fs/ocfs2/xattr.c

/* -*- mode: c; c-basic-offset: 8; -*-
 * vim: noexpandtab sw=8 ts=8 sts=0:
 *
 * xattr.c
 *
 * Copyright (C) 2004, 2008 Oracle.  All rights reserved.
 *
 * CREDITS:
 * Lots of code in this file is copy from linux/fs/ext3/xattr.c.
 * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de>
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License version 2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 */

#include <linux/capability.h>
#include <linux/fs.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
#include <linux/uio.h>
#include <linux/sched.h>
#include <linux/splice.h>
#include <linux/mount.h>
#include <linux/writeback.h>
#include <linux/falloc.h>
#include <linux/sort.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/string.h>
#include <linux/security.h>

#define MLOG_MASK_PREFIX ML_XATTR
#include <cluster/masklog.h>

#include "ocfs2.h"
#include "alloc.h"
#include "blockcheck.h"
#include "dlmglue.h"
#include "file.h"
#include "symlink.h"
#include "sysfile.h"
#include "inode.h"
#include "journal.h"
#include "ocfs2_fs.h"
#include "suballoc.h"
#include "uptodate.h"
#include "buffer_head_io.h"
#include "super.h"
#include "xattr.h"
#include "refcounttree.h"
#include "acl.h"

struct ocfs2_xattr_def_value_root {
	struct ocfs2_xattr_value_root	xv;
	struct ocfs2_extent_rec		er;
};

struct ocfs2_xattr_bucket {
	/* The inode these xattrs are associated with */
	struct inode *bu_inode;

	/* The actual buffers that make up the bucket */
	struct buffer_head *bu_bhs[OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET];

	/* How many blocks make up one bucket for this filesystem */
	int bu_blocks;
};

struct ocfs2_xattr_set_ctxt {
	handle_t *handle;
	struct ocfs2_alloc_context *meta_ac;
	struct ocfs2_alloc_context *data_ac;
	struct ocfs2_cached_dealloc_ctxt dealloc;
};

#define OCFS2_XATTR_ROOT_SIZE	(sizeof(struct ocfs2_xattr_def_value_root))
#define OCFS2_XATTR_INLINE_SIZE	80
#define OCFS2_XATTR_HEADER_GAP	4
#define OCFS2_XATTR_FREE_IN_IBODY	(OCFS2_MIN_XATTR_INLINE_SIZE \
					 - sizeof(struct ocfs2_xattr_header) \
					 - OCFS2_XATTR_HEADER_GAP)
#define OCFS2_XATTR_FREE_IN_BLOCK(ptr)	((ptr)->i_sb->s_blocksize \
					 - sizeof(struct ocfs2_xattr_block) \
					 - sizeof(struct ocfs2_xattr_header) \
					 - OCFS2_XATTR_HEADER_GAP)

static struct ocfs2_xattr_def_value_root def_xv = {
	.xv.xr_list.l_count = cpu_to_le16(1),
};

struct xattr_handler *ocfs2_xattr_handlers[] = {
	&ocfs2_xattr_user_handler,
#ifdef CONFIG_OCFS2_FS_POSIX_ACL
	&ocfs2_xattr_acl_access_handler,
	&ocfs2_xattr_acl_default_handler,
#endif
	&ocfs2_xattr_trusted_handler,
	&ocfs2_xattr_security_handler,
	NULL
};

static struct xattr_handler *ocfs2_xattr_handler_map[OCFS2_XATTR_MAX] = {
	[OCFS2_XATTR_INDEX_USER]	= &ocfs2_xattr_user_handler,
#ifdef CONFIG_OCFS2_FS_POSIX_ACL
	[OCFS2_XATTR_INDEX_POSIX_ACL_ACCESS]
					= &ocfs2_xattr_acl_access_handler,
	[OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT]
					= &ocfs2_xattr_acl_default_handler,
#endif
	[OCFS2_XATTR_INDEX_TRUSTED]	= &ocfs2_xattr_trusted_handler,
	[OCFS2_XATTR_INDEX_SECURITY]	= &ocfs2_xattr_security_handler,
};

struct ocfs2_xattr_info {
	int name_index;
	const char *name;
	const void *value;
	size_t value_len;
};

struct ocfs2_xattr_search {
	struct buffer_head *inode_bh;
	/*
	 * xattr_bh point to the block buffer head which has extended attribute
	 * when extended attribute in inode, xattr_bh is equal to inode_bh.
	 */
	struct buffer_head *xattr_bh;
	struct ocfs2_xattr_header *header;
	struct ocfs2_xattr_bucket *bucket;
	void *base;
	void *end;
	struct ocfs2_xattr_entry *here;
	int not_found;
};

static int ocfs2_xattr_bucket_get_name_value(struct super_block *sb,
					     struct ocfs2_xattr_header *xh,
					     int index,
					     int *block_off,
					     int *new_offset);

static int ocfs2_xattr_block_find(struct inode *inode,
				  int name_index,
				  const char *name,
				  struct ocfs2_xattr_search *xs);
static int ocfs2_xattr_index_block_find(struct inode *inode,
					struct buffer_head *root_bh,
					int name_index,
					const char *name,
					struct ocfs2_xattr_search *xs);

static int ocfs2_xattr_tree_list_index_block(struct inode *inode,
					struct buffer_head *blk_bh,
					char *buffer,
					size_t buffer_size);

static int ocfs2_xattr_create_index_block(struct inode *inode,
					  struct ocfs2_xattr_search *xs,
					  struct ocfs2_xattr_set_ctxt *ctxt);

static int ocfs2_xattr_set_entry_index_block(struct inode *inode,
					     struct ocfs2_xattr_info *xi,
					     struct ocfs2_xattr_search *xs,
					     struct ocfs2_xattr_set_ctxt *ctxt);

typedef int (xattr_tree_rec_func)(struct inode *inode,
				  struct buffer_head *root_bh,
				  u64 blkno, u32 cpos, u32 len, void *para);
static int ocfs2_iterate_xattr_index_block(struct inode *inode,
					   struct buffer_head *root_bh,
					   xattr_tree_rec_func *rec_func,
					   void *para);
static int ocfs2_delete_xattr_in_bucket(struct inode *inode,
					struct ocfs2_xattr_bucket *bucket,
					void *para);
static int ocfs2_rm_xattr_cluster(struct inode *inode,
				  struct buffer_head *root_bh,
				  u64 blkno,
				  u32 cpos,
				  u32 len,
				  void *para);

static int ocfs2_mv_xattr_buckets(struct inode *inode, handle_t *handle,
				  u64 src_blk, u64 last_blk, u64 to_blk,
				  unsigned int start_bucket,
				  u32 *first_hash);
static int ocfs2_prepare_refcount_xattr(struct inode *inode,
					struct ocfs2_dinode *di,
					struct ocfs2_xattr_info *xi,
					struct ocfs2_xattr_search *xis,
					struct ocfs2_xattr_search *xbs,
					struct ocfs2_refcount_tree **ref_tree,
					int *meta_need,
					int *credits);
static int ocfs2_get_xattr_tree_value_root(struct super_block *sb,
					   struct ocfs2_xattr_bucket *bucket,
					   int offset,
					   struct ocfs2_xattr_value_root **xv,
					   struct buffer_head **bh);
static int ocfs2_xattr_security_set(struct inode *inode, const char *name,
				    const void *value, size_t size, int flags);

static inline u16 ocfs2_xattr_buckets_per_cluster(struct ocfs2_super *osb)
{
	return (1 << osb->s_clustersize_bits) / OCFS2_XATTR_BUCKET_SIZE;
}

static inline u16 ocfs2_blocks_per_xattr_bucket(struct super_block *sb)
{
	return OCFS2_XATTR_BUCKET_SIZE / (1 << sb->s_blocksize_bits);
}

static inline u16 ocfs2_xattr_max_xe_in_bucket(struct super_block *sb)
{
	u16 len = sb->s_blocksize -
		 offsetof(struct ocfs2_xattr_header, xh_entries);

	return len / sizeof(struct ocfs2_xattr_entry);
}

#define bucket_blkno(_b) ((_b)->bu_bhs[0]->b_blocknr)
#define bucket_block(_b, _n) ((_b)->bu_bhs[(_n)]->b_data)
#define bucket_xh(_b) ((struct ocfs2_xattr_header *)bucket_block((_b), 0))

static struct ocfs2_xattr_bucket *ocfs2_xattr_bucket_new(struct inode *inode)
{
	struct ocfs2_xattr_bucket *bucket;
	int blks = ocfs2_blocks_per_xattr_bucket(inode->i_sb);

	BUG_ON(blks > OCFS2_XATTR_MAX_BLOCKS_PER_BUCKET);

	bucket = kzalloc(sizeof(struct ocfs2_xattr_bucket), GFP_NOFS);
	if (bucket) {
		bucket->bu_inode = inode;
		bucket->bu_blocks = blks;
	}

	return bucket;
}

static void ocfs2_xattr_bucket_relse(struct ocfs2_xattr_bucket *bucket)
{
	int i;

	for (i = 0; i < bucket->bu_blocks; i++) {
		brelse(bucket->bu_bhs[i]);
		bucket->bu_bhs[i] = NULL;
	}
}

static void ocfs2_xattr_bucket_free(struct ocfs2_xattr_bucket *bucket)
{
	if (bucket) {
		ocfs2_xattr_bucket_relse(bucket);
		bucket->bu_inode = NULL;
		kfree(bucket);
	}
}

/*
 * A bucket that has never been written to disk doesn't need to be
 * read.  We just need the buffer_heads.  Don't call this for
 * buckets that are already on disk.  ocfs2_read_xattr_bucket() initializes
 * them fully.
 */
static int ocfs2_init_xattr_bucket(struct ocfs2_xattr_bucket *bucket,
				   u64 xb_blkno)
{
	int i, rc = 0;

	for (i = 0; i < bucket->bu_blocks; i++) {
		bucket->bu_bhs[i] = sb_getblk(bucket->bu_inode->i_sb,
					      xb_blkno + i);
		if (!bucket->bu_bhs[i]) {
			rc = -EIO;
			mlog_errno(rc);
			break;
		}

		if (!ocfs2_buffer_uptodate(INODE_CACHE(bucket->bu_inode),
					   bucket->bu_bhs[i]))
			ocfs2_set_new_buffer_uptodate(INODE_CACHE(bucket->bu_inode),
						      bucket->bu_bhs[i]);
	}

	if (rc)
		ocfs2_xattr_bucket_relse(bucket);
	return rc;
}

/* Read the xattr bucket at xb_blkno */
static int ocfs2_read_xattr_bucket(struct ocfs2_xattr_bucket *bucket,
				   u64 xb_blkno)
{
	int rc;

	rc = ocfs2_read_blocks(INODE_CACHE(bucket->bu_inode), xb_blkno,
			       bucket->bu_blocks, bucket->bu_bhs, 0,
			       NULL);
	if (!rc) {
		spin_lock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock);
		rc = ocfs2_validate_meta_ecc_bhs(bucket->bu_inode->i_sb,
						 bucket->bu_bhs,
						 bucket->bu_blocks,
						 &bucket_xh(bucket)->xh_check);
		spin_unlock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock);
		if (rc)
			mlog_errno(rc);
	}

	if (rc)
		ocfs2_xattr_bucket_relse(bucket);
	return rc;
}

static int ocfs2_xattr_bucket_journal_access(handle_t *handle,
					     struct ocfs2_xattr_bucket *bucket,
					     int type)
{
	int i, rc = 0;

	for (i = 0; i < bucket->bu_blocks; i++) {
		rc = ocfs2_journal_access(handle,
					  INODE_CACHE(bucket->bu_inode),
					  bucket->bu_bhs[i], type);
		if (rc) {
			mlog_errno(rc);
			break;
		}
	}

	return rc;
}

static void ocfs2_xattr_bucket_journal_dirty(handle_t *handle,
					     struct ocfs2_xattr_bucket *bucket)
{
	int i;

	spin_lock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock);
	ocfs2_compute_meta_ecc_bhs(bucket->bu_inode->i_sb,
				   bucket->bu_bhs, bucket->bu_blocks,
				   &bucket_xh(bucket)->xh_check);
	spin_unlock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock);

	for (i = 0; i < bucket->bu_blocks; i++)
		ocfs2_journal_dirty(handle, bucket->bu_bhs[i]);
}

static void ocfs2_xattr_bucket_copy_data(struct ocfs2_xattr_bucket *dest,
					 struct ocfs2_xattr_bucket *src)
{
	int i;
	int blocksize = src->bu_inode->i_sb->s_blocksize;

	BUG_ON(dest->bu_blocks != src->bu_blocks);
	BUG_ON(dest->bu_inode != src->bu_inode);

	for (i = 0; i < src->bu_blocks; i++) {
		memcpy(bucket_block(dest, i), bucket_block(src, i),
		       blocksize);
	}
}

static int ocfs2_validate_xattr_block(struct super_block *sb,
				      struct buffer_head *bh)
{
	int rc;
	struct ocfs2_xattr_block *xb =
		(struct ocfs2_xattr_block *)bh->b_data;

	mlog(0, "Validating xattr block %llu\n",
	     (unsigned long long)bh->b_blocknr);

	BUG_ON(!buffer_uptodate(bh));

	/*
	 * If the ecc fails, we return the error but otherwise
	 * leave the filesystem running.  We know any error is
	 * local to this block.
	 */
	rc = ocfs2_validate_meta_ecc(sb, bh->b_data, &xb->xb_check);
	if (rc)
		return rc;

	/*
	 * Errors after here are fatal
	 */

	if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) {
		ocfs2_error(sb,
			    "Extended attribute block #%llu has bad "
			    "signature %.*s",
			    (unsigned long long)bh->b_blocknr, 7,
			    xb->xb_signature);
		return -EINVAL;
	}

	if (le64_to_cpu(xb->xb_blkno) != bh->b_blocknr) {
		ocfs2_error(sb,
			    "Extended attribute block #%llu has an "
			    "invalid xb_blkno of %llu",
			    (unsigned long long)bh->b_blocknr,
			    (unsigned long long)le64_to_cpu(xb->xb_blkno));
		return -EINVAL;
	}

	if (le32_to_cpu(xb->xb_fs_generation) != OCFS2_SB(sb)->fs_generation) {
		ocfs2_error(sb,
			    "Extended attribute block #%llu has an invalid "
			    "xb_fs_generation of #%u",
			    (unsigned long long)bh->b_blocknr,
			    le32_to_cpu(xb->xb_fs_generation));
		return -EINVAL;
	}

	return 0;
}

static int ocfs2_read_xattr_block(struct inode *inode, u64 xb_blkno,
				  struct buffer_head **bh)
{
	int rc;
	struct buffer_head *tmp = *bh;

	rc = ocfs2_read_block(INODE_CACHE(inode), xb_blkno, &tmp,
			      ocfs2_validate_xattr_block);

	/* If ocfs2_read_block() got us a new bh, pass it up. */
	if (!rc && !*bh)
		*bh = tmp;

	return rc;
}

static inline const char *ocfs2_xattr_prefix(int name_index)
{
	struct xattr_handler *handler = NULL;

	if (name_index > 0 && name_index < OCFS2_XATTR_MAX)
		handler = ocfs2_xattr_handler_map[name_index];

	return handler ? handler->prefix : NULL;
}

static u32 ocfs2_xattr_name_hash(struct inode *inode,
				 const char *name,
				 int name_len)
{
	/* Get hash value of uuid from super block */
	u32 hash = OCFS2_SB(inode->i_sb)->uuid_hash;
	int i;

	/* hash extended attribute name */
	for (i = 0; i < name_len; i++) {
		hash = (hash << OCFS2_HASH_SHIFT) ^
		       (hash >> (8*sizeof(hash) - OCFS2_HASH_SHIFT)) ^
		       *name++;
	}

	return hash;
}

/*
 * ocfs2_xattr_hash_entry()
 *
 * Compute the hash of an extended attribute.
 */
static void ocfs2_xattr_hash_entry(struct inode *inode,
				   struct ocfs2_xattr_header *header,
				   struct ocfs2_xattr_entry *entry)
{
	u32 hash = 0;
	char *name = (char *)header + le16_to_cpu(entry->xe_name_offset);

	hash = ocfs2_xattr_name_hash(inode, name, entry->xe_name_len);
	entry->xe_name_hash = cpu_to_le32(hash);

	return;
}

static int ocfs2_xattr_entry_real_size(int name_len, size_t value_len)
{
	int size = 0;

	if (value_len <= OCFS2_XATTR_INLINE_SIZE)
		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_SIZE(value_len);
	else
		size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
	size += sizeof(struct ocfs2_xattr_entry);

	return size;
}

int ocfs2_calc_security_init(struct inode *dir,
			     struct ocfs2_security_xattr_info *si,
			     int *want_clusters,
			     int *xattr_credits,
			     struct ocfs2_alloc_context **xattr_ac)
{
	int ret = 0;
	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
	int s_size = ocfs2_xattr_entry_real_size(strlen(si->name),
						 si->value_len);

	/*
	 * The max space of security xattr taken inline is
	 * 256(name) + 80(value) + 16(entry) = 352 bytes,
	 * So reserve one metadata block for it is ok.
	 */
	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE ||
	    s_size > OCFS2_XATTR_FREE_IN_IBODY) {
		ret = ocfs2_reserve_new_metadata_blocks(osb, 1, xattr_ac);
		if (ret) {
			mlog_errno(ret);
			return ret;
		}
		*xattr_credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
	}

	/* reserve clusters for xattr value which will be set in B tree*/
	if (si->value_len > OCFS2_XATTR_INLINE_SIZE) {
		int new_clusters = ocfs2_clusters_for_bytes(dir->i_sb,
							    si->value_len);

		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
							   new_clusters);
		*want_clusters += new_clusters;
	}
	return ret;
}

int ocfs2_calc_xattr_init(struct inode *dir,
			  struct buffer_head *dir_bh,
			  int mode,
			  struct ocfs2_security_xattr_info *si,
			  int *want_clusters,
			  int *xattr_credits,
			  int *want_meta)
{
	int ret = 0;
	struct ocfs2_super *osb = OCFS2_SB(dir->i_sb);
	int s_size = 0, a_size = 0, acl_len = 0, new_clusters;

	if (si->enable)
		s_size = ocfs2_xattr_entry_real_size(strlen(si->name),
						     si->value_len);

	if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) {
		acl_len = ocfs2_xattr_get_nolock(dir, dir_bh,
					OCFS2_XATTR_INDEX_POSIX_ACL_DEFAULT,
					"", NULL, 0);
		if (acl_len > 0) {
			a_size = ocfs2_xattr_entry_real_size(0, acl_len);
			if (S_ISDIR(mode))
				a_size <<= 1;
		} else if (acl_len != 0 && acl_len != -ENODATA) {
			mlog_errno(ret);
			return ret;
		}
	}

	if (!(s_size + a_size))
		return ret;

	/*
	 * The max space of security xattr taken inline is
	 * 256(name) + 80(value) + 16(entry) = 352 bytes,
	 * The max space of acl xattr taken inline is
	 * 80(value) + 16(entry) * 2(if directory) = 192 bytes,
	 * when blocksize = 512, may reserve one more cluser for
	 * xattr bucket, otherwise reserve one metadata block
	 * for them is ok.
	 * If this is a new directory with inline data,
	 * we choose to reserve the entire inline area for
	 * directory contents and force an external xattr block.
	 */
	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE ||
	    (S_ISDIR(mode) && ocfs2_supports_inline_data(osb)) ||
	    (s_size + a_size) > OCFS2_XATTR_FREE_IN_IBODY) {
		*want_meta = *want_meta + 1;
		*xattr_credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
	}

	if (dir->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE &&
	    (s_size + a_size) > OCFS2_XATTR_FREE_IN_BLOCK(dir)) {
		*want_clusters += 1;
		*xattr_credits += ocfs2_blocks_per_xattr_bucket(dir->i_sb);
	}

	/*
	 * reserve credits and clusters for xattrs which has large value
	 * and have to be set outside
	 */
	if (si->enable && si->value_len > OCFS2_XATTR_INLINE_SIZE) {
		new_clusters = ocfs2_clusters_for_bytes(dir->i_sb,
							si->value_len);
		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
							   new_clusters);
		*want_clusters += new_clusters;
	}
	if (osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL &&
	    acl_len > OCFS2_XATTR_INLINE_SIZE) {
		/* for directory, it has DEFAULT and ACCESS two types of acls */
		new_clusters = (S_ISDIR(mode) ? 2 : 1) *
				ocfs2_clusters_for_bytes(dir->i_sb, acl_len);
		*xattr_credits += ocfs2_clusters_to_blocks(dir->i_sb,
							   new_clusters);
		*want_clusters += new_clusters;
	}

	return ret;
}

static int ocfs2_xattr_extend_allocation(struct inode *inode,
					 u32 clusters_to_add,
					 struct ocfs2_xattr_value_buf *vb,
					 struct ocfs2_xattr_set_ctxt *ctxt)
{
	int status = 0;
	handle_t *handle = ctxt->handle;
	enum ocfs2_alloc_restarted why;
	u32 prev_clusters, logical_start = le32_to_cpu(vb->vb_xv->xr_clusters);
	struct ocfs2_extent_tree et;

	mlog(0, "(clusters_to_add for xattr= %u)\n", clusters_to_add);

	ocfs2_init_xattr_value_extent_tree(&et, INODE_CACHE(inode), vb);

	status = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
			      OCFS2_JOURNAL_ACCESS_WRITE);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	prev_clusters = le32_to_cpu(vb->vb_xv->xr_clusters);
	status = ocfs2_add_clusters_in_btree(handle,
					     &et,
					     &logical_start,
					     clusters_to_add,
					     0,
					     ctxt->data_ac,
					     ctxt->meta_ac,
					     &why);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_journal_dirty(handle, vb->vb_bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	clusters_to_add -= le32_to_cpu(vb->vb_xv->xr_clusters) - prev_clusters;

	/*
	 * We should have already allocated enough space before the transaction,
	 * so no need to restart.
	 */
	BUG_ON(why != RESTART_NONE || clusters_to_add);

leave:

	return status;
}

static int __ocfs2_remove_xattr_range(struct inode *inode,
				      struct ocfs2_xattr_value_buf *vb,
				      u32 cpos, u32 phys_cpos, u32 len,
				      unsigned int ext_flags,
				      struct ocfs2_xattr_set_ctxt *ctxt)
{
	int ret;
	u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
	handle_t *handle = ctxt->handle;
	struct ocfs2_extent_tree et;

	ocfs2_init_xattr_value_extent_tree(&et, INODE_CACHE(inode), vb);

	ret = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
			    OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_remove_extent(handle, &et, cpos, len, ctxt->meta_ac,
				  &ctxt->dealloc);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	le32_add_cpu(&vb->vb_xv->xr_clusters, -len);

	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	if (ext_flags & OCFS2_EXT_REFCOUNTED)
		ret = ocfs2_decrease_refcount(inode, handle,
					ocfs2_blocks_to_clusters(inode->i_sb,
								 phys_blkno),
					len, ctxt->meta_ac, &ctxt->dealloc, 1);
	else
		ret = ocfs2_cache_cluster_dealloc(&ctxt->dealloc,
						  phys_blkno, len);
	if (ret)
		mlog_errno(ret);

out:
	return ret;
}

static int ocfs2_xattr_shrink_size(struct inode *inode,
				   u32 old_clusters,
				   u32 new_clusters,
				   struct ocfs2_xattr_value_buf *vb,
				   struct ocfs2_xattr_set_ctxt *ctxt)
{
	int ret = 0;
	unsigned int ext_flags;
	u32 trunc_len, cpos, phys_cpos, alloc_size;
	u64 block;

	if (old_clusters <= new_clusters)
		return 0;

	cpos = new_clusters;
	trunc_len = old_clusters - new_clusters;
	while (trunc_len) {
		ret = ocfs2_xattr_get_clusters(inode, cpos, &phys_cpos,
					       &alloc_size,
					       &vb->vb_xv->xr_list, &ext_flags);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		if (alloc_size > trunc_len)
			alloc_size = trunc_len;

		ret = __ocfs2_remove_xattr_range(inode, vb, cpos,
						 phys_cpos, alloc_size,
						 ext_flags, ctxt);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		block = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
		ocfs2_remove_xattr_clusters_from_cache(INODE_CACHE(inode),
						       block, alloc_size);
		cpos += alloc_size;
		trunc_len -= alloc_size;
	}

out:
	return ret;
}

static int ocfs2_xattr_value_truncate(struct inode *inode,
				      struct ocfs2_xattr_value_buf *vb,
				      int len,
				      struct ocfs2_xattr_set_ctxt *ctxt)
{
	int ret;
	u32 new_clusters = ocfs2_clusters_for_bytes(inode->i_sb, len);
	u32 old_clusters = le32_to_cpu(vb->vb_xv->xr_clusters);

	if (new_clusters == old_clusters)
		return 0;

	if (new_clusters > old_clusters)
		ret = ocfs2_xattr_extend_allocation(inode,
						    new_clusters - old_clusters,
						    vb, ctxt);
	else
		ret = ocfs2_xattr_shrink_size(inode,
					      old_clusters, new_clusters,
					      vb, ctxt);

	return ret;
}

static int ocfs2_xattr_list_entry(char *buffer, size_t size,
				  size_t *result, const char *prefix,
				  const char *name, int name_len)
{
	char *p = buffer + *result;
	int prefix_len = strlen(prefix);
	int total_len = prefix_len + name_len + 1;

	*result += total_len;

	/* we are just looking for how big our buffer needs to be */
	if (!size)
		return 0;

	if (*result > size)
		return -ERANGE;

	memcpy(p, prefix, prefix_len);
	memcpy(p + prefix_len, name, name_len);
	p[prefix_len + name_len] = '\0';

	return 0;
}

static int ocfs2_xattr_list_entries(struct inode *inode,
				    struct ocfs2_xattr_header *header,
				    char *buffer, size_t buffer_size)
{
	size_t result = 0;
	int i, type, ret;
	const char *prefix, *name;

	for (i = 0 ; i < le16_to_cpu(header->xh_count); i++) {
		struct ocfs2_xattr_entry *entry = &header->xh_entries[i];
		type = ocfs2_xattr_get_type(entry);
		prefix = ocfs2_xattr_prefix(type);

		if (prefix) {
			name = (const char *)header +
				le16_to_cpu(entry->xe_name_offset);

			ret = ocfs2_xattr_list_entry(buffer, buffer_size,
						     &result, prefix, name,
						     entry->xe_name_len);
			if (ret)
				return ret;
		}
	}

	return result;
}

int ocfs2_has_inline_xattr_value_outside(struct inode *inode,
					 struct ocfs2_dinode *di)
{
	struct ocfs2_xattr_header *xh;
	int i;

	xh = (struct ocfs2_xattr_header *)
		 ((void *)di + inode->i_sb->s_blocksize -
		 le16_to_cpu(di->i_xattr_inline_size));

	for (i = 0; i < le16_to_cpu(xh->xh_count); i++)
		if (!ocfs2_xattr_is_local(&xh->xh_entries[i]))
			return 1;

	return 0;
}

static int ocfs2_xattr_ibody_list(struct inode *inode,
				  struct ocfs2_dinode *di,
				  char *buffer,
				  size_t buffer_size)
{
	struct ocfs2_xattr_header *header = NULL;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	int ret = 0;

	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL))
		return ret;

	header = (struct ocfs2_xattr_header *)
		 ((void *)di + inode->i_sb->s_blocksize -
		 le16_to_cpu(di->i_xattr_inline_size));

	ret = ocfs2_xattr_list_entries(inode, header, buffer, buffer_size);

	return ret;
}

static int ocfs2_xattr_block_list(struct inode *inode,
				  struct ocfs2_dinode *di,
				  char *buffer,
				  size_t buffer_size)
{
	struct buffer_head *blk_bh = NULL;
	struct ocfs2_xattr_block *xb;
	int ret = 0;

	if (!di->i_xattr_loc)
		return ret;

	ret = ocfs2_read_xattr_block(inode, le64_to_cpu(di->i_xattr_loc),
				     &blk_bh);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}

	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
		struct ocfs2_xattr_header *header = &xb->xb_attrs.xb_header;
		ret = ocfs2_xattr_list_entries(inode, header,
					       buffer, buffer_size);
	} else
		ret = ocfs2_xattr_tree_list_index_block(inode, blk_bh,
						   buffer, buffer_size);

	brelse(blk_bh);

	return ret;
}

ssize_t ocfs2_listxattr(struct dentry *dentry,
			char *buffer,
			size_t size)
{
	int ret = 0, i_ret = 0, b_ret = 0;
	struct buffer_head *di_bh = NULL;
	struct ocfs2_dinode *di = NULL;
	struct ocfs2_inode_info *oi = OCFS2_I(dentry->d_inode);

	if (!ocfs2_supports_xattr(OCFS2_SB(dentry->d_sb)))
		return -EOPNOTSUPP;

	if (!(oi->ip_dyn_features & OCFS2_HAS_XATTR_FL))
		return ret;

	ret = ocfs2_inode_lock(dentry->d_inode, &di_bh, 0);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}

	di = (struct ocfs2_dinode *)di_bh->b_data;

	down_read(&oi->ip_xattr_sem);
	i_ret = ocfs2_xattr_ibody_list(dentry->d_inode, di, buffer, size);
	if (i_ret < 0)
		b_ret = 0;
	else {
		if (buffer) {
			buffer += i_ret;
			size -= i_ret;
		}
		b_ret = ocfs2_xattr_block_list(dentry->d_inode, di,
					       buffer, size);
		if (b_ret < 0)
			i_ret = 0;
	}
	up_read(&oi->ip_xattr_sem);
	ocfs2_inode_unlock(dentry->d_inode, 0);

	brelse(di_bh);

	return i_ret + b_ret;
}

static int ocfs2_xattr_find_entry(int name_index,
				  const char *name,
				  struct ocfs2_xattr_search *xs)
{
	struct ocfs2_xattr_entry *entry;
	size_t name_len;
	int i, cmp = 1;

	if (name == NULL)
		return -EINVAL;

	name_len = strlen(name);
	entry = xs->here;
	for (i = 0; i < le16_to_cpu(xs->header->xh_count); i++) {
		cmp = name_index - ocfs2_xattr_get_type(entry);
		if (!cmp)
			cmp = name_len - entry->xe_name_len;
		if (!cmp)
			cmp = memcmp(name, (xs->base +
				     le16_to_cpu(entry->xe_name_offset)),
				     name_len);
		if (cmp == 0)
			break;
		entry += 1;
	}
	xs->here = entry;

	return cmp ? -ENODATA : 0;
}

static int ocfs2_xattr_get_value_outside(struct inode *inode,
					 struct ocfs2_xattr_value_root *xv,
					 void *buffer,
					 size_t len)
{
	u32 cpos, p_cluster, num_clusters, bpc, clusters;
	u64 blkno;
	int i, ret = 0;
	size_t cplen, blocksize;
	struct buffer_head *bh = NULL;
	struct ocfs2_extent_list *el;

	el = &xv->xr_list;
	clusters = le32_to_cpu(xv->xr_clusters);
	bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
	blocksize = inode->i_sb->s_blocksize;

	cpos = 0;
	while (cpos < clusters) {
		ret = ocfs2_xattr_get_clusters(inode, cpos, &p_cluster,
					       &num_clusters, el, NULL);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		blkno = ocfs2_clusters_to_blocks(inode->i_sb, p_cluster);
		/* Copy ocfs2_xattr_value */
		for (i = 0; i < num_clusters * bpc; i++, blkno++) {
			ret = ocfs2_read_block(INODE_CACHE(inode), blkno,
					       &bh, NULL);
			if (ret) {
				mlog_errno(ret);
				goto out;
			}

			cplen = len >= blocksize ? blocksize : len;
			memcpy(buffer, bh->b_data, cplen);
			len -= cplen;
			buffer += cplen;

			brelse(bh);
			bh = NULL;
			if (len == 0)
				break;
		}
		cpos += num_clusters;
	}
out:
	return ret;
}

static int ocfs2_xattr_ibody_get(struct inode *inode,
				 int name_index,
				 const char *name,
				 void *buffer,
				 size_t buffer_size,
				 struct ocfs2_xattr_search *xs)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
	struct ocfs2_xattr_value_root *xv;
	size_t size;
	int ret = 0;

	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL))
		return -ENODATA;

	xs->end = (void *)di + inode->i_sb->s_blocksize;
	xs->header = (struct ocfs2_xattr_header *)
			(xs->end - le16_to_cpu(di->i_xattr_inline_size));
	xs->base = (void *)xs->header;
	xs->here = xs->header->xh_entries;

	ret = ocfs2_xattr_find_entry(name_index, name, xs);
	if (ret)
		return ret;
	size = le64_to_cpu(xs->here->xe_value_size);
	if (buffer) {
		if (size > buffer_size)
			return -ERANGE;
		if (ocfs2_xattr_is_local(xs->here)) {
			memcpy(buffer, (void *)xs->base +
			       le16_to_cpu(xs->here->xe_name_offset) +
			       OCFS2_XATTR_SIZE(xs->here->xe_name_len), size);
		} else {
			xv = (struct ocfs2_xattr_value_root *)
				(xs->base + le16_to_cpu(
				 xs->here->xe_name_offset) +
				OCFS2_XATTR_SIZE(xs->here->xe_name_len));
			ret = ocfs2_xattr_get_value_outside(inode, xv,
							    buffer, size);
			if (ret < 0) {
				mlog_errno(ret);
				return ret;
			}
		}
	}

	return size;
}

static int ocfs2_xattr_block_get(struct inode *inode,
				 int name_index,
				 const char *name,
				 void *buffer,
				 size_t buffer_size,
				 struct ocfs2_xattr_search *xs)
{
	struct ocfs2_xattr_block *xb;
	struct ocfs2_xattr_value_root *xv;
	size_t size;
	int ret = -ENODATA, name_offset, name_len, i;
	int uninitialized_var(block_off);

	xs->bucket = ocfs2_xattr_bucket_new(inode);
	if (!xs->bucket) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto cleanup;
	}

	ret = ocfs2_xattr_block_find(inode, name_index, name, xs);
	if (ret) {
		mlog_errno(ret);
		goto cleanup;
	}

	if (xs->not_found) {
		ret = -ENODATA;
		goto cleanup;
	}

	xb = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data;
	size = le64_to_cpu(xs->here->xe_value_size);
	if (buffer) {
		ret = -ERANGE;
		if (size > buffer_size)
			goto cleanup;

		name_offset = le16_to_cpu(xs->here->xe_name_offset);
		name_len = OCFS2_XATTR_SIZE(xs->here->xe_name_len);
		i = xs->here - xs->header->xh_entries;

		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
			ret = ocfs2_xattr_bucket_get_name_value(inode->i_sb,
								bucket_xh(xs->bucket),
								i,
								&block_off,
								&name_offset);
			xs->base = bucket_block(xs->bucket, block_off);
		}
		if (ocfs2_xattr_is_local(xs->here)) {
			memcpy(buffer, (void *)xs->base +
			       name_offset + name_len, size);
		} else {
			xv = (struct ocfs2_xattr_value_root *)
				(xs->base + name_offset + name_len);
			ret = ocfs2_xattr_get_value_outside(inode, xv,
							    buffer, size);
			if (ret < 0) {
				mlog_errno(ret);
				goto cleanup;
			}
		}
	}
	ret = size;
cleanup:
	ocfs2_xattr_bucket_free(xs->bucket);

	brelse(xs->xattr_bh);
	xs->xattr_bh = NULL;
	return ret;
}

int ocfs2_xattr_get_nolock(struct inode *inode,
			   struct buffer_head *di_bh,
			   int name_index,
			   const char *name,
			   void *buffer,
			   size_t buffer_size)
{
	int ret;
	struct ocfs2_dinode *di = NULL;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_xattr_search xis = {
		.not_found = -ENODATA,
	};
	struct ocfs2_xattr_search xbs = {
		.not_found = -ENODATA,
	};

	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
		return -EOPNOTSUPP;

	if (!(oi->ip_dyn_features & OCFS2_HAS_XATTR_FL))
		ret = -ENODATA;

	xis.inode_bh = xbs.inode_bh = di_bh;
	di = (struct ocfs2_dinode *)di_bh->b_data;

	down_read(&oi->ip_xattr_sem);
	ret = ocfs2_xattr_ibody_get(inode, name_index, name, buffer,
				    buffer_size, &xis);
	if (ret == -ENODATA && di->i_xattr_loc)
		ret = ocfs2_xattr_block_get(inode, name_index, name, buffer,
					    buffer_size, &xbs);
	up_read(&oi->ip_xattr_sem);

	return ret;
}

/* ocfs2_xattr_get()
 *
 * Copy an extended attribute into the buffer provided.
 * Buffer is NULL to compute the size of buffer required.
 */
static int ocfs2_xattr_get(struct inode *inode,
			   int name_index,
			   const char *name,
			   void *buffer,
			   size_t buffer_size)
{
	int ret;
	struct buffer_head *di_bh = NULL;

	ret = ocfs2_inode_lock(inode, &di_bh, 0);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}
	ret = ocfs2_xattr_get_nolock(inode, di_bh, name_index,
				     name, buffer, buffer_size);

	ocfs2_inode_unlock(inode, 0);

	brelse(di_bh);

	return ret;
}

static int __ocfs2_xattr_set_value_outside(struct inode *inode,
					   handle_t *handle,
					   struct ocfs2_xattr_value_buf *vb,
					   const void *value,
					   int value_len)
{
	int ret = 0, i, cp_len;
	u16 blocksize = inode->i_sb->s_blocksize;
	u32 p_cluster, num_clusters;
	u32 cpos = 0, bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
	u32 clusters = ocfs2_clusters_for_bytes(inode->i_sb, value_len);
	u64 blkno;
	struct buffer_head *bh = NULL;
	unsigned int ext_flags;
	struct ocfs2_xattr_value_root *xv = vb->vb_xv;

	BUG_ON(clusters > le32_to_cpu(xv->xr_clusters));

	while (cpos < clusters) {
		ret = ocfs2_xattr_get_clusters(inode, cpos, &p_cluster,
					       &num_clusters, &xv->xr_list,
					       &ext_flags);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		BUG_ON(ext_flags & OCFS2_EXT_REFCOUNTED);

		blkno = ocfs2_clusters_to_blocks(inode->i_sb, p_cluster);

		for (i = 0; i < num_clusters * bpc; i++, blkno++) {
			ret = ocfs2_read_block(INODE_CACHE(inode), blkno,
					       &bh, NULL);
			if (ret) {
				mlog_errno(ret);
				goto out;
			}

			ret = ocfs2_journal_access(handle,
						   INODE_CACHE(inode),
						   bh,
						   OCFS2_JOURNAL_ACCESS_WRITE);
			if (ret < 0) {
				mlog_errno(ret);
				goto out;
			}

			cp_len = value_len > blocksize ? blocksize : value_len;
			memcpy(bh->b_data, value, cp_len);
			value_len -= cp_len;
			value += cp_len;
			if (cp_len < blocksize)
				memset(bh->b_data + cp_len, 0,
				       blocksize - cp_len);

			ret = ocfs2_journal_dirty(handle, bh);
			if (ret < 0) {
				mlog_errno(ret);
				goto out;
			}
			brelse(bh);
			bh = NULL;

			/*
			 * XXX: do we need to empty all the following
			 * blocks in this cluster?
			 */
			if (!value_len)
				break;
		}
		cpos += num_clusters;
	}
out:
	brelse(bh);

	return ret;
}

static int ocfs2_xattr_cleanup(struct inode *inode,
			       handle_t *handle,
			       struct ocfs2_xattr_info *xi,
			       struct ocfs2_xattr_search *xs,
			       struct ocfs2_xattr_value_buf *vb,
			       size_t offs)
{
	int ret = 0;
	size_t name_len = strlen(xi->name);
	void *val = xs->base + offs;
	size_t size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;

	ret = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
			    OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}
	/* Decrease xattr count */
	le16_add_cpu(&xs->header->xh_count, -1);
	/* Remove the xattr entry and tree root which has already be set*/
	memset((void *)xs->here, 0, sizeof(struct ocfs2_xattr_entry));
	memset(val, 0, size);

	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
	if (ret < 0)
		mlog_errno(ret);
out:
	return ret;
}

static int ocfs2_xattr_update_entry(struct inode *inode,
				    handle_t *handle,
				    struct ocfs2_xattr_info *xi,
				    struct ocfs2_xattr_search *xs,
				    struct ocfs2_xattr_value_buf *vb,
				    size_t offs)
{
	int ret;

	ret = vb->vb_access(handle, INODE_CACHE(inode), vb->vb_bh,
			    OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	xs->here->xe_name_offset = cpu_to_le16(offs);
	xs->here->xe_value_size = cpu_to_le64(xi->value_len);
	if (xi->value_len <= OCFS2_XATTR_INLINE_SIZE)
		ocfs2_xattr_set_local(xs->here, 1);
	else
		ocfs2_xattr_set_local(xs->here, 0);
	ocfs2_xattr_hash_entry(inode, xs->header, xs->here);

	ret = ocfs2_journal_dirty(handle, vb->vb_bh);
	if (ret < 0)
		mlog_errno(ret);
out:
	return ret;
}

/*
 * ocfs2_xattr_set_value_outside()
 *
 * Set large size value in B tree.
 */
static int ocfs2_xattr_set_value_outside(struct inode *inode,
					 struct ocfs2_xattr_info *xi,
					 struct ocfs2_xattr_search *xs,
					 struct ocfs2_xattr_set_ctxt *ctxt,
					 struct ocfs2_xattr_value_buf *vb,
					 size_t offs)
{
	size_t name_len = strlen(xi->name);
	void *val = xs->base + offs;
	struct ocfs2_xattr_value_root *xv = NULL;
	size_t size = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
	int ret = 0;

	memset(val, 0, size);
	memcpy(val, xi->name, name_len);
	xv = (struct ocfs2_xattr_value_root *)
		(val + OCFS2_XATTR_SIZE(name_len));
	xv->xr_clusters = 0;
	xv->xr_last_eb_blk = 0;
	xv->xr_list.l_tree_depth = 0;
	xv->xr_list.l_count = cpu_to_le16(1);
	xv->xr_list.l_next_free_rec = 0;
	vb->vb_xv = xv;

	ret = ocfs2_xattr_value_truncate(inode, vb, xi->value_len, ctxt);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}
	ret = ocfs2_xattr_update_entry(inode, ctxt->handle, xi, xs, vb, offs);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}
	ret = __ocfs2_xattr_set_value_outside(inode, ctxt->handle, vb,
					      xi->value, xi->value_len);
	if (ret < 0)
		mlog_errno(ret);

	return ret;
}

/*
 * ocfs2_xattr_set_entry_local()
 *
 * Set, replace or remove extended attribute in local.
 */
static void ocfs2_xattr_set_entry_local(struct inode *inode,
					struct ocfs2_xattr_info *xi,
					struct ocfs2_xattr_search *xs,
					struct ocfs2_xattr_entry *last,
					size_t min_offs)
{
	size_t name_len = strlen(xi->name);
	int i;

	if (xi->value && xs->not_found) {
		/* Insert the new xattr entry. */
		le16_add_cpu(&xs->header->xh_count, 1);
		ocfs2_xattr_set_type(last, xi->name_index);
		ocfs2_xattr_set_local(last, 1);
		last->xe_name_len = name_len;
	} else {
		void *first_val;
		void *val;
		size_t offs, size;

		first_val = xs->base + min_offs;
		offs = le16_to_cpu(xs->here->xe_name_offset);
		val = xs->base + offs;

		if (le64_to_cpu(xs->here->xe_value_size) >
		    OCFS2_XATTR_INLINE_SIZE)
			size = OCFS2_XATTR_SIZE(name_len) +
				OCFS2_XATTR_ROOT_SIZE;
		else
			size = OCFS2_XATTR_SIZE(name_len) +
			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));

		if (xi->value && size == OCFS2_XATTR_SIZE(name_len) +
				OCFS2_XATTR_SIZE(xi->value_len)) {
			/* The old and the new value have the
			   same size. Just replace the value. */
			ocfs2_xattr_set_local(xs->here, 1);
			xs->here->xe_value_size = cpu_to_le64(xi->value_len);
			/* Clear value bytes. */
			memset(val + OCFS2_XATTR_SIZE(name_len),
			       0,
			       OCFS2_XATTR_SIZE(xi->value_len));
			memcpy(val + OCFS2_XATTR_SIZE(name_len),
			       xi->value,
			       xi->value_len);
			return;
		}
		/* Remove the old name+value. */
		memmove(first_val + size, first_val, val - first_val);
		memset(first_val, 0, size);
		xs->here->xe_name_hash = 0;
		xs->here->xe_name_offset = 0;
		ocfs2_xattr_set_local(xs->here, 1);
		xs->here->xe_value_size = 0;

		min_offs += size;

		/* Adjust all value offsets. */
		last = xs->header->xh_entries;
		for (i = 0 ; i < le16_to_cpu(xs->header->xh_count); i++) {
			size_t o = le16_to_cpu(last->xe_name_offset);

			if (o < offs)
				last->xe_name_offset = cpu_to_le16(o + size);
			last += 1;
		}

		if (!xi->value) {
			/* Remove the old entry. */
			last -= 1;
			memmove(xs->here, xs->here + 1,
				(void *)last - (void *)xs->here);
			memset(last, 0, sizeof(struct ocfs2_xattr_entry));
			le16_add_cpu(&xs->header->xh_count, -1);
		}
	}
	if (xi->value) {
		/* Insert the new name+value. */
		size_t size = OCFS2_XATTR_SIZE(name_len) +
				OCFS2_XATTR_SIZE(xi->value_len);
		void *val = xs->base + min_offs - size;

		xs->here->xe_name_offset = cpu_to_le16(min_offs - size);
		memset(val, 0, size);
		memcpy(val, xi->name, name_len);
		memcpy(val + OCFS2_XATTR_SIZE(name_len),
		       xi->value,
		       xi->value_len);
		xs->here->xe_value_size = cpu_to_le64(xi->value_len);
		ocfs2_xattr_set_local(xs->here, 1);
		ocfs2_xattr_hash_entry(inode, xs->header, xs->here);
	}

	return;
}

/*
 * ocfs2_xattr_set_entry()
 *
 * Set extended attribute entry into inode or block.
 *
 * If extended attribute value size > OCFS2_XATTR_INLINE_SIZE,
 * We first insert tree root(ocfs2_xattr_value_root) with set_entry_local(),
 * then set value in B tree with set_value_outside().
 */
static int ocfs2_xattr_set_entry(struct inode *inode,
				 struct ocfs2_xattr_info *xi,
				 struct ocfs2_xattr_search *xs,
				 struct ocfs2_xattr_set_ctxt *ctxt,
				 int flag)
{
	struct ocfs2_xattr_entry *last;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
	size_t min_offs = xs->end - xs->base, name_len = strlen(xi->name);
	size_t size_l = 0;
	handle_t *handle = ctxt->handle;
	int free, i, ret;
	struct ocfs2_xattr_info xi_l = {
		.name_index = xi->name_index,
		.name = xi->name,
		.value = xi->value,
		.value_len = xi->value_len,
	};
	struct ocfs2_xattr_value_buf vb = {
		.vb_bh = xs->xattr_bh,
		.vb_access = ocfs2_journal_access_di,
	};

	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
		BUG_ON(xs->xattr_bh == xs->inode_bh);
		vb.vb_access = ocfs2_journal_access_xb;
	} else
		BUG_ON(xs->xattr_bh != xs->inode_bh);

	/* Compute min_offs, last and free space. */
	last = xs->header->xh_entries;

	for (i = 0 ; i < le16_to_cpu(xs->header->xh_count); i++) {
		size_t offs = le16_to_cpu(last->xe_name_offset);
		if (offs < min_offs)
			min_offs = offs;
		last += 1;
	}

	free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP;
	if (free < 0)
		return -EIO;

	if (!xs->not_found) {
		size_t size = 0;
		if (ocfs2_xattr_is_local(xs->here))
			size = OCFS2_XATTR_SIZE(name_len) +
			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));
		else
			size = OCFS2_XATTR_SIZE(name_len) +
				OCFS2_XATTR_ROOT_SIZE;
		free += (size + sizeof(struct ocfs2_xattr_entry));
	}
	/* Check free space in inode or block */
	if (xi->value && xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
		if (free < sizeof(struct ocfs2_xattr_entry) +
			   OCFS2_XATTR_SIZE(name_len) +
			   OCFS2_XATTR_ROOT_SIZE) {
			ret = -ENOSPC;
			goto out;
		}
		size_l = OCFS2_XATTR_SIZE(name_len) + OCFS2_XATTR_ROOT_SIZE;
		xi_l.value = (void *)&def_xv;
		xi_l.value_len = OCFS2_XATTR_ROOT_SIZE;
	} else if (xi->value) {
		if (free < sizeof(struct ocfs2_xattr_entry) +
			   OCFS2_XATTR_SIZE(name_len) +
			   OCFS2_XATTR_SIZE(xi->value_len)) {
			ret = -ENOSPC;
			goto out;
		}
	}

	if (!xs->not_found) {
		/* For existing extended attribute */
		size_t size = OCFS2_XATTR_SIZE(name_len) +
			OCFS2_XATTR_SIZE(le64_to_cpu(xs->here->xe_value_size));
		size_t offs = le16_to_cpu(xs->here->xe_name_offset);
		void *val = xs->base + offs;

		if (ocfs2_xattr_is_local(xs->here) && size == size_l) {
			/* Replace existing local xattr with tree root */
			ret = ocfs2_xattr_set_value_outside(inode, xi, xs,
							    ctxt, &vb, offs);
			if (ret < 0)
				mlog_errno(ret);
			goto out;
		} else if (!ocfs2_xattr_is_local(xs->here)) {
			/* For existing xattr which has value outside */
			vb.vb_xv = (struct ocfs2_xattr_value_root *)
				(val + OCFS2_XATTR_SIZE(name_len));

			if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
				/*
				 * If new value need set outside also,
				 * first truncate old value to new value,
				 * then set new value with set_value_outside().
				 */
				ret = ocfs2_xattr_value_truncate(inode,
								 &vb,
								 xi->value_len,
								 ctxt);
				if (ret < 0) {
					mlog_errno(ret);
					goto out;
				}

				ret = ocfs2_xattr_update_entry(inode,
							       handle,
							       xi,
							       xs,
							       &vb,
							       offs);
				if (ret < 0) {
					mlog_errno(ret);
					goto out;
				}

				ret = __ocfs2_xattr_set_value_outside(inode,
								handle,
								&vb,
								xi->value,
								xi->value_len);
				if (ret < 0)
					mlog_errno(ret);
				goto out;
			} else {
				/*
				 * If new value need set in local,
				 * just trucate old value to zero.
				 */
				 ret = ocfs2_xattr_value_truncate(inode,
								  &vb,
								  0,
								  ctxt);
				if (ret < 0)
					mlog_errno(ret);
			}
		}
	}

	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), xs->inode_bh,
				      OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
		ret = vb.vb_access(handle, INODE_CACHE(inode), vb.vb_bh,
				   OCFS2_JOURNAL_ACCESS_WRITE);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}
	}

	/*
	 * Set value in local, include set tree root in local.
	 * This is the first step for value size >INLINE_SIZE.
	 */
	ocfs2_xattr_set_entry_local(inode, &xi_l, xs, last, min_offs);

	if (!(flag & OCFS2_INLINE_XATTR_FL)) {
		ret = ocfs2_journal_dirty(handle, xs->xattr_bh);
		if (ret < 0) {
			mlog_errno(ret);
			goto out;
		}
	}

	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL) &&
	    (flag & OCFS2_INLINE_XATTR_FL)) {
		struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
		unsigned int xattrsize = osb->s_xattr_inline_size;

		/*
		 * Adjust extent record count or inline data size
		 * to reserve space for extended attribute.
		 */
		if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
			struct ocfs2_inline_data *idata = &di->id2.i_data;
			le16_add_cpu(&idata->id_count, -xattrsize);
		} else if (!(ocfs2_inode_is_fast_symlink(inode))) {
			struct ocfs2_extent_list *el = &di->id2.i_list;
			le16_add_cpu(&el->l_count, -(xattrsize /
					sizeof(struct ocfs2_extent_rec)));
		}
		di->i_xattr_inline_size = cpu_to_le16(xattrsize);
	}
	/* Update xattr flag */
	spin_lock(&oi->ip_lock);
	oi->ip_dyn_features |= flag;
	di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features);
	spin_unlock(&oi->ip_lock);

	ret = ocfs2_journal_dirty(handle, xs->inode_bh);
	if (ret < 0)
		mlog_errno(ret);

	if (!ret && xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
		/*
		 * Set value outside in B tree.
		 * This is the second step for value size > INLINE_SIZE.
		 */
		size_t offs = le16_to_cpu(xs->here->xe_name_offset);
		ret = ocfs2_xattr_set_value_outside(inode, xi, xs, ctxt,
						    &vb, offs);
		if (ret < 0) {
			int ret2;

			mlog_errno(ret);
			/*
			 * If set value outside failed, we have to clean
			 * the junk tree root we have already set in local.
			 */
			ret2 = ocfs2_xattr_cleanup(inode, ctxt->handle,
						   xi, xs, &vb, offs);
			if (ret2 < 0)
				mlog_errno(ret2);
		}
	}
out:
	return ret;
}

/*
 * In xattr remove, if it is stored outside and refcounted, we may have
 * the chance to split the refcount tree. So need the allocators.
 */
static int ocfs2_lock_xattr_remove_allocators(struct inode *inode,
					struct ocfs2_xattr_value_root *xv,
					struct ocfs2_caching_info *ref_ci,
					struct buffer_head *ref_root_bh,
					struct ocfs2_alloc_context **meta_ac,
					int *ref_credits)
{
	int ret, meta_add = 0;
	u32 p_cluster, num_clusters;
	unsigned int ext_flags;

	*ref_credits = 0;
	ret = ocfs2_xattr_get_clusters(inode, 0, &p_cluster,
				       &num_clusters,
				       &xv->xr_list,
				       &ext_flags);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	if (!(ext_flags & OCFS2_EXT_REFCOUNTED))
		goto out;

	ret = ocfs2_refcounted_xattr_delete_need(inode, ref_ci,
						 ref_root_bh, xv,
						 &meta_add, ref_credits);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_reserve_new_metadata_blocks(OCFS2_SB(inode->i_sb),
						meta_add, meta_ac);
	if (ret)
		mlog_errno(ret);

out:
	return ret;
}

static int ocfs2_remove_value_outside(struct inode*inode,
				      struct ocfs2_xattr_value_buf *vb,
				      struct ocfs2_xattr_header *header,
				      struct ocfs2_caching_info *ref_ci,
				      struct buffer_head *ref_root_bh)
{
	int ret = 0, i, ref_credits;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct ocfs2_xattr_set_ctxt ctxt = { NULL, NULL, };
	void *val;

	ocfs2_init_dealloc_ctxt(&ctxt.dealloc);

	for (i = 0; i < le16_to_cpu(header->xh_count); i++) {
		struct ocfs2_xattr_entry *entry = &header->xh_entries[i];

		if (ocfs2_xattr_is_local(entry))
			continue;

		val = (void *)header +
			le16_to_cpu(entry->xe_name_offset);
		vb->vb_xv = (struct ocfs2_xattr_value_root *)
			(val + OCFS2_XATTR_SIZE(entry->xe_name_len));

		ret = ocfs2_lock_xattr_remove_allocators(inode, vb->vb_xv,
							 ref_ci, ref_root_bh,
							 &ctxt.meta_ac,
							 &ref_credits);

		ctxt.handle = ocfs2_start_trans(osb, ref_credits +
					ocfs2_remove_extent_credits(osb->sb));
		if (IS_ERR(ctxt.handle)) {
			ret = PTR_ERR(ctxt.handle);
			mlog_errno(ret);
			break;
		}

		ret = ocfs2_xattr_value_truncate(inode, vb, 0, &ctxt);
		if (ret < 0) {
			mlog_errno(ret);
			break;
		}

		ocfs2_commit_trans(osb, ctxt.handle);
		if (ctxt.meta_ac) {
			ocfs2_free_alloc_context(ctxt.meta_ac);
			ctxt.meta_ac = NULL;
		}
	}

	if (ctxt.meta_ac)
		ocfs2_free_alloc_context(ctxt.meta_ac);
	ocfs2_schedule_truncate_log_flush(osb, 1);
	ocfs2_run_deallocs(osb, &ctxt.dealloc);
	return ret;
}

static int ocfs2_xattr_ibody_remove(struct inode *inode,
				    struct buffer_head *di_bh,
				    struct ocfs2_caching_info *ref_ci,
				    struct buffer_head *ref_root_bh)
{

	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
	struct ocfs2_xattr_header *header;
	int ret;
	struct ocfs2_xattr_value_buf vb = {
		.vb_bh = di_bh,
		.vb_access = ocfs2_journal_access_di,
	};

	header = (struct ocfs2_xattr_header *)
		 ((void *)di + inode->i_sb->s_blocksize -
		 le16_to_cpu(di->i_xattr_inline_size));

	ret = ocfs2_remove_value_outside(inode, &vb, header,
					 ref_ci, ref_root_bh);

	return ret;
}

struct ocfs2_rm_xattr_bucket_para {
	struct ocfs2_caching_info *ref_ci;
	struct buffer_head *ref_root_bh;
};

static int ocfs2_xattr_block_remove(struct inode *inode,
				    struct buffer_head *blk_bh,
				    struct ocfs2_caching_info *ref_ci,
				    struct buffer_head *ref_root_bh)
{
	struct ocfs2_xattr_block *xb;
	int ret = 0;
	struct ocfs2_xattr_value_buf vb = {
		.vb_bh = blk_bh,
		.vb_access = ocfs2_journal_access_xb,
	};
	struct ocfs2_rm_xattr_bucket_para args = {
		.ref_ci = ref_ci,
		.ref_root_bh = ref_root_bh,
	};

	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
		struct ocfs2_xattr_header *header = &(xb->xb_attrs.xb_header);
		ret = ocfs2_remove_value_outside(inode, &vb, header,
						 ref_ci, ref_root_bh);
	} else
		ret = ocfs2_iterate_xattr_index_block(inode,
						blk_bh,
						ocfs2_rm_xattr_cluster,
						&args);

	return ret;
}

static int ocfs2_xattr_free_block(struct inode *inode,
				  u64 block,
				  struct ocfs2_caching_info *ref_ci,
				  struct buffer_head *ref_root_bh)
{
	struct inode *xb_alloc_inode;
	struct buffer_head *xb_alloc_bh = NULL;
	struct buffer_head *blk_bh = NULL;
	struct ocfs2_xattr_block *xb;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	handle_t *handle;
	int ret = 0;
	u64 blk, bg_blkno;
	u16 bit;

	ret = ocfs2_read_xattr_block(inode, block, &blk_bh);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_xattr_block_remove(inode, blk_bh, ref_ci, ref_root_bh);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;
	blk = le64_to_cpu(xb->xb_blkno);
	bit = le16_to_cpu(xb->xb_suballoc_bit);
	bg_blkno = ocfs2_which_suballoc_group(blk, bit);

	xb_alloc_inode = ocfs2_get_system_file_inode(osb,
				EXTENT_ALLOC_SYSTEM_INODE,
				le16_to_cpu(xb->xb_suballoc_slot));
	if (!xb_alloc_inode) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto out;
	}
	mutex_lock(&xb_alloc_inode->i_mutex);

	ret = ocfs2_inode_lock(xb_alloc_inode, &xb_alloc_bh, 1);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_mutex;
	}

	handle = ocfs2_start_trans(osb, OCFS2_SUBALLOC_FREE);
	if (IS_ERR(handle)) {
		ret = PTR_ERR(handle);
		mlog_errno(ret);
		goto out_unlock;
	}

	ret = ocfs2_free_suballoc_bits(handle, xb_alloc_inode, xb_alloc_bh,
				       bit, bg_blkno, 1);
	if (ret < 0)
		mlog_errno(ret);

	ocfs2_commit_trans(osb, handle);
out_unlock:
	ocfs2_inode_unlock(xb_alloc_inode, 1);
	brelse(xb_alloc_bh);
out_mutex:
	mutex_unlock(&xb_alloc_inode->i_mutex);
	iput(xb_alloc_inode);
out:
	brelse(blk_bh);
	return ret;
}

/*
 * ocfs2_xattr_remove()
 *
 * Free extended attribute resources associated with this inode.
 */
int ocfs2_xattr_remove(struct inode *inode, struct buffer_head *di_bh)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
	struct ocfs2_refcount_tree *ref_tree = NULL;
	struct buffer_head *ref_root_bh = NULL;
	struct ocfs2_caching_info *ref_ci = NULL;
	handle_t *handle;
	int ret;

	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
		return 0;

	if (!(oi->ip_dyn_features & OCFS2_HAS_XATTR_FL))
		return 0;

	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL) {
		ret = ocfs2_lock_refcount_tree(OCFS2_SB(inode->i_sb),
					       le64_to_cpu(di->i_refcount_loc),
					       1, &ref_tree, &ref_root_bh);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}
		ref_ci = &ref_tree->rf_ci;

	}

	if (oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL) {
		ret = ocfs2_xattr_ibody_remove(inode, di_bh,
					       ref_ci, ref_root_bh);
		if (ret < 0) {
			mlog_errno(ret);
			goto out;
		}
	}

	if (di->i_xattr_loc) {
		ret = ocfs2_xattr_free_block(inode,
					     le64_to_cpu(di->i_xattr_loc),
					     ref_ci, ref_root_bh);
		if (ret < 0) {
			mlog_errno(ret);
			goto out;
		}
	}

	handle = ocfs2_start_trans((OCFS2_SB(inode->i_sb)),
				   OCFS2_INODE_UPDATE_CREDITS);
	if (IS_ERR(handle)) {
		ret = PTR_ERR(handle);
		mlog_errno(ret);
		goto out;
	}
	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh,
				      OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret) {
		mlog_errno(ret);
		goto out_commit;
	}

	di->i_xattr_loc = 0;

	spin_lock(&oi->ip_lock);
	oi->ip_dyn_features &= ~(OCFS2_INLINE_XATTR_FL | OCFS2_HAS_XATTR_FL);
	di->i_dyn_features = cpu_to_le16(oi->ip_dyn_features);
	spin_unlock(&oi->ip_lock);

	ret = ocfs2_journal_dirty(handle, di_bh);
	if (ret < 0)
		mlog_errno(ret);
out_commit:
	ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
out:
	if (ref_tree)
		ocfs2_unlock_refcount_tree(OCFS2_SB(inode->i_sb), ref_tree, 1);
	brelse(ref_root_bh);
	return ret;
}

static int ocfs2_xattr_has_space_inline(struct inode *inode,
					struct ocfs2_dinode *di)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	unsigned int xattrsize = OCFS2_SB(inode->i_sb)->s_xattr_inline_size;
	int free;

	if (xattrsize < OCFS2_MIN_XATTR_INLINE_SIZE)
		return 0;

	if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
		struct ocfs2_inline_data *idata = &di->id2.i_data;
		free = le16_to_cpu(idata->id_count) - le64_to_cpu(di->i_size);
	} else if (ocfs2_inode_is_fast_symlink(inode)) {
		free = ocfs2_fast_symlink_chars(inode->i_sb) -
			le64_to_cpu(di->i_size);
	} else {
		struct ocfs2_extent_list *el = &di->id2.i_list;
		free = (le16_to_cpu(el->l_count) -
			le16_to_cpu(el->l_next_free_rec)) *
			sizeof(struct ocfs2_extent_rec);
	}
	if (free >= xattrsize)
		return 1;

	return 0;
}

/*
 * ocfs2_xattr_ibody_find()
 *
 * Find extended attribute in inode block and
 * fill search info into struct ocfs2_xattr_search.
 */
static int ocfs2_xattr_ibody_find(struct inode *inode,
				  int name_index,
				  const char *name,
				  struct ocfs2_xattr_search *xs)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
	int ret;
	int has_space = 0;

	if (inode->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE)
		return 0;

	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL)) {
		down_read(&oi->ip_alloc_sem);
		has_space = ocfs2_xattr_has_space_inline(inode, di);
		up_read(&oi->ip_alloc_sem);
		if (!has_space)
			return 0;
	}

	xs->xattr_bh = xs->inode_bh;
	xs->end = (void *)di + inode->i_sb->s_blocksize;
	if (oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL)
		xs->header = (struct ocfs2_xattr_header *)
			(xs->end - le16_to_cpu(di->i_xattr_inline_size));
	else
		xs->header = (struct ocfs2_xattr_header *)
			(xs->end - OCFS2_SB(inode->i_sb)->s_xattr_inline_size);
	xs->base = (void *)xs->header;
	xs->here = xs->header->xh_entries;

	/* Find the named attribute. */
	if (oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL) {
		ret = ocfs2_xattr_find_entry(name_index, name, xs);
		if (ret && ret != -ENODATA)
			return ret;
		xs->not_found = ret;
	}

	return 0;
}

/*
 * ocfs2_xattr_ibody_set()
 *
 * Set, replace or remove an extended attribute into inode block.
 *
 */
static int ocfs2_xattr_ibody_set(struct inode *inode,
				 struct ocfs2_xattr_info *xi,
				 struct ocfs2_xattr_search *xs,
				 struct ocfs2_xattr_set_ctxt *ctxt)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
	int ret;

	if (inode->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE)
		return -ENOSPC;

	down_write(&oi->ip_alloc_sem);
	if (!(oi->ip_dyn_features & OCFS2_INLINE_XATTR_FL)) {
		if (!ocfs2_xattr_has_space_inline(inode, di)) {
			ret = -ENOSPC;
			goto out;
		}
	}

	ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
				(OCFS2_INLINE_XATTR_FL | OCFS2_HAS_XATTR_FL));
out:
	up_write(&oi->ip_alloc_sem);

	return ret;
}

/*
 * ocfs2_xattr_block_find()
 *
 * Find extended attribute in external block and
 * fill search info into struct ocfs2_xattr_search.
 */
static int ocfs2_xattr_block_find(struct inode *inode,
				  int name_index,
				  const char *name,
				  struct ocfs2_xattr_search *xs)
{
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data;
	struct buffer_head *blk_bh = NULL;
	struct ocfs2_xattr_block *xb;
	int ret = 0;

	if (!di->i_xattr_loc)
		return ret;

	ret = ocfs2_read_xattr_block(inode, le64_to_cpu(di->i_xattr_loc),
				     &blk_bh);
	if (ret < 0) {
		mlog_errno(ret);
		return ret;
	}

	xs->xattr_bh = blk_bh;
	xb = (struct ocfs2_xattr_block *)blk_bh->b_data;

	if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) {
		xs->header = &xb->xb_attrs.xb_header;
		xs->base = (void *)xs->header;
		xs->end = (void *)(blk_bh->b_data) + blk_bh->b_size;
		xs->here = xs->header->xh_entries;

		ret = ocfs2_xattr_find_entry(name_index, name, xs);
	} else
		ret = ocfs2_xattr_index_block_find(inode, blk_bh,
						   name_index,
						   name, xs);

	if (ret && ret != -ENODATA) {
		xs->xattr_bh = NULL;
		goto cleanup;
	}
	xs->not_found = ret;
	return 0;
cleanup:
	brelse(blk_bh);

	return ret;
}

static int ocfs2_create_xattr_block(handle_t *handle,
				    struct inode *inode,
				    struct buffer_head *inode_bh,
				    struct ocfs2_alloc_context *meta_ac,
				    struct buffer_head **ret_bh,
				    int indexed)
{
	int ret;
	u16 suballoc_bit_start;
	u32 num_got;
	u64 first_blkno;
	struct ocfs2_dinode *di =  (struct ocfs2_dinode *)inode_bh->b_data;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct buffer_head *new_bh = NULL;
	struct ocfs2_xattr_block *xblk;

	ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), inode_bh,
				      OCFS2_JOURNAL_ACCESS_CREATE);
	if (ret < 0) {
		mlog_errno(ret);
		goto end;
	}

	ret = ocfs2_claim_metadata(osb, handle, meta_ac, 1,
				   &suballoc_bit_start, &num_got,
				   &first_blkno);
	if (ret < 0) {
		mlog_errno(ret);
		goto end;
	}

	new_bh = sb_getblk(inode->i_sb, first_blkno);
	ocfs2_set_new_buffer_uptodate(INODE_CACHE(inode), new_bh);

	ret = ocfs2_journal_access_xb(handle, INODE_CACHE(inode),
				      new_bh,
				      OCFS2_JOURNAL_ACCESS_CREATE);
	if (ret < 0) {
		mlog_errno(ret);
		goto end;
	}

	/* Initialize ocfs2_xattr_block */
	xblk = (struct ocfs2_xattr_block *)new_bh->b_data;
	memset(xblk, 0, inode->i_sb->s_blocksize);
	strcpy((void *)xblk, OCFS2_XATTR_BLOCK_SIGNATURE);
	xblk->xb_suballoc_slot = cpu_to_le16(osb->slot_num);
	xblk->xb_suballoc_bit = cpu_to_le16(suballoc_bit_start);
	xblk->xb_fs_generation = cpu_to_le32(osb->fs_generation);
	xblk->xb_blkno = cpu_to_le64(first_blkno);

	if (indexed) {
		struct ocfs2_xattr_tree_root *xr = &xblk->xb_attrs.xb_root;
		xr->xt_clusters = cpu_to_le32(1);
		xr->xt_last_eb_blk = 0;
		xr->xt_list.l_tree_depth = 0;
		xr->xt_list.l_count = cpu_to_le16(
					ocfs2_xattr_recs_per_xb(inode->i_sb));
		xr->xt_list.l_next_free_rec = cpu_to_le16(1);
		xblk->xb_flags = cpu_to_le16(OCFS2_XATTR_INDEXED);
	}

	ret = ocfs2_journal_dirty(handle, new_bh);
	if (ret < 0) {
		mlog_errno(ret);
		goto end;
	}
	di->i_xattr_loc = cpu_to_le64(first_blkno);
	ocfs2_journal_dirty(handle, inode_bh);

	*ret_bh = new_bh;
	new_bh = NULL;

end:
	brelse(new_bh);
	return ret;
}

/*
 * ocfs2_xattr_block_set()
 *
 * Set, replace or remove an extended attribute into external block.
 *
 */
static int ocfs2_xattr_block_set(struct inode *inode,
				 struct ocfs2_xattr_info *xi,
				 struct ocfs2_xattr_search *xs,
				 struct ocfs2_xattr_set_ctxt *ctxt)
{
	struct buffer_head *new_bh = NULL;
	handle_t *handle = ctxt->handle;
	struct ocfs2_xattr_block *xblk = NULL;
	int ret;

	if (!xs->xattr_bh) {
		ret = ocfs2_create_xattr_block(handle, inode, xs->inode_bh,
					       ctxt->meta_ac, &new_bh, 0);
		if (ret) {
			mlog_errno(ret);
			goto end;
		}

		xs->xattr_bh = new_bh;
		xblk = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data;
		xs->header = &xblk->xb_attrs.xb_header;
		xs->base = (void *)xs->header;
		xs->end = (void *)xblk + inode->i_sb->s_blocksize;
		xs->here = xs->header->xh_entries;
	} else
		xblk = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data;

	if (!(le16_to_cpu(xblk->xb_flags) & OCFS2_XATTR_INDEXED)) {
		/* Set extended attribute into external block */
		ret = ocfs2_xattr_set_entry(inode, xi, xs, ctxt,
					    OCFS2_HAS_XATTR_FL);
		if (!ret || ret != -ENOSPC)
			goto end;

		ret = ocfs2_xattr_create_index_block(inode, xs, ctxt);
		if (ret)
			goto end;
	}

	ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs, ctxt);

end:

	return ret;
}

/* Check whether the new xattr can be inserted into the inode. */
static int ocfs2_xattr_can_be_in_inode(struct inode *inode,
				       struct ocfs2_xattr_info *xi,
				       struct ocfs2_xattr_search *xs)
{
	u64 value_size;
	struct ocfs2_xattr_entry *last;
	int free, i;
	size_t min_offs = xs->end - xs->base;

	if (!xs->header)
		return 0;

	last = xs->header->xh_entries;

	for (i = 0; i < le16_to_cpu(xs->header->xh_count); i++) {
		size_t offs = le16_to_cpu(last->xe_name_offset);
		if (offs < min_offs)
			min_offs = offs;
		last += 1;
	}

	free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP;
	if (free < 0)
		return 0;

	BUG_ON(!xs->not_found);

	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
		value_size = OCFS2_XATTR_ROOT_SIZE;
	else
		value_size = OCFS2_XATTR_SIZE(xi->value_len);

	if (free >= sizeof(struct ocfs2_xattr_entry) +
		   OCFS2_XATTR_SIZE(strlen(xi->name)) + value_size)
		return 1;

	return 0;
}

static int ocfs2_calc_xattr_set_need(struct inode *inode,
				     struct ocfs2_dinode *di,
				     struct ocfs2_xattr_info *xi,
				     struct ocfs2_xattr_search *xis,
				     struct ocfs2_xattr_search *xbs,
				     int *clusters_need,
				     int *meta_need,
				     int *credits_need)
{
	int ret = 0, old_in_xb = 0;
	int clusters_add = 0, meta_add = 0, credits = 0;
	struct buffer_head *bh = NULL;
	struct ocfs2_xattr_block *xb = NULL;
	struct ocfs2_xattr_entry *xe = NULL;
	struct ocfs2_xattr_value_root *xv = NULL;
	char *base = NULL;
	int name_offset, name_len = 0;
	u32 new_clusters = ocfs2_clusters_for_bytes(inode->i_sb,
						    xi->value_len);
	u64 value_size;

	/*
	 * Calculate the clusters we need to write.
	 * No matter whether we replace an old one or add a new one,
	 * we need this for writing.
	 */
	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE)
		credits += new_clusters *
			   ocfs2_clusters_to_blocks(inode->i_sb, 1);

	if (xis->not_found && xbs->not_found) {
		credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);

		if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
			clusters_add += new_clusters;
			credits += ocfs2_calc_extend_credits(inode->i_sb,
							&def_xv.xv.xr_list,
							new_clusters);
		}

		goto meta_guess;
	}

	if (!xis->not_found) {
		xe = xis->here;
		name_offset = le16_to_cpu(xe->xe_name_offset);
		name_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
		base = xis->base;
		credits += OCFS2_INODE_UPDATE_CREDITS;
	} else {
		int i, block_off = 0;
		xb = (struct ocfs2_xattr_block *)xbs->xattr_bh->b_data;
		xe = xbs->here;
		name_offset = le16_to_cpu(xe->xe_name_offset);
		name_len = OCFS2_XATTR_SIZE(xe->xe_name_len);
		i = xbs->here - xbs->header->xh_entries;
		old_in_xb = 1;

		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
			ret = ocfs2_xattr_bucket_get_name_value(inode->i_sb,
							bucket_xh(xbs->bucket),
							i, &block_off,
							&name_offset);
			base = bucket_block(xbs->bucket, block_off);
			credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
		} else {
			base = xbs->base;
			credits += OCFS2_XATTR_BLOCK_UPDATE_CREDITS;
		}
	}

	/*
	 * delete a xattr doesn't need metadata and cluster allocation.
	 * so just calculate the credits and return.
	 *
	 * The credits for removing the value tree will be extended
	 * by ocfs2_remove_extent itself.
	 */
	if (!xi->value) {
		if (!ocfs2_xattr_is_local(xe))
			credits += ocfs2_remove_extent_credits(inode->i_sb);

		goto out;
	}

	/* do cluster allocation guess first. */
	value_size = le64_to_cpu(xe->xe_value_size);

	if (old_in_xb) {
		/*
		 * In xattr set, we always try to set the xe in inode first,
		 * so if it can be inserted into inode successfully, the old
		 * one will be removed from the xattr block, and this xattr
		 * will be inserted into inode as a new xattr in inode.
		 */
		if (ocfs2_xattr_can_be_in_inode(inode, xi, xis)) {
			clusters_add += new_clusters;
			credits += ocfs2_remove_extent_credits(inode->i_sb) +
				    OCFS2_INODE_UPDATE_CREDITS;
			if (!ocfs2_xattr_is_local(xe))
				credits += ocfs2_calc_extend_credits(
							inode->i_sb,
							&def_xv.xv.xr_list,
							new_clusters);
			goto out;
		}
	}

	if (xi->value_len > OCFS2_XATTR_INLINE_SIZE) {
		/* the new values will be stored outside. */
		u32 old_clusters = 0;

		if (!ocfs2_xattr_is_local(xe)) {
			old_clusters =	ocfs2_clusters_for_bytes(inode->i_sb,
								 value_size);
			xv = (struct ocfs2_xattr_value_root *)
			     (base + name_offset + name_len);
			value_size = OCFS2_XATTR_ROOT_SIZE;
		} else
			xv = &def_xv.xv;

		if (old_clusters >= new_clusters) {
			credits += ocfs2_remove_extent_credits(inode->i_sb);
			goto out;
		} else {
			meta_add += ocfs2_extend_meta_needed(&xv->xr_list);
			clusters_add += new_clusters - old_clusters;
			credits += ocfs2_calc_extend_credits(inode->i_sb,
							     &xv->xr_list,
							     new_clusters -
							     old_clusters);
			if (value_size >= OCFS2_XATTR_ROOT_SIZE)
				goto out;
		}
	} else {
		/*
		 * Now the new value will be stored inside. So if the new
		 * value is smaller than the size of value root or the old
		 * value, we don't need any allocation, otherwise we have
		 * to guess metadata allocation.
		 */
		if ((ocfs2_xattr_is_local(xe) && value_size >= xi->value_len) ||
		    (!ocfs2_xattr_is_local(xe) &&
		     OCFS2_XATTR_ROOT_SIZE >= xi->value_len))
			goto out;
	}

meta_guess:
	/* calculate metadata allocation. */
	if (di->i_xattr_loc) {
		if (!xbs->xattr_bh) {
			ret = ocfs2_read_xattr_block(inode,
						     le64_to_cpu(di->i_xattr_loc),
						     &bh);
			if (ret) {
				mlog_errno(ret);
				goto out;
			}

			xb = (struct ocfs2_xattr_block *)bh->b_data;
		} else
			xb = (struct ocfs2_xattr_block *)xbs->xattr_bh->b_data;

		/*
		 * If there is already an xattr tree, good, we can calculate
		 * like other b-trees. Otherwise we may have the chance of
		 * create a tree, the credit calculation is borrowed from
		 * ocfs2_calc_extend_credits with root_el = NULL. And the
		 * new tree will be cluster based, so no meta is needed.
		 */
		if (le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED) {
			struct ocfs2_extent_list *el =
				 &xb->xb_attrs.xb_root.xt_list;
			meta_add += ocfs2_extend_meta_needed(el);
			credits += ocfs2_calc_extend_credits(inode->i_sb,
							     el, 1);
		} else
			credits += OCFS2_SUBALLOC_ALLOC + 1;

		/*
		 * This cluster will be used either for new bucket or for
		 * new xattr block.
		 * If the cluster size is the same as the bucket size, one
		 * more is needed since we may need to extend the bucket
		 * also.
		 */
		clusters_add += 1;
		credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
		if (OCFS2_XATTR_BUCKET_SIZE ==
			OCFS2_SB(inode->i_sb)->s_clustersize) {
			credits += ocfs2_blocks_per_xattr_bucket(inode->i_sb);
			clusters_add += 1;
		}
	} else {
		meta_add += 1;
		credits += OCFS2_XATTR_BLOCK_CREATE_CREDITS;
	}
out:
	if (clusters_need)
		*clusters_need = clusters_add;
	if (meta_need)
		*meta_need = meta_add;
	if (credits_need)
		*credits_need = credits;
	brelse(bh);
	return ret;
}

static int ocfs2_init_xattr_set_ctxt(struct inode *inode,
				     struct ocfs2_dinode *di,
				     struct ocfs2_xattr_info *xi,
				     struct ocfs2_xattr_search *xis,
				     struct ocfs2_xattr_search *xbs,
				     struct ocfs2_xattr_set_ctxt *ctxt,
				     int extra_meta,
				     int *credits)
{
	int clusters_add, meta_add, ret;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);

	memset(ctxt, 0, sizeof(struct ocfs2_xattr_set_ctxt));

	ocfs2_init_dealloc_ctxt(&ctxt->dealloc);

	ret = ocfs2_calc_xattr_set_need(inode, di, xi, xis, xbs,
					&clusters_add, &meta_add, credits);
	if (ret) {
		mlog_errno(ret);
		return ret;
	}

	meta_add += extra_meta;
	mlog(0, "Set xattr %s, reserve meta blocks = %d, clusters = %d, "
	     "credits = %d\n", xi->name, meta_add, clusters_add, *credits);

	if (meta_add) {
		ret = ocfs2_reserve_new_metadata_blocks(osb, meta_add,
							&ctxt->meta_ac);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}
	}

	if (clusters_add) {
		ret = ocfs2_reserve_clusters(osb, clusters_add, &ctxt->data_ac);
		if (ret)
			mlog_errno(ret);
	}
out:
	if (ret) {
		if (ctxt->meta_ac) {
			ocfs2_free_alloc_context(ctxt->meta_ac);
			ctxt->meta_ac = NULL;
		}

		/*
		 * We cannot have an error and a non null ctxt->data_ac.
		 */
	}

	return ret;
}

static int __ocfs2_xattr_set_handle(struct inode *inode,
				    struct ocfs2_dinode *di,
				    struct ocfs2_xattr_info *xi,
				    struct ocfs2_xattr_search *xis,
				    struct ocfs2_xattr_search *xbs,
				    struct ocfs2_xattr_set_ctxt *ctxt)
{
	int ret = 0, credits, old_found;

	if (!xi->value) {
		/* Remove existing extended attribute */
		if (!xis->not_found)
			ret = ocfs2_xattr_ibody_set(inode, xi, xis, ctxt);
		else if (!xbs->not_found)
			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
	} else {
		/* We always try to set extended attribute into inode first*/
		ret = ocfs2_xattr_ibody_set(inode, xi, xis, ctxt);
		if (!ret && !xbs->not_found) {
			/*
			 * If succeed and that extended attribute existing in
			 * external block, then we will remove it.
			 */
			xi->value = NULL;
			xi->value_len = 0;

			old_found = xis->not_found;
			xis->not_found = -ENODATA;
			ret = ocfs2_calc_xattr_set_need(inode,
							di,
							xi,
							xis,
							xbs,
							NULL,
							NULL,
							&credits);
			xis->not_found = old_found;
			if (ret) {
				mlog_errno(ret);
				goto out;
			}

			ret = ocfs2_extend_trans(ctxt->handle, credits +
					ctxt->handle->h_buffer_credits);
			if (ret) {
				mlog_errno(ret);
				goto out;
			}
			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
		} else if (ret == -ENOSPC) {
			if (di->i_xattr_loc && !xbs->xattr_bh) {
				ret = ocfs2_xattr_block_find(inode,
							     xi->name_index,
							     xi->name, xbs);
				if (ret)
					goto out;

				old_found = xis->not_found;
				xis->not_found = -ENODATA;
				ret = ocfs2_calc_xattr_set_need(inode,
								di,
								xi,
								xis,
								xbs,
								NULL,
								NULL,
								&credits);
				xis->not_found = old_found;
				if (ret) {
					mlog_errno(ret);
					goto out;
				}

				ret = ocfs2_extend_trans(ctxt->handle, credits +
					ctxt->handle->h_buffer_credits);
				if (ret) {
					mlog_errno(ret);
					goto out;
				}
			}
			/*
			 * If no space in inode, we will set extended attribute
			 * into external block.
			 */
			ret = ocfs2_xattr_block_set(inode, xi, xbs, ctxt);
			if (ret)
				goto out;
			if (!xis->not_found) {
				/*
				 * If succeed and that extended attribute
				 * existing in inode, we will remove it.
				 */
				xi->value = NULL;
				xi->value_len = 0;
				xbs->not_found = -ENODATA;
				ret = ocfs2_calc_xattr_set_need(inode,
								di,
								xi,
								xis,
								xbs,
								NULL,
								NULL,
								&credits);
				if (ret) {
					mlog_errno(ret);
					goto out;
				}

				ret = ocfs2_extend_trans(ctxt->handle, credits +
						ctxt->handle->h_buffer_credits);
				if (ret) {
					mlog_errno(ret);
					goto out;
				}
				ret = ocfs2_xattr_ibody_set(inode, xi,
							    xis, ctxt);
			}
		}
	}

	if (!ret) {
		/* Update inode ctime. */
		ret = ocfs2_journal_access_di(ctxt->handle, INODE_CACHE(inode),
					      xis->inode_bh,
					      OCFS2_JOURNAL_ACCESS_WRITE);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		inode->i_ctime = CURRENT_TIME;
		di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
		di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
		ocfs2_journal_dirty(ctxt->handle, xis->inode_bh);
	}
out:
	return ret;
}

/*
 * This function only called duing creating inode
 * for init security/acl xattrs of the new inode.
 * All transanction credits have been reserved in mknod.
 */
int ocfs2_xattr_set_handle(handle_t *handle,
			   struct inode *inode,
			   struct buffer_head *di_bh,
			   int name_index,
			   const char *name,
			   const void *value,
			   size_t value_len,
			   int flags,
			   struct ocfs2_alloc_context *meta_ac,
			   struct ocfs2_alloc_context *data_ac)
{
	struct ocfs2_dinode *di;
	int ret;

	struct ocfs2_xattr_info xi = {
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,
	};

	struct ocfs2_xattr_search xis = {
		.not_found = -ENODATA,
	};

	struct ocfs2_xattr_search xbs = {
		.not_found = -ENODATA,
	};

	struct ocfs2_xattr_set_ctxt ctxt = {
		.handle = handle,
		.meta_ac = meta_ac,
		.data_ac = data_ac,
	};

	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
		return -EOPNOTSUPP;

	/*
	 * In extreme situation, may need xattr bucket when
	 * block size is too small. And we have already reserved
	 * the credits for bucket in mknod.
	 */
	if (inode->i_sb->s_blocksize == OCFS2_MIN_BLOCKSIZE) {
		xbs.bucket = ocfs2_xattr_bucket_new(inode);
		if (!xbs.bucket) {
			mlog_errno(-ENOMEM);
			return -ENOMEM;
		}
	}

	xis.inode_bh = xbs.inode_bh = di_bh;
	di = (struct ocfs2_dinode *)di_bh->b_data;

	down_write(&OCFS2_I(inode)->ip_xattr_sem);

	ret = ocfs2_xattr_ibody_find(inode, name_index, name, &xis);
	if (ret)
		goto cleanup;
	if (xis.not_found) {
		ret = ocfs2_xattr_block_find(inode, name_index, name, &xbs);
		if (ret)
			goto cleanup;
	}

	ret = __ocfs2_xattr_set_handle(inode, di, &xi, &xis, &xbs, &ctxt);

cleanup:
	up_write(&OCFS2_I(inode)->ip_xattr_sem);
	brelse(xbs.xattr_bh);
	ocfs2_xattr_bucket_free(xbs.bucket);

	return ret;
}

/*
 * ocfs2_xattr_set()
 *
 * Set, replace or remove an extended attribute for this inode.
 * value is NULL to remove an existing extended attribute, else either
 * create or replace an extended attribute.
 */
int ocfs2_xattr_set(struct inode *inode,
		    int name_index,
		    const char *name,
		    const void *value,
		    size_t value_len,
		    int flags)
{
	struct buffer_head *di_bh = NULL;
	struct ocfs2_dinode *di;
	int ret, credits, ref_meta = 0, ref_credits = 0;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct inode *tl_inode = osb->osb_tl_inode;
	struct ocfs2_xattr_set_ctxt ctxt = { NULL, NULL, };
	struct ocfs2_refcount_tree *ref_tree = NULL;

	struct ocfs2_xattr_info xi = {
		.name_index = name_index,
		.name = name,
		.value = value,
		.value_len = value_len,
	};

	struct ocfs2_xattr_search xis = {
		.not_found = -ENODATA,
	};

	struct ocfs2_xattr_search xbs = {
		.not_found = -ENODATA,
	};

	if (!ocfs2_supports_xattr(OCFS2_SB(inode->i_sb)))
		return -EOPNOTSUPP;

	/*
	 * Only xbs will be used on indexed trees.  xis doesn't need a
	 * bucket.
	 */
	xbs.bucket = ocfs2_xattr_bucket_new(inode);
	if (!xbs.bucket) {
		mlog_errno(-ENOMEM);
		return -ENOMEM;
	}

	ret = ocfs2_inode_lock(inode, &di_bh, 1);
	if (ret < 0) {
		mlog_errno(ret);
		goto cleanup_nolock;
	}
	xis.inode_bh = xbs.inode_bh = di_bh;
	di = (struct ocfs2_dinode *)di_bh->b_data;

	down_write(&OCFS2_I(inode)->ip_xattr_sem);
	/*
	 * Scan inode and external block to find the same name
	 * extended attribute and collect search infomation.
	 */
	ret = ocfs2_xattr_ibody_find(inode, name_index, name, &xis);
	if (ret)
		goto cleanup;
	if (xis.not_found) {
		ret = ocfs2_xattr_block_find(inode, name_index, name, &xbs);
		if (ret)
			goto cleanup;
	}

	if (xis.not_found && xbs.not_found) {
		ret = -ENODATA;
		if (flags & XATTR_REPLACE)
			goto cleanup;
		ret = 0;
		if (!value)
			goto cleanup;
	} else {
		ret = -EEXIST;
		if (flags & XATTR_CREATE)
			goto cleanup;
	}

	/* Check whether the value is refcounted and do some prepartion. */
	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL &&
	    (!xis.not_found || !xbs.not_found)) {
		ret = ocfs2_prepare_refcount_xattr(inode, di, &xi,
						   &xis, &xbs, &ref_tree,
						   &ref_meta, &ref_credits);
		if (ret) {
			mlog_errno(ret);
			goto cleanup;
		}
	}

	mutex_lock(&tl_inode->i_mutex);

	if (ocfs2_truncate_log_needs_flush(osb)) {
		ret = __ocfs2_flush_truncate_log(osb);
		if (ret < 0) {
			mutex_unlock(&tl_inode->i_mutex);
			mlog_errno(ret);
			goto cleanup;
		}
	}
	mutex_unlock(&tl_inode->i_mutex);

	ret = ocfs2_init_xattr_set_ctxt(inode, di, &xi, &xis,
					&xbs, &ctxt, ref_meta, &credits);
	if (ret) {
		mlog_errno(ret);
		goto cleanup;
	}

	/* we need to update inode's ctime field, so add credit for it. */
	credits += OCFS2_INODE_UPDATE_CREDITS;
	ctxt.handle = ocfs2_start_trans(osb, credits + ref_credits);
	if (IS_ERR(ctxt.handle)) {
		ret = PTR_ERR(ctxt.handle);
		mlog_errno(ret);
		goto cleanup;
	}

	ret = __ocfs2_xattr_set_handle(inode, di, &xi, &xis, &xbs, &ctxt);

	ocfs2_commit_trans(osb, ctxt.handle);

	if (ctxt.data_ac)
		ocfs2_free_alloc_context(ctxt.data_ac);
	if (ctxt.meta_ac)
		ocfs2_free_alloc_context(ctxt.meta_ac);
	if (ocfs2_dealloc_has_cluster(&ctxt.dealloc))
		ocfs2_schedule_truncate_log_flush(osb, 1);