[pandora-kernel.git] / fs / ext4 / extents.c

/*
 * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
 * Written by Alex Tomas <alex@clusterfs.com>
 *
 * Architecture independence:
 *   Copyright (c) 2005, Bull S.A.
 *   Written by Pierre Peiffer <pierre.peiffer@bull.net>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public Licens
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-
 */

/*
 * Extents support for EXT4
 *
 * TODO:
 *   - ext4*_error() should be used in some situations
 *   - analyze all BUG()/BUG_ON(), use -EIO where appropriate
 *   - smart tree reduction
 */

#include <linux/module.h>
#include <linux/fs.h>
#include <linux/time.h>
#include <linux/jbd2.h>
#include <linux/highuid.h>
#include <linux/pagemap.h>
#include <linux/quotaops.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/falloc.h>
#include <asm/uaccess.h>
#include <linux/fiemap.h>
#include "ext4_jbd2.h"

#include <trace/events/ext4.h>

/*
 * used by extent splitting.
 */
#define EXT4_EXT_MAY_ZEROOUT    0x1  /* safe to zeroout if split fails \
                                        due to ENOSPC */
#define EXT4_EXT_MARK_UNINIT1   0x2  /* mark first half uninitialized */
#define EXT4_EXT_MARK_UNINIT2   0x4  /* mark second half uninitialized */

#define EXT4_EXT_DATA_VALID1    0x8  /* first half contains valid data */
#define EXT4_EXT_DATA_VALID2    0x10 /* second half contains valid data */

static int ext4_split_extent(handle_t *handle,
                                struct inode *inode,
                                struct ext4_ext_path *path,
                                struct ext4_map_blocks *map,
                                int split_flag,
                                int flags);

static int ext4_split_extent_at(handle_t *handle,
                             struct inode *inode,
                             struct ext4_ext_path *path,
                             ext4_lblk_t split,
                             int split_flag,
                             int flags);

static int ext4_ext_truncate_extend_restart(handle_t *handle,
                                            struct inode *inode,
                                            int needed)
{
        int err;

        if (!ext4_handle_valid(handle))
                return 0;
        if (handle->h_buffer_credits > needed)
                return 0;
        err = ext4_journal_extend(handle, needed);
        if (err <= 0)
                return err;
        err = ext4_truncate_restart_trans(handle, inode, needed);
        if (err == 0)
                err = -EAGAIN;

        return err;
}

/*
 * could return:
 *  - EROFS
 *  - ENOMEM
 */
static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
                                struct ext4_ext_path *path)
{
        if (path->p_bh) {
                /* path points to block */
                return ext4_journal_get_write_access(handle, path->p_bh);
        }
        /* path points to leaf/index in inode body */
        /* we use in-core data, no need to protect them */
        return 0;
}

/*
 * could return:
 *  - EROFS
 *  - ENOMEM
 *  - EIO
 */
#define ext4_ext_dirty(handle, inode, path) \
                __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path))
static int __ext4_ext_dirty(const char *where, unsigned int line,
                            handle_t *handle, struct inode *inode,
                            struct ext4_ext_path *path)
{
        int err;
        if (path->p_bh) {
                /* path points to block */
                err = __ext4_handle_dirty_metadata(where, line, handle,
                                                   inode, path->p_bh);
        } else {
                /* path points to leaf/index in inode body */
                err = ext4_mark_inode_dirty(handle, inode);
        }
        return err;
}

static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
                              struct ext4_ext_path *path,
                              ext4_lblk_t block)
{
        if (path) {
                int depth = path->p_depth;
                struct ext4_extent *ex;

                /*
                 * Try to predict block placement assuming that we are
                 * filling in a file which will eventually be
                 * non-sparse --- i.e., in the case of libbfd writing
                 * an ELF object sections out-of-order but in a way
                 * the eventually results in a contiguous object or
                 * executable file, or some database extending a table
                 * space file.  However, this is actually somewhat
                 * non-ideal if we are writing a sparse file such as
                 * qemu or KVM writing a raw image file that is going
                 * to stay fairly sparse, since it will end up
                 * fragmenting the file system's free space.  Maybe we
                 * should have some hueristics or some way to allow
                 * userspace to pass a hint to file system,
                 * especially if the latter case turns out to be
                 * common.
                 */
                ex = path[depth].p_ext;
                if (ex) {
                        ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex);
                        ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block);

                        if (block > ext_block)
                                return ext_pblk + (block - ext_block);
                        else
                                return ext_pblk - (ext_block - block);
                }

                /* it looks like index is empty;
                 * try to find starting block from index itself */
                if (path[depth].p_bh)
                        return path[depth].p_bh->b_blocknr;
        }

        /* OK. use inode's group */
        return ext4_inode_to_goal_block(inode);
}

/*
 * Allocation for a meta data block
 */
static ext4_fsblk_t
ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
                        struct ext4_ext_path *path,
                        struct ext4_extent *ex, int *err, unsigned int flags)
{
        ext4_fsblk_t goal, newblock;

        goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
        newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
                                        NULL, err);
        return newblock;
}

static inline int ext4_ext_space_block(struct inode *inode, int check)
{
        int size;

        size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
                        / sizeof(struct ext4_extent);
#ifdef AGGRESSIVE_TEST
        if (!check && size > 6)
                size = 6;
#endif
        return size;
}

static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
{
        int size;

        size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
                        / sizeof(struct ext4_extent_idx);
#ifdef AGGRESSIVE_TEST
        if (!check && size > 5)
                size = 5;
#endif
        return size;
}

static inline int ext4_ext_space_root(struct inode *inode, int check)
{
        int size;

        size = sizeof(EXT4_I(inode)->i_data);
        size -= sizeof(struct ext4_extent_header);
        size /= sizeof(struct ext4_extent);
#ifdef AGGRESSIVE_TEST
        if (!check && size > 3)
                size = 3;
#endif
        return size;
}

static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
{
        int size;

        size = sizeof(EXT4_I(inode)->i_data);
        size -= sizeof(struct ext4_extent_header);
        size /= sizeof(struct ext4_extent_idx);
#ifdef AGGRESSIVE_TEST
        if (!check && size > 4)
                size = 4;
#endif
        return size;
}

/*
 * Calculate the number of metadata blocks needed
 * to allocate @blocks
 * Worse case is one block per extent
 */
int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
{
        struct ext4_inode_info *ei = EXT4_I(inode);
        int idxs;

        idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
                / sizeof(struct ext4_extent_idx));

        /*
         * If the new delayed allocation block is contiguous with the
         * previous da block, it can share index blocks with the
         * previous block, so we only need to allocate a new index
         * block every idxs leaf blocks.  At ldxs**2 blocks, we need
         * an additional index block, and at ldxs**3 blocks, yet
         * another index blocks.
         */
        if (ei->i_da_metadata_calc_len &&
            ei->i_da_metadata_calc_last_lblock+1 == lblock) {
                int num = 0;

                if ((ei->i_da_metadata_calc_len % idxs) == 0)
                        num++;
                if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
                        num++;
                if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
                        num++;
                        ei->i_da_metadata_calc_len = 0;
                } else
                        ei->i_da_metadata_calc_len++;
                ei->i_da_metadata_calc_last_lblock++;
                return num;
        }

        /*
         * In the worst case we need a new set of index blocks at
         * every level of the inode's extent tree.
         */
        ei->i_da_metadata_calc_len = 1;
        ei->i_da_metadata_calc_last_lblock = lblock;
        return ext_depth(inode) + 1;
}

static int
ext4_ext_max_entries(struct inode *inode, int depth)
{
        int max;

        if (depth == ext_depth(inode)) {
                if (depth == 0)
                        max = ext4_ext_space_root(inode, 1);
                else
                        max = ext4_ext_space_root_idx(inode, 1);
        } else {
                if (depth == 0)
                        max = ext4_ext_space_block(inode, 1);
                else
                        max = ext4_ext_space_block_idx(inode, 1);
        }

        return max;
}

static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
{
        ext4_fsblk_t block = ext4_ext_pblock(ext);
        int len = ext4_ext_get_actual_len(ext);
        ext4_lblk_t lblock = le32_to_cpu(ext->ee_block);
        ext4_lblk_t last = lblock + len - 1;

        if (len == 0 || lblock > last)
                return 0;
        return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
}

static int ext4_valid_extent_idx(struct inode *inode,
                                struct ext4_extent_idx *ext_idx)
{
        ext4_fsblk_t block = ext4_idx_pblock(ext_idx);

        return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
}

static int ext4_valid_extent_entries(struct inode *inode,
                                struct ext4_extent_header *eh,
                                int depth)
{
        unsigned short entries;
        if (eh->eh_entries == 0)
                return 1;

        entries = le16_to_cpu(eh->eh_entries);

        if (depth == 0) {
                /* leaf entries */
                struct ext4_extent *ext = EXT_FIRST_EXTENT(eh);
                struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
                ext4_fsblk_t pblock = 0;
                ext4_lblk_t lblock = 0;
                ext4_lblk_t prev = 0;
                int len = 0;
                while (entries) {
                        if (!ext4_valid_extent(inode, ext))
                                return 0;

                        /* Check for overlapping extents */
                        lblock = le32_to_cpu(ext->ee_block);
                        len = ext4_ext_get_actual_len(ext);
                        if ((lblock <= prev) && prev) {
                                pblock = ext4_ext_pblock(ext);
                                es->s_last_error_block = cpu_to_le64(pblock);
                                return 0;
                        }
                        ext++;
                        entries--;
                        prev = lblock + len - 1;
                }
        } else {
                struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh);
                while (entries) {
                        if (!ext4_valid_extent_idx(inode, ext_idx))
                                return 0;
                        ext_idx++;
                        entries--;
                }
        }
        return 1;
}

static int __ext4_ext_check(const char *function, unsigned int line,
                            struct inode *inode, struct ext4_extent_header *eh,
                            int depth)
{
        const char *error_msg;
        int max = 0;

        if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
                error_msg = "invalid magic";
                goto corrupted;
        }
        if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
                error_msg = "unexpected eh_depth";
                goto corrupted;
        }
        if (unlikely(eh->eh_max == 0)) {
                error_msg = "invalid eh_max";
                goto corrupted;
        }
        max = ext4_ext_max_entries(inode, depth);
        if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
                error_msg = "too large eh_max";
                goto corrupted;
        }
        if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
                error_msg = "invalid eh_entries";
                goto corrupted;
        }
        if (!ext4_valid_extent_entries(inode, eh, depth)) {
                error_msg = "invalid extent entries";
                goto corrupted;
        }
        return 0;

corrupted:
        ext4_error_inode(inode, function, line, 0,
                        "bad header/extent: %s - magic %x, "
                        "entries %u, max %u(%u), depth %u(%u)",
                        error_msg, le16_to_cpu(eh->eh_magic),
                        le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
                        max, le16_to_cpu(eh->eh_depth), depth);

        return -EIO;
}

#define ext4_ext_check(inode, eh, depth)        \
        __ext4_ext_check(__func__, __LINE__, inode, eh, depth)

int ext4_ext_check_inode(struct inode *inode)
{
        return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
}

#ifdef EXT_DEBUG
static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
{
        int k, l = path->p_depth;

        ext_debug("path:");
        for (k = 0; k <= l; k++, path++) {
                if (path->p_idx) {
                  ext_debug("  %d->%llu", le32_to_cpu(path->p_idx->ei_block),
                            ext4_idx_pblock(path->p_idx));
                } else if (path->p_ext) {
                        ext_debug("  %d:[%d]%d:%llu ",
                                  le32_to_cpu(path->p_ext->ee_block),
                                  ext4_ext_is_uninitialized(path->p_ext),
                                  ext4_ext_get_actual_len(path->p_ext),
                                  ext4_ext_pblock(path->p_ext));
                } else
                        ext_debug("  []");
        }
        ext_debug("\n");
}

static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
{
        int depth = ext_depth(inode);
        struct ext4_extent_header *eh;
        struct ext4_extent *ex;
        int i;

        if (!path)
                return;

        eh = path[depth].p_hdr;
        ex = EXT_FIRST_EXTENT(eh);

        ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);

        for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
                ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
                          ext4_ext_is_uninitialized(ex),
                          ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
        }
        ext_debug("\n");
}

static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path,
                        ext4_fsblk_t newblock, int level)
{
        int depth = ext_depth(inode);
        struct ext4_extent *ex;

        if (depth != level) {
                struct ext4_extent_idx *idx;
                idx = path[level].p_idx;
                while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
                        ext_debug("%d: move %d:%llu in new index %llu\n", level,
                                        le32_to_cpu(idx->ei_block),
                                        ext4_idx_pblock(idx),
                                        newblock);
                        idx++;
                }

                return;
        }

        ex = path[depth].p_ext;
        while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
                ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
                                le32_to_cpu(ex->ee_block),
                                ext4_ext_pblock(ex),
                                ext4_ext_is_uninitialized(ex),
                                ext4_ext_get_actual_len(ex),
                                newblock);
                ex++;
        }
}

#else
#define ext4_ext_show_path(inode, path)
#define ext4_ext_show_leaf(inode, path)
#define ext4_ext_show_move(inode, path, newblock, level)
#endif

void ext4_ext_drop_refs(struct ext4_ext_path *path)
{
        int depth = path->p_depth;
        int i;

        for (i = 0; i <= depth; i++, path++)
                if (path->p_bh) {
                        brelse(path->p_bh);
                        path->p_bh = NULL;
                }
}

/*
 * ext4_ext_binsearch_idx:
 * binary search for the closest index of the given block
 * the header must be checked before calling this
 */
static void
ext4_ext_binsearch_idx(struct inode *inode,
                        struct ext4_ext_path *path, ext4_lblk_t block)
{
        struct ext4_extent_header *eh = path->p_hdr;
        struct ext4_extent_idx *r, *l, *m;


        ext_debug("binsearch for %u(idx):  ", block);

        l = EXT_FIRST_INDEX(eh) + 1;
        r = EXT_LAST_INDEX(eh);
        while (l <= r) {
                m = l + (r - l) / 2;
                if (block < le32_to_cpu(m->ei_block))
                        r = m - 1;
                else
                        l = m + 1;
                ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
                                m, le32_to_cpu(m->ei_block),
                                r, le32_to_cpu(r->ei_block));
        }

        path->p_idx = l - 1;
        ext_debug("  -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
                  ext4_idx_pblock(path->p_idx));

#ifdef CHECK_BINSEARCH
        {
                struct ext4_extent_idx *chix, *ix;
                int k;

                chix = ix = EXT_FIRST_INDEX(eh);
                for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
                  if (k != 0 &&
                      le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
                                printk(KERN_DEBUG "k=%d, ix=0x%p, "
                                       "first=0x%p\n", k,
                                       ix, EXT_FIRST_INDEX(eh));
                                printk(KERN_DEBUG "%u <= %u\n",
                                       le32_to_cpu(ix->ei_block),
                                       le32_to_cpu(ix[-1].ei_block));
                        }
                        BUG_ON(k && le32_to_cpu(ix->ei_block)
                                           <= le32_to_cpu(ix[-1].ei_block));
                        if (block < le32_to_cpu(ix->ei_block))
                                break;
                        chix = ix;
                }
                BUG_ON(chix != path->p_idx);
        }
#endif

}

/*
 * ext4_ext_binsearch:
 * binary search for closest extent of the given block
 * the header must be checked before calling this
 */
static void
ext4_ext_binsearch(struct inode *inode,
                struct ext4_ext_path *path, ext4_lblk_t block)
{
        struct ext4_extent_header *eh = path->p_hdr;
        struct ext4_extent *r, *l, *m;

        if (eh->eh_entries == 0) {
                /*
                 * this leaf is empty:
                 * we get such a leaf in split/add case
                 */
                return;
        }

        ext_debug("binsearch for %u:  ", block);

        l = EXT_FIRST_EXTENT(eh) + 1;
        r = EXT_LAST_EXTENT(eh);

        while (l <= r) {
                m = l + (r - l) / 2;
                if (block < le32_to_cpu(m->ee_block))
                        r = m - 1;
                else
                        l = m + 1;
                ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
                                m, le32_to_cpu(m->ee_block),
                                r, le32_to_cpu(r->ee_block));
        }

        path->p_ext = l - 1;
        ext_debug("  -> %d:%llu:[%d]%d ",
                        le32_to_cpu(path->p_ext->ee_block),
                        ext4_ext_pblock(path->p_ext),
                        ext4_ext_is_uninitialized(path->p_ext),
                        ext4_ext_get_actual_len(path->p_ext));

#ifdef CHECK_BINSEARCH
        {
                struct ext4_extent *chex, *ex;
                int k;

                chex = ex = EXT_FIRST_EXTENT(eh);
                for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
                        BUG_ON(k && le32_to_cpu(ex->ee_block)
                                          <= le32_to_cpu(ex[-1].ee_block));
                        if (block < le32_to_cpu(ex->ee_block))
                                break;
                        chex = ex;
                }
                BUG_ON(chex != path->p_ext);
        }
#endif

}

int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
{
        struct ext4_extent_header *eh;

        eh = ext_inode_hdr(inode);
        eh->eh_depth = 0;
        eh->eh_entries = 0;
        eh->eh_magic = EXT4_EXT_MAGIC;
        eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
        ext4_mark_inode_dirty(handle, inode);
        ext4_ext_invalidate_cache(inode);
        return 0;
}

struct ext4_ext_path *
ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
                                        struct ext4_ext_path *path)
{
        struct ext4_extent_header *eh;
        struct buffer_head *bh;
        short int depth, i, ppos = 0, alloc = 0;
        int ret;

        eh = ext_inode_hdr(inode);
        depth = ext_depth(inode);

        /* account possible depth increase */
        if (!path) {
                path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
                                GFP_NOFS);
                if (!path)
                        return ERR_PTR(-ENOMEM);
                alloc = 1;
        }
        path[0].p_hdr = eh;
        path[0].p_bh = NULL;

        i = depth;
        /* walk through the tree */
        while (i) {
                int need_to_validate = 0;

                ext_debug("depth %d: num %d, max %d\n",
                          ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));

                ext4_ext_binsearch_idx(inode, path + ppos, block);
                path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
                path[ppos].p_depth = i;
                path[ppos].p_ext = NULL;

                bh = sb_getblk_gfp(inode->i_sb, path[ppos].p_block,
                                   __GFP_MOVABLE | GFP_NOFS);
                if (unlikely(!bh)) {
                        ret = -ENOMEM;
                        goto err;
                }
                if (!bh_uptodate_or_lock(bh)) {
                        trace_ext4_ext_load_extent(inode, block,
                                                path[ppos].p_block);
                        ret = bh_submit_read(bh);
                        if (ret < 0) {
                                put_bh(bh);
                                goto err;
                        }
                        /* validate the extent entries */
                        need_to_validate = 1;
                }
                eh = ext_block_hdr(bh);
                ppos++;
                if (unlikely(ppos > depth)) {
                        put_bh(bh);
                        EXT4_ERROR_INODE(inode,
                                         "ppos %d > depth %d", ppos, depth);
                        ret = -EIO;
                        goto err;
                }
                path[ppos].p_bh = bh;
                path[ppos].p_hdr = eh;
                i--;

                ret = need_to_validate ? ext4_ext_check(inode, eh, i) : 0;
                if (ret < 0)
                        goto err;
        }

        path[ppos].p_depth = i;
        path[ppos].p_ext = NULL;
        path[ppos].p_idx = NULL;

        /* find extent */
        ext4_ext_binsearch(inode, path + ppos, block);
        /* if not an empty leaf */
        if (path[ppos].p_ext)
                path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);

        ext4_ext_show_path(inode, path);

        return path;

err:
        ext4_ext_drop_refs(path);
        if (alloc)
                kfree(path);
        return ERR_PTR(ret);
}

/*
 * ext4_ext_insert_index:
 * insert new index [@logical;@ptr] into the block at @curp;
 * check where to insert: before @curp or after @curp
 */
static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
                                 struct ext4_ext_path *curp,
                                 int logical, ext4_fsblk_t ptr)
{
        struct ext4_extent_idx *ix;
        int len, err;

        err = ext4_ext_get_access(handle, inode, curp);
        if (err)
                return err;

        if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
                EXT4_ERROR_INODE(inode,
                                 "logical %d == ei_block %d!",
                                 logical, le32_to_cpu(curp->p_idx->ei_block));
                return -EIO;
        }

        if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
                             >= le16_to_cpu(curp->p_hdr->eh_max))) {
                EXT4_ERROR_INODE(inode,
                                 "eh_entries %d >= eh_max %d!",
                                 le16_to_cpu(curp->p_hdr->eh_entries),
                                 le16_to_cpu(curp->p_hdr->eh_max));
                return -EIO;
        }

        if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
                /* insert after */
                ext_debug("insert new index %d after: %llu\n", logical, ptr);
                ix = curp->p_idx + 1;
        } else {
                /* insert before */
                ext_debug("insert new index %d before: %llu\n", logical, ptr);
                ix = curp->p_idx;
        }

        len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1;
        BUG_ON(len < 0);
        if (len > 0) {
                ext_debug("insert new index %d: "
                                "move %d indices from 0x%p to 0x%p\n",
                                logical, len, ix, ix + 1);
                memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx));
        }

        if (unlikely(ix > EXT_MAX_INDEX(curp->p_hdr))) {
                EXT4_ERROR_INODE(inode, "ix > EXT_MAX_INDEX!");
                return -EIO;
        }

        ix->ei_block = cpu_to_le32(logical);
        ext4_idx_store_pblock(ix, ptr);
        le16_add_cpu(&curp->p_hdr->eh_entries, 1);

        if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
                EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
                return -EIO;
        }

        err = ext4_ext_dirty(handle, inode, curp);
        ext4_std_error(inode->i_sb, err);

        return err;
}

/*
 * ext4_ext_split:
 * inserts new subtree into the path, using free index entry
 * at depth @at:
 * - allocates all needed blocks (new leaf and all intermediate index blocks)
 * - makes decision where to split
 * - moves remaining extents and index entries (right to the split point)
 *   into the newly allocated blocks
 * - initializes subtree
 */
static int ext4_ext_split(handle_t *handle, struct inode *inode,
                          unsigned int flags,
                          struct ext4_ext_path *path,
                          struct ext4_extent *newext, int at)
{
        struct buffer_head *bh = NULL;
        int depth = ext_depth(inode);
        struct ext4_extent_header *neh;
        struct ext4_extent_idx *fidx;
        int i = at, k, m, a;
        ext4_fsblk_t newblock, oldblock;
        __le32 border;
        ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
        int err = 0;

        /* make decision: where to split? */
        /* FIXME: now decision is simplest: at current extent */

        /* if current leaf will be split, then we should use
         * border from split point */
        if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
                EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
                return -EIO;
        }
        if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
                border = path[depth].p_ext[1].ee_block;
                ext_debug("leaf will be split."
                                " next leaf starts at %d\n",
                                  le32_to_cpu(border));
        } else {
                border = newext->ee_block;
                ext_debug("leaf will be added."
                                " next leaf starts at %d\n",
                                le32_to_cpu(border));
        }

        /*
         * If error occurs, then we break processing
         * and mark filesystem read-only. index won't
         * be inserted and tree will be in consistent
         * state. Next mount will repair buffers too.
         */

        /*
         * Get array to track all allocated blocks.
         * We need this to handle errors and free blocks
         * upon them.
         */
        ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
        if (!ablocks)
                return -ENOMEM;

        /* allocate all needed blocks */
        ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
        for (a = 0; a < depth - at; a++) {
                newblock = ext4_ext_new_meta_block(handle, inode, path,
                                                   newext, &err, flags);
                if (newblock == 0)
                        goto cleanup;
                ablocks[a] = newblock;
        }

        /* initialize new leaf */
        newblock = ablocks[--a];
        if (unlikely(newblock == 0)) {
                EXT4_ERROR_INODE(inode, "newblock == 0!");
                err = -EIO;
                goto cleanup;
        }
        bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS);
        if (!bh) {
                err = -ENOMEM;
                goto cleanup;
        }
        lock_buffer(bh);

        err = ext4_journal_get_create_access(handle, bh);
        if (err)
                goto cleanup;

        neh = ext_block_hdr(bh);
        neh->eh_entries = 0;
        neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
        neh->eh_magic = EXT4_EXT_MAGIC;
        neh->eh_depth = 0;

        /* move remainder of path[depth] to the new leaf */
        if (unlikely(path[depth].p_hdr->eh_entries !=
                     path[depth].p_hdr->eh_max)) {
                EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
                                 path[depth].p_hdr->eh_entries,
                                 path[depth].p_hdr->eh_max);
                err = -EIO;
                goto cleanup;
        }
        /* start copy from next extent */
        m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++;
        ext4_ext_show_move(inode, path, newblock, depth);
        if (m) {
                struct ext4_extent *ex;
                ex = EXT_FIRST_EXTENT(neh);
                memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m);
                le16_add_cpu(&neh->eh_entries, m);
        }

        set_buffer_uptodate(bh);
        unlock_buffer(bh);

        err = ext4_handle_dirty_metadata(handle, inode, bh);
        if (err)
                goto cleanup;
        brelse(bh);
        bh = NULL;

        /* correct old leaf */
        if (m) {
                err = ext4_ext_get_access(handle, inode, path + depth);
                if (err)
                        goto cleanup;
                le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
                err = ext4_ext_dirty(handle, inode, path + depth);
                if (err)
                        goto cleanup;

        }

        /* create intermediate indexes */
        k = depth - at - 1;
        if (unlikely(k < 0)) {
                EXT4_ERROR_INODE(inode, "k %d < 0!", k);
                err = -EIO;
                goto cleanup;
        }
        if (k)
                ext_debug("create %d intermediate indices\n", k);
        /* insert new index into current index block */
        /* current depth stored in i var */
        i = depth - 1;
        while (k--) {
                oldblock = newblock;
                newblock = ablocks[--a];
                bh = sb_getblk(inode->i_sb, newblock);
                if (!bh) {
                        err = -ENOMEM;
                        goto cleanup;
                }
                lock_buffer(bh);

                err = ext4_journal_get_create_access(handle, bh);
                if (err)
                        goto cleanup;

                neh = ext_block_hdr(bh);
                neh->eh_entries = cpu_to_le16(1);
                neh->eh_magic = EXT4_EXT_MAGIC;
                neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
                neh->eh_depth = cpu_to_le16(depth - i);
                fidx = EXT_FIRST_INDEX(neh);
                fidx->ei_block = border;
                ext4_idx_store_pblock(fidx, oldblock);

                ext_debug("int.index at %d (block %llu): %u -> %llu\n",
                                i, newblock, le32_to_cpu(border), oldblock);

                /* move remainder of path[i] to the new index block */
                if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
                                        EXT_LAST_INDEX(path[i].p_hdr))) {
                        EXT4_ERROR_INODE(inode,
                                         "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
                                         le32_to_cpu(path[i].p_ext->ee_block));
                        err = -EIO;
                        goto cleanup;
                }
                /* start copy indexes */
                m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
                ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
                                EXT_MAX_INDEX(path[i].p_hdr));
                ext4_ext_show_move(inode, path, newblock, i);
                if (m) {
                        memmove(++fidx, path[i].p_idx,
                                sizeof(struct ext4_extent_idx) * m);
                        le16_add_cpu(&neh->eh_entries, m);
                }
                set_buffer_uptodate(bh);
                unlock_buffer(bh);

                err = ext4_handle_dirty_metadata(handle, inode, bh);
                if (err)
                        goto cleanup;
                brelse(bh);
                bh = NULL;

                /* correct old index */
                if (m) {
                        err = ext4_ext_get_access(handle, inode, path + i);
                        if (err)
                                goto cleanup;
                        le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
                        err = ext4_ext_dirty(handle, inode, path + i);
                        if (err)
                                goto cleanup;
                }

                i--;
        }

        /* insert new index */
        err = ext4_ext_insert_index(handle, inode, path + at,
                                    le32_to_cpu(border), newblock);

cleanup:
        if (bh) {
                if (buffer_locked(bh))
                        unlock_buffer(bh);
                brelse(bh);
        }

        if (err) {
                /* free all allocated blocks in error case */
                for (i = 0; i < depth; i++) {
                        if (!ablocks[i])
                                continue;
                        ext4_free_blocks(handle, inode, NULL, ablocks[i], 1,
                                         EXT4_FREE_BLOCKS_METADATA);
                }
        }
        kfree(ablocks);

        return err;
}

/*
 * ext4_ext_grow_indepth:
 * implements tree growing procedure:
 * - allocates new block
 * - moves top-level data (index block or leaf) into the new block
 * - initializes new top-level, creating index that points to the
 *   just created block
 */
static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
                                 unsigned int flags,
                                 struct ext4_extent *newext)
{
        struct ext4_extent_header *neh;
        struct buffer_head *bh;
        ext4_fsblk_t newblock;
        int err = 0;

        newblock = ext4_ext_new_meta_block(handle, inode, NULL,
                newext, &err, flags);
        if (newblock == 0)
                return err;

        bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS);
        if (!bh)
                return -ENOMEM;
        lock_buffer(bh);

        err = ext4_journal_get_create_access(handle, bh);
        if (err) {
                unlock_buffer(bh);
                goto out;
        }

        /* move top-level index/leaf into new block */
        memmove(bh->b_data, EXT4_I(inode)->i_data,
                sizeof(EXT4_I(inode)->i_data));

        /* set size of new block */
        neh = ext_block_hdr(bh);
        /* old root could have indexes or leaves
         * so calculate e_max right way */
        if (ext_depth(inode))
                neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
        else
                neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
        neh->eh_magic = EXT4_EXT_MAGIC;
        set_buffer_uptodate(bh);
        unlock_buffer(bh);

        err = ext4_handle_dirty_metadata(handle, inode, bh);
        if (err)
                goto out;

        /* Update top-level index: num,max,pointer */
        neh = ext_inode_hdr(inode);
        neh->eh_entries = cpu_to_le16(1);
        ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock);
        if (neh->eh_depth == 0) {
                /* Root extent block becomes index block */
                neh->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
                EXT_FIRST_INDEX(neh)->ei_block =
                        EXT_FIRST_EXTENT(neh)->ee_block;
        }
        ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
                  le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
                  le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
                  ext4_idx_pblock(EXT_FIRST_INDEX(neh)));

        neh->eh_depth = cpu_to_le16(le16_to_cpu(neh->eh_depth) + 1);
        ext4_mark_inode_dirty(handle, inode);
out:
        brelse(bh);

        return err;
}

/*
 * ext4_ext_create_new_leaf:
 * finds empty index and adds new leaf.
 * if no free index is found, then it requests in-depth growing.
 */
static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
                                    unsigned int flags,
                                    struct ext4_ext_path *path,
                                    struct ext4_extent *newext)
{
        struct ext4_ext_path *curp;
        int depth, i, err = 0;

repeat:
        i = depth = ext_depth(inode);

        /* walk up to the tree and look for free index entry */
        curp = path + depth;
        while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
                i--;
                curp--;
        }

        /* we use already allocated block for index block,
         * so subsequent data blocks should be contiguous */
        if (EXT_HAS_FREE_INDEX(curp)) {
                /* if we found index with free entry, then use that
                 * entry: create all needed subtree and add new leaf */
                err = ext4_ext_split(handle, inode, flags, path, newext, i);
                if (err)
                        goto out;

                /* refill path */
                ext4_ext_drop_refs(path);
                path = ext4_ext_find_extent(inode,
                                    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
                                    path);
                if (IS_ERR(path))
                        err = PTR_ERR(path);
        } else {
                /* tree is full, time to grow in depth */
                err = ext4_ext_grow_indepth(handle, inode, flags, newext);
                if (err)
                        goto out;

                /* refill path */
                ext4_ext_drop_refs(path);
                path = ext4_ext_find_extent(inode,
                                   (ext4_lblk_t)le32_to_cpu(newext->ee_block),
                                    path);
                if (IS_ERR(path)) {
                        err = PTR_ERR(path);
                        goto out;
                }

                /*
                 * only first (depth 0 -> 1) produces free space;
                 * in all other cases we have to split the grown tree
                 */
                depth = ext_depth(inode);
                if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
                        /* now we need to split */
                        goto repeat;
                }
        }

out:
        return err;
}

/*
 * search the closest allocated block to the left for *logical
 * and returns it at @logical + it's physical address at @phys
 * if *logical is the smallest allocated block, the function
 * returns 0 at @phys
 * return value contains 0 (success) or error code
 */
static int ext4_ext_search_left(struct inode *inode,
                                struct ext4_ext_path *path,
                                ext4_lblk_t *logical, ext4_fsblk_t *phys)
{
        struct ext4_extent_idx *ix;
        struct ext4_extent *ex;
        int depth, ee_len;

        if (unlikely(path == NULL)) {
                EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
                return -EIO;
        }
        depth = path->p_depth;
        *phys = 0;

        if (depth == 0 && path->p_ext == NULL)
                return 0;

        /* usually extent in the path covers blocks smaller
         * then *logical, but it can be that extent is the
         * first one in the file */

        ex = path[depth].p_ext;
        ee_len = ext4_ext_get_actual_len(ex);
        if (*logical < le32_to_cpu(ex->ee_block)) {
                if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
                        EXT4_ERROR_INODE(inode,
                                         "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
                                         *logical, le32_to_cpu(ex->ee_block));
                        return -EIO;
                }
                while (--depth >= 0) {
                        ix = path[depth].p_idx;
                        if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
                                EXT4_ERROR_INODE(inode,
                                  "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
                                  ix != NULL ? le32_to_cpu(ix->ei_block) : 0,
                                  EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
                le32_to_cpu(EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block) : 0,
                                  depth);
                                return -EIO;
                        }
                }
                return 0;
        }

        if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
                EXT4_ERROR_INODE(inode,
                                 "logical %d < ee_block %d + ee_len %d!",
                                 *logical, le32_to_cpu(ex->ee_block), ee_len);
                return -EIO;
        }

        *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
        *phys = ext4_ext_pblock(ex) + ee_len - 1;
        return 0;
}

/*
 * search the closest allocated block to the right for *logical
 * and returns it at @logical + it's physical address at @phys
 * if *logical is the largest allocated block, the function
 * returns 0 at @phys
 * return value contains 0 (success) or error code
 */
static int ext4_ext_search_right(struct inode *inode,
                                 struct ext4_ext_path *path,
                                 ext4_lblk_t *logical, ext4_fsblk_t *phys,
                                 struct ext4_extent **ret_ex)
{
        struct buffer_head *bh = NULL;
        struct ext4_extent_header *eh;
        struct ext4_extent_idx *ix;
        struct ext4_extent *ex;
        ext4_fsblk_t block;
        int depth;      /* Note, NOT eh_depth; depth from top of tree */
        int ee_len;

        if (unlikely(path == NULL)) {
                EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
                return -EIO;
        }
        depth = path->p_depth;
        *phys = 0;

        if (depth == 0 && path->p_ext == NULL)
                return 0;

        /* usually extent in the path covers blocks smaller
         * then *logical, but it can be that extent is the
         * first one in the file */

        ex = path[depth].p_ext;
        ee_len = ext4_ext_get_actual_len(ex);
        if (*logical < le32_to_cpu(ex->ee_block)) {
                if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
                        EXT4_ERROR_INODE(inode,
                                         "first_extent(path[%d].p_hdr) != ex",
                                         depth);
                        return -EIO;
                }
                while (--depth >= 0) {
                        ix = path[depth].p_idx;
                        if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
                                EXT4_ERROR_INODE(inode,
                                                 "ix != EXT_FIRST_INDEX *logical %d!",
                                                 *logical);
                                return -EIO;
                        }
                }
                goto found_extent;
        }

        if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
                EXT4_ERROR_INODE(inode,
                                 "logical %d < ee_block %d + ee_len %d!",
                                 *logical, le32_to_cpu(ex->ee_block), ee_len);
                return -EIO;
        }

        if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
                /* next allocated block in this leaf */
                ex++;
                goto found_extent;
        }

        /* go up and search for index to the right */
        while (--depth >= 0) {
                ix = path[depth].p_idx;
                if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
                        goto got_index;
        }

        /* we've gone up to the root and found no index to the right */
        return 0;

got_index:
        /* we've found index to the right, let's
         * follow it and find the closest allocated
         * block to the right */
        ix++;
        block = ext4_idx_pblock(ix);
        while (++depth < path->p_depth) {
                bh = sb_bread(inode->i_sb, block);
                if (bh == NULL)
                        return -EIO;
                eh = ext_block_hdr(bh);
                /* subtract from p_depth to get proper eh_depth */
                if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
                        put_bh(bh);
                        return -EIO;
                }
                ix = EXT_FIRST_INDEX(eh);
                block = ext4_idx_pblock(ix);
                put_bh(bh);
        }

        bh = sb_bread(inode->i_sb, block);
        if (bh == NULL)
                return -EIO;
        eh = ext_block_hdr(bh);
        if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
                put_bh(bh);
                return -EIO;
        }
        ex = EXT_FIRST_EXTENT(eh);
found_extent:
        *logical = le32_to_cpu(ex->ee_block);
        *phys = ext4_ext_pblock(ex);
        *ret_ex = ex;
        if (bh)
                put_bh(bh);
        return 0;
}

/*
 * ext4_ext_next_allocated_block:
 * returns allocated block in subsequent extent or EXT_MAX_BLOCKS.
 * NOTE: it considers block number from index entry as
 * allocated block. Thus, index entries have to be consistent
 * with leaves.
 */
static ext4_lblk_t
ext4_ext_next_allocated_block(struct ext4_ext_path *path)
{
        int depth;

        BUG_ON(path == NULL);
        depth = path->p_depth;

        if (depth == 0 && path->p_ext == NULL)
                return EXT_MAX_BLOCKS;

        while (depth >= 0) {
                if (depth == path->p_depth) {
                        /* leaf */
                        if (path[depth].p_ext &&
                                path[depth].p_ext !=
                                        EXT_LAST_EXTENT(path[depth].p_hdr))
                          return le32_to_cpu(path[depth].p_ext[1].ee_block);
                } else {
                        /* index */
                        if (path[depth].p_idx !=
                                        EXT_LAST_INDEX(path[depth].p_hdr))
                          return le32_to_cpu(path[depth].p_idx[1].ei_block);
                }
                depth--;
        }

        return EXT_MAX_BLOCKS;
}

/*
 * ext4_ext_next_leaf_block:
 * returns first allocated block from next leaf or EXT_MAX_BLOCKS
 */
static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path)
{
        int depth;

        BUG_ON(path == NULL);
        depth = path->p_depth;

        /* zero-tree has no leaf blocks at all */
        if (depth == 0)
                return EXT_MAX_BLOCKS;

        /* go to index block */
        depth--;

        while (depth >= 0) {
                if (path[depth].p_idx !=
                                EXT_LAST_INDEX(path[depth].p_hdr))
                        return (ext4_lblk_t)
                                le32_to_cpu(path[depth].p_idx[1].ei_block);
                depth--;
        }

        return EXT_MAX_BLOCKS;
}

/*
 * ext4_ext_correct_indexes:
 * if leaf gets modified and modified extent is first in the leaf,
 * then we have to correct all indexes above.
 * TODO: do we need to correct tree in all cases?
 */
static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
                                struct ext4_ext_path *path)
{
        struct ext4_extent_header *eh;
        int depth = ext_depth(inode);
        struct ext4_extent *ex;
        __le32 border;
        int k, err = 0;

        eh = path[depth].p_hdr;
        ex = path[depth].p_ext;

        if (unlikely(ex == NULL || eh == NULL)) {
                EXT4_ERROR_INODE(inode,
                                 "ex %p == NULL or eh %p == NULL", ex, eh);
                return -EIO;
        }

        if (depth == 0) {
                /* there is no tree at all */
                return 0;
        }

        if (ex != EXT_FIRST_EXTENT(eh)) {
                /* we correct tree if first leaf got modified only */
                return 0;
        }

        /*
         * TODO: we need correction if border is smaller than current one
         */
        k = depth - 1;
        border = path[depth].p_ext->ee_block;
        err = ext4_ext_get_access(handle, inode, path + k);
        if (err)
                return err;
        path[k].p_idx->ei_block = border;
        err = ext4_ext_dirty(handle, inode, path + k);
        if (err)
                return err;

        while (k--) {
                /* change all left-side indexes */
                if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
                        break;
                err = ext4_ext_get_access(handle, inode, path + k);
                if (err)
                        break;
                path[k].p_idx->ei_block = border;
                err = ext4_ext_dirty(handle, inode, path + k);
                if (err)
                        break;
        }

        return err;
}

int
ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
                                struct ext4_extent *ex2)
{
        unsigned short ext1_ee_len, ext2_ee_len, max_len;

        /*
         * Make sure that either both extents are uninitialized, or
         * both are _not_.
         */
        if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
                return 0;

        if (ext4_ext_is_uninitialized(ex1))
                max_len = EXT_UNINIT_MAX_LEN;
        else
                max_len = EXT_INIT_MAX_LEN;

        ext1_ee_len = ext4_ext_get_actual_len(ex1);
        ext2_ee_len = ext4_ext_get_actual_len(ex2);

        if (le32_to_cpu(ex1->ee_block) + ext1_ee_len !=
                        le32_to_cpu(ex2->ee_block))
                return 0;

        /*
         * To allow future support for preallocated extents to be added
         * as an RO_COMPAT feature, refuse to merge to extents if
         * this can result in the top bit of ee_len being set.
         */
        if (ext1_ee_len + ext2_ee_len > max_len)
                return 0;
#ifdef AGGRESSIVE_TEST
        if (ext1_ee_len >= 4)
                return 0;
#endif

        if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
                return 1;
        return 0;
}

/*
 * This function tries to merge the "ex" extent to the next extent in the tree.
 * It always tries to merge towards right. If you want to merge towards
 * left, pass "ex - 1" as argument instead of "ex".
 * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns
 * 1 if they got merged.
 */
static int ext4_ext_try_to_merge_right(struct inode *inode,
                                 struct ext4_ext_path *path,
                                 struct ext4_extent *ex)
{
        struct ext4_extent_header *eh;
        unsigned int depth, len;
        int merge_done = 0;
        int uninitialized = 0;

        depth = ext_depth(inode);
        BUG_ON(path[depth].p_hdr == NULL);
        eh = path[depth].p_hdr;

        while (ex < EXT_LAST_EXTENT(eh)) {
                if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
                        break;
                /* merge with next extent! */
                if (ext4_ext_is_uninitialized(ex))
                        uninitialized = 1;
                ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
                                + ext4_ext_get_actual_len(ex + 1));
                if (uninitialized)
                        ext4_ext_mark_uninitialized(ex);

                if (ex + 1 < EXT_LAST_EXTENT(eh)) {
                        len = (EXT_LAST_EXTENT(eh) - ex - 1)
                                * sizeof(struct ext4_extent);
                        memmove(ex + 1, ex + 2, len);
                }
                le16_add_cpu(&eh->eh_entries, -1);
                merge_done = 1;
                WARN_ON(eh->eh_entries == 0);
                if (!eh->eh_entries)
                        EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
        }

        return merge_done;
}

/*
 * This function tries to merge the @ex extent to neighbours in the tree.
 * return 1 if merge left else 0.
 */
static int ext4_ext_try_to_merge(struct inode *inode,
                                  struct ext4_ext_path *path,
                                  struct ext4_extent *ex) {
        struct ext4_extent_header *eh;
        unsigned int depth;
        int merge_done = 0;
        int ret = 0;

        depth = ext_depth(inode);
        BUG_ON(path[depth].p_hdr == NULL);
        eh = path[depth].p_hdr;

        if (ex > EXT_FIRST_EXTENT(eh))
                merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1);

        if (!merge_done)
                ret = ext4_ext_try_to_merge_right(inode, path, ex);

        return ret;
}

/*
 * check if a portion of the "newext" extent overlaps with an
 * existing extent.
 *
 * If there is an overlap discovered, it updates the length of the newext
 * such that there will be no overlap, and then returns 1.
 * If there is no overlap found, it returns 0.
 */
static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi,
                                           struct inode *inode,
                                           struct ext4_extent *newext,
                                           struct ext4_ext_path *path)
{
        ext4_lblk_t b1, b2;
        unsigned int depth, len1;
        unsigned int ret = 0;

        b1 = le32_to_cpu(newext->ee_block);
        len1 = ext4_ext_get_actual_len(newext);
        depth = ext_depth(inode);
        if (!path[depth].p_ext)
                goto out;
        b2 = EXT4_LBLK_CMASK(sbi, le32_to_cpu(path[depth].p_ext->ee_block));

        /*
         * get the next allocated block if the extent in the path
         * is before the requested block(s)
         */
        if (b2 < b1) {
                b2 = ext4_ext_next_allocated_block(path);
                if (b2 == EXT_MAX_BLOCKS)
                        goto out;
                b2 = EXT4_LBLK_CMASK(sbi, b2);
        }

        /* check for wrap through zero on extent logical start block*/
        if (b1 + len1 < b1) {
                len1 = EXT_MAX_BLOCKS - b1;
                newext->ee_len = cpu_to_le16(len1);
                ret = 1;
        }

        /* check for overlap */
        if (b1 + len1 > b2) {
                newext->ee_len = cpu_to_le16(b2 - b1);
                ret = 1;
        }
out:
        return ret;
}

/*
 * ext4_ext_insert_extent:
 * tries to merge requsted extent into the existing extent or
 * inserts requested extent as new one into the tree,
 * creating new leaf in the no-space case.
 */
int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
                                struct ext4_ext_path *path,
                                struct ext4_extent *newext, int flag)
{
        struct ext4_extent_header *eh;
        struct ext4_extent *ex, *fex;
        struct ext4_extent *nearex; /* nearest extent */
        struct ext4_ext_path *npath = NULL;
        int depth, len, err;
        ext4_lblk_t next;
        unsigned uninitialized = 0;
        int flags = 0;

        if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
                EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
                return -EIO;
        }
        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        if (unlikely(path[depth].p_hdr == NULL)) {
                EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
                return -EIO;
        }

        /* try to insert block into found extent and return */
        if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
                && ext4_can_extents_be_merged(inode, ex, newext)) {
                ext_debug("append [%d]%d block to %u:[%d]%d (from %llu)\n",
                          ext4_ext_is_uninitialized(newext),
                          ext4_ext_get_actual_len(newext),
                          le32_to_cpu(ex->ee_block),
                          ext4_ext_is_uninitialized(ex),
                          ext4_ext_get_actual_len(ex),
                          ext4_ext_pblock(ex));
                err = ext4_ext_get_access(handle, inode, path + depth);
                if (err)
                        return err;

                /*
                 * ext4_can_extents_be_merged should have checked that either
                 * both extents are uninitialized, or both aren't. Thus we
                 * need to check only one of them here.
                 */
                if (ext4_ext_is_uninitialized(ex))
                        uninitialized = 1;
                ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
                                        + ext4_ext_get_actual_len(newext));
                if (uninitialized)
                        ext4_ext_mark_uninitialized(ex);
                eh = path[depth].p_hdr;
                nearex = ex;
                goto merge;
        }

        depth = ext_depth(inode);
        eh = path[depth].p_hdr;
        if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max))
                goto has_space;

        /* probably next leaf has space for us? */
        fex = EXT_LAST_EXTENT(eh);
        next = EXT_MAX_BLOCKS;
        if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block))
                next = ext4_ext_next_leaf_block(path);
        if (next != EXT_MAX_BLOCKS) {
                ext_debug("next leaf block - %u\n", next);
                BUG_ON(npath != NULL);
                npath = ext4_ext_find_extent(inode, next, NULL);
                if (IS_ERR(npath))
                        return PTR_ERR(npath);
                BUG_ON(npath->p_depth != path->p_depth);
                eh = npath[depth].p_hdr;
                if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
                        ext_debug("next leaf isn't full(%d)\n",
                                  le16_to_cpu(eh->eh_entries));
                        path = npath;
                        goto has_space;
                }
                ext_debug("next leaf has no free space(%d,%d)\n",
                          le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
        }

        /*
         * There is no free space in the found leaf.
         * We're gonna add a new leaf in the tree.
         */
        if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT)
                flags = EXT4_MB_USE_ROOT_BLOCKS;
        err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext);
        if (err)
                goto cleanup;
        depth = ext_depth(inode);
        eh = path[depth].p_hdr;

has_space:
        nearex = path[depth].p_ext;

        err = ext4_ext_get_access(handle, inode, path + depth);
        if (err)
                goto cleanup;

        if (!nearex) {
                /* there is no extent in this leaf, create first one */
                ext_debug("first extent in the leaf: %u:%llu:[%d]%d\n",
                                le32_to_cpu(newext->ee_block),
                                ext4_ext_pblock(newext),
                                ext4_ext_is_uninitialized(newext),
                                ext4_ext_get_actual_len(newext));
                nearex = EXT_FIRST_EXTENT(eh);
        } else {
                if (le32_to_cpu(newext->ee_block)
                           > le32_to_cpu(nearex->ee_block)) {
                        /* Insert after */
                        ext_debug("insert %u:%llu:[%d]%d before: "
                                        "nearest %p\n",
                                        le32_to_cpu(newext->ee_block),
                                        ext4_ext_pblock(newext),
                                        ext4_ext_is_uninitialized(newext),
                                        ext4_ext_get_actual_len(newext),
                                        nearex);
                        nearex++;
                } else {
                        /* Insert before */
                        BUG_ON(newext->ee_block == nearex->ee_block);
                        ext_debug("insert %u:%llu:[%d]%d after: "
                                        "nearest %p\n",
                                        le32_to_cpu(newext->ee_block),
                                        ext4_ext_pblock(newext),
                                        ext4_ext_is_uninitialized(newext),
                                        ext4_ext_get_actual_len(newext),
                                        nearex);
                }
                len = EXT_LAST_EXTENT(eh) - nearex + 1;
                if (len > 0) {
                        ext_debug("insert %u:%llu:[%d]%d: "
                                        "move %d extents from 0x%p to 0x%p\n",
                                        le32_to_cpu(newext->ee_block),
                                        ext4_ext_pblock(newext),
                                        ext4_ext_is_uninitialized(newext),
                                        ext4_ext_get_actual_len(newext),
                                        len, nearex, nearex + 1);
                        memmove(nearex + 1, nearex,
                                len * sizeof(struct ext4_extent));
                }
        }

        le16_add_cpu(&eh->eh_entries, 1);
        path[depth].p_ext = nearex;
        nearex->ee_block = newext->ee_block;
        ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
        nearex->ee_len = newext->ee_len;

merge:
        /* try to merge extents to the right */
        if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
                ext4_ext_try_to_merge(inode, path, nearex);

        /* try to merge extents to the left */

        /* time to correct all indexes above */
        err = ext4_ext_correct_indexes(handle, inode, path);
        if (err)
                goto cleanup;

        err = ext4_ext_dirty(handle, inode, path + depth);

cleanup:
        if (npath) {
                ext4_ext_drop_refs(npath);
                kfree(npath);
        }
        ext4_ext_invalidate_cache(inode);
        return err;
}

static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
                               ext4_lblk_t num, ext_prepare_callback func,
                               void *cbdata)
{
        struct ext4_ext_path *path = NULL;
        struct ext4_ext_cache cbex;
        struct ext4_extent *ex;
        ext4_lblk_t next, start = 0, end = 0;
        ext4_lblk_t last = block + num;
        int depth, exists, err = 0;

        BUG_ON(func == NULL);
        BUG_ON(inode == NULL);

        while (block < last && block != EXT_MAX_BLOCKS) {
                num = last - block;
                /* find extent for this block */
                down_read(&EXT4_I(inode)->i_data_sem);
                path = ext4_ext_find_extent(inode, block, path);
                up_read(&EXT4_I(inode)->i_data_sem);
                if (IS_ERR(path)) {
                        err = PTR_ERR(path);
                        path = NULL;
                        break;
                }

                depth = ext_depth(inode);
                if (unlikely(path[depth].p_hdr == NULL)) {
                        EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
                        err = -EIO;
                        break;
                }
                ex = path[depth].p_ext;
                next = ext4_ext_next_allocated_block(path);

                exists = 0;
                if (!ex) {
                        /* there is no extent yet, so try to allocate
                         * all requested space */
                        start = block;
                        end = block + num;
                } else if (le32_to_cpu(ex->ee_block) > block) {
                        /* need to allocate space before found extent */
                        start = block;
                        end = le32_to_cpu(ex->ee_block);
                        if (block + num < end)
                                end = block + num;
                } else if (block >= le32_to_cpu(ex->ee_block)
                                        + ext4_ext_get_actual_len(ex)) {
                        /* need to allocate space after found extent */
                        start = block;
                        end = block + num;
                        if (end >= next)
                                end = next;
                } else if (block >= le32_to_cpu(ex->ee_block)) {
                        /*
                         * some part of requested space is covered
                         * by found extent
                         */
                        start = block;
                        end = le32_to_cpu(ex->ee_block)
                                + ext4_ext_get_actual_len(ex);
                        if (block + num < end)
                                end = block + num;
                        exists = 1;
                } else {
                        BUG();
                }
                BUG_ON(end <= start);

                if (!exists) {
                        cbex.ec_block = start;
                        cbex.ec_len = end - start;
                        cbex.ec_start = 0;
                } else {
                        cbex.ec_block = le32_to_cpu(ex->ee_block);
                        cbex.ec_len = ext4_ext_get_actual_len(ex);
                        cbex.ec_start = ext4_ext_pblock(ex);
                }

                if (unlikely(cbex.ec_len == 0)) {
                        EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
                        err = -EIO;
                        break;
                }
                err = func(inode, next, &cbex, ex, cbdata);
                ext4_ext_drop_refs(path);

                if (err < 0)
                        break;

                if (err == EXT_REPEAT)
                        continue;
                else if (err == EXT_BREAK) {
                        err = 0;
                        break;
                }

                if (ext_depth(inode) != depth) {
                        /* depth was changed. we have to realloc path */
                        kfree(path);
                        path = NULL;
                }

                block = cbex.ec_block + cbex.ec_len;
        }

        if (path) {
                ext4_ext_drop_refs(path);
                kfree(path);
        }

        return err;
}

static void
ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
                        __u32 len, ext4_fsblk_t start)
{
        struct ext4_ext_cache *cex;
        BUG_ON(len == 0);
        spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
        trace_ext4_ext_put_in_cache(inode, block, len, start);
        cex = &EXT4_I(inode)->i_cached_extent;
        cex->ec_block = block;
        cex->ec_len = len;
        cex->ec_start = start;
        spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
}

/*
 * ext4_ext_put_gap_in_cache:
 * calculate boundaries of the gap that the requested block fits into
 * and cache this gap
 */
static void
ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
                                ext4_lblk_t block)
{
        int depth = ext_depth(inode);
        unsigned long len;
        ext4_lblk_t lblock;
        struct ext4_extent *ex;

        ex = path[depth].p_ext;
        if (ex == NULL) {
                /* there is no extent yet, so gap is [0;-] */
                lblock = 0;
                len = EXT_MAX_BLOCKS;
                ext_debug("cache gap(whole file):");
        } else if (block < le32_to_cpu(ex->ee_block)) {
                lblock = block;
                len = le32_to_cpu(ex->ee_block) - block;
                ext_debug("cache gap(before): %u [%u:%u]",
                                block,
                                le32_to_cpu(ex->ee_block),
                                 ext4_ext_get_actual_len(ex));
        } else if (block >= le32_to_cpu(ex->ee_block)
                        + ext4_ext_get_actual_len(ex)) {
                ext4_lblk_t next;
                lblock = le32_to_cpu(ex->ee_block)
                        + ext4_ext_get_actual_len(ex);

                next = ext4_ext_next_allocated_block(path);
                ext_debug("cache gap(after): [%u:%u] %u",
                                le32_to_cpu(ex->ee_block),
                                ext4_ext_get_actual_len(ex),
                                block);
                BUG_ON(next == lblock);
                len = next - lblock;
        } else {
                lblock = len = 0;
                BUG();
        }

        ext_debug(" -> %u:%lu\n", lblock, len);
        ext4_ext_put_in_cache(inode, lblock, len, 0);
}

/*
 * ext4_ext_check_cache()
 * Checks to see if the given block is in the cache.
 * If it is, the cached extent is stored in the given
 * cache extent pointer.  If the cached extent is a hole,
 * this routine should be used instead of
 * ext4_ext_in_cache if the calling function needs to
 * know the size of the hole.
 *
 * @inode: The files inode
 * @block: The block to look for in the cache
 * @ex:    Pointer where the cached extent will be stored
 *         if it contains block
 *
 * Return 0 if cache is invalid; 1 if the cache is valid
 */
static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
        struct ext4_ext_cache *ex){
        struct ext4_ext_cache *cex;
        struct ext4_sb_info *sbi;
        int ret = 0;

        /*
         * We borrow i_block_reservation_lock to protect i_cached_extent
         */
        spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
        cex = &EXT4_I(inode)->i_cached_extent;
        sbi = EXT4_SB(inode->i_sb);

        /* has cache valid data? */
        if (cex->ec_len == 0)
                goto errout;

        if (in_range(block, cex->ec_block, cex->ec_len)) {
                memcpy(ex, cex, sizeof(struct ext4_ext_cache));
                ext_debug("%u cached by %u:%u:%llu\n",
                                block,
                                cex->ec_block, cex->ec_len, cex->ec_start);
                ret = 1;
        }
errout:
        trace_ext4_ext_in_cache(inode, block, ret);
        spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
        return ret;
}

/*
 * ext4_ext_in_cache()
 * Checks to see if the given block is in the cache.
 * If it is, the cached extent is stored in the given
 * extent pointer.
 *
 * @inode: The files inode
 * @block: The block to look for in the cache
 * @ex:    Pointer where the cached extent will be stored
 *         if it contains block
 *
 * Return 0 if cache is invalid; 1 if the cache is valid
 */
static int
ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
                        struct ext4_extent *ex)
{
        struct ext4_ext_cache cex;
        int ret = 0;

        if (ext4_ext_check_cache(inode, block, &cex)) {
                ex->ee_block = cpu_to_le32(cex.ec_block);
                ext4_ext_store_pblock(ex, cex.ec_start);
                ex->ee_len = cpu_to_le16(cex.ec_len);
                ret = 1;
        }

        return ret;
}


/*
 * ext4_ext_rm_idx:
 * removes index from the index block.
 */
static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
                        struct ext4_ext_path *path, int depth)
{
        int err;
        ext4_fsblk_t leaf;

        /* free index block */
        depth--;
        path = path + depth;
        leaf = ext4_idx_pblock(path->p_idx);
        if (unlikely(path->p_hdr->eh_entries == 0)) {
                EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
                return -EIO;
        }
        err = ext4_ext_get_access(handle, inode, path);
        if (err)
                return err;

        if (path->p_idx != EXT_LAST_INDEX(path->p_hdr)) {
                int len = EXT_LAST_INDEX(path->p_hdr) - path->p_idx;
                len *= sizeof(struct ext4_extent_idx);
                memmove(path->p_idx, path->p_idx + 1, len);
        }

        le16_add_cpu(&path->p_hdr->eh_entries, -1);
        err = ext4_ext_dirty(handle, inode, path);
        if (err)
                return err;
        ext_debug("index is empty, remove it, free block %llu\n", leaf);
        trace_ext4_ext_rm_idx(inode, leaf);

        ext4_free_blocks(handle, inode, NULL, leaf, 1,
                         EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);

        while (--depth >= 0) {
                if (path->p_idx != EXT_FIRST_INDEX(path->p_hdr))
                        break;
                path--;
                err = ext4_ext_get_access(handle, inode, path);
                if (err)
                        break;
                path->p_idx->ei_block = (path+1)->p_idx->ei_block;
                err = ext4_ext_dirty(handle, inode, path);
                if (err)
                        break;
        }
        return err;
}

/*
 * ext4_ext_calc_credits_for_single_extent:
 * This routine returns max. credits that needed to insert an extent
 * to the extent tree.
 * When pass the actual path, the caller should calculate credits
 * under i_data_sem.
 */
int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
                                                struct ext4_ext_path *path)
{
        if (path) {
                int depth = ext_depth(inode);
                int ret = 0;

                /* probably there is space in leaf? */
                if (le16_to_cpu(path[depth].p_hdr->eh_entries)
                                < le16_to_cpu(path[depth].p_hdr->eh_max)) {

                        /*
                         *  There are some space in the leaf tree, no
                         *  need to account for leaf block credit
                         *
                         *  bitmaps and block group descriptor blocks
                         *  and other metadata blocks still need to be
                         *  accounted.
                         */
                        /* 1 bitmap, 1 block group descriptor */
                        ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
                        return ret;
                }
        }

        return ext4_chunk_trans_blocks(inode, nrblocks);
}

/*
 * How many index/leaf blocks need to change/allocate to modify nrblocks?
 *
 * if nrblocks are fit in a single extent (chunk flag is 1), then
 * in the worse case, each tree level index/leaf need to be changed
 * if the tree split due to insert a new extent, then the old tree
 * index/leaf need to be updated too
 *
 * If the nrblocks are discontiguous, they could cause
 * the whole tree split more than once, but this is really rare.
 */
int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
{
        int index;
        int depth = ext_depth(inode);

        if (chunk)
                index = depth * 2;
        else
                index = depth * 3;

        return index;
}

static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
                              struct ext4_extent *ex,
                              ext4_fsblk_t *partial_cluster,
                              ext4_lblk_t from, ext4_lblk_t to)
{
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        unsigned short ee_len =  ext4_ext_get_actual_len(ex);
        ext4_fsblk_t pblk;
        int flags = EXT4_FREE_BLOCKS_FORGET;

        if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
                flags |= EXT4_FREE_BLOCKS_METADATA;
        /*
         * For bigalloc file systems, we never free a partial cluster
         * at the beginning of the extent.  Instead, we make a note
         * that we tried freeing the cluster, and check to see if we
         * need to free it on a subsequent call to ext4_remove_blocks,
         * or at the end of the ext4_truncate() operation.
         */
        flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;

        trace_ext4_remove_blocks(inode, ex, from, to, *partial_cluster);
        /*
         * If we have a partial cluster, and it's different from the
         * cluster of the last block, we need to explicitly free the
         * partial cluster here.
         */
        pblk = ext4_ext_pblock(ex) + ee_len - 1;
        if (*partial_cluster && (EXT4_B2C(sbi, pblk) != *partial_cluster)) {
                ext4_free_blocks(handle, inode, NULL,
                                 EXT4_C2B(sbi, *partial_cluster),
                                 sbi->s_cluster_ratio, flags);
                *partial_cluster = 0;
        }

#ifdef EXTENTS_STATS
        {
                struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
                spin_lock(&sbi->s_ext_stats_lock);
                sbi->s_ext_blocks += ee_len;
                sbi->s_ext_extents++;
                if (ee_len < sbi->s_ext_min)
                        sbi->s_ext_min = ee_len;
                if (ee_len > sbi->s_ext_max)
                        sbi->s_ext_max = ee_len;
                if (ext_depth(inode) > sbi->s_depth_max)
                        sbi->s_depth_max = ext_depth(inode);
                spin_unlock(&sbi->s_ext_stats_lock);
        }
#endif
        if (from >= le32_to_cpu(ex->ee_block)
            && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
                /* tail removal */
                ext4_lblk_t num;

                num = le32_to_cpu(ex->ee_block) + ee_len - from;
                pblk = ext4_ext_pblock(ex) + ee_len - num;
                ext_debug("free last %u blocks starting %llu\n", num, pblk);
                ext4_free_blocks(handle, inode, NULL, pblk, num, flags);
                /*
                 * If the block range to be freed didn't start at the
                 * beginning of a cluster, and we removed the entire
                 * extent, save the partial cluster here, since we
                 * might need to delete if we determine that the
                 * truncate operation has removed all of the blocks in
                 * the cluster.
                 */
                if (EXT4_PBLK_COFF(sbi, pblk) &&
                    (ee_len == num))
                        *partial_cluster = EXT4_B2C(sbi, pblk);
                else
                        *partial_cluster = 0;
        } else if (from == le32_to_cpu(ex->ee_block)
                   && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
                /* head removal */
                ext4_lblk_t num;
                ext4_fsblk_t start;

                num = to - from;
                start = ext4_ext_pblock(ex);

                ext_debug("free first %u blocks starting %llu\n", num, start);
                ext4_free_blocks(handle, inode, NULL, start, num, flags);

        } else {
                printk(KERN_INFO "strange request: removal(2) "
                                "%u-%u from %u:%u\n",
                                from, to, le32_to_cpu(ex->ee_block), ee_len);
        }
        return 0;
}


/*
 * ext4_ext_rm_leaf() Removes the extents associated with the
 * blocks appearing between "start" and "end", and splits the extents
 * if "start" and "end" appear in the same extent
 *
 * @handle: The journal handle
 * @inode:  The files inode
 * @path:   The path to the leaf
 * @start:  The first block to remove
 * @end:   The last block to remove
 */
static int
ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
                 struct ext4_ext_path *path, ext4_fsblk_t *partial_cluster,
                 ext4_lblk_t start, ext4_lblk_t end)
{
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        int err = 0, correct_index = 0;
        int depth = ext_depth(inode), credits;
        struct ext4_extent_header *eh;
        ext4_lblk_t a, b;
        unsigned num;
        ext4_lblk_t ex_ee_block;
        unsigned short ex_ee_len;
        unsigned uninitialized = 0;
        struct ext4_extent *ex;

        /* the header must be checked already in ext4_ext_remove_space() */
        ext_debug("truncate since %u in leaf to %u\n", start, end);
        if (!path[depth].p_hdr)
                path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
        eh = path[depth].p_hdr;
        if (unlikely(path[depth].p_hdr == NULL)) {
                EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
                return -EIO;
        }
        /* find where to start removing */
        ex = EXT_LAST_EXTENT(eh);

        ex_ee_block = le32_to_cpu(ex->ee_block);
        ex_ee_len = ext4_ext_get_actual_len(ex);

        /*
         * If we're starting with an extent other than the last one in the
         * node, we need to see if it shares a cluster with the extent to
         * the right (towards the end of the file). If its leftmost cluster
         * is this extent's rightmost cluster and it is not cluster aligned,
         * we'll mark it as a partial that is not to be deallocated.
         */

        if (ex != EXT_LAST_EXTENT(eh)) {
                ext4_fsblk_t current_pblk, right_pblk;
                long long current_cluster, right_cluster;

                current_pblk = ext4_ext_pblock(ex) + ex_ee_len - 1;
                current_cluster = (long long)EXT4_B2C(sbi, current_pblk);
                right_pblk = ext4_ext_pblock(ex + 1);
                right_cluster = (long long)EXT4_B2C(sbi, right_pblk);
                if (current_cluster == right_cluster &&
                        EXT4_PBLK_COFF(sbi, right_pblk))
                        *partial_cluster = -right_cluster;
        }

        trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster);

        while (ex >= EXT_FIRST_EXTENT(eh) &&
                        ex_ee_block + ex_ee_len > start) {

                if (ext4_ext_is_uninitialized(ex))
                        uninitialized = 1;
                else
                        uninitialized = 0;

                ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
                         uninitialized, ex_ee_len);
                path[depth].p_ext = ex;

                a = ex_ee_block > start ? ex_ee_block : start;
                b = ex_ee_block+ex_ee_len - 1 < end ?
                        ex_ee_block+ex_ee_len - 1 : end;

                ext_debug("  border %u:%u\n", a, b);

                /* If this extent is beyond the end of the hole, skip it */
                if (end < ex_ee_block) {
                        ex--;
                        ex_ee_block = le32_to_cpu(ex->ee_block);
                        ex_ee_len = ext4_ext_get_actual_len(ex);
                        continue;
                } else if (b != ex_ee_block + ex_ee_len - 1) {
                        EXT4_ERROR_INODE(inode,"  bad truncate %u:%u\n",
                                         start, end);
                        err = -EIO;
                        goto out;
                } else if (a != ex_ee_block) {
                        /* remove tail of the extent */
                        num = a - ex_ee_block;
                } else {
                        /* remove whole extent: excellent! */
                        num = 0;
                }
                /*
                 * 3 for leaf, sb, and inode plus 2 (bmap and group
                 * descriptor) for each block group; assume two block
                 * groups plus ex_ee_len/blocks_per_block_group for
                 * the worst case
                 */
                credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb));
                if (ex == EXT_FIRST_EXTENT(eh)) {
                        correct_index = 1;
                        credits += (ext_depth(inode)) + 1;
                }
                credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);

                err = ext4_ext_truncate_extend_restart(handle, inode, credits);
                if (err)
                        goto out;

                err = ext4_ext_get_access(handle, inode, path + depth);
                if (err)
                        goto out;

                err = ext4_remove_blocks(handle, inode, ex, partial_cluster,
                                         a, b);
                if (err)
                        goto out;

                if (num == 0)
                        /* this extent is removed; mark slot entirely unused */
                        ext4_ext_store_pblock(ex, 0);

                ex->ee_len = cpu_to_le16(num);
                /*
                 * Do not mark uninitialized if all the blocks in the
                 * extent have been removed.
                 */
                if (uninitialized && num)
                        ext4_ext_mark_uninitialized(ex);
                /*
                 * If the extent was completely released,
                 * we need to remove it from the leaf
                 */
                if (num == 0) {
                        if (end != EXT_MAX_BLOCKS - 1) {
                                /*
                                 * For hole punching, we need to scoot all the
                                 * extents up when an extent is removed so that
                                 * we dont have blank extents in the middle
                                 */
                                memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) *
                                        sizeof(struct ext4_extent));

                                /* Now get rid of the one at the end */
                                memset(EXT_LAST_EXTENT(eh), 0,
                                        sizeof(struct ext4_extent));
                        }
                        le16_add_cpu(&eh->eh_entries, -1);
                } else
                        *partial_cluster = 0;

                err = ext4_ext_dirty(handle, inode, path + depth);
                if (err)
                        goto out;

                ext_debug("new extent: %u:%u:%llu\n", ex_ee_block, num,
                                ext4_ext_pblock(ex));
                ex--;
                ex_ee_block = le32_to_cpu(ex->ee_block);
                ex_ee_len = ext4_ext_get_actual_len(ex);
        }

        if (correct_index && eh->eh_entries)
                err = ext4_ext_correct_indexes(handle, inode, path);

        /*
         * If there is still a entry in the leaf node, check to see if
         * it references the partial cluster.  This is the only place
         * where it could; if it doesn't, we can free the cluster.
         */
        if (*partial_cluster && ex >= EXT_FIRST_EXTENT(eh) &&
            (EXT4_B2C(sbi, ext4_ext_pblock(ex) + ex_ee_len - 1) !=
             *partial_cluster)) {
                int flags = EXT4_FREE_BLOCKS_FORGET;

                if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
                        flags |= EXT4_FREE_BLOCKS_METADATA;

                ext4_free_blocks(handle, inode, NULL,
                                 EXT4_C2B(sbi, *partial_cluster),
                                 sbi->s_cluster_ratio, flags);
                *partial_cluster = 0;
        }

        /* if this leaf is free, then we should
         * remove it from index block above */
        if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
                err = ext4_ext_rm_idx(handle, inode, path, depth);

out:
        return err;
}

/*
 * ext4_ext_more_to_rm:
 * returns 1 if current index has to be freed (even partial)
 */
static int
ext4_ext_more_to_rm(struct ext4_ext_path *path)
{
        BUG_ON(path->p_idx == NULL);

        if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
                return 0;

        /*
         * if truncate on deeper level happened, it wasn't partial,
         * so we have to consider current index for truncation
         */
        if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
                return 0;
        return 1;
}

static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
                                 ext4_lblk_t end)
{
        struct super_block *sb = inode->i_sb;
        int depth = ext_depth(inode);
        struct ext4_ext_path *path = NULL;
        ext4_fsblk_t partial_cluster = 0;
        handle_t *handle;
        int i = 0, err;

        ext_debug("truncate since %u to %u\n", start, end);

        /* probably first extent we're gonna free will be last in block */
        handle = ext4_journal_start(inode, depth + 1);
        if (IS_ERR(handle))
                return PTR_ERR(handle);

again:
        ext4_ext_invalidate_cache(inode);

        trace_ext4_ext_remove_space(inode, start, depth);

        /*
         * Check if we are removing extents inside the extent tree. If that
         * is the case, we are going to punch a hole inside the extent tree
         * so we have to check whether we need to split the extent covering
         * the last block to remove so we can easily remove the part of it
         * in ext4_ext_rm_leaf().
         */
        if (end < EXT_MAX_BLOCKS - 1) {
                struct ext4_extent *ex;
                ext4_lblk_t ee_block;

                /* find extent for this block */
                path = ext4_ext_find_extent(inode, end, NULL);
                if (IS_ERR(path)) {
                        ext4_journal_stop(handle);
                        return PTR_ERR(path);
                }
                depth = ext_depth(inode);
                ex = path[depth].p_ext;
                if (!ex) {
                        ext4_ext_drop_refs(path);
                        kfree(path);
                        path = NULL;
                        goto cont;
                }

                ee_block = le32_to_cpu(ex->ee_block);

                /*
                 * See if the last block is inside the extent, if so split
                 * the extent at 'end' block so we can easily remove the
                 * tail of the first part of the split extent in
                 * ext4_ext_rm_leaf().
                 */
                if (end >= ee_block &&
                    end < ee_block + ext4_ext_get_actual_len(ex) - 1) {
                        int split_flag = 0;

                        if (ext4_ext_is_uninitialized(ex))
                                split_flag = EXT4_EXT_MARK_UNINIT1 |
                                             EXT4_EXT_MARK_UNINIT2;

                        /*
                         * Split the extent in two so that 'end' is the last
                         * block in the first new extent
                         */
                        err = ext4_split_extent_at(handle, inode, path,
                                                end + 1, split_flag,
                                                EXT4_GET_BLOCKS_PRE_IO |
                                                EXT4_GET_BLOCKS_PUNCH_OUT_EXT);

                        if (err < 0)
                                goto out;
                }
        }
cont:

        /*
         * We start scanning from right side, freeing all the blocks
         * after i_size and walking into the tree depth-wise.
         */
        depth = ext_depth(inode);
        if (path) {
                int k = i = depth;
                while (--k > 0)
                        path[k].p_block =
                                le16_to_cpu(path[k].p_hdr->eh_entries)+1;
        } else {
                path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1),
                               GFP_NOFS);
                if (path == NULL) {
                        ext4_journal_stop(handle);
                        return -ENOMEM;
                }
                path[0].p_depth = depth;
                path[0].p_hdr = ext_inode_hdr(inode);
                i = 0;

                if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
                        err = -EIO;
                        goto out;
                }
        }
        err = 0;

        while (i >= 0 && err == 0) {
                if (i == depth) {
                        /* this is leaf block */
                        err = ext4_ext_rm_leaf(handle, inode, path,
                                               &partial_cluster, start,
                                               end);
                        /* root level has p_bh == NULL, brelse() eats this */
                        brelse(path[i].p_bh);
                        path[i].p_bh = NULL;
                        i--;
                        continue;
                }

                /* this is index block */
                if (!path[i].p_hdr) {
                        ext_debug("initialize header\n");
                        path[i].p_hdr = ext_block_hdr(path[i].p_bh);
                }

                if (!path[i].p_idx) {
                        /* this level hasn't been touched yet */
                        path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
                        path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
                        ext_debug("init index ptr: hdr 0x%p, num %d\n",
                                  path[i].p_hdr,
                                  le16_to_cpu(path[i].p_hdr->eh_entries));
                } else {
                        /* we were already here, see at next index */
                        path[i].p_idx--;
                }

                ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
                                i, EXT_FIRST_INDEX(path[i].p_hdr),
                                path[i].p_idx);
                if (ext4_ext_more_to_rm(path + i)) {
                        struct buffer_head *bh;
                        /* go to the next level */
                        ext_debug("move to level %d (block %llu)\n",
                                  i + 1, ext4_idx_pblock(path[i].p_idx));
                        memset(path + i + 1, 0, sizeof(*path));
                        bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
                        if (!bh) {
                                /* should we reset i_size? */
                                err = -EIO;
                                break;
                        }
                        if (WARN_ON(i + 1 > depth)) {
                                err = -EIO;
                                break;
                        }
                        if (ext4_ext_check(inode, ext_block_hdr(bh),
                                                        depth - i - 1)) {
                                err = -EIO;
                                break;
                        }
                        path[i + 1].p_bh = bh;

                        /* save actual number of indexes since this
                         * number is changed at the next iteration */
                        path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
                        i++;
                } else {
                        /* we finished processing this index, go up */
                        if (path[i].p_hdr->eh_entries == 0 && i > 0) {
                                /* index is empty, remove it;
                                 * handle must be already prepared by the
                                 * truncatei_leaf() */
                                err = ext4_ext_rm_idx(handle, inode, path, i);
                        }
                        /* root level has p_bh == NULL, brelse() eats this */
                        brelse(path[i].p_bh);
                        path[i].p_bh = NULL;
                        i--;
                        ext_debug("return to level %d\n", i);
                }
        }

        trace_ext4_ext_remove_space_done(inode, start, depth, partial_cluster,
                        path->p_hdr->eh_entries);

        /* If we still have something in the partial cluster and we have removed
         * even the first extent, then we should free the blocks in the partial
         * cluster as well. */
        if (partial_cluster && path->p_hdr->eh_entries == 0) {
                int flags = EXT4_FREE_BLOCKS_FORGET;

                if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
                        flags |= EXT4_FREE_BLOCKS_METADATA;

                ext4_free_blocks(handle, inode, NULL,
                                 EXT4_C2B(EXT4_SB(sb), partial_cluster),
                                 EXT4_SB(sb)->s_cluster_ratio, flags);
                partial_cluster = 0;
        }

        /* TODO: flexible tree reduction should be here */
        if (path->p_hdr->eh_entries == 0) {
                /*
                 * truncate to zero freed all the tree,
                 * so we need to correct eh_depth
                 */
                err = ext4_ext_get_access(handle, inode, path);
                if (err == 0) {
                        ext_inode_hdr(inode)->eh_depth = 0;
                        ext_inode_hdr(inode)->eh_max =
                                cpu_to_le16(ext4_ext_space_root(inode, 0));
                        err = ext4_ext_dirty(handle, inode, path);
                }
        }
out:
        ext4_ext_drop_refs(path);
        kfree(path);
        if (err == -EAGAIN) {
                path = NULL;
                goto again;
        }
        ext4_journal_stop(handle);

        return err;
}

/*
 * called at mount time
 */
void ext4_ext_init(struct super_block *sb)
{
        /*
         * possible initialization would be here
         */

        if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
#if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
                printk(KERN_INFO "EXT4-fs: file extents enabled");
#ifdef AGGRESSIVE_TEST
                printk(", aggressive tests");
#endif
#ifdef CHECK_BINSEARCH
                printk(", check binsearch");
#endif
#ifdef EXTENTS_STATS
                printk(", stats");
#endif
                printk("\n");
#endif
#ifdef EXTENTS_STATS
                spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
                EXT4_SB(sb)->s_ext_min = 1 << 30;
                EXT4_SB(sb)->s_ext_max = 0;
#endif
        }
}

/*
 * called at umount time
 */
void ext4_ext_release(struct super_block *sb)
{
        if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
                return;

#ifdef EXTENTS_STATS
        if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
                struct ext4_sb_info *sbi = EXT4_SB(sb);
                printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
                        sbi->s_ext_blocks, sbi->s_ext_extents,
                        sbi->s_ext_blocks / sbi->s_ext_extents);
                printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
                        sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
        }
#endif
}

/* FIXME!! we need to try to merge to left or right after zero-out  */
static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
{
        ext4_fsblk_t ee_pblock;
        unsigned int ee_len;
        int ret;

        ee_len    = ext4_ext_get_actual_len(ex);
        ee_pblock = ext4_ext_pblock(ex);

        ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
        if (ret > 0)
                ret = 0;

        return ret;
}

/*
 * ext4_split_extent_at() splits an extent at given block.
 *
 * @handle: the journal handle
 * @inode: the file inode
 * @path: the path to the extent
 * @split: the logical block where the extent is splitted.
 * @split_flags: indicates if the extent could be zeroout if split fails, and
 *               the states(init or uninit) of new extents.
 * @flags: flags used to insert new extent to extent tree.
 *
 *
 * Splits extent [a, b] into two extents [a, @split) and [@split, b], states
 * of which are deterimined by split_flag.
 *
 * There are two cases:
 *  a> the extent are splitted into two extent.
 *  b> split is not needed, and just mark the extent.
 *
 * return 0 on success.
 */
static int ext4_split_extent_at(handle_t *handle,
                             struct inode *inode,
                             struct ext4_ext_path *path,
                             ext4_lblk_t split,
                             int split_flag,
                             int flags)
{
        ext4_fsblk_t newblock;
        ext4_lblk_t ee_block;
        struct ext4_extent *ex, newex, orig_ex;
        struct ext4_extent *ex2 = NULL;
        unsigned int ee_len, depth;
        int err = 0;

        BUG_ON((split_flag & (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2)) ==
               (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2));

        ext_debug("ext4_split_extents_at: inode %lu, logical"
                "block %llu\n", inode->i_ino, (unsigned long long)split);

        ext4_ext_show_leaf(inode, path);

        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        ee_block = le32_to_cpu(ex->ee_block);
        ee_len = ext4_ext_get_actual_len(ex);
        newblock = split - ee_block + ext4_ext_pblock(ex);

        BUG_ON(split < ee_block || split >= (ee_block + ee_len));

        err = ext4_ext_get_access(handle, inode, path + depth);
        if (err)
                goto out;

        if (split == ee_block) {
                /*
                 * case b: block @split is the block that the extent begins with
                 * then we just change the state of the extent, and splitting
                 * is not needed.
                 */
                if (split_flag & EXT4_EXT_MARK_UNINIT2)
                        ext4_ext_mark_uninitialized(ex);
                else
                        ext4_ext_mark_initialized(ex);

                if (!(flags & EXT4_GET_BLOCKS_PRE_IO))
                        ext4_ext_try_to_merge(inode, path, ex);

                err = ext4_ext_dirty(handle, inode, path + depth);
                goto out;
        }

        /* case a */
        memcpy(&orig_ex, ex, sizeof(orig_ex));
        ex->ee_len = cpu_to_le16(split - ee_block);
        if (split_flag & EXT4_EXT_MARK_UNINIT1)
                ext4_ext_mark_uninitialized(ex);

        /*
         * path may lead to new leaf, not to original leaf any more
         * after ext4_ext_insert_extent() returns,
         */
        err = ext4_ext_dirty(handle, inode, path + depth);
        if (err)
                goto fix_extent_len;

        ex2 = &newex;
        ex2->ee_block = cpu_to_le32(split);
        ex2->ee_len   = cpu_to_le16(ee_len - (split - ee_block));
        ext4_ext_store_pblock(ex2, newblock);
        if (split_flag & EXT4_EXT_MARK_UNINIT2)
                ext4_ext_mark_uninitialized(ex2);

        err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
        if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
                if (split_flag & (EXT4_EXT_DATA_VALID1|EXT4_EXT_DATA_VALID2)) {
                        if (split_flag & EXT4_EXT_DATA_VALID1)
                                err = ext4_ext_zeroout(inode, ex2);
                        else
                                err = ext4_ext_zeroout(inode, ex);
                } else
                        err = ext4_ext_zeroout(inode, &orig_ex);

                if (err)
                        goto fix_extent_len;
                /* update the extent length and mark as initialized */
                ex->ee_len = cpu_to_le16(ee_len);
                ext4_ext_try_to_merge(inode, path, ex);
                err = ext4_ext_dirty(handle, inode, path + depth);
                goto out;
        } else if (err)
                goto fix_extent_len;

out:
        ext4_ext_show_leaf(inode, path);
        return err;

fix_extent_len:
        ex->ee_len = orig_ex.ee_len;
        ext4_ext_dirty(handle, inode, path + depth);
        return err;
}

/*
 * ext4_split_extents() splits an extent and mark extent which is covered
 * by @map as split_flags indicates
 *
 * It may result in splitting the extent into multiple extents (upto three)
 * There are three possibilities:
 *   a> There is no split required
 *   b> Splits in two extents: Split is happening at either end of the extent
 *   c> Splits in three extents: Somone is splitting in middle of the extent
 *
 */
static int ext4_split_extent(handle_t *handle,
                              struct inode *inode,
                              struct ext4_ext_path *path,
                              struct ext4_map_blocks *map,
                              int split_flag,
                              int flags)
{
        ext4_lblk_t ee_block;
        struct ext4_extent *ex;
        unsigned int ee_len, depth;
        int err = 0;
        int uninitialized;
        int split_flag1, flags1;
        int allocated = map->m_len;

        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        ee_block = le32_to_cpu(ex->ee_block);
        ee_len = ext4_ext_get_actual_len(ex);
        uninitialized = ext4_ext_is_uninitialized(ex);

        if (map->m_lblk + map->m_len < ee_block + ee_len) {
                split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT;
                flags1 = flags | EXT4_GET_BLOCKS_PRE_IO;
                if (uninitialized)
                        split_flag1 |= EXT4_EXT_MARK_UNINIT1 |
                                       EXT4_EXT_MARK_UNINIT2;
                if (split_flag & EXT4_EXT_DATA_VALID2)
                        split_flag1 |= EXT4_EXT_DATA_VALID1;
                err = ext4_split_extent_at(handle, inode, path,
                                map->m_lblk + map->m_len, split_flag1, flags1);
                if (err)
                        goto out;
        } else {
                allocated = ee_len - (map->m_lblk - ee_block);
        }

        ext4_ext_drop_refs(path);
        path = ext4_ext_find_extent(inode, map->m_lblk, path);
        if (IS_ERR(path))
                return PTR_ERR(path);

        if (map->m_lblk >= ee_block) {
                split_flag1 = split_flag & (EXT4_EXT_MAY_ZEROOUT |
                                            EXT4_EXT_DATA_VALID2);
                if (uninitialized)
                        split_flag1 |= EXT4_EXT_MARK_UNINIT1;
                if (split_flag & EXT4_EXT_MARK_UNINIT2)
                        split_flag1 |= EXT4_EXT_MARK_UNINIT2;
                err = ext4_split_extent_at(handle, inode, path,
                                map->m_lblk, split_flag1, flags);
                if (err)
                        goto out;
        }

        ext4_ext_show_leaf(inode, path);
out:
        return err ? err : allocated;
}

#define EXT4_EXT_ZERO_LEN 7
/*
 * This function is called by ext4_ext_map_blocks() if someone tries to write
 * to an uninitialized extent. It may result in splitting the uninitialized
 * extent into multiple extents (up to three - one initialized and two
 * uninitialized).
 * There are three possibilities:
 *   a> There is no split required: Entire extent should be initialized
 *   b> Splits in two extents: Write is happening at either end of the extent
 *   c> Splits in three extents: Somone is writing in middle of the extent
 *
 * Pre-conditions:
 *  - The extent pointed to by 'path' is uninitialized.
 *  - The extent pointed to by 'path' contains a superset
 *    of the logical span [map->m_lblk, map->m_lblk + map->m_len).
 *
 * Post-conditions on success:
 *  - the returned value is the number of blocks beyond map->l_lblk
 *    that are allocated and initialized.
 *    It is guaranteed to be >= map->m_len.
 */
static int ext4_ext_convert_to_initialized(handle_t *handle,
                                           struct inode *inode,
                                           struct ext4_map_blocks *map,
                                           struct ext4_ext_path *path)
{
        struct ext4_extent_header *eh;
        struct ext4_map_blocks split_map;
        struct ext4_extent zero_ex;
        struct ext4_extent *ex;
        ext4_lblk_t ee_block, eof_block;
        unsigned int ee_len, depth;
        int allocated;
        int err = 0;
        int split_flag = 0;

        ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
                "block %llu, max_blocks %u\n", inode->i_ino,
                (unsigned long long)map->m_lblk, map->m_len);

        eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
                inode->i_sb->s_blocksize_bits;
        if (eof_block < map->m_lblk + map->m_len)
                eof_block = map->m_lblk + map->m_len;

        depth = ext_depth(inode);
        eh = path[depth].p_hdr;
        ex = path[depth].p_ext;
        ee_block = le32_to_cpu(ex->ee_block);
        ee_len = ext4_ext_get_actual_len(ex);
        allocated = ee_len - (map->m_lblk - ee_block);

        trace_ext4_ext_convert_to_initialized_enter(inode, map, ex);

        /* Pre-conditions */
        BUG_ON(!ext4_ext_is_uninitialized(ex));
        BUG_ON(!in_range(map->m_lblk, ee_block, ee_len));

        /*
         * Attempt to transfer newly initialized blocks from the currently
         * uninitialized extent to its left neighbor. This is much cheaper
         * than an insertion followed by a merge as those involve costly
         * memmove() calls. This is the common case in steady state for
         * workloads doing fallocate(FALLOC_FL_KEEP_SIZE) followed by append
         * writes.
         *
         * Limitations of the current logic:
         *  - L1: we only deal with writes at the start of the extent.
         *    The approach could be extended to writes at the end
         *    of the extent but this scenario was deemed less common.
         *  - L2: we do not deal with writes covering the whole extent.
         *    This would require removing the extent if the transfer
         *    is possible.
         *  - L3: we only attempt to merge with an extent stored in the
         *    same extent tree node.
         */
        if ((map->m_lblk == ee_block) &&        /*L1*/
                (map->m_len < ee_len) &&        /*L2*/
                (ex > EXT_FIRST_EXTENT(eh))) {  /*L3*/
                struct ext4_extent *prev_ex;
                ext4_lblk_t prev_lblk;
                ext4_fsblk_t prev_pblk, ee_pblk;
                unsigned int prev_len, write_len;

                prev_ex = ex - 1;
                prev_lblk = le32_to_cpu(prev_ex->ee_block);
                prev_len = ext4_ext_get_actual_len(prev_ex);
                prev_pblk = ext4_ext_pblock(prev_ex);
                ee_pblk = ext4_ext_pblock(ex);
                write_len = map->m_len;

                /*
                 * A transfer of blocks from 'ex' to 'prev_ex' is allowed
                 * upon those conditions:
                 * - C1: prev_ex is initialized,
                 * - C2: prev_ex is logically abutting ex,
                 * - C3: prev_ex is physically abutting ex,
                 * - C4: prev_ex can receive the additional blocks without
                 *   overflowing the (initialized) length limit.
                 */
                if ((!ext4_ext_is_uninitialized(prev_ex)) &&            /*C1*/
                        ((prev_lblk + prev_len) == ee_block) &&         /*C2*/
                        ((prev_pblk + prev_len) == ee_pblk) &&          /*C3*/
                        (prev_len < (EXT_INIT_MAX_LEN - write_len))) {  /*C4*/
                        err = ext4_ext_get_access(handle, inode, path + depth);
                        if (err)
                                goto out;

                        trace_ext4_ext_convert_to_initialized_fastpath(inode,
                                map, ex, prev_ex);

                        /* Shift the start of ex by 'write_len' blocks */
                        ex->ee_block = cpu_to_le32(ee_block + write_len);
                        ext4_ext_store_pblock(ex, ee_pblk + write_len);
                        ex->ee_len = cpu_to_le16(ee_len - write_len);
                        ext4_ext_mark_uninitialized(ex); /* Restore the flag */

                        /* Extend prev_ex by 'write_len' blocks */
                        prev_ex->ee_len = cpu_to_le16(prev_len + write_len);

                        /* Mark the block containing both extents as dirty */
                        ext4_ext_dirty(handle, inode, path + depth);

                        /* Update path to point to the right extent */
                        path[depth].p_ext = prev_ex;

                        /* Result: number of initialized blocks past m_lblk */
                        allocated = write_len;
                        goto out;
                }
        }

        WARN_ON(map->m_lblk < ee_block);
        /*
         * It is safe to convert extent to initialized via explicit
         * zeroout only if extent is fully insde i_size or new_size.
         */
        split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;

        /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
        if (ee_len <= 2*EXT4_EXT_ZERO_LEN &&
            (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
                err = ext4_ext_zeroout(inode, ex);
                if (err)
                        goto out;

                err = ext4_ext_get_access(handle, inode, path + depth);
                if (err)
                        goto out;
                ext4_ext_mark_initialized(ex);
                ext4_ext_try_to_merge(inode, path, ex);
                err = ext4_ext_dirty(handle, inode, path + depth);
                goto out;
        }

        /*
         * four cases:
         * 1. split the extent into three extents.
         * 2. split the extent into two extents, zeroout the first half.
         * 3. split the extent into two extents, zeroout the second half.
         * 4. split the extent into two extents with out zeroout.
         */
        split_map.m_lblk = map->m_lblk;
        split_map.m_len = map->m_len;

        if (allocated > map->m_len) {
                if (allocated <= EXT4_EXT_ZERO_LEN &&
                    (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
                        /* case 3 */
                        zero_ex.ee_block =
                                         cpu_to_le32(map->m_lblk);
                        zero_ex.ee_len = cpu_to_le16(allocated);
                        ext4_ext_store_pblock(&zero_ex,
                                ext4_ext_pblock(ex) + map->m_lblk - ee_block);
                        err = ext4_ext_zeroout(inode, &zero_ex);
                        if (err)
                                goto out;
                        split_map.m_lblk = map->m_lblk;
                        split_map.m_len = allocated;
                } else if ((map->m_lblk - ee_block + map->m_len <
                           EXT4_EXT_ZERO_LEN) &&
                           (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
                        /* case 2 */
                        if (map->m_lblk != ee_block) {
                                zero_ex.ee_block = ex->ee_block;
                                zero_ex.ee_len = cpu_to_le16(map->m_lblk -
                                                        ee_block);
                                ext4_ext_store_pblock(&zero_ex,
                                                      ext4_ext_pblock(ex));
                                err = ext4_ext_zeroout(inode, &zero_ex);
                                if (err)
                                        goto out;
                        }

                        split_map.m_lblk = ee_block;
                        split_map.m_len = map->m_lblk - ee_block + map->m_len;
                        allocated = map->m_len;
                }
        }

        allocated = ext4_split_extent(handle, inode, path,
                                       &split_map, split_flag, 0);
        if (allocated < 0)
                err = allocated;

out:
        return err ? err : allocated;
}

/*
 * This function is called by ext4_ext_map_blocks() from
 * ext4_get_blocks_dio_write() when DIO to write
 * to an uninitialized extent.
 *
 * Writing to an uninitialized extent may result in splitting the uninitialized
 * extent into multiple /initialized uninitialized extents (up to three)
 * There are three possibilities:
 *   a> There is no split required: Entire extent should be uninitialized
 *   b> Splits in two extents: Write is happening at either end of the extent
 *   c> Splits in three extents: Somone is writing in middle of the extent
 *
 * One of more index blocks maybe needed if the extent tree grow after
 * the uninitialized extent split. To prevent ENOSPC occur at the IO
 * complete, we need to split the uninitialized extent before DIO submit
 * the IO. The uninitialized extent called at this time will be split
 * into three uninitialized extent(at most). After IO complete, the part
 * being filled will be convert to initialized by the end_io callback function
 * via ext4_convert_unwritten_extents().
 *
 * Returns the size of uninitialized extent to be written on success.
 */
static int ext4_split_unwritten_extents(handle_t *handle,
                                        struct inode *inode,
                                        struct ext4_map_blocks *map,
                                        struct ext4_ext_path *path,
                                        int flags)
{
        ext4_lblk_t eof_block;
        ext4_lblk_t ee_block;
        struct ext4_extent *ex;
        unsigned int ee_len;
        int split_flag = 0, depth;

        ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
                "block %llu, max_blocks %u\n", inode->i_ino,
                (unsigned long long)map->m_lblk, map->m_len);

        eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
                inode->i_sb->s_blocksize_bits;
        if (eof_block < map->m_lblk + map->m_len)
                eof_block = map->m_lblk + map->m_len;
        /*
         * It is safe to convert extent to initialized via explicit
         * zeroout only if extent is fully insde i_size or new_size.
         */
        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        ee_block = le32_to_cpu(ex->ee_block);
        ee_len = ext4_ext_get_actual_len(ex);

        split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0;
        split_flag |= EXT4_EXT_MARK_UNINIT2;
        if (flags & EXT4_GET_BLOCKS_CONVERT)
                split_flag |= EXT4_EXT_DATA_VALID2;
        flags |= EXT4_GET_BLOCKS_PRE_IO;
        return ext4_split_extent(handle, inode, path, map, split_flag, flags);
}

static int ext4_convert_unwritten_extents_endio(handle_t *handle,
                                                struct inode *inode,
                                                struct ext4_map_blocks *map,
                                                struct ext4_ext_path *path)
{
        struct ext4_extent *ex;
        ext4_lblk_t ee_block;
        unsigned int ee_len;
        int depth;
        int err = 0;

        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        ee_block = le32_to_cpu(ex->ee_block);
        ee_len = ext4_ext_get_actual_len(ex);

        ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
                "block %llu, max_blocks %u\n", inode->i_ino,
                  (unsigned long long)ee_block, ee_len);

        /* If extent is larger than requested then split is required */
        if (ee_block != map->m_lblk || ee_len > map->m_len) {
                err = ext4_split_unwritten_extents(handle, inode, map, path,
                                                   EXT4_GET_BLOCKS_CONVERT);
                if (err < 0)
                        goto out;
                ext4_ext_drop_refs(path);
                path = ext4_ext_find_extent(inode, map->m_lblk, path);
                if (IS_ERR(path)) {
                        err = PTR_ERR(path);
                        goto out;
                }
                depth = ext_depth(inode);
                ex = path[depth].p_ext;
        }

        err = ext4_ext_get_access(handle, inode, path + depth);
        if (err)
                goto out;
        /* first mark the extent as initialized */
        ext4_ext_mark_initialized(ex);

        /* note: ext4_ext_correct_indexes() isn't needed here because
         * borders are not changed
         */
        ext4_ext_try_to_merge(inode, path, ex);

        /* Mark modified extent as dirty */
        err = ext4_ext_dirty(handle, inode, path + depth);
out:
        ext4_ext_show_leaf(inode, path);
        return err;
}

static void unmap_underlying_metadata_blocks(struct block_device *bdev,
                        sector_t block, int count)
{
        int i;
        for (i = 0; i < count; i++)
                unmap_underlying_metadata(bdev, block + i);
}

/*
 * Handle EOFBLOCKS_FL flag, clearing it if necessary
 */
static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
                              ext4_lblk_t lblk,
                              struct ext4_ext_path *path,
                              unsigned int len)
{
        int i, depth;
        struct ext4_extent_header *eh;
        struct ext4_extent *last_ex;

        if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
                return 0;

        depth = ext_depth(inode);
        eh = path[depth].p_hdr;

        if (unlikely(!eh->eh_entries)) {
                EXT4_ERROR_INODE(inode, "eh->eh_entries == 0 and "
                                 "EOFBLOCKS_FL set");
                return -EIO;
        }
        last_ex = EXT_LAST_EXTENT(eh);
        /*
         * We should clear the EOFBLOCKS_FL flag if we are writing the
         * last block in the last extent in the file.  We test this by
         * first checking to see if the caller to
         * ext4_ext_get_blocks() was interested in the last block (or
         * a block beyond the last block) in the current extent.  If
         * this turns out to be false, we can bail out from this
         * function immediately.
         */
        if (lblk + len < le32_to_cpu(last_ex->ee_block) +
            ext4_ext_get_actual_len(last_ex))
                return 0;
        /*
         * If the caller does appear to be planning to write at or
         * beyond the end of the current extent, we then test to see
         * if the current extent is the last extent in the file, by
         * checking to make sure it was reached via the rightmost node
         * at each level of the tree.
         */
        for (i = depth-1; i >= 0; i--)
                if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
                        return 0;
        ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
        return ext4_mark_inode_dirty(handle, inode);
}

/**
 * ext4_find_delalloc_range: find delayed allocated block in the given range.
 *
 * Goes through the buffer heads in the range [lblk_start, lblk_end] and returns
 * whether there are any buffers marked for delayed allocation. It returns '1'
 * on the first delalloc'ed buffer head found. If no buffer head in the given
 * range is marked for delalloc, it returns 0.
 * lblk_start should always be <= lblk_end.
 * search_hint_reverse is to indicate that searching in reverse from lblk_end to
 * lblk_start might be more efficient (i.e., we will likely hit the delalloc'ed
 * block sooner). This is useful when blocks are truncated sequentially from
 * lblk_start towards lblk_end.
 */
static int ext4_find_delalloc_range(struct inode *inode,
                                    ext4_lblk_t lblk_start,
                                    ext4_lblk_t lblk_end,
                                    int search_hint_reverse)
{
        struct address_space *mapping = inode->i_mapping;
        struct buffer_head *head, *bh = NULL;
        struct page *page;
        ext4_lblk_t i, pg_lblk;
        pgoff_t index;

        /* reverse search wont work if fs block size is less than page size */
        if (inode->i_blkbits < PAGE_CACHE_SHIFT)
                search_hint_reverse = 0;

        if (search_hint_reverse)
                i = lblk_end;
        else
                i = lblk_start;

        index = i >> (PAGE_CACHE_SHIFT - inode->i_blkbits);

        while ((i >= lblk_start) && (i <= lblk_end)) {
                page = find_get_page(mapping, index);
                if (!page)
                        goto nextpage;

                if (!page_has_buffers(page))
                        goto nextpage;

                head = page_buffers(page);
                if (!head)
                        goto nextpage;

                bh = head;
                pg_lblk = index << (PAGE_CACHE_SHIFT -
                                                inode->i_blkbits);
                do {
                        if (unlikely(pg_lblk < lblk_start)) {
                                /*
                                 * This is possible when fs block size is less
                                 * than page size and our cluster starts/ends in
                                 * middle of the page. So we need to skip the
                                 * initial few blocks till we reach the 'lblk'
                                 */
                                pg_lblk++;
                                continue;
                        }

                        /* Check if the buffer is delayed allocated and that it
                         * is not yet mapped. (when da-buffers are mapped during
                         * their writeout, their da_mapped bit is set.)
                         */
                        if (buffer_delay(bh) && !buffer_da_mapped(bh)) {
                                page_cache_release(page);
                                trace_ext4_find_delalloc_range(inode,
                                                lblk_start, lblk_end,
                                                search_hint_reverse,
                                                1, i);
                                return 1;
                        }
                        if (search_hint_reverse)
                                i--;
                        else
                                i++;
                } while ((i >= lblk_start) && (i <= lblk_end) &&
                                ((bh = bh->b_this_page) != head));
nextpage:
                if (page)
                        page_cache_release(page);
                /*
                 * Move to next page. 'i' will be the first lblk in the next
                 * page.
                 */
                if (search_hint_reverse)
                        index--;
                else
                        index++;
                i = index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
        }

        trace_ext4_find_delalloc_range(inode, lblk_start, lblk_end,
                                        search_hint_reverse, 0, 0);
        return 0;
}

int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk,
                               int search_hint_reverse)
{
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        ext4_lblk_t lblk_start, lblk_end;
        lblk_start = EXT4_LBLK_CMASK(sbi, lblk);
        lblk_end = lblk_start + sbi->s_cluster_ratio - 1;

        return ext4_find_delalloc_range(inode, lblk_start, lblk_end,
                                        search_hint_reverse);
}

/**
 * Determines how many complete clusters (out of those specified by the 'map')
 * are under delalloc and were reserved quota for.
 * This function is called when we are writing out the blocks that were
 * originally written with their allocation delayed, but then the space was