_entries !=
 884                     path[depth].p_hdr->eh_max)) {
 885                EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
 886                                 path[depth].p_hdr->eh_entries,
 887                                 path[depth].p_hdr->eh_max);
 888                err = -EIO;
 889                goto cleanup;
 890        }
 891        /* start copy from next extent */
 892        m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++;
 893        ext4_ext_show_move(inode, path, newblock, depth);
 894        if (m) {
 895                struct ext4_extent *ex;
 896                ex = EXT_FIRST_EXTENT(neh);
 897                memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m);
 898                le16_add_cpu(&neh->eh_entries, m);
 899        }
 900
 901        set_buffer_uptodate(bh);
 902        unlock_buffer(bh);
 903
 904        err = ext4_handle_dirty_metadata(handle, inode, bh);
 905        if (err)
 906                goto cleanup;
 907        brelse(bh);
 908        bh = NULL;
 909
 910        /* correct old leaf */
 911        if (m) {
 912                err = ext4_ext_get_access(handle, inode, path + depth);
 913                if (err)
 914                        goto cleanup;
 915                le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
 916                err = ext4_ext_dirty(handle, inode, path + depth);
 917                if (err)
 918                        goto cleanup;
 919
 920        }
 921
 922        /* create intermediate indexes */
 923        k = depth - at - 1;
 924        if (unlikely(k < 0)) {
 925                EXT4_ERROR_INODE(inode, "k %d < 0!", k);
 926                err = -EIO;
 927                goto cleanup;
 928        }
 929        if (k)
 930                ext_debug("create %d intermediate indices\n", k);
 931        /* insert new index into current index block */
 932        /* current depth stored in i var */
 933        i = depth - 1;
 934        while (k--) {
 935                oldblock = newblock;
 936                newblock = ablocks[--a];
 937                bh = sb_getblk(inode->i_sb, newblock);
 938                if (!bh) {
 939                        err = -EIO;
 940                        goto cleanup;
 941                }
 942                lock_buffer(bh);
 943
 944                err = ext4_journal_get_create_access(handle, bh);
 945                if (err)
 946                        goto cleanup;
 947
 948                neh = ext_block_hdr(bh);
 949                neh->eh_entries = cpu_to_le16(1);
 950                neh->eh_magic = EXT4_EXT_MAGIC;
 951                neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
 952                neh->eh_depth = cpu_to_le16(depth - i);
 953                fidx = EXT_FIRST_INDEX(neh);
 954                fidx->ei_block = border;
 955                ext4_idx_store_pblock(fidx, oldblock);
 956
 957                ext_debug("int.index at %d (block %llu): %u -> %llu\n",
 958                                i, newblock, le32_to_cpu(border), oldblock);
 959
 960                /* move remainder of path[i] to the new index block */
 961                if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
 962                                        EXT_LAST_INDEX(path[i].p_hdr))) {
 963                        EXT4_ERROR_INODE(inode,
 964                                         "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
 965                                         le32_to_cpu(path[i].p_ext->ee_block));
 966                        err = -EIO;
 967                        goto cleanup;
 968                }
 969                /* start copy indexes */
 970                m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
 971                ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
 972                                EXT_MAX_INDEX(path[i].p_hdr));
 973                ext4_ext_show_move(inode, path, newblock, i);
 974                if (m) {
 975                        memmove(++fidx, path[i].p_idx,
 976                                sizeof(struct ext4_extent_idx) * m);
 977                        le16_add_cpu(&neh->eh_entries, m);
 978                }
 979                set_buffer_uptodate(bh);
 980                unlock_buffer(bh);
 981
 982                err = ext4_handle_dirty_metadata(handle, inode, bh);
 983                if (err)
 984                        goto cleanup;
 985                brelse(bh);
 986                bh = NULL;
 987
 988                /* correct old index */
 989                if (m) {
 990                        err = ext4_ext_get_access(handle, inode, path + i);
 991                        if (err)
 992                                goto cleanup;
 993                        le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
 994                        err = ext4_ext_dirty(handle, inode, path + i);
 995                        if (err)
 996                                goto cleanup;
 997                }
 998
 999                i--;
1000        }
1001
1002        /* insert new index */
1003        err = ext4_ext_insert_index(handle, inode, path + at,
1004                                    le32_to_cpu(border), newblock);
1005
1006cleanup:
1007        if (bh) {
1008                if (buffer_locked(bh))
1009                        unlock_buffer(bh);
1010                brelse(bh);
1011        }
1012
1013        if (err) {
1014                /* free all allocated blocks in error case */
1015                for (i = 0; i < depth; i++) {
1016                        if (!ablocks[i])
1017                                continue;
1018                        ext4_free_blocks(handle, inode, NULL, ablocks[i], 1,
1019                                         EXT4_FREE_BLOCKS_METADATA);
1020                }
1021        }
1022        kfree(ablocks);
1023
1024        return err;
1025}
1026
1027/*
1028 * ext4_ext_grow_indepth:
1029 * implements tree growing procedure:
1030 * - allocates new block
1031 * - moves top-level data (index block or leaf) into the new block
1032 * - initializes new top-level, creating index that points to the
1033 *   just created block
1034 */
1035static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
1036                                 unsigned int flags,
1037                                 struct ext4_extent *newext)
1038{
1039        struct ext4_extent_header *neh;
1040        struct buffer_head *bh;
1041        ext4_fsblk_t newblock;
1042        int err = 0;
1043
1044        newblock = ext4_ext_new_meta_block(handle, inode, NULL,
1045                newext, &err, flags);
1046        if (newblock == 0)
1047                return err;
1048
1049        bh = sb_getblk(inode->i_sb, newblock);
1050        if (!bh) {
1051                err = -EIO;
1052                ext4_std_error(inode->i_sb, err);
1053                return err;
1054        }
1055        lock_buffer(bh);
1056
1057        err = ext4_journal_get_create_access(handle, bh);
1058        if (err) {
1059                unlock_buffer(bh);
1060                goto out;
1061        }
1062
1063        /* move top-level index/leaf into new block */
1064        memmove(bh->b_data, EXT4_I(inode)->i_data,
1065                sizeof(EXT4_I(inode)->i_data));
1066
1067        /* set size of new block */
1068        neh = ext_block_hdr(bh);
1069        /* old root could have indexes or leaves
1070         * so calculate e_max right way */
1071        if (ext_depth(inode))
1072                neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
1073        else
1074                neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
1075        neh->eh_magic = EXT4_EXT_MAGIC;
1076        set_buffer_uptodate(bh);
1077        unlock_buffer(bh);
1078
1079        err = ext4_handle_dirty_metadata(handle, inode, bh);
1080        if (err)
1081                goto out;
1082
1083        /* Update top-level index: num,max,pointer */
1084        neh = ext_inode_hdr(inode);
1085        neh->eh_entries = cpu_to_le16(1);
1086        ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock);
1087        if (neh->eh_depth == 0) {
1088                /* Root extent block becomes index block */
1089                neh->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
1090                EXT_FIRST_INDEX(neh)->ei_block =
1091                        EXT_FIRST_EXTENT(neh)->ee_block;
1092        }
1093        ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
1094                  le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
1095                  le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
1096                  ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
1097
1098        neh->eh_depth = cpu_to_le16(le16_to_cpu(neh->eh_depth) + 1);
1099        ext4_mark_inode_dirty(handle, inode);
1100out:
1101        brelse(bh);
1102
1103        return err;
1104}
1105
1106/*
1107 * ext4_ext_create_new_leaf:
1108 * finds empty index and adds new leaf.
1109 * if no free index is found, then it requests in-depth growing.
1110 */
1111static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
1112                                    unsigned int flags,
1113                                    struct ext4_ext_path *path,
1114                                    struct ext4_extent *newext)
1115{
1116        struct ext4_ext_path *curp;
1117        int depth, i, err = 0;
1118
1119repeat:
1120        i = depth = ext_depth(inode);
1121
1122        /* walk up to the tree and look for free index entry */
1123        curp = path + depth;
1124        while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
1125                i--;
1126                curp--;
1127        }
1128
1129        /* we use already allocated block for index block,
1130         * so subsequent data blocks should be contiguous */
1131        if (EXT_HAS_FREE_INDEX(curp)) {
1132                /* if we found index with free entry, then use that
1133                 * entry: create all needed subtree and add new leaf */
1134                err = ext4_ext_split(handle, inode, flags, path, newext, i);
1135                if (err)
1136                        goto out;
1137
1138                /* refill path */
1139                ext4_ext_drop_refs(path);
1140                path = ext4_ext_find_extent(inode,
1141                                    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1142                                    path);
1143                if (IS_ERR(path))
1144                        err = PTR_ERR(path);
1145        } else {
1146                /* tree is full, time to grow in depth */
1147                err = ext4_ext_grow_indepth(handle, inode, flags, newext);
1148                if (err)
1149                        goto out;
1150
1151                /* refill path */
1152                ext4_ext_drop_refs(path);
1153                path = ext4_ext_find_extent(inode,
1154                                   (ext4_lblk_t)le32_to_cpu(newext->ee_block),
1155                                    path);
1156                if (IS_ERR(path)) {
1157                        err = PTR_ERR(path);
1158                        goto out;
1159                }
1160
1161                /*
1162                 * only first (depth 0 -> 1) produces free space;
1163                 * in all other cases we have to split the grown tree
1164                 */
1165                depth = ext_depth(inode);
1166                if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
1167                        /* now we need to split */
1168                        goto repeat;
1169                }
1170        }
1171
1172out:
1173        return err;
1174}
1175
1176/*
1177 * search the closest allocated block to the left for *logical
1178 * and returns it at @logical + it's physical address at @phys
1179 * if *logical is the smallest allocated block, the function
1180 * returns 0 at @phys
1181 * return value contains 0 (success) or error code
1182 */
1183static int ext4_ext_search_left(struct inode *inode,
1184                                struct ext4_ext_path *path,
1185                                ext4_lblk_t *logical, ext4_fsblk_t *phys)
1186{
1187        struct ext4_extent_idx *ix;
1188        struct ext4_extent *ex;
1189        int depth, ee_len;
1190
1191        if (unlikely(path == NULL)) {
1192                EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1193                return -EIO;
1194        }
1195        depth = path->p_depth;
1196        *phys = 0;
1197
1198        if (depth == 0 && path->p_ext == NULL)
1199                return 0;
1200
1201        /* usually extent in the path covers blocks smaller
1202         * then *logical, but it can be that extent is the
1203         * first one in the file */
1204
1205        ex = path[depth].p_ext;
1206        ee_len = ext4_ext_get_actual_len(ex);
1207        if (*logical < le32_to_cpu(ex->ee_block)) {
1208                if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1209                        EXT4_ERROR_INODE(inode,
1210                                         "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
1211                                         *logical, le32_to_cpu(ex->ee_block));
1212                        return -EIO;
1213                }
1214                while (--depth >= 0) {
1215                        ix = path[depth].p_idx;
1216                        if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1217                                EXT4_ERROR_INODE(inode,
1218                                  "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
1219                                  ix != NULL ? le32_to_cpu(ix->ei_block) : 0,
1220                                  EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
1221                le32_to_cpu(EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block) : 0,
1222                                  depth);
1223                                return -EIO;
1224                        }
1225                }
1226                return 0;
1227        }
1228
1229        if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1230                EXT4_ERROR_INODE(inode,
1231                                 "logical %d < ee_block %d + ee_len %d!",
1232                                 *logical, le32_to_cpu(ex->ee_block), ee_len);
1233                return -EIO;
1234        }
1235
1236        *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
1237        *phys = ext4_ext_pblock(ex) + ee_len - 1;
1238        return 0;
1239}
1240
1241/*
1242 * search the closest allocated block to the right for *logical
1243 * and returns it at @logical + it's physical address at @phys
1244 * if *logical is the largest allocated block, the function
1245 * returns 0 at @phys
1246 * return value contains 0 (success) or error code
1247 */
1248static int ext4_ext_search_right(struct inode *inode,
1249                                 struct ext4_ext_path *path,
1250                                 ext4_lblk_t *logical, ext4_fsblk_t *phys,
1251                                 struct ext4_extent **ret_ex)
1252{
1253        struct buffer_head *bh = NULL;
1254        struct ext4_extent_header *eh;
1255        struct ext4_extent_idx *ix;
1256        struct ext4_extent *ex;
1257        ext4_fsblk_t block;
1258        int depth;      /* Note, NOT eh_depth; depth from top of tree */
1259        int ee_len;
1260
1261        if (unlikely(path == NULL)) {
1262                EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
1263                return -EIO;
1264        }
1265        depth = path->p_depth;
1266        *phys = 0;
1267
1268        if (depth == 0 && path->p_ext == NULL)
1269                return 0;
1270
1271        /* usually extent in the path covers blocks smaller
1272         * then *logical, but it can be that extent is the
1273         * first one in the file */
1274
1275        ex = path[depth].p_ext;
1276        ee_len = ext4_ext_get_actual_len(ex);
1277        if (*logical < le32_to_cpu(ex->ee_block)) {
1278                if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
1279                        EXT4_ERROR_INODE(inode,
1280                                         "first_extent(path[%d].p_hdr) != ex",
1281                                         depth);
1282                        return -EIO;
1283                }
1284                while (--depth >= 0) {
1285                        ix = path[depth].p_idx;
1286                        if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
1287                                EXT4_ERROR_INODE(inode,
1288                                                 "ix != EXT_FIRST_INDEX *logical %d!",
1289                                                 *logical);
1290                                return -EIO;
1291                        }
1292                }
1293                goto found_extent;
1294        }
1295
1296        if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
1297                EXT4_ERROR_INODE(inode,
1298                                 "logical %d < ee_block %d + ee_len %d!",
1299                                 *logical, le32_to_cpu(ex->ee_block), ee_len);
1300                return -EIO;
1301        }
1302
1303        if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
1304                /* next allocated block in this leaf */
1305                ex++;
1306                goto found_extent;
1307        }
1308
1309        /* go up and search for index to the right */
1310        while (--depth >= 0) {
1311                ix = path[depth].p_idx;
1312                if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
1313                        goto got_index;
1314        }
1315
1316        /* we've gone up to the root and found no index to the right */
1317        return 0;
1318
1319got_index:
1320        /* we've found index to the right, let's
1321         * follow it and find the closest allocated
1322         * block to the right */
1323        ix++;
1324        block = ext4_idx_pblock(ix);
1325        while (++depth < path->p_depth) {
1326                bh = sb_bread(inode->i_sb, block);
1327                if (bh == NULL)
1328                        return -EIO;
1329                eh = ext_block_hdr(bh);
1330                /* subtract from p_depth to get proper eh_depth */
1331                if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
1332                        put_bh(bh);
1333                        return -EIO;
1334                }
1335                ix = EXT_FIRST_INDEX(eh);
1336                block = ext4_idx_pblock(ix);
1337                put_bh(bh);
1338        }
1339
1340        bh = sb_bread(inode->i_sb, block);
1341        if (bh == NULL)
1342                return -EIO;
1343        eh = ext_block_hdr(bh);
1344        if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
1345                put_bh(bh);
1346                return -EIO;
1347        }
1348        ex = EXT_FIRST_EXTENT(eh);
1349found_extent:
1350        *logical = le32_to_cpu(ex->ee_block);
1351        *phys = ext4_ext_pblock(ex);
1352        *ret_ex = ex;
1353        if (bh)
1354                put_bh(bh);
1355        return 0;
1356}
1357
1358/*
1359 * ext4_ext_next_allocated_block:
1360 * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
1361 * NOTE: it considers block number from index entry as
1362 * allocated block. Thus, index entries have to be consistent
1363 * with leaves.
1364 */
1365static ext4_lblk_t
1366ext4_ext_next_allocated_block(struct ext4_ext_path *path)
1367{
1368        int depth;
1369
1370        BUG_ON(path == NULL);
1371        depth = path->p_depth;
1372
1373        if (depth == 0 && path->p_ext == NULL)
1374                return EXT_MAX_BLOCKS;
1375
1376        while (depth >= 0) {
1377                if (depth == path->p_depth) {
1378                        /* leaf */
1379                        if (path[depth].p_ext &&
1380                                path[depth].p_ext !=
1381                                        EXT_LAST_EXTENT(path[depth].p_hdr))
1382                          return le32_to_cpu(path[depth].p_ext[1].ee_block);
1383                } else {
1384                        /* index */
1385                        if (path[depth].p_idx !=
1386                                        EXT_LAST_INDEX(path[depth].p_hdr))
1387                          return le32_to_cpu(path[depth].p_idx[1].ei_block);
1388                }
1389                depth--;
1390        }
1391
1392        return EXT_MAX_BLOCKS;
1393}
1394
1395/*
1396 * ext4_ext_next_leaf_block:
1397 * returns first allocated block from next leaf or EXT_MAX_BLOCKS
1398 */
1399static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path)
1400{
1401        int depth;
1402
1403        BUG_ON(path == NULL);
1404        depth = path->p_depth;
1405
1406        /* zero-tree has no leaf blocks at all */
1407        if (depth == 0)
1408                return EXT_MAX_BLOCKS;
1409
1410        /* go to index block */
1411        depth--;
1412
1413        while (depth >= 0) {
1414                if (path[depth].p_idx !=
1415                                EXT_LAST_INDEX(path[depth].p_hdr))
1416                        return (ext4_lblk_t)
1417                                le32_to_cpu(path[depth].p_idx[1].ei_block);
1418                depth--;
1419        }
1420
1421        return EXT_MAX_BLOCKS;
1422}
1423
1424/*
1425 * ext4_ext_correct_indexes:
1426 * if leaf gets modified and modified extent is first in the leaf,
1427 * then we have to correct all indexes above.
1428 * TODO: do we need to correct tree in all cases?
1429 */
1430static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
1431                                struct ext4_ext_path *path)
1432{
1433        struct ext4_extent_header *eh;
1434        int depth = ext_depth(inode);
1435        struct ext4_extent *ex;
1436        __le32 border;
1437        int k, err = 0;
1438
1439        eh = path[depth].p_hdr;
1440        ex = path[depth].p_ext;
1441
1442        if (unlikely(ex == NULL || eh == NULL)) {
1443                EXT4_ERROR_INODE(inode,
1444                                 "ex %p == NULL or eh %p == NULL", ex, eh);
1445                return -EIO;
1446        }
1447
1448        if (depth == 0) {
1449                /* there is no tree at all */
1450                return 0;
1451        }
1452
1453        if (ex != EXT_FIRST_EXTENT(eh)) {
1454                /* we correct tree if first leaf got modified only */
1455                return 0;
1456        }
1457
1458        /*
1459         * TODO: we need correction if border is smaller than current one
1460         */
1461        k = depth - 1;
1462        border = path[depth].p_ext->ee_block;
1463        err = ext4_ext_get_access(handle, inode, path + k);
1464        if (err)
1465                return err;
1466        path[k].p_idx->ei_block = border;
1467        err = ext4_ext_dirty(handle, inode, path + k);
1468        if (err)
1469                return err;
1470
1471        while (k--) {
1472                /* change all left-side indexes */
1473                if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
1474                        break;
1475                err = ext4_ext_get_access(handle, inode, path + k);
1476                if (err)
1477                        break;
1478                path[k].p_idx->ei_block = border;
1479                err = ext4_ext_dirty(handle, inode, path + k);
1480                if (err)
1481                        break;
1482        }
1483
1484        return err;
1485}
1486
1487int
1488ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
1489                                struct ext4_extent *ex2)
1490{
1491        unsigned short ext1_ee_len, ext2_ee_len, max_len;
1492
1493        /*
1494         * Make sure that either both extents are uninitialized, or
1495         * both are _not_.
1496         */
1497        if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
1498                return 0;
1499
1500        if (ext4_ext_is_uninitialized(ex1))
1501                max_len = EXT_UNINIT_MAX_LEN;
1502        else
1503                max_len = EXT_INIT_MAX_LEN;
1504
1505        ext1_ee_len = ext4_ext_get_actual_len(ex1);
1506        ext2_ee_len = ext4_ext_get_actual_len(ex2);
1507
1508        if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
1509                        le32_to_cpu(ex2->ee_block))
1510                return 0;
1511
1512        /*
1513         * To allow future support for preallocated extents to be added
1514         * as an RO_COMPAT feature, refuse to merge to extents if
1515         * this can result in the top bit of ee_len being set.
1516         */
1517        if (ext1_ee_len + ext2_ee_len > max_len)
1518                return 0;
1519#ifdef AGGRESSIVE_TEST
1520        if (ext1_ee_len >= 4)
1521                return 0;
1522#endif
1523
1524        if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
1525                return 1;
1526        return 0;
1527}
1528
1529/*
1530 * This function tries to merge the "ex" extent to the next extent in the tree.
1531 * It always tries to merge towards right. If you want to merge towards
1532 * left, pass "ex - 1" as argument instead of "ex".
1533 * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
1534 * 1 if they got merged.
1535 */
1536static int ext4_ext_try_to_merge_right(struct inode *inode,
1537                                 struct ext4_ext_path *path,
1538                                 struct ext4_extent *ex)
1539{
1540        struct ext4_extent_header *eh;
1541        unsigned int depth, len;
1542        int merge_done = 0;
1543        int uninitialized = 0;
1544
1545        depth = ext_depth(inode);
1546        BUG_ON(path[depth].p_hdr == NULL);
1547        eh = path[depth].p_hdr;
1548
1549        while (ex < EXT_LAST_EXTENT(eh)) {
1550                if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
1551                        break;
1552                /* merge with next extent! */
1553                if (ext4_ext_is_uninitialized(ex))
1554                        uninitialized = 1;
1555                ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1556                                + ext4_ext_get_actual_len(ex + 1));
1557                if (uninitialized)
1558                        ext4_ext_mark_uninitialized(ex);
1559
1560                if (ex + 1 < EXT_LAST_EXTENT(eh)) {
1561                        len = (EXT_LAST_EXTENT(eh) - ex - 1)
1562                                * sizeof(struct ext4_extent);
1563                        memmove(ex + 1, ex + 2, len);
1564                }
1565                le16_add_cpu(&eh->eh_entries, -1);
1566                merge_done = 1;
1567                WARN_ON(eh->eh_entries == 0);
1568                if (!eh->eh_entries)
1569                        EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
1570        }
1571
1572        return merge_done;
1573}
1574
1575/*
1576 * This function tries to merge the @ex extent to neighbours in the tree.
1577 * return 1 if merge left else 0.
1578 */
1579static int ext4_ext_try_to_merge(struct inode *inode,
1580                                  struct ext4_ext_path *path,
1581                                  struct ext4_extent *ex) {
1582        struct ext4_extent_header *eh;
1583        unsigned int depth;
1584        int merge_done = 0;
1585        int ret = 0;
1586
1587        depth = ext_depth(inode);
1588        BUG_ON(path[depth].p_hdr == NULL);
1589        eh = path[depth].p_hdr;
1590
1591        if (ex > EXT_FIRST_EXTENT(eh))
1592                merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1);
1593
1594        if (!merge_done)
1595                ret = ext4_ext_try_to_merge_right(inode, path, ex);
1596
1597        return ret;
1598}
1599
1600/*
1601 * check if a portion of the "newext" extent overlaps with an
1602 * existing extent.
1603 *
1604 * If there is an overlap discovered, it updates the length of the newext
1605 * such that there will be no overlap, and then returns 1.
1606 * If there is no overlap found, it returns 0.
1607 */
1608static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi,
1609                                           struct inode *inode,
1610                                           struct ext4_extent *newext,
1611                                           struct ext4_ext_path *path)
1612{
1613        ext4_lblk_t b1, b2;
1614        unsigned int depth, len1;
1615        unsigned int ret = 0;
1616
1617        b1 = le32_to_cpu(newext->ee_block);
1618        len1 = ext4_ext_get_actual_len(newext);
1619        depth = ext_depth(inode);
1620        if (!path[depth].p_ext)
1621                goto out;
1622        b2 = le32_to_cpu(path[depth].p_ext->ee_block);
1623        b2 &= ~(sbi->s_cluster_ratio - 1);
1624
1625        /*
1626         * get the next allocated block if the extent in the path
1627         * is before the requested block(s)
1628         */
1629        if (b2 < b1) {
1630                b2 = ext4_ext_next_allocated_block(path);
1631                if (b2 == EXT_MAX_BLOCKS)
1632                        goto out;
1633                b2 &= ~(sbi->s_cluster_ratio - 1);
1634        }
1635
1636        /* check for wrap through zero on extent logical start block*/
1637        if (b1 + len1 < b1) {
1638                len1 = EXT_MAX_BLOCKS - b1;
1639                newext->ee_len = cpu_to_le16(len1);
1640                ret = 1;
1641        }
1642
1643        /* check for overlap */
1644        if (b1 + len1 > b2) {
1645                newext->ee_len = cpu_to_le16(b2 - b1);
1646                ret = 1;
1647        }
1648out:
1649        return ret;
1650}
1651
1652/*
1653 * ext4_ext_insert_extent:
1654 * tries to merge requsted extent into the existing extent or
1655 * inserts requested extent as new one into the tree,
1656 * creating new leaf in the no-space case.
1657 */
1658int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
1659                                struct ext4_ext_path *path,
1660                                struct ext4_extent *newext, int flag)
1661{
1662        struct ext4_extent_header *eh;
1663        struct ext4_extent *ex, *fex;
1664        struct ext4_extent *nearex; /* nearest extent */
1665        struct ext4_ext_path *npath = NULL;
1666        int depth, len, err;
1667        ext4_lblk_t next;
1668        unsigned uninitialized = 0;
1669        int flags = 0;
1670
1671        if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
1672                EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
1673                return -EIO;
1674        }
1675        depth = ext_depth(inode);
1676        ex = path[depth].p_ext;
1677        if (unlikely(path[depth].p_hdr == NULL)) {
1678                EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1679                return -EIO;
1680        }
1681
1682        /* try to insert block into found extent and return */
1683        if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
1684                && ext4_can_extents_be_merged(inode, ex, newext)) {
1685                ext_debug("append [%d]%d block to %u:[%d]%d (from %llu)\n",
1686                          ext4_ext_is_uninitialized(newext),
1687                          ext4_ext_get_actual_len(newext),
1688                          le32_to_cpu(ex->ee_block),
1689                          ext4_ext_is_uninitialized(ex),
1690                          ext4_ext_get_actual_len(ex),
1691                          ext4_ext_pblock(ex));
1692                err = ext4_ext_get_access(handle, inode, path + depth);
1693                if (err)
1694                        return err;
1695
1696                /*
1697                 * ext4_can_extents_be_merged should have checked that either
1698                 * both extents are uninitialized, or both aren't. Thus we
1699                 * need to check only one of them here.
1700                 */
1701                if (ext4_ext_is_uninitialized(ex))
1702                        uninitialized = 1;
1703                ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
1704                                        + ext4_ext_get_actual_len(newext));
1705                if (uninitialized)
1706                        ext4_ext_mark_uninitialized(ex);
1707                eh = path[depth].p_hdr;
1708                nearex = ex;
1709                goto merge;
1710        }
1711
1712        depth = ext_depth(inode);
1713        eh = path[depth].p_hdr;
1714        if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
1715                goto has_space;
1716
1717        /* probably next leaf has space for us? */
1718        fex = EXT_LAST_EXTENT(eh);
1719        next = EXT_MAX_BLOCKS;
1720        if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block))
1721                next = ext4_ext_next_leaf_block(path);
1722        if (next != EXT_MAX_BLOCKS) {
1723                ext_debug("next leaf block - %u\n", next);
1724                BUG_ON(npath != NULL);
1725                npath = ext4_ext_find_extent(inode, next, NULL);
1726                if (IS_ERR(npath))
1727                        return PTR_ERR(npath);
1728                BUG_ON(npath->p_depth != path->p_depth);
1729                eh = npath[depth].p_hdr;
1730                if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
1731                        ext_debug("next leaf isn't full(%d)\n",
1732                                  le16_to_cpu(eh->eh_entries));
1733                        path = npath;
1734                        goto has_space;
1735                }
1736                ext_debug("next leaf has no free space(%d,%d)\n",
1737                          le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
1738        }
1739
1740        /*
1741         * There is no free space in the found leaf.
1742         * We're gonna add a new leaf in the tree.
1743         */
1744        if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT)
1745                flags = EXT4_MB_USE_ROOT_BLOCKS;
1746        err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
1747        if (err)
1748                goto cleanup;
1749        depth = ext_depth(inode);
1750        eh = path[depth].p_hdr;
1751
1752has_space:
1753        nearex = path[depth].p_ext;
1754
1755        err = ext4_ext_get_access(handle, inode, path + depth);
1756        if (err)
1757                goto cleanup;
1758
1759        if (!nearex) {
1760                /* there is no extent in this leaf, create first one */
1761                ext_debug("first extent in the leaf: %u:%llu:[%d]%d\n",
1762                                le32_to_cpu(newext->ee_block),
1763                                ext4_ext_pblock(newext),
1764                                ext4_ext_is_uninitialized(newext),
1765                                ext4_ext_get_actual_len(newext));
1766                nearex = EXT_FIRST_EXTENT(eh);
1767        } else {
1768                if (le32_to_cpu(newext->ee_block)
1769                           > le32_to_cpu(nearex->ee_block)) {
1770                        /* Insert after */
1771                        ext_debug("insert %u:%llu:[%d]%d before: "
1772                                        "nearest %p\n",
1773                                        le32_to_cpu(newext->ee_block),
1774                                        ext4_ext_pblock(newext),
1775                                        ext4_ext_is_uninitialized(newext),
1776                                        ext4_ext_get_actual_len(newext),
1777                                        nearex);
1778                        nearex++;
1779                } else {
1780                        /* Insert before */
1781                        BUG_ON(newext->ee_block == nearex->ee_block);
1782                        ext_debug("insert %u:%llu:[%d]%d after: "
1783                                        "nearest %p\n",
1784                                        le32_to_cpu(newext->ee_block),
1785                                        ext4_ext_pblock(newext),
1786                                        ext4_ext_is_uninitialized(newext),
1787                                        ext4_ext_get_actual_len(newext),
1788                                        nearex);
1789                }
1790                len = EXT_LAST_EXTENT(eh) - nearex + 1;
1791                if (len > 0) {
1792                        ext_debug("insert %u:%llu:[%d]%d: "
1793                                        "move %d extents from 0x%p to 0x%p\n",
1794                                        le32_to_cpu(newext->ee_block),
1795                                        ext4_ext_pblock(newext),
1796                                        ext4_ext_is_uninitialized(newext),
1797                                        ext4_ext_get_actual_len(newext),
1798                                        len, nearex, nearex + 1);
1799                        memmove(nearex + 1, nearex,
1800                                len * sizeof(struct ext4_extent));
1801                }
1802        }
1803
1804        le16_add_cpu(&eh->eh_entries, 1);
1805        path[depth].p_ext = nearex;
1806        nearex->ee_block = newext->ee_block;
1807        ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
1808        nearex->ee_len = newext->ee_len;
1809
1810merge:
1811        /* try to merge extents to the right */
1812        if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
1813                ext4_ext_try_to_merge(inode, path, nearex);
1814
1815        /* try to merge extents to the left */
1816
1817        /* time to correct all indexes above */
1818        err = ext4_ext_correct_indexes(handle, inode, path);
1819        if (err)
1820                goto cleanup;
1821
1822        err = ext4_ext_dirty(handle, inode, path + depth);
1823
1824cleanup:
1825        if (npath) {
1826                ext4_ext_drop_refs(npath);
1827                kfree(npath);
1828        }
1829        ext4_ext_invalidate_cache(inode);
1830        return err;
1831}
1832
1833static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
1834                               ext4_lblk_t num, ext_prepare_callback func,
1835                               void *cbdata)
1836{
1837        struct ext4_ext_path *path = NULL;
1838        struct ext4_ext_cache cbex;
1839        struct ext4_extent *ex;
1840        ext4_lblk_t next, start = 0, end = 0;
1841        ext4_lblk_t last = block + num;
1842        int depth, exists, err = 0;
1843
1844        BUG_ON(func == NULL);
1845        BUG_ON(inode == NULL);
1846
1847        while (block < last && block != EXT_MAX_BLOCKS) {
1848                num = last - block;
1849                /* find extent for this block */
1850                down_read(&EXT4_I(inode)->i_data_sem);
1851                path = ext4_ext_find_extent(inode, block, path);
1852                up_read(&EXT4_I(inode)->i_data_sem);
1853                if (IS_ERR(path)) {
1854                        err = PTR_ERR(path);
1855                        path = NULL;
1856                        break;
1857                }
1858
1859                depth = ext_depth(inode);
1860                if (unlikely(path[depth].p_hdr == NULL)) {
1861                        EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
1862                        err = -EIO;
1863                        break;
1864                }
1865                ex = path[depth].p_ext;
1866                next = ext4_ext_next_allocated_block(path);
1867
1868                exists = 0;
1869                if (!ex) {
1870                        /* there is no extent yet, so try to allocate
1871                         * all requested space */
1872                        start = block;
1873                        end = block + num;
1874                } else if (le32_to_cpu(ex->ee_block) > block) {
1875                        /* need to allocate space before found extent */
1876                        start = block;
1877                        end = le32_to_cpu(ex->ee_block);
1878                        if (block + num < end)
1879                                end = block + num;
1880                } else if (block >= le32_to_cpu(ex->ee_block)
1881                                        + ext4_ext_get_actual_len(ex)) {
1882                        /* need to allocate space after found extent */
1883                        start = block;
1884                        end = block + num;
1885                        if (end >= next)
1886                                end = next;
1887                } else if (block >= le32_to_cpu(ex->ee_block)) {
1888                        /*
1889                         * some part of requested space is covered
1890                         * by found extent
1891                         */
1892                        start = block;
1893                        end = le32_to_cpu(ex->ee_block)
1894                                + ext4_ext_get_actual_len(ex);
1895                        if (block + num < end)
1896                                end = block + num;
1897                        exists = 1;
1898                } else {
1899                        BUG();
1900                }
1901                BUG_ON(end <= start);
1902
1903                if (!exists) {
1904                        cbex.ec_block = start;
1905                        cbex.ec_len = end - start;
1906                        cbex.ec_start = 0;
1907                } else {
1908                        cbex.ec_block = le32_to_cpu(ex->ee_block);
1909                        cbex.ec_len = ext4_ext_get_actual_len(ex);
1910                        cbex.ec_start = ext4_ext_pblock(ex);
1911                }
1912
1913                if (unlikely(cbex.ec_len == 0)) {
1914                        EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
1915                        err = -EIO;
1916                        break;
1917                }
1918                err = func(inode, next, &cbex, ex, cbdata);
1919                ext4_ext_drop_refs(path);
1920
1921                if (err < 0)
1922                        break;
1923
1924                if (err == EXT_REPEAT)
1925                        continue;
1926                else if (err == EXT_BREAK) {
1927                        err = 0;
1928                        break;
1929                }
1930
1931                if (ext_depth(inode) != depth) {
1932                        /* depth was changed. we have to realloc path */
1933                        kfree(path);
1934                        path = NULL;
1935                }
1936
1937                block = cbex.ec_block + cbex.ec_len;
1938        }
1939
1940        if (path) {
1941                ext4_ext_drop_refs(path);
1942                kfree(path);
1943        }
1944
1945        return err;
1946}
1947
1948static void
1949ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
1950                        __u32 len, ext4_fsblk_t start)
1951{
1952        struct ext4_ext_cache *cex;
1953        BUG_ON(len == 0);
1954        spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
1955        trace_ext4_ext_put_in_cache(inode, block, len, start);
1956        cex = &EXT4_I(inode)->i_cached_extent;
1957        cex->ec_block = block;
1958        cex->ec_len = len;
1959        cex->ec_start = start;
1960        spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
1961}
1962
1963/*
1964 * ext4_ext_put_gap_in_cache:
1965 * calculate boundaries of the gap that the requested block fits into
1966 * and cache this gap
1967 */
1968static void
1969ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
1970                                ext4_lblk_t block)
1971{
1972        int depth = ext_depth(inode);
1973        unsigned long len;
1974        ext4_lblk_t lblock;
1975        struct ext4_extent *ex;
1976
1977        ex = path[depth].p_ext;
1978        if (ex == NULL) {
1979                /* there is no extent yet, so gap is [0;-] */
1980                lblock = 0;
1981                len = EXT_MAX_BLOCKS;
1982                ext_debug("cache gap(whole file):");
1983        } else if (block < le32_to_cpu(ex->ee_block)) {
1984                lblock = block;
1985                len = le32_to_cpu(ex->ee_block) - block;
1986                ext_debug("cache gap(before): %u [%u:%u]",
1987                                block,
1988                                le32_to_cpu(ex->ee_block),
1989                                 ext4_ext_get_actual_len(ex));
1990        } else if (block >= le32_to_cpu(ex->ee_block)
1991                        + ext4_ext_get_actual_len(ex)) {
1992                ext4_lblk_t next;
1993                lblock = le32_to_cpu(ex->ee_block)
1994                        + ext4_ext_get_actual_len(ex);
1995
1996                next = ext4_ext_next_allocated_block(path);
1997                ext_debug("cache gap(after): [%u:%u] %u",
1998                                le32_to_cpu(ex->ee_block),
1999                                ext4_ext_get_actual_len(ex),
2000                                block);
2001                BUG_ON(next == lblock);
2002                len = next - lblock;
2003        } else {
2004                lblock = len = 0;
2005                BUG();
2006        }
2007
2008        ext_debug(" -> %u:%lu\n", lblock, len);
2009        ext4_ext_put_in_cache(inode, lblock, len, 0);
2010}
2011
2012/*
2013 * ext4_ext_check_cache()
2014 * Checks to see if the given block is in the cache.
2015 * If it is, the cached extent is stored in the given
2016 * cache extent pointer.  If the cached extent is a hole,
2017 * this routine should be used instead of
2018 * ext4_ext_in_cache if the calling function needs to
2019 * know the size of the hole.
2020 *
2021 * @inode: The files inode
2022 * @block: The block to look for in the cache
2023 * @ex:    Pointer where the cached extent will be stored
2024 *         if it contains block
2025 *
2026 * Return 0 if cache is invalid; 1 if the cache is valid
2027 */
2028static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
2029        struct ext4_ext_cache *ex){
2030        struct ext4_ext_cache *cex;
2031        struct ext4_sb_info *sbi;
2032        int ret = 0;
2033
2034        /*
2035         * We borrow i_block_reservation_lock to protect i_cached_extent
2036         */
2037        spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
2038        cex = &EXT4_I(inode)->i_cached_extent;
2039        sbi = EXT4_SB(inode->i_sb);
2040
2041        /* has cache valid data? */
2042        if (cex->ec_len == 0)
2043                goto errout;
2044
2045        if (in_range(block, cex->ec_block, cex->ec_len)) {
2046                memcpy(ex, cex, sizeof(struct ext4_ext_cache));
2047                ext_debug("%u cached by %u:%u:%llu\n",
2048                                block,
2049                                cex->ec_block, cex->ec_len, cex->ec_start);
2050                ret = 1;
2051        }
2052errout:
2053        if (!ret)
2054                sbi->extent_cache_misses++;
2055        else
2056                sbi->extent_cache_hits++;
2057        trace_ext4_ext_in_cache(inode, block, ret);
2058        spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
2059        return ret;
2060}
2061
2062/*
2063 * ext4_ext_in_cache()
2064 * Checks to see if the given block is in the cache.
2065 * If it is, the cached extent is stored in the given
2066 * extent pointer.
2067 *
2068 * @inode: The files inode
2069 * @block: The block to look for in the cache
2070 * @ex:    Pointer where the cached extent will be stored
2071 *         if it contains block
2072 *
2073 * Return 0 if cache is invalid; 1 if the cache is valid
2074 */
2075static int
2076ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
2077                        struct ext4_extent *ex)
2078{
2079        struct ext4_ext_cache cex;
2080        int ret = 0;
2081
2082        if (ext4_ext_check_cache(inode, block, &cex)) {
2083                ex->ee_block = cpu_to_le32(cex.ec_block);
2084                ext4_ext_store_pblock(ex, cex.ec_start);
2085                ex->ee_len = cpu_to_le16(cex.ec_len);
2086                ret = 1;
2087        }
2088
2089        return ret;
2090}
2091
2092
2093/*
2094 * ext4_ext_rm_idx:
2095 * removes index from the index block.
2096 */
2097static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
2098                        struct ext4_ext_path *path)
2099{
2100        int err;
2101        ext4_fsblk_t leaf;
2102
2103        /* free index block */
2104        path--;
2105        leaf = ext4_idx_pblock(path->p_idx);
2106        if (unlikely(path->p_hdr->eh_entries == 0)) {
2107                EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
2108                return -EIO;
2109        }
2110        err = ext4_ext_get_access(handle, inode, path);
2111        if (err)
2112                return err;
2113
2114        if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) {
2115                int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx;
2116                len *= sizeof(struct ext4_extent_idx);
2117                memmove(path->p_idx, path->p_idx + 1, len);
2118        }
2119
2120        le16_add_cpu(&path->p_hdr->eh_entries, -1);
2121        err = ext4_ext_dirty(handle, inode, path);
2122        if (err)
2123                return err;
2124        ext_debug("index is empty, remove it, free block %llu\n", leaf);
2125        trace_ext4_ext_rm_idx(inode, leaf);
2126
2127        ext4_free_blocks(handle, inode, NULL, leaf, 1,
2128                         EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
2129        return err;
2130}
2131
2132/*
2133 * ext4_ext_calc_credits_for_single_extent:
2134 * This routine returns max. credits that needed to insert an extent
2135 * to the extent tree.
2136 * When pass the actual path, the caller should calculate credits
2137 * under i_data_sem.
2138 */
2139int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
2140                                                struct ext4_ext_path *path)
2141{
2142        if (path) {
2143                int depth = ext_depth(inode);
2144                int ret = 0;
2145
2146                /* probably there is space in leaf? */
2147                if (le16_to_cpu(path[depth].p_hdr->eh_entries)
2148                                < le16_to_cpu(path[depth].p_hdr->eh_max)) {
2149
2150                        /*
2151                         *  There are some space in the leaf tree, no
2152                         *  need to account for leaf block credit
2153                         *
2154                         *  bitmaps and block group descriptor blocks
2155                         *  and other metadata blocks still need to be
2156                         *  accounted.
2157                         */
2158                        /* 1 bitmap, 1 block group descriptor */
2159                        ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
2160                        return ret;
2161                }
2162        }
2163
2164        return ext4_chunk_trans_blocks(inode, nrblocks);
2165}
2166
2167/*
2168 * How many index/leaf blocks need to change/allocate to modify nrblocks?
2169 *
2170 * if nrblocks are fit in a single extent (chunk flag is 1), then
2171 * in the worse case, each tree level index/leaf need to be changed
2172 * if the tree split due to insert a new extent, then the old tree
2173 * index/leaf need to be updated too
2174 *
2175 * If the nrblocks are discontiguous, they could cause
2176 * the whole tree split more than once, but this is really rare.
2177 */
2178int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
2179{
2180        int index;
2181        int depth = ext_depth(inode);
2182
2183        if (chunk)
2184                index = depth * 2;
2185        else
2186                index = depth * 3;
2187
2188        return index;
2189}
2190
2191static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
2192                              struct ext4_extent *ex,
2193                              ext4_fsblk_t *partial_cluster,
2194                              ext4_lblk_t from, ext4_lblk_t to)
2195{
2196        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2197        unsigned short ee_len =  ext4_ext_get_actual_len(ex);
2198        ext4_fsblk_t pblk;
2199        int flags = EXT4_FREE_BLOCKS_FORGET;
2200
2201        if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2202                flags |= EXT4_FREE_BLOCKS_METADATA;
2203        /*
2204         * For bigalloc file systems, we never free a partial cluster
2205         * at the beginning of the extent.  Instead, we make a note
2206         * that we tried freeing the cluster, and check to see if we
2207         * need to free it on a subsequent call to ext4_remove_blocks,
2208         * or at the end of the ext4_truncate() operation.
2209         */
2210        flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;
2211
2212        trace_ext4_remove_blocks(inode, ex, from, to, *partial_cluster);
2213        /*
2214         * If we have a partial cluster, and it's different from the
2215         * cluster of the last block, we need to explicitly free the
2216         * partial cluster here.
2217         */
2218        pblk = ext4_ext_pblock(ex) + ee_len - 1;
2219        if (*partial_cluster && (EXT4_B2C(sbi, pblk) != *partial_cluster)) {
2220                ext4_free_blocks(handle, inode, NULL,
2221                                 EXT4_C2B(sbi, *partial_cluster),
2222                                 sbi->s_cluster_ratio, flags);
2223                *partial_cluster = 0;
2224        }
2225
2226#ifdef EXTENTS_STATS
2227        {
2228                struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2229                spin_lock(&sbi->s_ext_stats_lock);
2230                sbi->s_ext_blocks += ee_len;
2231                sbi->s_ext_extents++;
2232                if (ee_len < sbi->s_ext_min)
2233                        sbi->s_ext_min = ee_len;
2234                if (ee_len > sbi->s_ext_max)
2235                        sbi->s_ext_max = ee_len;
2236                if (ext_depth(inode) > sbi->s_depth_max)
2237                        sbi->s_depth_max = ext_depth(inode);
2238                spin_unlock(&sbi->s_ext_stats_lock);
2239        }
2240#endif
2241        if (from >= le32_to_cpu(ex->ee_block)
2242            && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
2243                /* tail removal */
2244                ext4_lblk_t num;
2245
2246                num = le32_to_cpu(ex->ee_block) + ee_len - from;
2247                pblk = ext4_ext_pblock(ex) + ee_len - num;
2248                ext_debug("free last %u blocks starting %llu\n", num, pblk);
2249                ext4_free_blocks(handle, inode, NULL, pblk, num, flags);
2250                /*
2251                 * If the block range to be freed didn't start at the
2252                 * beginning of a cluster, and we removed the entire
2253                 * extent, save the partial cluster here, since we
2254                 * might need to delete if we determine that the
2255                 * truncate operation has removed all of the blocks in
2256                 * the cluster.
2257                 */
2258                if (pblk & (sbi->s_cluster_ratio - 1) &&
2259                    (ee_len == num))
2260                        *partial_cluster = EXT4_B2C(sbi, pblk);
2261                else
2262                        *partial_cluster = 0;
2263        } else if (from == le32_to_cpu(ex->ee_block)
2264                   && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
2265                /* head removal */
2266                ext4_lblk_t num;
2267                ext4_fsblk_t start;
2268
2269                num = to - from;
2270                start = ext4_ext_pblock(ex);
2271
2272                ext_debug("free first %u blocks starting %llu\n", num, start);
2273                ext4_free_blocks(handle, inode, NULL, start, num, flags);
2274
2275        } else {
2276                printk(KERN_INFO "strange request: removal(2) "
2277                                "%u-%u from %u:%u\n",
2278                                from, to, le32_to_cpu(ex->ee_block), ee_len);
2279        }
2280        return 0;
2281}
2282
2283
2284/*
2285 * ext4_ext_rm_leaf() Removes the extents associated with the
2286 * blocks appearing between "start" and "end", and splits the extents
2287 * if "start" and "end" appear in the same extent
2288 *
2289 * @handle: The journal handle
2290 * @inode:  The files inode
2291 * @path:   The path to the leaf
2292 * @start:  The first block to remove
2293 * @end:   The last block to remove
2294 */
2295static int
2296ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
2297                 struct ext4_ext_path *path, ext4_fsblk_t *partial_cluster,
2298                 ext4_lblk_t start, ext4_lblk_t end)
2299{
2300        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2301        int err = 0, correct_index = 0;
2302        int depth = ext_depth(inode), credits;
2303        struct ext4_extent_header *eh;
2304        ext4_lblk_t a, b;
2305        unsigned num;
2306        ext4_lblk_t ex_ee_block;
2307        unsigned short ex_ee_len;
2308        unsigned uninitialized = 0;
2309        struct ext4_extent *ex;
2310
2311        /* the header must be checked already in ext4_ext_remove_space() */
2312        ext_debug("truncate since %u in leaf\n", start);
2313        if (!path[depth].p_hdr)
2314                path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
2315        eh = path[depth].p_hdr;
2316        if (unlikely(path[depth].p_hdr == NULL)) {
2317                EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
2318                return -EIO;
2319        }
2320        /* find where to start removing */
2321        ex = EXT_LAST_EXTENT(eh);
2322
2323        ex_ee_block = le32_to_cpu(ex->ee_block);
2324        ex_ee_len = ext4_ext_get_actual_len(ex);
2325
2326        trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster);
2327
2328        while (ex >= EXT_FIRST_EXTENT(eh) &&
2329                        ex_ee_block + ex_ee_len > start) {
2330
2331                if (ext4_ext_is_uninitialized(ex))
2332                        uninitialized = 1;
2333                else
2334                        uninitialized = 0;
2335
2336                ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
2337                         uninitialized, ex_ee_len);
2338                path[depth].p_ext = ex;
2339
2340                a = ex_ee_block > start ? ex_ee_block : start;
2341                b = ex_ee_block+ex_ee_len - 1 < end ?
2342                        ex_ee_block+ex_ee_len - 1 : end;
2343
2344                ext_debug("  border %u:%u\n", a, b);
2345
2346                /* If this extent is beyond the end of the hole, skip it */
2347                if (end <= ex_ee_block) {
2348                        ex--;
2349                        ex_ee_block = le32_to_cpu(ex->ee_block);
2350                        ex_ee_len = ext4_ext_get_actual_len(ex);
2351                        continue;
2352                } else if (b != ex_ee_block + ex_ee_len - 1) {
2353                        EXT4_ERROR_INODE(inode,"  bad truncate %u:%u\n",
2354                                         start, end);
2355                        err = -EIO;
2356                        goto out;
2357                } else if (a != ex_ee_block) {
2358                        /* remove tail of the extent */
2359                        num = a - ex_ee_block;
2360                } else {
2361                        /* remove whole extent: excellent! */
2362                        num = 0;
2363                }
2364                /*
2365                 * 3 for leaf, sb, and inode plus 2 (bmap and group
2366                 * descriptor) for each block group; assume two block
2367                 * groups plus ex_ee_len/blocks_per_block_group for
2368                 * the worst case
2369                 */
2370                credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb));
2371                if (ex == EXT_FIRST_EXTENT(eh)) {
2372                        correct_index = 1;
2373                        credits += (ext_depth(inode)) + 1;
2374                }
2375                credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
2376
2377                err = ext4_ext_truncate_extend_restart(handle, inode, credits);
2378                if (err)
2379                        goto out;
2380
2381                err = ext4_ext_get_access(handle, inode, path + depth);
2382                if (err)
2383                        goto out;
2384
2385                err = ext4_remove_blocks(handle, inode, ex, partial_cluster,
2386                                         a, b);
2387                if (err)
2388                        goto out;
2389
2390                if (num == 0)
2391                        /* this extent is removed; mark slot entirely unused */
2392                        ext4_ext_store_pblock(ex, 0);
2393
2394                ex->ee_len = cpu_to_le16(num);
2395                /*
2396                 * Do not mark uninitialized if all the blocks in the
2397                 * extent have been removed.
2398                 */
2399                if (uninitialized && num)
2400                        ext4_ext_mark_uninitialized(ex);
2401                /*
2402                 * If the extent was completely released,
2403                 * we need to remove it from the leaf
2404                 */
2405                if (num == 0) {
2406                        if (end != EXT_MAX_BLOCKS - 1) {
2407                                /*
2408                                 * For hole punching, we need to scoot all the
2409                                 * extents up when an extent is removed so that
2410                                 * we dont have blank extents in the middle
2411                                 */
2412                                memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) *
2413                                        sizeof(struct ext4_extent));
2414
2415                                /* Now get rid of the one at the end */
2416                                memset(EXT_LAST_EXTENT(eh), 0,
2417                                        sizeof(struct ext4_extent));
2418                        }
2419                        le16_add_cpu(&eh->eh_entries, -1);
2420                } else
2421                        *partial_cluster = 0;
2422
2423                err = ext4_ext_dirty(handle, inode, path + depth);
2424                if (err)
2425                        goto out;
2426
2427                ext_debug("new extent: %u:%u:%llu\n", ex_ee_block, num,
2428                                ext4_ext_pblock(ex));
2429                ex--;
2430                ex_ee_block = le32_to_cpu(ex->ee_block);
2431                ex_ee_len = ext4_ext_get_actual_len(ex);
2432        }
2433
2434        if (correct_index && eh->eh_entries)
2435                err = ext4_ext_correct_indexes(handle, inode, path);
2436
2437        /*
2438         * If there is still a entry in the leaf node, check to see if
2439         * it references the partial cluster.  This is the only place
2440         * where it could; if it doesn't, we can free the cluster.
2441         */
2442        if (*partial_cluster && ex >= EXT_FIRST_EXTENT(eh) &&
2443            (EXT4_B2C(sbi, ext4_ext_pblock(ex) + ex_ee_len - 1) !=
2444             *partial_cluster)) {
2445                int flags = EXT4_FREE_BLOCKS_FORGET;
2446
2447                if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2448                        flags |= EXT4_FREE_BLOCKS_METADATA;
2449
2450                ext4_free_blocks(handle, inode, NULL,
2451                                 EXT4_C2B(sbi, *partial_cluster),
2452                                 sbi->s_cluster_ratio, flags);
2453                *partial_cluster = 0;
2454        }
2455
2456        /* if this leaf is free, then we should
2457         * remove it from index block above */
2458        if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
2459                err = ext4_ext_rm_idx(handle, inode, path + depth);
2460
2461out:
2462        return err;
2463}
2464
2465/*
2466 * ext4_ext_more_to_rm:
2467 * returns 1 if current index has to be freed (even partial)
2468 */
2469static int
2470ext4_ext_more_to_rm(struct ext4_ext_path *path)
2471{
2472        BUG_ON(path->p_idx == NULL);
2473
2474        if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
2475                return 0;
2476
2477        /*
2478         * if truncate on deeper level happened, it wasn't partial,
2479         * so we have to consider current index for truncation
2480         */
2481        if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
2482                return 0;
2483        return 1;
2484}
2485
2486static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
2487{
2488        struct super_block *sb = inode->i_sb;
2489        int depth = ext_depth(inode);
2490        struct ext4_ext_path *path;
2491        ext4_fsblk_t partial_cluster = 0;
2492        handle_t *handle;
2493        int i, err;
2494
2495        ext_debug("truncate since %u\n", start);
2496
2497        /* probably first extent we're gonna free will be last in block */
2498        handle = ext4_journal_start(inode, depth + 1);
2499        if (IS_ERR(handle))
2500                return PTR_ERR(handle);
2501
2502again:
2503        ext4_ext_invalidate_cache(inode);
2504
2505        trace_ext4_ext_remove_space(inode, start, depth);
2506
2507        /*
2508         * We start scanning from right side, freeing all the blocks
2509         * after i_size and walking into the tree depth-wise.
2510         */
2511        depth = ext_depth(inode);
2512        path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
2513        if (path == NULL) {
2514                ext4_journal_stop(handle);
2515                return -ENOMEM;
2516        }
2517        path[0].p_depth = depth;
2518        path[0].p_hdr = ext_inode_hdr(inode);
2519        if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
2520                err = -EIO;
2521                goto out;
2522        }
2523        i = err = 0;
2524
2525        while (i >= 0 && err == 0) {
2526                if (i == depth) {
2527                        /* this is leaf block */
2528                        err = ext4_ext_rm_leaf(handle, inode, path,
2529                                               &partial_cluster, start,
2530                                               EXT_MAX_BLOCKS - 1);
2531                        /* root level has p_bh == NULL, brelse() eats this */
2532                        brelse(path[i].p_bh);
2533                        path[i].p_bh = NULL;
2534                        i--;
2535                        continue;
2536                }
2537
2538                /* this is index block */
2539                if (!path[i].p_hdr) {
2540                        ext_debug("initialize header\n");
2541                        path[i].p_hdr = ext_block_hdr(path[i].p_bh);
2542                }
2543
2544                if (!path[i].p_idx) {
2545                        /* this level hasn't been touched yet */
2546                        path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2547                        path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
2548                        ext_debug("init index ptr: hdr 0x%p, num %d\n",
2549                                  path[i].p_hdr,
2550                                  le16_to_cpu(path[i].p_hdr->eh_entries));
2551                } else {
2552                        /* we were already here, see at next index */
2553                        path[i].p_idx--;
2554                }
2555
2556                ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
2557                                i, EXT_FIRST_INDEX(path[i].p_hdr),
2558                                path[i].p_idx);
2559                if (ext4_ext_more_to_rm(path + i)) {
2560                        struct buffer_head *bh;
2561                        /* go to the next level */
2562                        ext_debug("move to level %d (block %llu)\n",
2563                                  i + 1, ext4_idx_pblock(path[i].p_idx));
2564                        memset(path + i + 1, 0, sizeof(*path));
2565                        bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
2566                        if (!bh) {
2567                                /* should we reset i_size? */
2568                                err = -EIO;
2569                                break;
2570                        }
2571                        if (WARN_ON(i + 1 > depth)) {
2572                                err = -EIO;
2573                                break;
2574                        }
2575                        if (ext4_ext_check(inode, ext_block_hdr(bh),
2576                                                        depth - i - 1)) {
2577                                err = -EIO;
2578                                break;
2579                        }
2580                        path[i + 1].p_bh = bh;
2581
2582                        /* save actual number of indexes since this
2583                         * number is changed at the next iteration */
2584                        path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
2585                        i++;
2586                } else {
2587                        /* we finished processing this index, go up */
2588                        if (path[i].p_hdr->eh_entries == 0 && i > 0) {
2589                                /* index is empty, remove it;
2590                                 * handle must be already prepared by the
2591                                 * truncatei_leaf() */
2592                                err = ext4_ext_rm_idx(handle, inode, path + i);
2593                        }
2594                        /* root level has p_bh == NULL, brelse() eats this */
2595                        brelse(path[i].p_bh);
2596                        path[i].p_bh = NULL;
2597                        i--;
2598                        ext_debug("return to level %d\n", i);
2599                }
2600        }
2601
2602        trace_ext4_ext_remove_space_done(inode, start, depth, partial_cluster,
2603                        path->p_hdr->eh_entries);
2604
2605        /* If we still have something in the partial cluster and we have removed
2606         * even the first extent, then we should free the blocks in the partial
2607         * cluster as well. */
2608        if (partial_cluster && path->p_hdr->eh_entries == 0) {
2609                int flags = EXT4_FREE_BLOCKS_FORGET;
2610
2611                if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
2612                        flags |= EXT4_FREE_BLOCKS_METADATA;
2613
2614                ext4_free_blocks(handle, inode, NULL,
2615                                 EXT4_C2B(EXT4_SB(sb), partial_cluster),
2616                                 EXT4_SB(sb)->s_cluster_ratio, flags);
2617                partial_cluster = 0;
2618        }
2619
2620        /* TODO: flexible tree reduction should be here */
2621        if (path->p_hdr->eh_entries == 0) {
2622                /*
2623                 * truncate to zero freed all the tree,
2624                 * so we need to correct eh_depth
2625                 */
2626                err = ext4_ext_get_access(handle, inode, path);
2627                if (err == 0) {
2628                        ext_inode_hdr(inode)->eh_depth = 0;
2629                        ext_inode_hdr(inode)->eh_max =
2630                                cpu_to_le16(ext4_ext_space_root(inode, 0));
2631                        err = ext4_ext_dirty(handle, inode, path);
2632                }
2633        }
2634out:
2635        ext4_ext_drop_refs(path);
2636        kfree(path);
2637        if (err == -EAGAIN)
2638                goto again;
2639        ext4_journal_stop(handle);
2640
2641        return err;
2642}
2643
2644/*
2645 * called at mount time
2646 */
2647void ext4_ext_init(struct super_block *sb)
2648{
2649        /*
2650         * possible initialization would be here
2651         */
2652
2653        if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
2654#if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
2655                printk(KERN_INFO "EXT4-fs: file extents enabled");
2656#ifdef AGGRESSIVE_TEST
2657                printk(", aggressive tests");
2658#endif
2659#ifdef CHECK_BINSEARCH
2660                printk(", check binsearch");
2661#endif
2662#ifdef EXTENTS_STATS
2663                printk(", stats");
2664#endif
2665                printk("\n");
2666#endif
2667#ifdef EXTENTS_STATS
2668                spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
2669                EXT4_SB(sb)->s_ext_min = 1 << 30;
2670                EXT4_SB(sb)->s_ext_max = 0;
2671#endif
2672        }
2673}
2674
2675/*
2676 * called at umount time
2677 */
2678void ext4_ext_release(struct super_block *sb)
2679{
2680        if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
2681                return;
2682
2683#ifdef EXTENTS_STATS
2684        if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
2685                struct ext4_sb_info *sbi = EXT4_SB(sb);
2686                printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
2687                        sbi->s_ext_blocks, sbi->s_ext_extents,
2688                        sbi->s_ext_blocks / sbi->s_ext_extents);
2689                printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
2690                        sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
2691        }
2692#endif
2693}
2694
2695/* FIXME!! we need to try to merge to left or right after zero-out  */
2696static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
2697{
2698        ext4_fsblk_t ee_pblock;
2699        unsigned int ee_len;
2700        int ret;
2701
2702        ee_len    = ext4_ext_get_actual_len(ex);
2703        ee_pblock = ext4_ext_pblock(ex);
2704
2705        ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
2706        if (ret > 0)
2707                ret = 0;
2708
2709        return ret;
2710}
2711
2712/*
2713 * used by extent splitting.
2714 */
2715#define EXT4_EXT_MAY_ZEROOUT    0x1  /* safe to zeroout if split fails \
2716                                        due to ENOSPC */
2717#define EXT4_EXT_MARK_UNINIT1   0x2  /* mark first half uninitialized */
2718#define EXT4_EXT_MARK_UNINIT2   0x4  /* mark second half uninitialized */
2719
2720/*
2721 * ext4_split_extent_at() splits an extent at given block.
2722 *
2723 * @handle: the journal handle
2724 * @inode: the file inode
2725 * @path: the path to the extent
2726 * @split: the logical block where the extent is splitted.
2727 * @split_flags: indicates if the extent could be zeroout if split fails, and
2728 *               the states(init or uninit) of new extents.
2729 * @flags: flags used to insert new extent to extent tree.
2730 *
2731 *
2732 * Splits extent [a, b] into two extents [a, @split) and [@split, b], states
2733 * of which are deterimined by split_flag.
2734 *
2735 * There are two cases:
2736 *  a> the extent are splitted into two extent.
2737 *  b> split is not needed, and just mark the extent.
2738 *
2739 * return 0 on success.
2740 */
2741static int ext4_split_extent_at(handle_t *handle,
2742                             struct inode *inode,
2743                             struct ext4_ext_path *path,
2744                             ext4_lblk_t split,
2745                             int split_flag,
2746                             int flags)
2747{
2748        ext4_fsblk_t newblock;
2749        ext4_lblk_t ee_block;
2750        struct ext4_extent *ex, newex, orig_ex;
2751        struct ext4_extent *ex2 = NULL;
2752        unsigned int ee_len, depth;
2753        int err = 0;
2754
2755        ext_debug("ext4_split_extents_at: inode %lu, logical"
2756                "block %llu\n", inode->i_ino, (unsigned long long)split);
2757
2758        ext4_ext_show_leaf(inode, path);
2759
2760        depth = ext_depth(inode);
2761        ex = path[depth].p_ext;
2762        ee_block = le32_to_cpu(ex->ee_block);
2763        ee_len = ext4_ext_get_actual_len(ex);
2764        newblock = split - ee_block + ext4_ext_pblock(ex);
2765
2766        BUG_ON(split < ee_block || split >= (ee_block + ee_len));
2767
2768        err = ext4_ext_get_access(handle, inode, path + depth);
2769        if (err)
2770                goto out;
2771
2772        if (split == ee_block) {
2773                /*
2774                 * case b: block @split is the block that the extent begins with
2775                 * then we just change the state of the extent, and splitting
2776                 * is not needed.
2777                 */
2778                if (split_flag & EXT4_EXT_MARK_UNINIT2)
2779                        ext4_ext_mark_uninitialized(ex);
2780                else
2781                        ext4_ext_mark_initialized(ex);
2782
2783                if (!(flags & EXT4_GET_BLOCKS_PRE_IO))
2784                        ext4_ext_try_to_merge(inode, path, ex);
2785
2786                err = ext4_ext_dirty(handle, inode, path + depth);
2787                goto out;
2788        }
2789
2790        /* case a */
2791        memcpy(&orig_ex, ex, sizeof(orig_ex));
2792        ex->ee_len = cpu_to_le16(split - ee_block);
2793        if (split_flag & EXT4_EXT_MARK_UNINIT1)
2794                ext4_ext_mark_uninitialized(ex);
2795
2796        /*
2797         * path may lead to new leaf, not to original leaf any more
2798         * after ext4_ext_insert_extent() returns,
2799         */
2800        err = ext4_ext_dirty(handle, inode, path + depth);
2801        if (err)
2802                goto fix_extent_len;
2803
2804        ex2 = &newex;
2805        ex2->ee_block = cpu_to_le32(split);
2806        ex2->ee_len   = cpu_to_le16(ee_len - (split - ee_block));
2807        ext4_ext_store_pblock(ex2, newblock);
2808        if (split_flag & EXT4_EXT_MARK_UNINIT2)
2809                ext4_ext_mark_uninitialized(ex2);
2810
2811        err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
2812        if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
2813                err = ext4_ext_zeroout(inode, &orig_ex);
2814                if (err)
2815                        goto fix_extent_len;
2816                /* update the extent length and mark as initialized */
2817                ex->ee_len = cpu_to_le32(ee_len);
2818                ext4_ext_try_to_merge(inode, path, ex);
2819                err = ext4_ext_dirty(handle, inode, path + depth);
2820                goto out;
2821        } else if (err)
2822                goto fix_extent_len;
2823
2824out:
2825        ext4_ext_show_leaf(inode, path);
2826        return err;
2827
2828fix_extent_len:
2829        ex->ee_len = orig_ex.ee_len;
2830        ext4_ext_dirty(handle, inode, path + depth);
2831        return err;
2832}
2833
2834/*
2835 * ext4_split_extents() splits an extent and mark extent which is covered
2836 * by @map as split_flags indicates
2837 *
2838 * It may result in splitting the extent into multiple extents (upto three)
2839 * There are three possibilities:
2840 *   a> There is no split required
2841 *   b> Splits in two extents: Split is happening at either end of the extent
2842 *   c> Splits in three extents: Somone is splitting in middle of the extent
2843 *
2844 */
2845static int ext4_split_extent(handle_t *handle,
2846                              struct inode *inode,
2847                              struct ext4_ext_path *path,
2848                              struct ext4_map_blocks *map,
2849                              int split_flag,
2850                              int flags)
2851{
2852        ext4_lblk_t ee_block;
2853        struct ext4_extent *ex;
2854        unsigned int ee_len, depth;
2855        int err = 0;
2856        int uninitialized;
2857        int split_flag1, flags1;
2858
2859        depth = ext_depth(inode);
2860        ex = path[depth].p_ext;
2861        ee_block = le32_to_cpu(ex->ee_block);
2862        ee_len = ext4_ext_get_actual_len(ex);
2863        uninitialized = ext4_ext_is_uninitialized(ex);
2864
2865        if (map->m_lblk + map->m_len < ee_block + ee_len) {
2866                split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2867                              EXT4_EXT_MAY_ZEROOUT : 0;
2868                flags1 = flags | EXT4_GET_BLOCKS_PRE_IO;
2869                if (uninitialized)
2870                        split_flag1 |= EXT4_EXT_MARK_UNINIT1 |
2871                                       EXT4_EXT_MARK_UNINIT2;
2872                err = ext4_split_extent_at(handle, inode, path,
2873                                map->m_lblk + map->m_len, split_flag1, flags1);
2874                if (err)
2875                        goto out;
2876        }
2877
2878        ext4_ext_drop_refs(path);
2879        path = ext4_ext_find_extent(inode, map->m_lblk, path);
2880        if (IS_ERR(path))
2881                return PTR_ERR(path);
2882
2883        if (map->m_lblk >= ee_block) {
2884                split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ?
2885                              EXT4_EXT_MAY_ZEROOUT : 0;
2886                if (uninitialized)
2887                        split_flag1 |= EXT4_EXT_MARK_UNINIT1;
2888                if (split_flag & EXT4_EXT_MARK_UNINIT2)
2889                        split_flag1 |= EXT4_EXT_MARK_UNINIT2;
2890                err = ext4_split_extent_at(handle, inode, path,
2891                                map->m_lblk, split_flag1, flags);
2892                if (err)
2893                        goto out;
2894        }
2895
2896        ext4_ext_show_leaf(inode, path);
2897out:
2898        return err ? err : map->m_len;
2899}
2900
2901#define EXT4_EXT_ZERO_LEN 7
2902/*
2903 * This function is called by ext4_ext_map_blocks() if someone tries to write
2904 * to an uninitialized extent. It may result in splitting the uninitialized
2905 * extent into multiple extents (up to three - one initialized and two
2906 * uninitialized).
2907 * There are three possibilities:
2908 *   a> There is no split required: Entire extent should be initialized
2909 *   b> Splits in two extents: Write is happening at either end of the extent
2910 *   c> Splits in three extents: Somone is writing in middle of the extent
2911 *
2912 * Pre-conditions:
2913 *  - The extent pointed to by 'path' is uninitialized.
2914 *  - The extent pointed to by 'path' contains a superset
2915 *    of the logical span [map->m_lblk, map->m_lblk + map->m_len).
2916 *
2917 * Post-conditions on success:
2918 *  - the returned value is the number of blocks beyond map->l_lblk
2919 *    that are allocated and initialized.
2920 *    It is guaranteed to be >= map->m_len.
2921 */
2922static int ext4_ext_convert_to_initialized(handle_t *handle,
2923                                           struct inode *inode,
2924                                           struct ext4_map_blocks *map,
2925                                           struct ext4_ext_path *path)
2926{
2927        struct ext4_extent_header *eh;
2928        struct ext4_map_blocks split_map;
2929        struct ext4_extent zero_ex;
2930        struct ext4_extent *ex;
2931        ext4_lblk_t ee_block, eof_block;
2932        unsigned int ee_len, depth;
2933        int allocated;
2934        int err = 0;
2935        int split_flag = 0;
2936
2937        ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
2938                "block %llu, max_blocks %u\n", inode->i_ino,
2939                (unsigned long long)map->m_lblk, map->m_len);
2940
2941        eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
2942                inode->i_sb->s_blocksize_bits;
2943        if (eof_block < map->m_lblk + map->m_len)
2944                eof_block = map->m_lblk + map->m_len;
2945
2946        depth = ext_depth(inode);
2947        eh = path[depth].p_hdr;
2948        ex = path[depth].p_ext;
2949        ee_block = le32_to_cpu(ex->ee_block);
2950        ee_len = ext4_ext_get_actual_len(ex);
2951        allocated = ee_len - (map->m_lblk - ee_block);
2952
2953        trace_ext4_ext_convert_to_initialized_enter(inode, map, ex);
2954
2955        /* Pre-conditions */
2956        BUG_ON(!ext4_ext_is_uninitialized(ex));
2957        BUG_ON(!in_range(map->m_lblk, ee_block, ee_len));
2958
2959        /*
2960         * Attempt to transfer newly initialized blocks from the currently
2961         * uninitialized extent to its left neighbor. This is much cheaper
2962         * than an insertion followed by a merge as those involve costly
2963         * memmove() calls. This is the common case in steady state for
2964         * workloads doing fallocate(FALLOC_FL_KEEP_SIZE) followed by append
2965         * writes.
2966         *
2967         * Limitations of the current logic:
2968         *  - L1: we only deal with writes at the start of the extent.
2969         *    The approach could be extended to writes at the end
2970         *    of the extent but this scenario was deemed less common.
2971         *  - L2: we do not deal with writes covering the whole extent.
2972         *    This would require removing the extent if the transfer
2973         *    is possible.
2974         *  - L3: we only attempt to merge with an extent stored in the
2975         *    same extent tree node.
2976         */
2977        if ((map->m_lblk == ee_block) &&        /*L1*/
2978                (map->m_len < ee_len) &&        /*L2*/
2979                (ex > EXT_FIRST_EXTENT(eh))) {  /*L3*/
2980                struct ext4_extent *prev_ex;
2981                ext4_lblk_t prev_lblk;
2982                ext4_fsblk_t prev_pblk, ee_pblk;
2983                unsigned int prev_len, write_len;
2984
2985                prev_ex = ex - 1;
2986                prev_lblk = le32_to_cpu(prev_ex->ee_block);
2987                prev_len = ext4_ext_get_actual_len(prev_ex);
2988                prev_pblk = ext4_ext_pblock(prev_ex);
2989                ee_pblk = ext4_ext_pblock(ex);
2990                write_len = map->m_len;
2991
2992                /*
2993                 * A transfer of blocks from 'ex' to 'prev_ex' is allowed
2994                 * upon those conditions:
2995                 * - C1: prev_ex is initialized,
2996                 * - C2: prev_ex is logically abutting ex,
2997                 * - C3: prev_ex is physically abutting ex,
2998                 * - C4: prev_ex can receive the additional blocks without
2999                 *   overflowing the (initialized) length limit.
3000                 */
3001                if ((!ext4_ext_is_uninitialized(prev_ex)) &&            /*C1*/
3002                        ((prev_lblk + prev_len) == ee_block) &&         /*C2*/
3003                        ((prev_pblk + prev_len) == ee_pblk) &&          /*C3*/
3004                        (prev_len < (EXT_INIT_MAX_LEN - write_len))) {  /*C4*/
3005                        err = ext4_ext_get_access(handle, inode, path + depth);
3006                        if (err)
3007                                goto out;
3008
3009                        trace_ext4_ext_convert_to_initialized_fastpath(inode,
3010                                map, ex, prev_ex);
3011
3012                        /* Shift the start of ex by 'write_len' blocks */
3013                        ex->ee_block = cpu_to_le32(ee_block + write_len);
3014                        ext4_ext_store_pblock(ex, ee_pblk + write_len);
3015                        ex->ee_len = cpu_to_le16(ee_len - write_len);
3016                        ext4_ext_mark_uninitialized(ex); /* Restore the flag */
3017
3018                        /* Extend prev_ex by 'write_len' blocks */
3019                        prev_ex->ee_len = cpu_to_le16(prev_len + write_len);
3020
3021                        /* Mark the block containing both extents as dirty */
3022                        ext4_ext_dirty(handle, inode, path + depth);
3023
3024                        /* Update path to point to the right extent */
3025                        path[depth].p_ext = prev_ex;
3026
3027                        /* Result: number of initialized blocks past m_lblk */
3028                        allocated = write_len;
3029                        goto out;
3030                }
3031        }
3032
3033        WARN_ON(map->m_lblk < ee_block);
3034        /*
3035         * It is safe to convert extent to initialized via explicit
3036         * zeroout only if extent is fully insde i_size or new_size.
3037         */
3038        split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3039
3040        /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
3041        if (ee_len <= 2*EXT4_EXT_ZERO_LEN &&
3042            (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3043                err = ext4_ext_zeroout(inode, ex);
3044                if (err)
3045                        goto out;
3046
3047                err = ext4_ext_get_access(handle, inode, path + depth);
3048                if (err)
3049                        goto out;
3050                ext4_ext_mark_initialized(ex);
3051                ext4_ext_try_to_merge(inode, path, ex);
3052                err = ext4_ext_dirty(handle, inode, path + depth);
3053                goto out;
3054        }
3055
3056        /*
3057         * four cases:
3058         * 1. split the extent into three extents.
3059         * 2. split the extent into two extents, zeroout the first half.
3060         * 3. split the extent into two extents, zeroout the second half.
3061         * 4. split the extent into two extents with out zeroout.
3062         */
3063        split_map.m_lblk = map->m_lblk;
3064        split_map.m_len = map->m_len;
3065
3066        if (allocated > map->m_len) {
3067                if (allocated <= EXT4_EXT_ZERO_LEN &&
3068                    (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3069                        /* case 3 */
3070                        zero_ex.ee_block =
3071                                         cpu_to_le32(map->m_lblk);
3072                        zero_ex.ee_len = cpu_to_le16(allocated);
3073                        ext4_ext_store_pblock(&zero_ex,
3074                                ext4_ext_pblock(ex) + map->m_lblk - ee_block);
3075                        err = ext4_ext_zeroout(inode, &zero_ex);
3076                        if (err)
3077                                goto out;
3078                        split_map.m_lblk = map->m_lblk;
3079                        split_map.m_len = allocated;
3080                } else if ((map->m_lblk - ee_block + map->m_len <
3081                           EXT4_EXT_ZERO_LEN) &&
3082                           (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
3083                        /* case 2 */
3084                        if (map->m_lblk != ee_block) {
3085                                zero_ex.ee_block = ex->ee_block;
3086                                zero_ex.ee_len = cpu_to_le16(map->m_lblk -
3087                                                        ee_block);
3088                                ext4_ext_store_pblock(&zero_ex,
3089                                                      ext4_ext_pblock(ex));
3090                                err = ext4_ext_zeroout(inode, &zero_ex);
3091                                if (err)
3092                                        goto out;
3093                        }
3094
3095                        split_map.m_lblk = ee_block;
3096                        split_map.m_len = map->m_lblk - ee_block + map->m_len;
3097                        allocated = map->m_len;
3098                }
3099        }
3100
3101        allocated = ext4_split_extent(handle, inode, path,
3102                                       &split_map, split_flag, 0);
3103        if (allocated < 0)
3104                err = allocated;
3105
3106out:
3107        return err ? err : allocated;
3108}
3109
3110/*
3111 * This function is called by ext4_ext_map_blocks() from
3112 * ext4_get_blocks_dio_write() when DIO to write
3113 * to an uninitialized extent.
3114 *
3115 * Writing to an uninitialized extent may result in splitting the uninitialized
3116 * extent into multiple /initialized uninitialized extents (up to three)
3117 * There are three possibilities:
3118 *   a> There is no split required: Entire extent should be uninitialized
3119 *   b> Splits in two extents: Write is happening at either end of the extent
3120 *   c> Splits in three extents: Somone is writing in middle of the extent
3121 *
3122 * One of more index blocks maybe needed if the extent tree grow after
3123 * the uninitialized extent split. To prevent ENOSPC occur at the IO
3124 * complete, we need to split the uninitialized extent before DIO submit
3125 * the IO. The uninitialized extent called at this time will be split
3126 * into three uninitialized extent(at most). After IO complete, the part
3127 * being filled will be convert to initialized by the end_io callback function
3128 * via ext4_convert_unwritten_extents().
3129 *
3130 * Returns the size of uninitialized extent to be written on success.
3131 */
3132static int ext4_split_unwritten_extents(handle_t *handle,
3133                                        struct inode *inode,
3134                                        struct ext4_map_blocks *map,
3135                                        struct ext4_ext_path *path,
3136                                        int flags)
3137{
3138        ext4_lblk_t eof_block;
3139        ext4_lblk_t ee_block;
3140        struct ext4_extent *ex;
3141        unsigned int ee_len;
3142        int split_flag = 0, depth;
3143
3144        ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
3145                "block %llu, max_blocks %u\n", inode->i_ino,
3146                (unsigned long long)map->m_lblk, map->m_len);
3147
3148        eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
3149                inode->i_sb->s_blocksize_bits;
3150        if (eof_block < map->m_lblk + map->m_len)
3151                eof_block = map->m_lblk + map->m_len;
3152        /*
3153         * It is safe to convert extent to initialized via explicit
3154         * zeroout only if extent is fully insde i_size or new_size.
3155         */
3156        depth = ext_depth(inode);
3157        ex = path[depth].p_ext;
3158        ee_block = le32_to_cpu(ex->ee_block);
3159        ee_len = ext4_ext_get_actual_len(ex);
3160
3161        split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
3162        split_flag |= EXT4_EXT_MARK_UNINIT2;
3163
3164        flags |= EXT4_GET_BLOCKS_PRE_IO;
3165        return ext4_split_extent(handle, inode, path, map, split_flag, flags);
3166}
3167
3168static int ext4_convert_unwritten_extents_endio(handle_t *handle,
3169                                              struct inode *inode,
3170                                              struct ext4_ext_path *path)
3171{
3172        struct ext4_extent *ex;
3173        int depth;
3174        int err = 0;
3175
3176        depth = ext_depth(inode);
3177        ex = path[depth].p_ext;
3178
3179        ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
3180                "block %llu, max_blocks %u\n", inode->i_ino,
3181                (unsigned long long)le32_to_cpu(ex->ee_block),
3182                ext4_ext_get_actual_len(ex));
3183
3184        err = ext4_ext_get_access(handle, inode, path + depth);
3185        if (err)
3186                goto out;
3187        /* first mark the extent as initialized */
3188        ext4_ext_mark_initialized(ex);
3189
3190        /* note: ext4_ext_correct_indexes() isn't needed here because
3191         * borders are not changed
3192         */
3193        ext4_ext_try_to_merge(inode, path, ex);
3194
3195        /* Mark modified extent as dirty */
3196        err = ext4_ext_dirty(handle, inode, path + depth);
3197out:
3198        ext4_ext_show_leaf(inode, path);
3199        return err;
3200}
3201
3202static void unmap_underlying_metadata_blocks(struct block_device *bdev,
3203                        sector_t block, int count)
3204{
3205        int i;
3206        for (i = 0; i < count; i++)
3207                unmap_underlying_metadata(bdev, block + i);
3208}
3209
3210/*
3211 * Handle EOFBLOCKS_FL flag, clearing it if necessary
3212 */
3213static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
3214                              ext4_lblk_t lblk,
3215                              struct ext4_ext_path *path,
3216                              unsigned int len)
3217{
3218        int i, depth;
3219        struct ext4_extent_header *eh;
3220        struct ext4_extent *last_ex;
3221
3222        if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
3223                return 0;
3224
3225        depth = ext_depth(inode);
3226        eh = path[depth].p_hdr;
3227
3228        if (unlikely(!eh->eh_entries)) {
3229                EXT4_ERROR_INODE(inode, "eh->eh_entries == 0 and "
3230                                 "EOFBLOCKS_FL set");
3231                return -EIO;
3232        }
3233        last_ex = EXT_LAST_EXTENT(eh);
3234        /*
3235         * We should clear the EOFBLOCKS_FL flag if we are writing the
3236         * last block in the last extent in the file.  We test this by
3237         * first checking to see if the caller to
3238         * ext4_ext_get_blocks() was interested in the last block (or
3239         * a block beyond the last block) in the current extent.  If
3240         * this turns out to be false, we can bail out from this
3241         * function immediately.
3242         */
3243        if (lblk + len < le32_to_cpu(last_ex->ee_block) +
3244            ext4_ext_get_actual_len(last_ex))
3245                return 0;
3246        /*
3247         * If the caller does appear to be planning to write at or
3248         * beyond the end of the current extent, we then test to see
3249         * if the current extent is the last extent in the file, by
3250         * checking to make sure it was reached via the rightmost node
3251         * at each level of the tree.
3252         */
3253        for (i = depth-1; i >= 0; i--)
3254                if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
3255                        return 0;
3256        ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3257        return ext4_mark_inode_dirty(handle, inode);
3258}
3259
3260/**
3261 * ext4_find_delalloc_range: find delayed allocated block in the given range.
3262 *
3263 * Goes through the buffer heads in the range [lblk_start, lblk_end] and returns
3264 * whether there are any buffers marked for delayed allocation. It returns '1'
3265 * on the first delalloc'ed buffer head found. If no buffer head in the given
3266 * range is marked for delalloc, it returns 0.
3267 * lblk_start should always be <= lblk_end.
3268 * search_hint_reverse is to indicate that searching in reverse from lblk_end to
3269 * lblk_start might be more efficient (i.e., we will likely hit the delalloc'ed
3270 * block sooner). This is useful when blocks are truncated sequentially from
3271 * lblk_start towards lblk_end.
3272 */
3273static int ext4_find_delalloc_range(struct inode *inode,
3274                                    ext4_lblk_t lblk_start,
3275                                    ext4_lblk_t lblk_end,
3276                                    int search_hint_reverse)
3277{
3278        struct address_space *mapping = inode->i_mapping;
3279        struct buffer_head *head, *bh = NULL;
3280        struct page *page;
3281        ext4_lblk_t i, pg_lblk;
3282        pgoff_t index;
3283
3284        /* reverse search wont work if fs block size is less than page size */
3285        if (inode->i_blkbits < PAGE_CACHE_SHIFT)
3286                search_hint_reverse = 0;
3287
3288        if (search_hint_reverse)
3289                i = lblk_end;
3290        else
3291                i = lblk_start;
3292
3293        index = i >> (PAGE_CACHE_SHIFT - inode->i_blkbits);
3294
3295        while ((i >= lblk_start) && (i <= lblk_end)) {
3296                page = find_get_page(mapping, index);
3297                if (!page)
3298                        goto nextpage;
3299
3300                if (!page_has_buffers(page))
3301                        goto nextpage;
3302
3303                head = page_buffers(page);
3304                if (!head)
3305                        goto nextpage;
3306
3307                bh = head;
3308                pg_lblk = index << (PAGE_CACHE_SHIFT -
3309                                                inode->i_blkbits);
3310                do {
3311                        if (unlikely(pg_lblk < lblk_start)) {
3312                                /*
3313                                 * This is possible when fs block size is less
3314                                 * than page size and our cluster starts/ends in
3315                                 * middle of the page. So we need to skip the
3316                                 * initial few blocks till we reach the 'lblk'
3317                                 */
3318                                pg_lblk++;
3319                                continue;
3320                        }
3321
3322                        /* Check if the buffer is delayed allocated and that it
3323                         * is not yet mapped. (when da-buffers are mapped during
3324                         * their writeout, their da_mapped bit is set.)
3325                         */
3326                        if (buffer_delay(bh) && !buffer_da_mapped(bh)) {
3327                                page_cache_release(page);
3328                                trace_ext4_find_delalloc_range(inode,
3329                                                lblk_start, lblk_end,
3330                                                search_hint_reverse,
3331                                                1, i);
3332                                return 1;
3333                        }
3334                        if (search_hint_reverse)
3335                                i--;
3336                        else
3337                                i++;
3338                } while ((i >= lblk_start) && (i <= lblk_end) &&
3339                                ((bh = bh->b_this_page) != head));
3340nextpage:
3341                if (page)
3342                        page_cache_release(page);
3343                /*
3344                 * Move to next page. 'i' will be the first lblk in the next
3345                 * page.
3346                 */
3347                if (search_hint_reverse)
3348                        index--;
3349                else
3350                        index++;
3351                i = index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
3352        }
3353
3354        trace_ext4_find_delalloc_range(inode, lblk_start, lblk_end,
3355                                        search_hint_reverse, 0, 0);
3356        return 0;
3357}
3358
3359int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk,
3360                               int search_hint_reverse)
3361{
3362        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3363        ext4_lblk_t lblk_start, lblk_end;
3364        lblk_start = lblk & (~(sbi->s_cluster_ratio - 1));
3365        lblk_end = lblk_start + sbi->s_cluster_ratio - 1;
3366
3367        return ext4_find_delalloc_range(inode, lblk_start, lblk_end,
3368                                        search_hint_reverse);
3369}
3370
3371/**
3372 * Determines how many complete clusters (out of those specified by the 'map')
3373 * are under delalloc and were reserved quota for.
3374 * This function is called when we are writing out the blocks that were
3375 * originally written with their allocation delayed, but then the space was
3376 * allocated using fallocate() before the delayed allocation could be resolved.
3377 * The cases to look for are:
3378 * ('=' indicated delayed allocated blocks
3379 *  '-' indicates non-delayed allocated blocks)
3380 * (a) partial clusters towards beginning and/or end outside of allocated range
3381 *     are not delalloc'ed.
3382 *      Ex:
3383 *      |----c---=|====c====|====c====|===-c----|
3384 *               |++++++ allocated ++++++|
3385 *      ==> 4 complete clusters in above example
3386 *
3387 * (b) partial cluster (outside of allocated range) towards either end is
3388 *     marked for delayed allocation. In this case, we will exclude that
3389 *     cluster.
3390 *      Ex:
3391 *      |----====c========|========c========|
3392 *           |++++++ allocated ++++++|
3393 *      ==> 1 complete clusters in above example
3394 *
3395 *      Ex:
3396 *      |================c================|
3397 *            |++++++ allocated ++++++|
3398 *      ==> 0 complete clusters in above example
3399 *
3400 * The ext4_da_update_reserve_space will be called only if we
3401 * determine here that there were some "entire" clusters that span
3402 * this 'allocated' range.
3403 * In the non-bigalloc case, this function will just end up returning num_blks
3404 * without ever calling ext4_find_delalloc_range.
3405 */
3406static unsigned int
3407get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start,
3408                           unsigned int num_blks)
3409{
3410        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3411        ext4_lblk_t alloc_cluster_start, alloc_cluster_end;
3412        ext4_lblk_t lblk_from, lblk_to, c_offset;
3413        unsigned int allocated_clusters = 0;
3414
3415        alloc_cluster_start = EXT4_B2C(sbi, lblk_start);
3416        alloc_cluster_end = EXT4_B2C(sbi, lblk_start + num_blks - 1);
3417
3418        /* max possible clusters for this allocation */
3419        allocated_clusters = alloc_cluster_end - alloc_cluster_start + 1;
3420
3421        trace_ext4_get_reserved_cluster_alloc(inode, lblk_start, num_blks);
3422
3423        /* Check towards left side */
3424        c_offset = lblk_start & (sbi->s_cluster_ratio - 1);
3425        if (c_offset) {
3426                lblk_from = lblk_start & (~(sbi->s_cluster_ratio - 1));
3427                lblk_to = lblk_from + c_offset - 1;
3428
3429                if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0))
3430                        allocated_clusters--;
3431        }
3432
3433        /* Now check towards right. */
3434        c_offset = (lblk_start + num_blks) & (sbi->s_cluster_ratio - 1);
3435        if (allocated_clusters && c_offset) {
3436                lblk_from = lblk_start + num_blks;
3437                lblk_to = lblk_from + (sbi->s_cluster_ratio - c_offset) - 1;
3438
3439                if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0))
3440                        allocated_clusters--;
3441        }
3442
3443        return allocated_clusters;
3444}
3445
3446static int
3447ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
3448                        struct ext4_map_blocks *map,
3449                        struct ext4_ext_path *path, int flags,
3450                        unsigned int allocated, ext4_fsblk_t newblock)
3451{
3452        int ret = 0;
3453        int err = 0;
3454        ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3455
3456        ext_debug("ext4_ext_handle_uninitialized_extents: inode %lu, logical"
3457                  "block %llu, max_blocks %u, flags %d, allocated %u",
3458                  inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
3459                  flags, allocated);
3460        ext4_ext_show_leaf(inode, path);
3461
3462        trace_ext4_ext_handle_uninitialized_extents(inode, map, allocated,
3463                                                    newblock);
3464
3465        /* get_block() before submit the IO, split the extent */
3466        if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
3467                ret = ext4_split_unwritten_extents(handle, inode, map,
3468                                                   path, flags);
3469                /*
3470                 * Flag the inode(non aio case) or end_io struct (aio case)
3471                 * that this IO needs to conversion to written when IO is
3472                 * completed
3473                 */
3474                if (io)
3475                        ext4_set_io_unwritten_flag(inode, io);
3476                else
3477                        ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
3478                if (ext4_should_dioread_nolock(inode))
3479                        map->m_flags |= EXT4_MAP_UNINIT;
3480                goto out;
3481        }
3482        /* IO end_io complete, convert the filled extent to written */
3483        if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
3484                ret = ext4_convert_unwritten_extents_endio(handle, inode,
3485                                                        path);
3486                if (ret >= 0) {
3487                        ext4_update_inode_fsync_trans(handle, inode, 1);
3488                        err = check_eofblocks_fl(handle, inode, map->m_lblk,
3489                                                 path, map->m_len);
3490                } else
3491                        err = ret;
3492                goto out2;
3493        }
3494        /* buffered IO case */
3495        /*
3496         * repeat fallocate creation request
3497         * we already have an unwritten extent
3498         */
3499        if (flags & EXT4_GET_BLOCKS_UNINIT_EXT)
3500                goto map_out;
3501
3502        /* buffered READ or buffered write_begin() lookup */
3503        if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3504                /*
3505                 * We have blocks reserved already.  We
3506                 * return allocated blocks so that delalloc
3507                 * won't do block reservation for us.  But
3508                 * the buffer head will be unmapped so that
3509                 * a read from the block returns 0s.
3510                 */
3511                map->m_flags |= EXT4_MAP_UNWRITTEN;
3512                goto out1;
3513        }
3514
3515        /* buffered write, writepage time, convert*/
3516        ret = ext4_ext_convert_to_initialized(handle, inode, map, path);
3517        if (ret >= 0)
3518                ext4_update_inode_fsync_trans(handle, inode, 1);
3519out:
3520        if (ret <= 0) {
3521                err = ret;
3522                goto out2;
3523        } else
3524                allocated = ret;
3525        map->m_flags |= EXT4_MAP_NEW;
3526        /*
3527         * if we allocated more blocks than requested
3528         * we need to make sure we unmap the extra block
3529         * allocated. The actual needed block will get
3530         * unmapped later when we find the buffer_head marked
3531         * new.
3532         */
3533        if (allocated > map->m_len) {
3534                unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
3535                                        newblock + map->m_len,
3536                                        allocated - map->m_len);
3537                allocated = map->m_len;
3538        }
3539
3540        /*
3541         * If we have done fallocate with the offset that is already
3542         * delayed allocated, we would have block reservation
3543         * and quota reservation done in the delayed write path.
3544         * But fallocate would have already updated quota and block
3545         * count for this offset. So cancel these reservation
3546         */
3547        if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
3548                unsigned int reserved_clusters;
3549                reserved_clusters = get_reserved_cluster_alloc(inode,
3550                                map->m_lblk, map->m_len);
3551                if (reserved_clusters)
3552                        ext4_da_update_reserve_space(inode,
3553                                                     reserved_clusters,
3554                                                     0);
3555        }
3556
3557map_out:
3558        map->m_flags |= EXT4_MAP_MAPPED;
3559        if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0) {
3560                err = check_eofblocks_fl(handle, inode, map->m_lblk, path,
3561                                         map->m_len);
3562                if (err < 0)
3563                        goto out2;
3564        }
3565out1:
3566        if (allocated > map->m_len)
3567                allocated = map->m_len;
3568        ext4_ext_show_leaf(inode, path);
3569        map->m_pblk = newblock;
3570        map->m_len = allocated;
3571out2:
3572        if (path) {
3573                ext4_ext_drop_refs(path);
3574                kfree(path);
3575        }
3576        return err ? err : allocated;
3577}
3578
3579/*
3580 * get_implied_cluster_alloc - check to see if the requested
3581 * allocation (in the map structure) overlaps with a cluster already
3582 * allocated in an extent.
3583 *      @sb     The filesystem superblock structure
3584 *      @map    The requested lblk->pblk mapping
3585 *      @ex     The extent structure which might contain an implied
3586 *                      cluster allocation
3587 *
3588 * This function is called by ext4_ext_map_blocks() after we failed to
3589 * find blocks that were already in the inode's extent tree.  Hence,
3590 * we know that the beginning of the requested region cannot overlap
3591 * the extent from the inode's extent tree.  There are three cases we
3592 * want to catch.  The first is this case:
3593 *
3594 *               |--- cluster # N--|
3595 *    |--- extent ---|  |---- requested region ---|
3596 *                      |==========|
3597 *
3598 * The second case that we need to test for is this one:
3599 *
3600 *   |--------- cluster # N ----------------|
3601 *         |--- requested region --|   |------- extent ----|
3602 *         |=======================|
3603 *
3604 * The third case is when the requested region lies between two extents
3605 * within the same cluster:
3606 *          |------------- cluster # N-------------|
3607 * |----- ex -----|                  |---- ex_right ----|
3608 *                  |------ requested region ------|
3609 *                  |================|
3610 *
3611 * In each of the above cases, we need to set the map->m_pblk and
3612 * map->m_len so it corresponds to the return the extent labelled as
3613 * "|====|" from cluster #N, since it is already in use for data in
3614 * cluster EXT4_B2C(sbi, map->m_lblk).  We will then return 1 to
3615 * signal to ext4_ext_map_blocks() that map->m_pblk should be treated
3616 * as a new "allocated" block region.  Otherwise, we will return 0 and
3617 * ext4_ext_map_blocks() will then allocate one or more new clusters
3618 * by calling ext4_mb_new_blocks().
3619 */
3620static int get_implied_cluster_alloc(struct super_block *sb,
3621                                     struct ext4_map_blocks *map,
3622                                     struct ext4_extent *ex,
3623                                     struct ext4_ext_path *path)
3624{
3625        struct ext4_sb_info *sbi = EXT4_SB(sb);
3626        ext4_lblk_t c_offset = map->m_lblk & (sbi->s_cluster_ratio-1);
3627        ext4_lblk_t ex_cluster_start, ex_cluster_end;
3628        ext4_lblk_t rr_cluster_start, rr_cluster_end;
3629        ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
3630        ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
3631        unsigned short ee_len = ext4_ext_get_actual_len(ex);
3632
3633        /* The extent passed in that we are trying to match */
3634        ex_cluster_start = EXT4_B2C(sbi, ee_block);
3635        ex_cluster_end = EXT4_B2C(sbi, ee_block + ee_len - 1);
3636
3637        /* The requested region passed into ext4_map_blocks() */
3638        rr_cluster_start = EXT4_B2C(sbi, map->m_lblk);
3639        rr_cluster_end = EXT4_B2C(sbi, map->m_lblk + map->m_len - 1);
3640
3641        if ((rr_cluster_start == ex_cluster_end) ||
3642            (rr_cluster_start == ex_cluster_start)) {
3643                if (rr_cluster_start == ex_cluster_end)
3644                        ee_start += ee_len - 1;
3645                map->m_pblk = (ee_start & ~(sbi->s_cluster_ratio - 1)) +
3646                        c_offset;
3647                map->m_len = min(map->m_len,
3648                                 (unsigned) sbi->s_cluster_ratio - c_offset);
3649                /*
3650                 * Check for and handle this case:
3651                 *
3652                 *   |--------- cluster # N-------------|
3653                 *                     |------- extent ----|
3654                 *         |--- requested region ---|
3655                 *         |===========|
3656                 */
3657
3658                if (map->m_lblk < ee_block)
3659                        map->m_len = min(map->m_len, ee_block - map->m_lblk);
3660
3661                /*
3662                 * Check for the case where there is already another allocated
3663                 * block to the right of 'ex' but before the end of the cluster.
3664                 *
3665                 *          |------------- cluster # N-------------|
3666                 * |----- ex -----|                  |---- ex_right ----|
3667                 *                  |------ requested region ------|
3668                 *                  |================|
3669                 */
3670                if (map->m_lblk > ee_block) {
3671                        ext4_lblk_t next = ext4_ext_next_allocated_block(path);
3672                        map->m_len = min(map->m_len, next - map->m_lblk);
3673                }
3674
3675                trace_ext4_get_implied_cluster_alloc_exit(sb, map, 1);
3676                return 1;
3677        }
3678
3679        trace_ext4_get_implied_cluster_alloc_exit(sb, map, 0);
3680        return 0;
3681}
3682
3683
3684/*
3685 * Block allocation/map/preallocation routine for extents based files
3686 *
3687 *
3688 * Need to be called with
3689 * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
3690 * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
3691 *
3692 * return > 0, number of of blocks already mapped/allocated
3693 *          if create == 0 and these are pre-allocated blocks
3694 *              buffer head is unmapped
3695 *          otherwise blocks are mapped
3696 *
3697 * return = 0, if plain look up failed (blocks have not been allocated)
3698 *          buffer head is unmapped
3699 *
3700 * return < 0, error case.
3701 */
3702int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
3703                        struct ext4_map_blocks *map, int flags)
3704{
3705        struct ext4_ext_path *path = NULL;
3706        struct ext4_extent newex, *ex, *ex2;
3707        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3708        ext4_fsblk_t newblock = 0;
3709        int free_on_err = 0, err = 0, depth, ret;
3710        unsigned int allocated = 0, offset = 0;
3711        unsigned int allocated_clusters = 0;
3712        unsigned int punched_out = 0;
3713        unsigned int result = 0;
3714        struct ext4_allocation_request ar;
3715        ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
3716        ext4_lblk_t cluster_offset;
3717
3718        ext_debug("blocks %u/%u requested for inode %lu\n",
3719                  map->m_lblk, map->m_len, inode->i_ino);
3720        trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
3721
3722        /* check in cache */
3723        if (!(flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) &&
3724                ext4_ext_in_cache(inode, map->m_lblk, &newex)) {
3725                if (!newex.ee_start_lo && !newex.ee_start_hi) {
3726                        if ((sbi->s_cluster_ratio > 1) &&
3727                            ext4_find_delalloc_cluster(inode, map->m_lblk, 0))
3728                                map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3729
3730                        if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3731                                /*
3732                                 * block isn't allocated yet and
3733                                 * user doesn't want to allocate it
3734                                 */
3735                                goto out2;
3736                        }
3737                        /* we should allocate requested block */
3738                } else {
3739                        /* block is already allocated */
3740                        if (sbi->s_cluster_ratio > 1)
3741                                map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3742                        newblock = map->m_lblk
3743                                   - le32_to_cpu(newex.ee_block)
3744                                   + ext4_ext_pblock(&newex);
3745                        /* number of remaining blocks in the extent */
3746                        allocated = ext4_ext_get_actual_len(&newex) -
3747                                (map->m_lblk - le32_to_cpu(newex.ee_block));
3748                        goto out;
3749                }
3750        }
3751
3752        /* find extent for this block */
3753        path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
3754        if (IS_ERR(path)) {
3755                err = PTR_ERR(path);
3756                path = NULL;
3757                goto out2;
3758        }
3759
3760        depth = ext_depth(inode);
3761
3762        /*
3763         * consistent leaf must not be empty;
3764         * this situation is possible, though, _during_ tree modification;
3765         * this is why assert can't be put in ext4_ext_find_extent()
3766         */
3767        if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
3768                EXT4_ERROR_INODE(inode, "bad extent address "
3769                                 "lblock: %lu, depth: %d pblock %lld",
3770                                 (unsigned long) map->m_lblk, depth,
3771                                 path[depth].p_block);
3772                err = -EIO;
3773                goto out2;
3774        }
3775
3776        ex = path[depth].p_ext;
3777        if (ex) {
3778                ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
3779                ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
3780                unsigned short ee_len;
3781
3782                /*
3783                 * Uninitialized extents are treated as holes, except that
3784                 * we split out initialized portions during a write.
3785                 */
3786                ee_len = ext4_ext_get_actual_len(ex);
3787
3788                trace_ext4_ext_show_extent(inode, ee_block, ee_start, ee_len);
3789
3790                /* if found extent covers block, simply return it */
3791                if (in_range(map->m_lblk, ee_block, ee_len)) {
3792                        struct ext4_map_blocks punch_map;
3793                        ext4_fsblk_t partial_cluster = 0;
3794
3795                        newblock = map->m_lblk - ee_block + ee_start;
3796                        /* number of remaining blocks in the extent */
3797                        allocated = ee_len - (map->m_lblk - ee_block);
3798                        ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
3799                                  ee_block, ee_len, newblock);
3800
3801                        if ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0) {
3802                                /*
3803                                 * Do not put uninitialized extent
3804                                 * in the cache
3805                                 */
3806                                if (!ext4_ext_is_uninitialized(ex)) {
3807                                        ext4_ext_put_in_cache(inode, ee_block,
3808                                                ee_len, ee_start);
3809                                        goto out;
3810                                }
3811                                ret = ext4_ext_handle_uninitialized_extents(
3812                                        handle, inode, map, path, flags,
3813                                        allocated, newblock);
3814                                return ret;
3815                        }
3816
3817                        /*
3818                         * Punch out the map length, but only to the
3819                         * end of the extent
3820                         */
3821                        punched_out = allocated < map->m_len ?
3822                                allocated : map->m_len;
3823
3824                        /*
3825                         * Sense extents need to be converted to
3826                         * uninitialized, they must fit in an
3827                         * uninitialized extent
3828                         */
3829                        if (punched_out > EXT_UNINIT_MAX_LEN)
3830                                punched_out = EXT_UNINIT_MAX_LEN;
3831
3832                        punch_map.m_lblk = map->m_lblk;
3833                        punch_map.m_pblk = newblock;
3834                        punch_map.m_len = punched_out;
3835                        punch_map.m_flags = 0;
3836
3837                        /* Check to see if the extent needs to be split */
3838                        if (punch_map.m_len != ee_len ||
3839                                punch_map.m_lblk != ee_block) {
3840
3841                                ret = ext4_split_extent(handle, inode,
3842                                path, &punch_map, 0,
3843                                EXT4_GET_BLOCKS_PUNCH_OUT_EXT |
3844                                EXT4_GET_BLOCKS_PRE_IO);
3845
3846                                if (ret < 0) {
3847                                        err = ret;
3848                                        goto out2;
3849                                }
3850                                /*
3851                                 * find extent for the block at
3852                                 * the start of the hole
3853                                 */
3854                                ext4_ext_drop_refs(path);
3855                                kfree(path);
3856
3857                                path = ext4_ext_find_extent(inode,
3858                                map->m_lblk, NULL);
3859                                if (IS_ERR(path)) {
3860                                        err = PTR_ERR(path);
3861                                        path = NULL;
3862                                        goto out2;
3863                                }
3864
3865                                depth = ext_depth(inode);
3866                                ex = path[depth].p_ext;
3867                                ee_len = ext4_ext_get_actual_len(ex);
3868                                ee_block = le32_to_cpu(ex->ee_block);
3869                                ee_start = ext4_ext_pblock(ex);
3870
3871                        }
3872
3873                        ext4_ext_mark_uninitialized(ex);
3874
3875                        ext4_ext_invalidate_cache(inode);
3876
3877                        err = ext4_ext_rm_leaf(handle, inode, path,
3878                                               &partial_cluster, map->m_lblk,
3879                                               map->m_lblk + punched_out);
3880
3881                        if (!err && path->p_hdr->eh_entries == 0) {
3882                                /*
3883                                 * Punch hole freed all of this sub tree,
3884                                 * so we need to correct eh_depth
3885                                 */
3886                                err = ext4_ext_get_access(handle, inode, path);
3887                                if (err == 0) {
3888                                        ext_inode_hdr(inode)->eh_depth = 0;
3889                                        ext_inode_hdr(inode)->eh_max =
3890                                        cpu_to_le16(ext4_ext_space_root(
3891                                                inode, 0));
3892
3893                                        err = ext4_ext_dirty(
3894                                                handle, inode, path);
3895                                }
3896                        }
3897
3898                        goto out2;
3899                }
3900        }
3901
3902        if ((sbi->s_cluster_ratio > 1) &&
3903            ext4_find_delalloc_cluster(inode, map->m_lblk, 0))
3904                map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3905
3906        /*
3907         * requested block isn't allocated yet;
3908         * we couldn't try to create block if create flag is zero
3909         */
3910        if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
3911                /*
3912                 * put just found gap into cache to speed up
3913                 * subsequent requests
3914                 */
3915                ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
3916                goto out2;
3917        }
3918
3919        /*
3920         * Okay, we need to do block allocation.
3921         */
3922        map->m_flags &= ~EXT4_MAP_FROM_CLUSTER;
3923        newex.ee_block = cpu_to_le32(map->m_lblk);
3924        cluster_offset = map->m_lblk & (sbi->s_cluster_ratio-1);
3925
3926        /*
3927         * If we are doing bigalloc, check to see if the extent returned
3928         * by ext4_ext_find_extent() implies a cluster we can use.
3929         */
3930        if (cluster_offset && ex &&
3931            get_implied_cluster_alloc(inode->i_sb, map, ex, path)) {
3932                ar.len = allocated = map->m_len;
3933                newblock = map->m_pblk;
3934                map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3935                goto got_allocated_blocks;
3936        }
3937
3938        /* find neighbour allocated blocks */
3939        ar.lleft = map->m_lblk;
3940        err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
3941        if (err)
3942                goto out2;
3943        ar.lright = map->m_lblk;
3944        ex2 = NULL;
3945        err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright, &ex2);
3946        if (err)
3947                goto out2;
3948
3949        /* Check if the extent after searching to the right implies a
3950         * cluster we can use. */
3951        if ((sbi->s_cluster_ratio > 1) && ex2 &&
3952            get_implied_cluster_alloc(inode->i_sb, map, ex2, path)) {
3953                ar.len = allocated = map->m_len;
3954                newblock = map->m_pblk;
3955                map->m_flags |= EXT4_MAP_FROM_CLUSTER;
3956                goto got_allocated_blocks;
3957        }
3958
3959        /*
3960         * See if request is beyond maximum number of blocks we can have in
3961         * a single extent. For an initialized extent this limit is
3962         * EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
3963         * EXT_UNINIT_MAX_LEN.
3964         */
3965        if (map->m_len > EXT_INIT_MAX_LEN &&
3966            !(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
3967                map->m_len = EXT_INIT_MAX_LEN;
3968        else if (map->m_len > EXT_UNINIT_MAX_LEN &&
3969                 (flags & EXT4_GET_BLOCKS_UNINIT_EXT))
3970                map->m_len = EXT_UNINIT_MAX_LEN;
3971
3972        /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */
3973        newex.ee_len = cpu_to_le16(map->m_len);
3974        err = ext4_ext_check_overlap(sbi, inode, &newex, path);
3975        if (err)
3976                allocated = ext4_ext_get_actual_len(&newex);
3977        else
3978                allocated = map->m_len;
3979
3980        /* allocate new block */
3981        ar.inode = inode;
3982        ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk);
3983        ar.logical = map->m_lblk;
3984        /*
3985         * We calculate the offset from the beginning of the cluster
3986         * for the logical block number, since when we allocate a
3987         * physical cluster, the physical block should start at the
3988         * same offset from the beginning of the cluster.  This is
3989         * needed so that future calls to get_implied_cluster_alloc()
3990         * work correctly.
3991         */
3992        offset = map->m_lblk & (sbi->s_cluster_ratio - 1);
3993        ar.len = EXT4_NUM_B2C(sbi, offset+allocated);
3994        ar.goal -= offset;
3995        ar.logical -= offset;
3996        if (S_ISREG(inode->i_mode))
3997                ar.flags = EXT4_MB_HINT_DATA;
3998        else
3999                /* disable in-core preallocation for non-regular files */
4000                ar.flags = 0;
4001        if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE)
4002                ar.flags |= EXT4_MB_HINT_NOPREALLOC;
4003        newblock = ext4_mb_new_blocks(handle, &ar, &err);
4004        if (!newblock)
4005                goto out2;
4006        ext_debug("allocate new block: goal %llu, found %llu/%u\n",
4007                  ar.goal, newblock, allocated);
4008        free_on_err = 1;
4009        allocated_clusters = ar.len;
4010        ar.len = EXT4_C2B(sbi, ar.len) - offset;
4011        if (ar.len > allocated)
4012                ar.len = allocated;
4013
4014got_allocated_blocks:
4015        /* try to insert new extent into found leaf and return */
4016        ext4_ext_store_pblock(&newex, newblock + offset);
4017        newex.ee_len = cpu_to_le16(ar.len);
4018        /* Mark uninitialized */
4019        if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
4020                ext4_ext_mark_uninitialized(&newex);
4021                /*
4022                 * io_end structure was created for every IO write to an
4023                 * uninitialized extent. To avoid unnecessary conversion,
4024                 * here we flag the IO that really needs the conversion.
4025                 * For non asycn direct IO case, flag the inode state
4026                 * that we need to perform conversion when IO is done.
4027                 */
4028                if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
4029                        if (io)
4030                                ext4_set_io_unwritten_flag(inode, io);
4031                        else
4032                                ext4_set_inode_state(inode,
4033                                                     EXT4_STATE_DIO_UNWRITTEN);
4034                }
4035                if (ext4_should_dioread_nolock(inode))
4036                        map->m_flags |= EXT4_MAP_UNINIT;
4037        }
4038
4039        err = 0;
4040        if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0)
4041                err = check_eofblocks_fl(handle, inode, map->m_lblk,
4042                                         path, ar.len);
4043        if (!err)
4044                err = ext4_ext_insert_extent(handle, inode, path,
4045                                             &newex, flags);
4046        if (err && free_on_err) {
4047                int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ?
4048                        EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0;
4049                /* free data blocks we just allocated */
4050                /* not a good idea to call discard here directly,
4051                 * but otherwise we'd need to call it every free() */
4052                ext4_discard_preallocations(inode);
4053                ext4_free_blocks(handle, inode, NULL, ext4_ext_pblock(&newex),
4054                                 ext4_ext_get_actual_len(&newex), fb_flags);
4055                goto out2;
4056        }
4057
4058        /* previous routine could use block we allocated */
4059        newblock = ext4_ext_pblock(&newex);
4060        allocated = ext4_ext_get_actual_len(&newex);
4061        if (allocated > map->m_len)
4062                allocated = map->m_len;
4063        map->m_flags |= EXT4_MAP_NEW;
4064
4065        /*
4066         * Update reserved blocks/metadata blocks after successful
4067         * block allocation which had been deferred till now.
4068         */
4069        if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
4070                unsigned int reserved_clusters;
4071                /*
4072                 * Check how many clusters we had reserved this allocated range
4073                 */
4074                reserved_clusters = get_reserved_cluster_alloc(inode,
4075                                                map->m_lblk, allocated);
4076                if (map->m_flags & EXT4_MAP_FROM_CLUSTER) {
4077                        if (reserved_clusters) {
4078                                /*
4079                                 * We have clusters reserved for this range.
4080                                 * But since we are not doing actual allocation
4081                                 * and are simply using blocks from previously
4082                                 * allocated cluster, we should release the
4083                                 * reservation and not claim quota.
4084                                 */
4085                                ext4_da_update_reserve_space(inode,
4086                                                reserved_clusters, 0);
4087                        }
4088                } else {
4089                        BUG_ON(allocated_clusters < reserved_clusters);
4090                        /* We will claim quota for all newly allocated blocks.*/
4091                        ext4_da_update_reserve_space(inode, allocated_clusters,
4092                                                        1);
4093                        if (reserved_clusters < allocated_clusters) {
4094                                struct ext4_inode_info *ei = EXT4_I(inode);
4095                                int reservation = allocated_clusters -
4096                                                  reserved_clusters;
4097                                /*
4098                                 * It seems we claimed few clusters outside of
4099                                 * the range of this allocation. We should give
4100                                 * it back to the reservation pool. This can
4101                                 * happen in the following case:
4102                                 *
4103                                 * * Suppose s_cluster_ratio is 4 (i.e., each
4104                                 *   cluster has 4 blocks. Thus, the clusters
4105                                 *   are [0-3],[4-7],[8-11]...
4106                                 * * First comes delayed allocation write for
4107                                 *   logical blocks 10 & 11. Since there were no
4108                                 *   previous delayed allocated blocks in the
4109                                 *   range [8-11], we would reserve 1 cluster
4110                                 *   for this write.
4111                                 * * Next comes write for logical blocks 3 to 8.
4112                                 *   In this case, we will reserve 2 clusters
4113                                 *   (for [0-3] and [4-7]; and not for [8-11] as
4114                                 *   that range has a delayed allocated blocks.
4115                                 *   Thus total reserved clusters now becomes 3.
4116                                 * * Now, during the delayed allocation writeout
4117                                 *   time, we will first write blocks [3-8] and
4118                                 *   allocate 3 clusters for writing these
4119                                 *   blocks. Also, we would claim all these
4120                                 *   three clusters above.
4121                                 * * Now when we come here to writeout the
4122                                 *   blocks [10-11], we would expect to claim
4123                                 *   the reservation of 1 cluster we had made
4124                                 *   (and we would claim it since there are no
4125                                 *   more delayed allocated blocks in the range
4126                                 *   [8-11]. But our reserved cluster count had
4127                                 *   already gone to 0.
4128                                 *
4129                                 *   Thus, at the step 4 above when we determine
4130                                 *   that there are still some unwritten delayed
4131                                 *   allocated blocks outside of our current
4132                                 *   block range, we should increment the
4133                                 *   reserved clusters count so that when the
4134                                 *   remaining blocks finally gets written, we
4135                                 *   could claim them.
4136                                 */
4137                                dquot_reserve_block(inode,
4138                                                EXT4_C2B(sbi, reservation));
4139                                spin_lock(&ei->i_block_reservation_lock);
4140                                ei->i_reserved_data_blocks += reservation;
4141                                spin_unlock(&ei->i_block_reservation_lock);
4142                        }
4143                }
4144        }
4145
4146        /*
4147         * Cache the extent and update transaction to commit on fdatasync only
4148         * when it is _not_ an uninitialized extent.
4149         */
4150        if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) {
4151                ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock);
4152                ext4_update_inode_fsync_trans(handle, inode, 1);
4153        } else
4154                ext4_update_inode_fsync_trans(handle, inode, 0);
4155out:
4156        if (allocated > map->m_len)
4157                allocated = map->m_len;
4158        ext4_ext_show_leaf(inode, path);
4159        map->m_flags |= EXT4_MAP_MAPPED;
4160        map->m_pblk = newblock;
4161        map->m_len = allocated;
4162out2:
4163        if (path) {
4164                ext4_ext_drop_refs(path);
4165                kfree(path);
4166        }
4167        result = (flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) ?
4168                        punched_out : allocated;
4169
4170        trace_ext4_ext_map_blocks_exit(inode, map->m_lblk,
4171                newblock, map->m_len, err ? err : result);
4172
4173        return err ? err : result;
4174}
4175
4176void ext4_ext_truncate(struct inode *inode)
4177{
4178        struct address_space *mapping = inode->i_mapping;
4179        struct super_block *sb = inode->i_sb;
4180        ext4_lblk_t last_block;
4181        handle_t *handle;
4182        loff_t page_len;
4183        int err = 0;
4184
4185        /*
4186         * finish any pending end_io work so we won't run the risk of
4187         * converting any truncated blocks to initialized later
4188         */
4189        ext4_flush_completed_IO(inode);
4190
4191        /*
4192         * probably first extent we're gonna free will be last in block
4193         */
4194        err = ext4_writepage_trans_blocks(inode);
4195        handle = ext4_journal_start(inode, err);
4196        if (IS_ERR(handle))
4197                return;
4198
4199        if (inode->i_size % PAGE_CACHE_SIZE != 0) {
4200                page_len = PAGE_CACHE_SIZE -
4201                        (inode->i_size & (PAGE_CACHE_SIZE - 1));
4202
4203                err = ext4_discard_partial_page_buffers(handle,
4204                        mapping, inode->i_size, page_len, 0);
4205
4206                if (err)
4207                        goto out_stop;
4208        }
4209
4210        if (ext4_orphan_add(handle, inode))
4211                goto out_stop;
4212
4213        down_write(&EXT4_I(inode)->i_data_sem);
4214        ext4_ext_invalidate_cache(inode);
4215
4216        ext4_discard_preallocations(inode);
4217
4218        /*
4219         * TODO: optimization is possible here.
4220         * Probably we need not scan at all,
4221         * because page truncation is enough.
4222         */
4223
4224        /* we have to know where to truncate from in crash case */
4225        EXT4_I(inode)->i_disksize = inode->i_size;
4226        ext4_mark_inode_dirty(handle, inode);
4227
4228        last_block = (inode->i_size + sb->s_blocksize - 1)
4229                        >> EXT4_BLOCK_SIZE_BITS(sb);
4230        err = ext4_ext_remove_space(inode, last_block);
4231
4232        /* In a multi-transaction truncate, we only make the final
4233         * transaction synchronous.
4234         */
4235        if (IS_SYNC(inode))
4236                ext4_handle_sync(handle);
4237
4238        up_write(&EXT4_I(inode)->i_data_sem);
4239
4240out_stop:
4241        /*
4242         * If this was a simple ftruncate() and the file will remain alive,
4243         * then we need to clear up the orphan record which we created above.
4244         * However, if this was a real unlink then we were called by
4245         * ext4_delete_inode(), and we allow that function to clean up the
4246         * orphan info for us.
4247         */
4248        if (inode->i_nlink)
4249                ext4_orphan_del(handle, inode);
4250
4251        inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4252        ext4_mark_inode_dirty(handle, inode);
4253        ext4_journal_stop(handle);
4254}
4255
4256static void ext4_falloc_update_inode(struct inode *inode,
4257                                int mode, loff_t new_size, int update_ctime)
4258{
4259        struct timespec now;
4260
4261        if (update_ctime) {
4262                now = current_fs_time(inode->i_sb);
4263                if (!timespec_equal(&inode->i_ctime, &now))
4264                        inode->i_ctime = now;
4265        }
4266        /*
4267         * Update only when preallocation was requested beyond
4268         * the file size.
4269         */
4270        if (!(mode & FALLOC_FL_KEEP_SIZE)) {
4271                if (new_size > i_size_read(inode))
4272                        i_size_write(inode, new_size);
4273                if (new_size > EXT4_I(inode)->i_disksize)
4274                        ext4_update_i_disksize(inode, new_size);
4275        } else {
4276                /*
4277                 * Mark that we allocate beyond EOF so the subsequent truncate
4278                 * can proceed even if the new size is the same as i_size.
4279                 */
4280                if (new_size > i_size_read(inode))
4281                        ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
4282        }
4283
4284}
4285
4286/*
4287 * preallocate space for a file. This implements ext4's fallocate file
4288 * operation, which gets called from sys_fallocate system call.
4289 * For block-mapped files, posix_fallocate should fall back to the method
4290 * of writing zeroes to the required new blocks (the same behavior which is
4291 * expected for file systems which do not support fallocate() system call).
4292 */
4293long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
4294{
4295        struct inode *inode = file->f_path.dentry->d_inode;
4296        handle_t *handle;
4297        loff_t new_size;
4298        unsigned int max_blocks;
4299        int ret = 0;
4300        int ret2 = 0;
4301        int retries = 0;
4302        int flags;
4303        struct ext4_map_blocks map;
4304        unsigned int credits, blkbits = inode->i_blkbits;
4305
4306        /*
4307         * currently supporting (pre)allocate mode for extent-based
4308         * files _only_
4309         */
4310        if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
4311                return -EOPNOTSUPP;
4312
4313        /* Return error if mode is not supported */
4314        if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
4315                return -EOPNOTSUPP;
4316
4317        if (mode & FALLOC_FL_PUNCH_HOLE)
4318                return ext4_punch_hole(file, offset, len);
4319
4320        trace_ext4_fallocate_enter(inode, offset, len, mode);
4321        map.m_lblk = offset >> blkbits;
4322        /*
4323         * We can't just convert len to max_blocks because
4324         * If blocksize = 4096 offset = 3072 and len = 2048
4325         */
4326        max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
4327                - map.m_lblk;
4328        /*
4329         * credits to insert 1 extent into extent tree
4330         */
4331        credits = ext4_chunk_trans_blocks(inode, max_blocks);
4332        mutex_lock(&inode->i_mutex);
4333        ret = inode_newsize_ok(inode, (len + offset));
4334        if (ret) {
4335                mutex_unlock(&inode->i_mutex);
4336                trace_ext4_fallocate_exit(inode, offset, max_blocks, ret);
4337                return ret;
4338        }
4339        flags = EXT4_GET_BLOCKS_CREATE_UNINIT_EXT;
4340        if (mode & FALLOC_FL_KEEP_SIZE)
4341                flags |= EXT4_GET_BLOCKS_KEEP_SIZE;
4342        /*
4343         * Don't normalize the request if it can fit in one extent so
4344         * that it doesn't get unnecessarily split into multiple
4345         * extents.
4346         */
4347        if (len <= EXT_UNINIT_MAX_LEN << blkbits)
4348                flags |= EXT4_GET_BLOCKS_NO_NORMALIZE;
4349retry:
4350        while (ret >= 0 && ret < max_blocks) {
4351                map.m_lblk = map.m_lblk + ret;
4352                map.m_len = max_blocks = max_blocks - ret;
4353                handle = ext4_journal_start(inode, credits);
4354                if (IS_ERR(handle)) {
4355                        ret = PTR_ERR(handle);
4356                        break;
4357                }
4358                ret = ext4_map_blocks(handle, inode, &map, flags);
4359                if (ret <= 0) {
4360#ifdef EXT4FS_DEBUG
4361                        WARN_ON(ret <= 0);
4362                        printk(KERN_ERR "%s: ext4_ext_map_blocks "
4363                                    "returned error inode#%lu, block=%u, "
4364                                    "max_blocks=%u", __func__,
4365                                    inode->i_ino, map.m_lblk, max_blocks);
4366#endif
4367                        ext4_mark_inode_dirty(handle, inode);
4368                        ret2 = ext4_journal_stop(handle);
4369                        break;
4370                }
4371                if ((map.m_lblk + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
4372                                                blkbits) >> blkbits))
4373                        new_size = offset + len;
4374                else
4375                        new_size = ((loff_t) map.m_lblk + ret) << blkbits;
4376
4377                ext4_falloc_update_inode(inode, mode, new_size,
4378                                         (map.m_flags & EXT4_MAP_NEW));
4379                ext4_mark_inode_dirty(handle, inode);
4380                ret2 = ext4_journal_stop(handle);
4381                if (ret2)
4382                        break;
4383        }
4384        if (ret == -ENOSPC &&
4385                        ext4_should_retry_alloc(inode->i_sb, &retries)) {
4386                ret = 0;
4387                goto retry;
4388        }
4389        mutex_unlock(&inode->i_mutex);
4390        trace_ext4_fallocate_exit(inode, offset, max_blocks,
4391                                ret > 0 ? ret2 : ret);
4392        return ret > 0 ? ret2 : ret;
4393}
4394
4395/*
4396 * This function convert a range of blocks to written extents
4397 * The caller of this function will pass the start offset and the size.
4398 * all unwritten extents within this range will be converted to
4399 * written extents.
4400 *
4401 * This function is called from the direct IO end io call back
4402 * function, to convert the fallocated extents after IO is completed.
4403 * Returns 0 on success.
4404 */
4405int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
4406                                    ssize_t len)
4407{
4408        handle_t *handle;
4409        unsigned int max_blocks;
4410        int ret = 0;
4411        int ret2 = 0;
4412        struct ext4_map_blocks map;
4413        unsigned int credits, blkbits = inode->i_blkbits;
4414
4415        map.m_lblk = offset >> blkbits;
4416        /*
4417         * We can't just convert len to max_blocks because
4418         * If blocksize = 4096 offset = 3072 and len = 2048
4419         */
4420        max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) -
4421                      map.m_lblk);
4422        /*
4423         * credits to insert 1 extent into extent tree
4424         */
4425        credits = ext4_chunk_trans_blocks(inode, max_blocks);
4426        while (ret >= 0 && ret < max_blocks) {
4427                map.m_lblk += ret;
4428                map.m_len = (max_blocks -= ret);
4429                handle = ext4_journal_start(inode, credits);
4430                if (IS_ERR(handle)) {
4431                        ret = PTR_ERR(handle);
4432                        break;
4433                }
4434                ret = ext4_map_blocks(handle, inode, &map,
4435                                      EXT4_GET_BLOCKS_IO_CONVERT_EXT);
4436                if (ret <= 0) {
4437                        WARN_ON(ret <= 0);
4438                        printk(KERN_ERR "%s: ext4_ext_map_blocks "
4439                                    "returned error inode#%lu, block=%u, "
4440                                    "max_blocks=%u", __func__,
4441                                    inode->i_ino, map.m_lblk, map.m_len);
4442                }
4443                ext4_mark_inode_dirty(handle, inode);
4444                ret2 = ext4_journal_stop(handle);
4445                if (ret <= 0 || ret2 )
4446                        break;
4447        }
4448        return ret > 0 ? ret2 : ret;
4449}
4450
4451/*
4452 * Callback function called for each extent to gather FIEMAP information.
4453 */
4454static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next,
4455                       struct ext4_ext_cache *newex, struct ext4_extent *ex,
4456                       void *data)
4457{
4458        __u64   logical;
4459        __u64   physical;
4460        __u64   length;
4461        __u32   flags = 0;
4462        int             ret = 0;
4463        struct fiemap_extent_info *fieinfo = data;
4464        unsigned char blksize_bits;
4465
4466        blksize_bits = inode->i_sb->s_blocksize_bits;
4467        logical = (__u64)newex->ec_block << blksize_bits;
4468
4469        if (newex->ec_start == 0) {
4470                /*
4471                 * No extent in extent-tree contains block @newex->ec_start,
4472                 * then the block may stay in 1)a hole or 2)delayed-extent.
4473                 *
4474                 * Holes or delayed-extents are processed as follows.
4475                 * 1. lookup dirty pages with specified range in pagecache.
4476                 *    If no page is got, then there is no delayed-extent and
4477                 *    return with EXT_CONTINUE.
4478                 * 2. find the 1st mapped buffer,
4479                 * 3. check if the mapped buffer is both in the request range
4480                 *    and a delayed buffer. If not, there is no delayed-extent,
4481                 *    then return.
4482                 * 4. a delayed-extent is found, the extent will be collected.
4483                 */
4484                ext4_lblk_t     end = 0;
4485                pgoff_t         last_offset;
4486                pgoff_t         offset;
4487                pgoff_t         index;
4488                pgoff_t         start_index = 0;
4489                struct page     **pages = NULL;
4490                struct buffer_head *bh = NULL;
4491                struct buffer_head *head = NULL;
4492                unsigned int nr_pages = PAGE_SIZE / sizeof(struct page *);
4493
4494                pages = kmalloc(PAGE_SIZE, GFP_KERNEL);
4495                if (pages == NULL)
4496                        return -ENOMEM;
4497
4498                offset = logical >> PAGE_SHIFT;
4499repeat:
4500                last_offset = offset;
4501                head = NULL;
4502                ret = find_get_pages_tag(inode->i_mapping, &offset,
4503                                        PAGECACHE_TAG_DIRTY, nr_pages, pages);
4504
4505                if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4506                        /* First time, try to find a mapped buffer. */
4507                        if (ret == 0) {
4508out:
4509                                for (index = 0; index < ret; index++)
4510                                        page_cache_release(pages[index]);
4511                                /* just a hole. */
4512                                kfree(pages);
4513                                return EXT_CONTINUE;
4514                        }
4515                        index = 0;
4516
4517next_page:
4518                        /* Try to find the 1st mapped buffer. */
4519                        end = ((__u64)pages[index]->index << PAGE_SHIFT) >>
4520                                  blksize_bits;
4521                        if (!page_has_buffers(pages[index]))
4522                                goto out;
4523                        head = page_buffers(pages[index]);
4524                        if (!head)
4525                                goto out;
4526
4527                        index++;
4528                        bh = head;
4529                        do {
4530                                if (end >= newex->ec_block +
4531                                        newex->ec_len)
4532                                        /* The buffer is out of
4533                                         * the request range.
4534                                         */
4535                                        goto out;
4536
4537                                if (buffer_mapped(bh) &&
4538                                    end >= newex->ec_block) {
4539                                        start_index = index - 1;
4540                                        /* get the 1st mapped buffer. */
4541                                        goto found_mapped_buffer;
4542                                }
4543
4544                                bh = bh->b_this_page;
4545                                end++;
4546                        } while (bh != head);
4547
4548                        /* No mapped buffer in the range found in this page,
4549                         * We need to look up next page.
4550                         */
4551                        if (index >= ret) {
4552                                /* There is no page left, but we need to limit
4553                                 * newex->ec_len.
4554                                 */
4555                                newex->ec_len = end - newex->ec_block;
4556                                goto out;
4557                        }
4558                        goto next_page;
4559                } else {
4560                        /*Find contiguous delayed buffers. */
4561                        if (ret > 0 && pages[0]->index == last_offset)
4562                                head = page_buffers(pages[0]);
4563                        bh = head;
4564                        index = 1;
4565                        start_index = 0;
4566                }
4567
4568found_mapped_buffer:
4569                if (bh != NULL && buffer_delay(bh)) {
4570                        /* 1st or contiguous delayed buffer found. */
4571                        if (!(flags & FIEMAP_EXTENT_DELALLOC)) {
4572                                /*
4573                                 * 1st delayed buffer found, record
4574                                 * the start of extent.
4575                                 */
4576                                flags |= FIEMAP_EXTENT_DELALLOC;
4577                                newex->ec_block = end;
4578                                logical = (__u64)end << blksize_bits;
4579                        }
4580                        /* Find contiguous delayed buffers. */
4581                        do {
4582                                if (!buffer_delay(bh))
4583                                        goto found_delayed_extent;
4584                                bh = bh->b_this_page;
4585                                end++;
4586                        } while (bh != head);
4587
4588                        for (; index < ret; index++) {
4589                                if (!page_has_buffers(pages[index])) {
4590                                        bh = NULL;
4591                                        break;
4592                                }
4593                                head = page_buffers(pages[index]);
4594                                if (!head) {
4595                                        bh = NULL;
4596                                        break;
4597                                }
4598
4599                                if (pages[index]->index !=
4600                                    pages[start_index]->index + index
4601                                    - start_index) {
4602                                        /* Blocks are not contiguous. */
4603                                        bh = NULL;
4604                                        break;
4605                                }
4606                                bh = head;
4607                                do {
4608                                        if (!buffer_delay(bh))
4609                                                /* Delayed-extent ends. */
4610                                                goto found_delayed_extent;
4611                                        bh = bh->b_this_page;
4612                                        end++;
4613                                } while (bh != head);
4614                        }
4615                } else if (!(flags & FIEMAP_EXTENT_DELALLOC))
4616                        /* a hole found. */
4617                        goto out;
4618
4619found_delayed_extent:
4620                newex->ec_len = min(end - newex->ec_block,
4621                                                (ext4_lblk_t)EXT_INIT_MAX_LEN);
4622                if (ret == nr_pages && bh != NULL &&
4623                        newex->ec_len < EXT_INIT_MAX_LEN &&
4624                        buffer_delay(bh)) {
4625                        /* Have not collected an extent and continue. */
4626                        for (index = 0; index < ret; index++)
4627                                page_cache_release(pages[index]);
4628                        goto repeat;
4629                }
4630
4631                for (index = 0; index < ret; index++)
4632                        page_cache_release(pages[index]);
4633                kfree(pages);
4634        }
4635
4636        physical = (__u64)newex->ec_start << blksize_bits;
4637        length =   (__u64)newex->ec_len << blksize_bits;
4638
4639        if (ex && ext4_ext_is_uninitialized(ex))
4640                flags |= FIEMAP_EXTENT_UNWRITTEN;
4641
4642        if (next == EXT_MAX_BLOCKS)
4643                flags |= FIEMAP_EXTENT_LAST;
4644
4645        ret = fiemap_fill_next_extent(fieinfo, logical, physical,
4646                                        length, flags);
4647        if (ret < 0)
4648                return ret;
4649        if (ret == 1)
4650                return EXT_BREAK;
4651        return EXT_CONTINUE;
4652}
4653/* fiemap flags we can handle specified here */
4654#define EXT4_FIEMAP_FLAGS       (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
4655
4656static int ext4_xattr_fiemap(struct inode *inode,
4657                                struct fiemap_extent_info *fieinfo)
4658{
4659        __u64 physical = 0;
4660        __u64 length;
4661        __u32 flags = FIEMAP_EXTENT_LAST;
4662        int blockbits = inode->i_sb->s_blocksize_bits;
4663        int error = 0;
4664
4665        /* in-inode? */
4666        if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
4667                struct ext4_iloc iloc;
4668                int offset;     /* offset of xattr in inode */
4669
4670                error = ext4_get_inode_loc(inode, &iloc);
4671                if (error)
4672                        return error;
4673                physical = iloc.bh->b_blocknr << blockbits;
4674                offset = EXT4_GOOD_OLD_INODE_SIZE +
4675                                EXT4_I(inode)->i_extra_isize;
4676                physical += offset;
4677                length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
4678                flags |= FIEMAP_EXTENT_DATA_INLINE;
4679                brelse(iloc.bh);
4680        } else { /* external block */
4681                physical = EXT4_I(inode)->i_file_acl << blockbits;
4682                length = inode->i_sb->s_blocksize;
4683        }
4684
4685        if (physical)
4686                error = fiemap_fill_next_extent(fieinfo, 0, physical,
4687                                                length, flags);
4688        return (error < 0 ? error : 0);
4689}
4690
4691/*
4692 * ext4_ext_punch_hole
4693 *
4694 * Punches a hole of "length" bytes in a file starting
4695 * at byte "offset"
4696 *
4697 * @inode:  The inode of the file to punch a hole in
4698 * @offset: The starting byte offset of the hole
4699 * @length: The length of the hole
4700 *
4701 * Returns the number of blocks removed or negative on err
4702 */
4703int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length)
4704{
4705        struct inode *inode = file->f_path.dentry->d_inode;
4706        struct super_block *sb = inode->i_sb;
4707        struct ext4_ext_cache cache_ex;
4708        ext4_lblk_t first_block, last_block, num_blocks, iblock, max_blocks;
4709        struct address_space *mapping = inode->i_mapping;
4710        struct ext4_map_blocks map;
4711        handle_t *handle;
4712        loff_t first_page, last_page, page_len;
4713        loff_t first_page_offset, last_page_offset;
4714        int ret, credits, blocks_released, err = 0;
4715
4716        /* No need to punch hole beyond i_size */
4717        if (offset >= inode->i_size)
4718                return 0;
4719
4720        /*
4721         * If the hole extends beyond i_size, set the hole
4722         * to end after the page that contains i_size
4723         */
4724        if (offset + length > inode->i_size) {
4725                length = inode->i_size +
4726                   PAGE_CACHE_SIZE - (inode->i_size & (PAGE_CACHE_SIZE - 1)) -
4727                   offset;
4728        }
4729
4730        first_block = (offset + sb->s_blocksize - 1) >>
4731                EXT4_BLOCK_SIZE_BITS(sb);
4732        last_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb);
4733
4734        first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
4735        last_page = (offset + length) >> PAGE_CACHE_SHIFT;
4736
4737        first_page_offset = first_page << PAGE_CACHE_SHIFT;
4738        last_page_offset = last_page << PAGE_CACHE_SHIFT;
4739
4740        /*
4741         * Write out all dirty pages to avoid race conditions
4742         * Then release them.
4743         */
4744        if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
4745                err = filemap_write_and_wait_range(mapping,
4746                        offset, offset + length - 1);
4747
4748                if (err)
4749                        return err;
4750        }
4751
4752        /* Now release the pages */
4753        if (last_page_offset > first_page_offset) {
4754                truncate_inode_pages_range(mapping, first_page_offset,
4755                                           last_page_offset-1);
4756        }
4757
4758        /* finish any pending end_io work */
4759        ext4_flush_completed_IO(inode);
4760
4761        credits = ext4_writepage_trans_blocks(inode);
4762        handle = ext4_journal_start(inode, credits);
4763        if (IS_ERR(handle))
4764                return PTR_ERR(handle);
4765
4766        err = ext4_orphan_add(handle, inode);
4767        if (err)
4768                goto out;
4769
4770        /*
4771         * Now we need to zero out the non-page-aligned data in the
4772         * pages at the start and tail of the hole, and unmap the buffer
4773         * heads for the block aligned regions of the page that were
4774         * completely zeroed.
4775         */
4776        if (first_page > last_page) {
4777                /*
4778                 * If the file space being truncated is contained within a page
4779                 * just zero out and unmap the middle of that page
4780                 */
4781                err = ext4_discard_partial_page_buffers(handle,
4782                        mapping, offset, length, 0);
4783
4784                if (err)
4785                        goto out;
4786        } else {
4787                /*
4788                 * zero out and unmap the partial page that contains
4789                 * the start of the hole
4790                 */
4791                page_len  = first_page_offset - offset;
4792                if (page_len > 0) {
4793                        err = ext4_discard_partial_page_buffers(handle, mapping,
4794                                                   offset, page_len, 0);
4795                        if (err)
4796                                goto out;
4797                }
4798
4799                /*
4800                 * zero out and unmap the partial page that contains
4801                 * the end of the hole
4802                 */
4803                page_len = offset + length - last_page_offset;
4804                if (page_len > 0) {
4805                        err = ext4_discard_partial_page_buffers(handle, mapping,
4806                                        last_page_offset, page_len, 0);
4807                        if (err)
4808                                goto out;
4809                }
4810        }
4811
4812
4813        /*
4814         * If i_size is contained in the last page, we need to
4815         * unmap and zero the partial page after i_size
4816         */
4817        if (inode->i_size >> PAGE_CACHE_SHIFT == last_page &&
4818           inode->i_size % PAGE_CACHE_SIZE != 0) {
4819
4820                page_len = PAGE_CACHE_SIZE -
4821                        (inode->i_size & (PAGE_CACHE_SIZE - 1));
4822
4823                if (page_len > 0) {
4824                        err = ext4_discard_partial_page_buffers(handle,
4825                          mapping, inode->i_size, page_len, 0);
4826
4827                        if (err)
4828                                goto out;
4829                }
4830        }
4831
4832        /* If there are no blocks to remove, return now */
4833        if (first_block >= last_block)
4834                goto out;
4835
4836        down_write(&EXT4_I(inode)->i_data_sem);
4837        ext4_ext_invalidate_cache(inode);
4838        ext4_discard_preallocations(inode);
4839
4840        /*
4841         * Loop over all the blocks and identify blocks
4842         * that need to be punched out
4843         */
4844        iblock = first_block;
4845        blocks_released = 0;
4846        while (iblock < last_block) {
4847                max_blocks = last_block - iblock;
4848                num_blocks = 1;
4849                memset(&map, 0, sizeof(map));
4850                map.m_lblk = iblock;
4851                map.m_len = max_blocks;
4852                ret = ext4_ext_map_blocks(handle, inode, &map,
4853                        EXT4_GET_BLOCKS_PUNCH_OUT_EXT);
4854
4855                if (ret > 0) {
4856                        blocks_released += ret;
4857                        num_blocks = ret;
4858                } else if (ret == 0) {
4859                        /*
4860                         * If map blocks could not find the block,
4861                         * then it is in a hole.  If the hole was
4862                         * not already cached, then map blocks should
4863                         * put it in the cache.  So we can get the hole
4864                         * out of the cache
4865                         */
4866                        memset(&cache_ex, 0, sizeof(cache_ex));
4867                        if ((ext4_ext_check_cache(inode, iblock, &cache_ex)) &&
4868                                !cache_ex.ec_start) {
4869
4870                                /* The hole is cached */
4871                                num_blocks = cache_ex.ec_block +
4872                                cache_ex.ec_len - iblock;
4873
4874                        } else {
4875                                /* The block could not be identified */
4876                                err = -EIO;
4877                                break;
4878                        }
4879                } else {
4880                        /* Map blocks error */
4881                        err = ret;
4882                        break;
4883                }
4884
4885                if (num_blocks == 0) {
4886                        /* This condition should never happen */
4887                        ext_debug("Block lookup failed");
4888                        err = -EIO;
4889                        break;
4890                }
4891
4892                iblock += num_blocks;
4893        }
4894
4895        if (blocks_released > 0) {
4896                ext4_ext_invalidate_cache(inode);
4897                ext4_discard_preallocations(inode);
4898        }
4899
4900        if (IS_SYNC(inode))
4901                ext4_handle_sync(handle);
4902
4903        up_write(&EXT4_I(inode)->i_data_sem);
4904
4905out:
4906        ext4_orphan_del(handle, inode);
4907        inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
4908        ext4_mark_inode_dirty(handle, inode);
4909        ext4_journal_stop(handle);
4910        return err;
4911}
4912int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
4913                __u64 start, __u64 len)
4914{
4915        ext4_lblk_t start_blk;
4916        int error = 0;
4917
4918        /* fallback to generic here if not in extents fmt */
4919        if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
4920                return generic_block_fiemap(inode, fieinfo, start, len,
4921                        ext4_get_block);
4922
4923        if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
4924                return -EBADR;
4925
4926        if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
4927                error = ext4_xattr_fiemap(inode, fieinfo);
4928        } else {
4929                ext4_lblk_t len_blks;
4930                __u64 last_blk;
4931
4932                start_blk = start >> inode->i_sb->s_blocksize_bits;
4933                last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
4934                if (last_blk >= EXT_MAX_BLOCKS)
4935                        last_blk = EXT_MAX_BLOCKS-1;
4936                len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
4937
4938                /*
4939                 * Walk the extent tree gathering extent information.
4940                 * ext4_ext_fiemap_cb will push extents back to user.
4941                 */
4942                error = ext4_ext_walk_space(inode, start_blk, len_blks,
4943                                          ext4_ext_fiemap_cb, fieinfo);
4944        }
4945
4946        return error;
4947}
4948
lxr.linux.no kindly hosted by Redpill Linpro AS, provider of Linux consulting and operations services since 1995.