// SPDX-License-Identifier: GPL-2.0
/*
 * 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>
 */

/*
 * 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/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/uaccess.h>
#include <linux/fiemap.h>
#include <linux/backing-dev.h>
#include "ext4_jbd2.h"
#include "ext4_extents.h"
#include "xattr.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_UNWRIT1   0x2  /* mark first half unwritten */
#define EXT4_EXT_MARK_UNWRIT2   0x4  /* mark second half unwritten */

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

static __le32 ext4_extent_block_csum(struct inode *inode,
                                     struct ext4_extent_header *eh)
{
        struct ext4_inode_info *ei = EXT4_I(inode);
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        __u32 csum;

        csum = ext4_chksum(sbi, ei->i_csum_seed, (__u8 *)eh,
                           EXT4_EXTENT_TAIL_OFFSET(eh));
        return cpu_to_le32(csum);
}

static int ext4_extent_block_csum_verify(struct inode *inode,
                                         struct ext4_extent_header *eh)
{
        struct ext4_extent_tail *et;

        if (!ext4_has_metadata_csum(inode->i_sb))
                return 1;

        et = find_ext4_extent_tail(eh);
        if (et->et_checksum != ext4_extent_block_csum(inode, eh))
                return 0;
        return 1;
}

static void ext4_extent_block_csum_set(struct inode *inode,
                                       struct ext4_extent_header *eh)
{
        struct ext4_extent_tail *et;

        if (!ext4_has_metadata_csum(inode->i_sb))
                return;

        et = find_ext4_extent_tail(eh);
        et->et_checksum = ext4_extent_block_csum(inode, eh);
}

static int ext4_split_extent(handle_t *handle,
                                struct inode *inode,
                                struct ext4_ext_path **ppath,
                                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 **ppath,
                             ext4_lblk_t split,
                             int split_flag,
                             int flags);

static int ext4_find_delayed_extent(struct inode *inode,
                                    struct extent_status *newes);

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;
        /*
         * If we need to extend the journal get a few extra blocks
         * while we're at it for efficiency's sake.
         */
        needed += 3;
        err = ext4_journal_extend(handle, needed - handle->h_buffer_credits);
        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 */
                BUFFER_TRACE(path->p_bh, "get_write_access");
                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
 */
int __ext4_ext_dirty(const char *where, unsigned int line, handle_t *handle,
                     struct inode *inode, struct ext4_ext_path *path)
{
        int err;

        WARN_ON(!rwsem_is_locked(&EXT4_I(inode)->i_data_sem));
        if (path->p_bh) {
                ext4_extent_block_csum_set(inode, ext_block_hdr(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;
}

static inline int
ext4_force_split_extent_at(handle_t *handle, struct inode *inode,
                           struct ext4_ext_path **ppath, ext4_lblk_t lblk,
                           int nofail)
{
        struct ext4_ext_path *path = *ppath;
        int unwritten = ext4_ext_is_unwritten(path[path->p_depth].p_ext);

        return ext4_split_extent_at(handle, inode, ppath, lblk, unwritten ?
                        EXT4_EXT_MARK_UNWRIT1|EXT4_EXT_MARK_UNWRIT2 : 0,
                        EXT4_EX_NOCACHE | EXT4_GET_BLOCKS_PRE_IO |
                        (nofail ? EXT4_GET_BLOCKS_METADATA_NOFAIL:0));
}

/*
 * 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);

        /*
         * We allow neither:
         *  - zero length
         *  - overflow/wrap-around
         */
        if (lblock + len <= lblock)
                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, ext4_fsblk_t pblk)
{
        const char *error_msg;
        int max = 0, err = -EFSCORRUPTED;

        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;
        }
        if (unlikely(depth > 32)) {
                error_msg = "too large eh_depth";
                goto corrupted;
        }
        /* Verify checksum on non-root extent tree nodes */
        if (ext_depth(inode) != depth &&
            !ext4_extent_block_csum_verify(inode, eh)) {
                error_msg = "extent tree corrupted";
                err = -EFSBADCRC;
                goto corrupted;
        }
        return 0;

corrupted:
        ext4_error_inode(inode, function, line, 0,
                         "pblk %llu bad header/extent: %s - magic %x, "
                         "entries %u, max %u(%u), depth %u(%u)",
                         (unsigned long long) pblk, 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 err;
}

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

int ext4_ext_check_inode(struct inode *inode)
{
        return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode), 0);
}

static struct buffer_head *
__read_extent_tree_block(const char *function, unsigned int line,
                         struct inode *inode, ext4_fsblk_t pblk, int depth,
                         int flags)
{
        struct buffer_head              *bh;
        int                             err;

        bh = sb_getblk_gfp(inode->i_sb, pblk, __GFP_MOVABLE | GFP_NOFS);
        if (unlikely(!bh))
                return ERR_PTR(-ENOMEM);

        if (!bh_uptodate_or_lock(bh)) {
                trace_ext4_ext_load_extent(inode, pblk, _RET_IP_);
                err = bh_submit_read(bh);
                if (err < 0)
                        goto errout;
        }
        if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE))
                return bh;
        err = __ext4_ext_check(function, line, inode,
                               ext_block_hdr(bh), depth, pblk);
        if (err)
                goto errout;
        set_buffer_verified(bh);
        /*
         * If this is a leaf block, cache all of its entries
         */
        if (!(flags & EXT4_EX_NOCACHE) && depth == 0) {
                struct ext4_extent_header *eh = ext_block_hdr(bh);
                struct ext4_extent *ex = EXT_FIRST_EXTENT(eh);
                ext4_lblk_t prev = 0;
                int i;

                for (i = le16_to_cpu(eh->eh_entries); i > 0; i--, ex++) {
                        unsigned int status = EXTENT_STATUS_WRITTEN;
                        ext4_lblk_t lblk = le32_to_cpu(ex->ee_block);
                        int len = ext4_ext_get_actual_len(ex);

                        if (prev && (prev != lblk))
                                ext4_es_cache_extent(inode, prev,
                                                     lblk - prev, ~0,
                                                     EXTENT_STATUS_HOLE);

                        if (ext4_ext_is_unwritten(ex))
                                status = EXTENT_STATUS_UNWRITTEN;
                        ext4_es_cache_extent(inode, lblk, len,
                                             ext4_ext_pblock(ex), status);
                        prev = lblk + len;
                }
        }
        return bh;
errout:
        put_bh(bh);
        return ERR_PTR(err);

}

#define read_extent_tree_block(inode, pblk, depth, flags)               \
        __read_extent_tree_block(__func__, __LINE__, (inode), (pblk),   \
                                 (depth), (flags))

/*
 * This function is called to cache a file's extent information in the
 * extent status tree
 */
int ext4_ext_precache(struct inode *inode)
{
        struct ext4_inode_info *ei = EXT4_I(inode);
        struct ext4_ext_path *path = NULL;
        struct buffer_head *bh;
        int i = 0, depth, ret = 0;

        if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
                return 0;       /* not an extent-mapped inode */

        down_read(&ei->i_data_sem);
        depth = ext_depth(inode);

        path = kcalloc(depth + 1, sizeof(struct ext4_ext_path),
                       GFP_NOFS);
        if (path == NULL) {
                up_read(&ei->i_data_sem);
                return -ENOMEM;
        }

        /* Don't cache anything if there are no external extent blocks */
        if (depth == 0)
                goto out;
        path[0].p_hdr = ext_inode_hdr(inode);
        ret = ext4_ext_check(inode, path[0].p_hdr, depth, 0);
        if (ret)
                goto out;
        path[0].p_idx = EXT_FIRST_INDEX(path[0].p_hdr);
        while (i >= 0) {
                /*
                 * If this is a leaf block or we've reached the end of
                 * the index block, go up
                 */
                if ((i == depth) ||
                    path[i].p_idx > EXT_LAST_INDEX(path[i].p_hdr)) {
                        brelse(path[i].p_bh);
                        path[i].p_bh = NULL;
                        i--;
                        continue;
                }
                bh = read_extent_tree_block(inode,
                                            ext4_idx_pblock(path[i].p_idx++),
                                            depth - i - 1,
                                            EXT4_EX_FORCE_CACHE);
                if (IS_ERR(bh)) {
                        ret = PTR_ERR(bh);
                        break;
                }
                i++;
                path[i].p_bh = bh;
                path[i].p_hdr = ext_block_hdr(bh);
                path[i].p_idx = EXT_FIRST_INDEX(path[i].p_hdr);
        }
        ext4_set_inode_state(inode, EXT4_STATE_EXT_PRECACHED);
out:
        up_read(&ei->i_data_sem);
        ext4_ext_drop_refs(path);
        kfree(path);
        return ret;
}

#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_unwritten(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_unwritten(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_unwritten(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, i;

        if (!path)
                return;
        depth = path->p_depth;
        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("  -> %u->%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_unwritten(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);
        return 0;
}

struct ext4_ext_path *
ext4_find_extent(struct inode *inode, ext4_lblk_t block,
                 struct ext4_ext_path **orig_path, int flags)
{
        struct ext4_extent_header *eh;
        struct buffer_head *bh;
        struct ext4_ext_path *path = orig_path ? *orig_path : NULL;
        short int depth, i, ppos = 0;
        int ret;

        eh = ext_inode_hdr(inode);
        depth = ext_depth(inode);
        if (depth < 0 || depth > EXT4_MAX_EXTENT_DEPTH) {
                EXT4_ERROR_INODE(inode, "inode has invalid extent depth: %d",
                                 depth);
                ret = -EFSCORRUPTED;
                goto err;
        }

        if (path) {
                ext4_ext_drop_refs(path);
                if (depth > path[0].p_maxdepth) {
                        kfree(path);
                        *orig_path = path = NULL;
                }
        }
        if (!path) {
                /* account possible depth increase */
                path = kcalloc(depth + 2, sizeof(struct ext4_ext_path),
                                GFP_NOFS);
                if (unlikely(!path))
                        return ERR_PTR(-ENOMEM);
                path[0].p_maxdepth = depth + 1;
        }
        path[0].p_hdr = eh;
        path[0].p_bh = NULL;

        i = depth;
        /* walk through the tree */
        while (i) {
                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 = read_extent_tree_block(inode, path[ppos].p_block, --i,
                                            flags);
                if (IS_ERR(bh)) {
                        ret = PTR_ERR(bh);
                        goto err;
                }

                eh = ext_block_hdr(bh);
                ppos++;
                path[ppos].p_bh = bh;
                path[ppos].p_hdr = eh;
        }

        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);
        kfree(path);
        if (orig_path)
                *orig_path = NULL;
        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 -EFSCORRUPTED;
        }

        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 -EFSCORRUPTED;
        }

        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 -EFSCORRUPTED;
        }

        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 -EFSCORRUPTED;
        }

        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 -EFSCORRUPTED;
        }
        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 = kcalloc(depth, sizeof(ext4_fsblk_t), 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 = -EFSCORRUPTED;
                goto cleanup;
        }
        bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS);
        if (unlikely(!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 = -EFSCORRUPTED;
                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);
        }

        ext4_extent_block_csum_set(inode, neh);
        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 = -EFSCORRUPTED;
                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 (unlikely(!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 = -EFSCORRUPTED;
                        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);
                }
                ext4_extent_block_csum_set(inode, neh);
                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_header *neh;
        struct buffer_head *bh;
        ext4_fsblk_t newblock, goal = 0;
        struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
        int err = 0;

        /* Try to prepend new index to old one */
        if (ext_depth(inode))
                goal = ext4_idx_pblock(EXT_FIRST_INDEX(ext_inode_hdr(inode)));
        if (goal > le32_to_cpu(es->s_first_data_block)) {
                flags |= EXT4_MB_HINT_TRY_GOAL;
                goal--;
        } else
                goal = ext4_inode_to_goal_block(inode);
        newblock = ext4_new_meta_blocks(handle, inode, goal, flags,
                                        NULL, &err);
        if (newblock == 0)
                return err;

        bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS);
        if (unlikely(!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;
        ext4_extent_block_csum_set(inode, neh);
        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)));

        le16_add_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 mb_flags,
                                    unsigned int gb_flags,
                                    struct ext4_ext_path **ppath,
                                    struct ext4_extent *newext)
{
        struct ext4_ext_path *path = *ppath;
        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, mb_flags, path, newext, i);
                if (err)
                        goto out;

                /* refill path */
                path = ext4_find_extent(inode,
                                    (ext4_lblk_t)le32_to_cpu(newext->ee_block),
                                    ppath, gb_flags);
                if (IS_ERR(path))
                        err = PTR_ERR(path);
        } else {
                /* tree is full, time to grow in depth */
                err = ext4_ext_grow_indepth(handle, inode, mb_flags);
                if (err)
                        goto out;

                /* refill path */
                path = ext4_find_extent(inode,
                                   (ext4_lblk_t)le32_to_cpu(newext->ee_block),
                                    ppath, gb_flags);
                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 -EFSCORRUPTED;
        }
        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 -EFSCORRUPTED;
                }
                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 -EFSCORRUPTED;
                        }
                }
                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 -EFSCORRUPTED;
        }

        *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 -EFSCORRUPTED;
        }
        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 -EFSCORRUPTED;
                }
                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 -EFSCORRUPTED;
                        }
                }
                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 -EFSCORRUPTED;
        }

        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) {
                /* subtract from p_depth to get proper eh_depth */
                bh = read_extent_tree_block(inode, block,
                                            path->p_depth - depth, 0);
                if (IS_ERR(bh))
                        return PTR_ERR(bh);
                eh = ext_block_hdr(bh);
                ix = EXT_FIRST_INDEX(eh);
                block = ext4_idx_pblock(ix);
                put_bh(bh);
        }

        bh = read_extent_tree_block(inode, block, path->p_depth - depth, 0);
        if (IS_ERR(bh))
                return PTR_ERR(bh);
        eh = ext_block_hdr(bh);
        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.
 */
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 -EFSCORRUPTED;
        }

        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;

        if (ext4_ext_is_unwritten(ex1) != ext4_ext_is_unwritten(ex2))
                return 0;

        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 > EXT_INIT_MAX_LEN)
                return 0;
        /*
         * The check for IO to unwritten extent is somewhat racy as we
         * increment i_unwritten / set EXT4_STATE_DIO_UNWRITTEN only after
         * dropping i_data_sem. But reserved blocks should save us in that
         * case.
         */
        if (ext4_ext_is_unwritten(ex1) &&
            (ext4_test_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN) ||
             atomic_read(&EXT4_I(inode)->i_unwritten) ||
             (ext1_ee_len + ext2_ee_len > EXT_UNWRITTEN_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, unwritten;

        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! */
                unwritten = ext4_ext_is_unwritten(ex);
                ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
                                + ext4_ext_get_actual_len(ex + 1));
                if (unwritten)
                        ext4_ext_mark_unwritten(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 does a very simple check to see if we can collapse
 * an extent tree with a single extent tree leaf block into the inode.
 */
static void ext4_ext_try_to_merge_up(handle_t *handle,
                                     struct inode *inode,
                                     struct ext4_ext_path *path)
{
        size_t s;
        unsigned max_root = ext4_ext_space_root(inode, 0);
        ext4_fsblk_t blk;

        if ((path[0].p_depth != 1) ||
            (le16_to_cpu(path[0].p_hdr->eh_entries) != 1) ||
            (le16_to_cpu(path[1].p_hdr->eh_entries) > max_root))
                return;

        /*
         * We need to modify the block allocation bitmap and the block
         * group descriptor to release the extent tree block.  If we
         * can't get the journal credits, give up.
         */
        if (ext4_journal_extend(handle, 2))
                return;

        /*
         * Copy the extent data up to the inode
         */
        blk = ext4_idx_pblock(path[0].p_idx);
        s = le16_to_cpu(path[1].p_hdr->eh_entries) *
                sizeof(struct ext4_extent_idx);
        s += sizeof(struct ext4_extent_header);

        path[1].p_maxdepth = path[0].p_maxdepth;
        memcpy(path[0].p_hdr, path[1].p_hdr, s);
        path[0].p_depth = 0;
        path[0].p_ext = EXT_FIRST_EXTENT(path[0].p_hdr) +
                (path[1].p_ext - EXT_FIRST_EXTENT(path[1].p_hdr));
        path[0].p_hdr->eh_max = cpu_to_le16(max_root);

        brelse(path[1].p_bh);
        ext4_free_blocks(handle, inode, NULL, blk, 1,
                         EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
}

/*
 * This function tries to merge the @ex extent to neighbours in the tree.
 * return 1 if merge left else 0.
 */
static void ext4_ext_try_to_merge(handle_t *handle,
                                  struct inode *inode,
                                  struct ext4_ext_path *path,
                                  struct ext4_extent *ex) {
        struct ext4_extent_header *eh;
        unsigned int depth;
        int merge_done = 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)
                (void) ext4_ext_try_to_merge_right(inode, path, ex);

        ext4_ext_try_to_merge_up(handle, inode, path);
}

/*
 * 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 **ppath,
                                struct ext4_extent *newext, int gb_flags)
{
        struct ext4_ext_path *path = *ppath;
        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;
        int mb_flags = 0, unwritten;

        if (gb_flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
                mb_flags |= EXT4_MB_DELALLOC_RESERVED;
        if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
                EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
                return -EFSCORRUPTED;
        }
        depth = ext_depth(inode);
        ex = path[depth].p_ext;
        eh = path[depth].p_hdr;
        if (unlikely(path[depth].p_hdr == NULL)) {
                EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
                return -EFSCORRUPTED;
        }

        /* try to insert block into found extent and return */
        if (ex && !(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) {

                /*
                 * Try to see whether we should rather test the extent on
                 * right from ex, or from the left of ex. This is because
                 * ext4_find_extent() can return either extent on the
                 * left, or on the right from the searched position. This
                 * will make merging more effective.
                 */
                if (ex < EXT_LAST_EXTENT(eh) &&
                    (le32_to_cpu(ex->ee_block) +
                    ext4_ext_get_actual_len(ex) <
                    le32_to_cpu(newext->ee_block))) {
                        ex += 1;
                        goto prepend;
                } else if ((ex > EXT_FIRST_EXTENT(eh)) &&
                           (le32_to_cpu(newext->ee_block) +
                           ext4_ext_get_actual_len(newext) <
                           le32_to_cpu(ex->ee_block)))
                        ex -= 1;

                /* Try to append newex to the ex */
                if (ext4_can_extents_be_merged(inode, ex, newext)) {
                        ext_debug("append [%d]%d block to %u:[%d]%d"
                                  "(from %llu)\n",
                                  ext4_ext_is_unwritten(newext),
                                  ext4_ext_get_actual_len(newext),
                                  le32_to_cpu(ex->ee_block),
                                  ext4_ext_is_unwritten(ex),
                                  ext4_ext_get_actual_len(ex),
                                  ext4_ext_pblock(ex));
                        err = ext4_ext_get_access(handle, inode,
                                                  path + depth);
                        if (err)
                                return err;
                        unwritten = ext4_ext_is_unwritten(ex);
                        ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
                                        + ext4_ext_get_actual_len(newext));
                        if (unwritten)
                                ext4_ext_mark_unwritten(ex);
                        eh = path[depth].p_hdr;
                        nearex = ex;

                        goto merge;
                }

prepend:
                /* Try to prepend newex to the ex */
                if (ext4_can_extents_be_merged(inode, newext, ex)) {
                        ext_debug("prepend %u[%d]%d block to %u:[%d]%d"
                                  "(from %llu)\n",
                                  le32_to_cpu(newext->ee_block),
                                  ext4_ext_is_unwritten(newext),
                                  ext4_ext_get_actual_len(newext),
                                  le32_to_cpu(ex->ee_block),
                                  ext4_ext_is_unwritten(ex),
                                  ext4_ext_get_actual_len(ex),
                                  ext4_ext_pblock(ex));
                        err = ext4_ext_get_access(handle, inode,
                                                  path + depth);
                        if (err)
                                return err;

                        unwritten = ext4_ext_is_unwritten(ex);
                        ex->ee_block = newext->ee_block;
                        ext4_ext_store_pblock(ex, ext4_ext_pblock(newext));
                        ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
                                        + ext4_ext_get_actual_len(newext));
                        if (unwritten)
                                ext4_ext_mark_unwritten(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_find_extent(inode, next, NULL, 0);
                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 (gb_flags & EXT4_GET_BLOCKS_METADATA_NOFAIL)
                mb_flags |= EXT4_MB_USE_RESERVED;
        err = ext4_ext_create_new_leaf(handle, inode, mb_flags, gb_flags,
                                       ppath, 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_unwritten(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_unwritten(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_unwritten(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_unwritten(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 */
        if (!(gb_flags & EXT4_GET_BLOCKS_PRE_IO))
                ext4_ext_try_to_merge(handle, inode, path, nearex);


        /* 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 + path->p_depth);

cleanup:
        ext4_ext_drop_refs(npath);
        kfree(npath);
        return err;
}

static int ext4_fill_fiemap_extents(struct inode *inode,
                                    ext4_lblk_t block, ext4_lblk_t num,
                                    struct fiemap_extent_info *fieinfo)
{
        struct ext4_ext_path *path = NULL;
        struct ext4_extent *ex;
        struct extent_status es;
        ext4_lblk_t next, next_del, start = 0, end = 0;
        ext4_lblk_t last = block + num;
        int exists, depth = 0, err = 0;
        unsigned int flags = 0;
        unsigned char blksize_bits = inode->i_sb->s_blocksize_bits;

        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_find_extent(inode, block, &path, 0);
                if (IS_ERR(path)) {
                        up_read(&EXT4_I(inode)->i_data_sem);
                        err = PTR_ERR(path);
                        path = NULL;
                        break;
                }

                depth = ext_depth(inode);
                if (unlikely(path[depth].p_hdr == NULL)) {
                        up_read(&EXT4_I(inode)->i_data_sem);
                        EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
                        err = -EFSCORRUPTED;
                        break;
                }
                ex = path[depth].p_ext;
                next = ext4_ext_next_allocated_block(path);

                flags = 0;
                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) {
                        es.es_lblk = start;
                        es.es_len = end - start;
                        es.es_pblk = 0;
                } else {
                        es.es_lblk = le32_to_cpu(ex->ee_block);
                        es.es_len = ext4_ext_get_actual_len(ex);
                        es.es_pblk = ext4_ext_pblock(ex);
                        if (ext4_ext_is_unwritten(ex))
                                flags |= FIEMAP_EXTENT_UNWRITTEN;
                }

                /*
                 * Find delayed extent and update es accordingly. We call
                 * it even in !exists case to find out whether es is the
                 * last existing extent or not.
                 */
                next_del = ext4_find_delayed_extent(inode, &es);
                if (!exists && next_del) {
                        exists = 1;
                        flags |= (FIEMAP_EXTENT_DELALLOC |
                                  FIEMAP_EXTENT_UNKNOWN);
                }
                up_read(&EXT4_I(inode)->i_data_sem);

                if (unlikely(es.es_len == 0)) {
                        EXT4_ERROR_INODE(inode, "es.es_len == 0");
                        err = -EFSCORRUPTED;
                        break;
                }

                /*
                 * This is possible iff next == next_del == EXT_MAX_BLOCKS.
                 * we need to check next == EXT_MAX_BLOCKS because it is
                 * possible that an extent is with unwritten and delayed
                 * status due to when an extent is delayed allocated and
                 * is allocated by fallocate status tree will track both of
                 * them in a extent.
                 *
                 * So we could return a unwritten and delayed extent, and
                 * its block is equal to 'next'.
                 */
                if (next == next_del && next == EXT_MAX_BLOCKS) {
                        flags |= FIEMAP_EXTENT_LAST;
                        if (unlikely(next_del != EXT_MAX_BLOCKS ||
                                     next != EXT_MAX_BLOCKS)) {
                                EXT4_ERROR_INODE(inode,
                                                 "next extent == %u, next "
                                                 "delalloc extent = %u",
                                                 next, next_del);
                                err = -EFSCORRUPTED;
                                break;
                        }
                }

                if (exists) {
                        err = fiemap_fill_next_extent(fieinfo,
                                (__u64)es.es_lblk << blksize_bits,
                                (__u64)es.es_pblk << blksize_bits,
                                (__u64)es.es_len << blksize_bits,
                                flags);
                        if (err < 0)
                                break;
                        if (err == 1) {
                                err = 0;
                                break;
                        }
                }

                block = es.es_lblk + es.es_len;
        }

        ext4_ext_drop_refs(path);
        kfree(path);
        return err;
}

/*
 * ext4_ext_determine_hole - determine hole around given block
 * @inode:      inode we lookup in
 * @path:       path in extent tree to @lblk
 * @lblk:       pointer to logical block around which we want to determine hole
 *
 * Determine hole length (and start if easily possible) around given logical
 * block. We don't try too hard to find the beginning of the hole but @path
 * actually points to extent before @lblk, we provide it.
 *
 * The function returns the length of a hole starting at @lblk. We update @lblk
 * to the beginning of the hole if we managed to find it.
 */
static ext4_lblk_t ext4_ext_determine_hole(struct inode *inode,
                                           struct ext4_ext_path *path,
                                           ext4_lblk_t *lblk)
{
        int depth = ext_depth(inode);
        struct ext4_extent *ex;
        ext4_lblk_t len;

        ex = path[depth].p_ext;
        if (ex == NULL) {
                /* there is no extent yet, so gap is [0;-] */
                *lblk = 0;
                len = EXT_MAX_BLOCKS;
        } else if (*lblk < le32_to_cpu(ex->ee_block)) {
                len = le32_to_cpu(ex->ee_block) - *lblk;
        } else if (*lblk >= le32_to_cpu(ex->ee_block)
                        + ext4_ext_get_actual_len(ex)) {
                ext4_lblk_t next;

                *lblk = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex);
                next = ext4_ext_next_allocated_block(path);
                BUG_ON(next == *lblk);
                len = next - *lblk;
        } else {
                BUG();
        }
        return len;
}

/*
 * 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, ext4_lblk_t hole_start,
                          ext4_lblk_t hole_len)
{
        struct extent_status es;

        ext4_es_find_extent_range(inode, &ext4_es_is_delayed, hole_start,
                                  hole_start + hole_len - 1, &es);
        if (es.es_len) {
                /* There's delayed extent containing lblock? */
                if (es.es_lblk <= hole_start)
                        return;
                hole_len = min(es.es_lblk - hole_start, hole_len);
        }
        ext_debug(" -> %u:%u\n", hole_start, hole_len);
        ext4_es_insert_extent(inode, hole_start, hole_len, ~0,
                              EXTENT_STATUS_HOLE);
}

/*
 * 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 -EFSCORRUPTED;
        }
        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 add @extents extents?
 *
 * If we add a single extent, then in the worse case, each tree level
 * index/leaf need to be changed in case of the tree split.
 *
 * If more extents are inserted, 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 extents)
{
        int index;
        int depth;

        /* If we are converting the inline data, only one is needed here. */
        if (ext4_has_inline_data(inode))
                return 1;

        depth = ext_depth(inode);

        if (extents <= 1)
                index = depth * 2;
        else
                index = depth * 3;

        return index;
}

static inline int get_default_free_blocks_flags(struct inode *inode)
{
        if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode) ||
            ext4_test_inode_flag(inode, EXT4_INODE_EA_INODE))
                return EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET;
        else if (ext4_should_journal_data(inode))
                return EXT4_FREE_BLOCKS_FORGET;
        return 0;
}

/*
 * ext4_rereserve_cluster - increment the reserved cluster count when
 *                          freeing a cluster with a pending reservation
 *
 * @inode - file containing the cluster
 * @lblk - logical block in cluster to be reserved
 *
 * Increments the reserved cluster count and adjusts quota in a bigalloc
 * file system when freeing a partial cluster containing at least one
 * delayed and unwritten block.  A partial cluster meeting that
 * requirement will have a pending reservation.  If so, the
 * RERESERVE_CLUSTER flag is used when calling ext4_free_blocks() to
 * defer reserved and allocated space accounting to a subsequent call
 * to this function.
 */
static void ext4_rereserve_cluster(struct inode *inode, ext4_lblk_t lblk)
{
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        struct ext4_inode_info *ei = EXT4_I(inode);

        dquot_reclaim_block(inode, EXT4_C2B(sbi, 1));

        spin_lock(&ei->i_block_reservation_lock);
        ei->i_reserved_data_blocks++;
        percpu_counter_add(&sbi->s_dirtyclusters_counter, 1);
        spin_unlock(&ei->i_block_reservation_lock);

        percpu_counter_add(&sbi->s_freeclusters_counter, 1);
        ext4_remove_pending(inode, lblk);
}

static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
                              struct ext4_extent *ex,
                              struct partial_cluster *partial,
                              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 last_pblk, pblk;
        ext4_lblk_t num;
        int flags;

        /* only extent tail removal is allowed */
        if (from < le32_to_cpu(ex->ee_block) ||
            to != le32_to_cpu(ex->ee_block) + ee_len - 1) {
                ext4_error(sbi->s_sb,
                           "strange request: removal(2) %u-%u from %u:%u",
                           from, to, le32_to_cpu(ex->ee_block), ee_len);
                return 0;
        }

#ifdef EXTENTS_STATS
        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

        trace_ext4_remove_blocks(inode, ex, from, to, partial);

        /*
         * if we have a partial cluster, and it's different from the
         * cluster of the last block in the extent, we free it
         */
        last_pblk = ext4_ext_pblock(ex) + ee_len - 1;

        if (partial->state != initial &&
            partial->pclu != EXT4_B2C(sbi, last_pblk)) {
                if (partial->state == tofree) {
                        flags = get_default_free_blocks_flags(inode);
                        if (ext4_is_pending(inode, partial->lblk))
                                flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
                        ext4_free_blocks(handle, inode, NULL,
                                         EXT4_C2B(sbi, partial->pclu),
                                         sbi->s_cluster_ratio, flags);
                        if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
                                ext4_rereserve_cluster(inode, partial->lblk);
                }
                partial->state = initial;
        }

        num = le32_to_cpu(ex->ee_block) + ee_len - from;
        pblk = ext4_ext_pblock(ex) + ee_len - num;

        /*
         * We free the partial cluster at the end of the extent (if any),
         * unless the cluster is used by another extent (partial_cluster
         * state is nofree).  If a partial cluster exists here, it must be
         * shared with the last block in the extent.
         */
        flags = get_default_free_blocks_flags(inode);

        /* partial, left end cluster aligned, right end unaligned */
        if ((EXT4_LBLK_COFF(sbi, to) != sbi->s_cluster_ratio - 1) &&
            (EXT4_LBLK_CMASK(sbi, to) >= from) &&
            (partial->state != nofree)) {
                if (ext4_is_pending(inode, to))
                        flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
                ext4_free_blocks(handle, inode, NULL,
                                 EXT4_PBLK_CMASK(sbi, last_pblk),
                                 sbi->s_cluster_ratio, flags);
                if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
                        ext4_rereserve_cluster(inode, to);
                partial->state = initial;
                flags = get_default_free_blocks_flags(inode);
        }

        flags |= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER;

        /*
         * For bigalloc file systems, we never free a partial cluster
         * at the beginning of the extent.  Instead, we check to see if we
         * need to free it on a subsequent call to ext4_remove_blocks,
         * or at the end of ext4_ext_rm_leaf or ext4_ext_remove_space.
         */
        flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;
        ext4_free_blocks(handle, inode, NULL, pblk, num, flags);

        /* reset the partial cluster if we've freed past it */
        if (partial->state != initial && partial->pclu != EXT4_B2C(sbi, pblk))
                partial->state = initial;

        /*
         * If we've freed the entire extent but the beginning is not left
         * cluster aligned and is not marked as ineligible for freeing we
         * record the partial cluster at the beginning of the extent.  It
         * wasn't freed by the preceding ext4_free_blocks() call, and we
         * need to look farther to the left to determine if it's to be freed
         * (not shared with another extent). Else, reset the partial
         * cluster - we're either  done freeing or the beginning of the
         * extent is left cluster aligned.
         */
        if (EXT4_LBLK_COFF(sbi, from) && num == ee_len) {
                if (partial->state == initial) {
                        partial->pclu = EXT4_B2C(sbi, pblk);
                        partial->lblk = from;
                        partial->state = tofree;
                }
        } else {
                partial->state = initial;
        }

        return 0;
}

/*
 * ext4_ext_rm_leaf() Removes the extents associated with the
 * blocks appearing between "start" and "end".  Both "start"
 * and "end" must appear in the same extent or EIO is returned.
 *
 * @handle: The journal handle
 * @inode:  The files inode
 * @path:   The path to the leaf
 * @partial_cluster: The cluster which we'll have to free if all extents
 *                   has been released from it.  However, if this value is
 *                   negative, it's a cluster just to the right of the
 *                   punched region and it must not be freed.
 * @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,
                 struct partial_cluster *partial,
                 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 unwritten = 0;
        struct ext4_extent *ex;
        ext4_fsblk_t pblk;

        /* 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 -EFSCORRUPTED;
        }
        /* find where to start removing */
        ex = path[depth].p_ext;
        if (!ex)
                ex = EXT_LAST_EXTENT(eh);

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

        trace_ext4_ext_rm_leaf(inode, start, ex, partial);

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

                if (ext4_ext_is_unwritten(ex))
                        unwritten = 1;
                else
                        unwritten = 0;

                ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
                          unwritten, 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) {
                        /*
                         * We're going to skip this extent and move to another,
                         * so note that its first cluster is in use to avoid
                         * freeing it when removing blocks.  Eventually, the
                         * right edge of the truncated/punched region will
                         * be just to the left.
                         */
                        if (sbi->s_cluster_ratio > 1) {
                                pblk = ext4_ext_pblock(ex);
                                partial->pclu = EXT4_B2C(sbi, pblk);
                                partial->state = nofree;
                        }
                        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,
                                         "can not handle truncate %u:%u "
                                         "on extent %u:%u",
                                         start, end, ex_ee_block,
                                         ex_ee_block + ex_ee_len - 1);
                        err = -EFSCORRUPTED;
                        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, 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 unwritten if all the blocks in the
                 * extent have been removed.
                 */
                if (unwritten && num)
                        ext4_ext_mark_unwritten(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);
                }

                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's a partial cluster and at least one extent remains in
         * the leaf, free the partial cluster if it isn't shared with the
         * current extent.  If it is shared with the current extent
         * we reset the partial cluster because we've reached the start of the
         * truncated/punched region and we're done removing blocks.
         */
        if (partial->state == tofree && ex >= EXT_FIRST_EXTENT(eh)) {
                pblk = ext4_ext_pblock(ex) + ex_ee_len - 1;
                if (partial->pclu != EXT4_B2C(sbi, pblk)) {
                        int flags = get_default_free_blocks_flags(inode);

                        if (ext4_is_pending(inode, partial->lblk))
                                flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
                        ext4_free_blocks(handle, inode, NULL,
                                         EXT4_C2B(sbi, partial->pclu),
                                         sbi->s_cluster_ratio, flags);
                        if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
                                ext4_rereserve_cluster(inode, partial->lblk);
                }
                partial->state = initial;
        }

        /* 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;
}

int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start,
                          ext4_lblk_t end)
{
        struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
        int depth = ext_depth(inode);
        struct ext4_ext_path *path = NULL;
        struct partial_cluster partial;
        handle_t *handle;
        int i = 0, err = 0;

        partial.pclu = 0;
        partial.lblk = 0;
        partial.state = initial;

        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, EXT4_HT_TRUNCATE, depth + 1);
        if (IS_ERR(handle))
                return PTR_ERR(handle);

again:
        trace_ext4_ext_remove_space(inode, start, end, 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, ex_end, lblk;
                ext4_fsblk_t pblk;

                /* find extent for or closest extent to this block */
                path = ext4_find_extent(inode, end, NULL, EXT4_EX_NOCACHE);
                if (IS_ERR(path)) {
                        ext4_journal_stop(handle);
                        return PTR_ERR(path);
                }
                depth = ext_depth(inode);
                /* Leaf not may not exist only if inode has no blocks at all */
                ex = path[depth].p_ext;
                if (!ex) {
                        if (depth) {
                                EXT4_ERROR_INODE(inode,
                                                 "path[%d].p_hdr == NULL",
                                                 depth);
                                err = -EFSCORRUPTED;
                        }
                        goto out;
                }

                ee_block = le32_to_cpu(ex->ee_block);
                ex_end = ee_block + ext4_ext_get_actual_len(ex) - 1;

                /*
                 * 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 < ex_end) {

                        /*
                         * If we're going to split the extent, note that
                         * the cluster containing the block after 'end' is
                         * in use to avoid freeing it when removing blocks.
                         */
                        if (sbi->s_cluster_ratio > 1) {
                                pblk = ext4_ext_pblock(ex) + end - ee_block + 2;
                                partial.pclu = EXT4_B2C(sbi, pblk);
                                partial.state = nofree;
                        }

                        /*
                         * Split the extent in two so that 'end' is the last
                         * block in the first new extent. Also we should not
                         * fail removing space due to ENOSPC so try to use
                         * reserved block if that happens.
                         */
                        err = ext4_force_split_extent_at(handle, inode, &path,
                                                         end + 1, 1);
                        if (err < 0)
                                goto out;

                } else if (sbi->s_cluster_ratio > 1 && end >= ex_end &&
                           partial.state == initial) {
                        /*
                         * If we're punching, there's an extent to the right.
                         * If the partial cluster hasn't been set, set it to
                         * that extent's first cluster and its state to nofree
                         * so it won't be freed should it contain blocks to be
                         * removed. If it's already set (tofree/nofree), we're
                         * retrying and keep the original partial cluster info
                         * so a cluster marked tofree as a result of earlier
                         * extent removal is not lost.
                         */
                        lblk = ex_end + 1;
                        err = ext4_ext_search_right(inode, path, &lblk, &pblk,
                                                    &ex);
                        if (err)
                                goto out;
                        if (pblk) {
                                partial.pclu = EXT4_B2C(sbi, pblk);
                                partial.state = nofree;
                        }
                }
        }
        /*
         * 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 = kcalloc(depth + 1, sizeof(struct ext4_ext_path),
                               GFP_NOFS);
                if (path == NULL) {
                        ext4_journal_stop(handle);
                        return -ENOMEM;
                }
                path[0].p_maxdepth = path[0].p_depth = depth;
                path[0].p_hdr = ext_inode_hdr(inode);