namei.c

linux 内核源代码

					ext4_error(sb, __FUNCTION__,
						   "error %d reading directory"
						   " #%lu offset %lu",
						   err, inode->i_ino, offset);
				offset += sb->s_blocksize;
				continue;
			}
			de = (struct ext4_dir_entry_2 *) bh->b_data;
		}
		if (!ext4_check_dir_entry("empty_dir", inode, de, bh, offset)) {
			de = (struct ext4_dir_entry_2 *)(bh->b_data +
							 sb->s_blocksize);
			offset = (offset | (sb->s_blocksize - 1)) + 1;
			continue;
		}
		if (le32_to_cpu(de->inode)) {
			brelse (bh);
			return 0;
		}
		offset += le16_to_cpu(de->rec_len);
		de = (struct ext4_dir_entry_2 *)
				((char *) de + le16_to_cpu(de->rec_len));
	}
	brelse (bh);
	return 1;
}

/* ext4_orphan_add() links an unlinked or truncated inode into a list of
 * such inodes, starting at the superblock, in case we crash before the
 * file is closed/deleted, or in case the inode truncate spans multiple
 * transactions and the last transaction is not recovered after a crash.
 *
 * At filesystem recovery time, we walk this list deleting unlinked
 * inodes and truncating linked inodes in ext4_orphan_cleanup().
 */
int ext4_orphan_add(handle_t *handle, struct inode *inode)
{
	struct super_block *sb = inode->i_sb;
	struct ext4_iloc iloc;
	int err = 0, rc;

	lock_super(sb);
	if (!list_empty(&EXT4_I(inode)->i_orphan))
		goto out_unlock;

	/* Orphan handling is only valid for files with data blocks
	 * being truncated, or files being unlinked. */

	/* @@@ FIXME: Observation from aviro:
	 * I think I can trigger J_ASSERT in ext4_orphan_add().  We block
	 * here (on lock_super()), so race with ext4_link() which might bump
	 * ->i_nlink. For, say it, character device. Not a regular file,
	 * not a directory, not a symlink and ->i_nlink > 0.
	 */
	J_ASSERT ((S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) ||
		S_ISLNK(inode->i_mode)) || inode->i_nlink == 0);

	BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get_write_access");
	err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh);
	if (err)
		goto out_unlock;

	err = ext4_reserve_inode_write(handle, inode, &iloc);
	if (err)
		goto out_unlock;

	/* Insert this inode at the head of the on-disk orphan list... */
	NEXT_ORPHAN(inode) = le32_to_cpu(EXT4_SB(sb)->s_es->s_last_orphan);
	EXT4_SB(sb)->s_es->s_last_orphan = cpu_to_le32(inode->i_ino);
	err = ext4_journal_dirty_metadata(handle, EXT4_SB(sb)->s_sbh);
	rc = ext4_mark_iloc_dirty(handle, inode, &iloc);
	if (!err)
		err = rc;

	/* Only add to the head of the in-memory list if all the
	 * previous operations succeeded.  If the orphan_add is going to
	 * fail (possibly taking the journal offline), we can't risk
	 * leaving the inode on the orphan list: stray orphan-list
	 * entries can cause panics at unmount time.
	 *
	 * This is safe: on error we're going to ignore the orphan list
	 * anyway on the next recovery. */
	if (!err)
		list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan);

	jbd_debug(4, "superblock will point to %lu\n", inode->i_ino);
	jbd_debug(4, "orphan inode %lu will point to %d\n",
			inode->i_ino, NEXT_ORPHAN(inode));
out_unlock:
	unlock_super(sb);
	ext4_std_error(inode->i_sb, err);
	return err;
}

/*
 * ext4_orphan_del() removes an unlinked or truncated inode from the list
 * of such inodes stored on disk, because it is finally being cleaned up.
 */
int ext4_orphan_del(handle_t *handle, struct inode *inode)
{
	struct list_head *prev;
	struct ext4_inode_info *ei = EXT4_I(inode);
	struct ext4_sb_info *sbi;
	unsigned long ino_next;
	struct ext4_iloc iloc;
	int err = 0;

	lock_super(inode->i_sb);
	if (list_empty(&ei->i_orphan)) {
		unlock_super(inode->i_sb);
		return 0;
	}

	ino_next = NEXT_ORPHAN(inode);
	prev = ei->i_orphan.prev;
	sbi = EXT4_SB(inode->i_sb);

	jbd_debug(4, "remove inode %lu from orphan list\n", inode->i_ino);

	list_del_init(&ei->i_orphan);

	/* If we're on an error path, we may not have a valid
	 * transaction handle with which to update the orphan list on
	 * disk, but we still need to remove the inode from the linked
	 * list in memory. */
	if (!handle)
		goto out;

	err = ext4_reserve_inode_write(handle, inode, &iloc);
	if (err)
		goto out_err;

	if (prev == &sbi->s_orphan) {
		jbd_debug(4, "superblock will point to %lu\n", ino_next);
		BUFFER_TRACE(sbi->s_sbh, "get_write_access");
		err = ext4_journal_get_write_access(handle, sbi->s_sbh);
		if (err)
			goto out_brelse;
		sbi->s_es->s_last_orphan = cpu_to_le32(ino_next);
		err = ext4_journal_dirty_metadata(handle, sbi->s_sbh);
	} else {
		struct ext4_iloc iloc2;
		struct inode *i_prev =
			&list_entry(prev, struct ext4_inode_info, i_orphan)->vfs_inode;

		jbd_debug(4, "orphan inode %lu will point to %lu\n",
			  i_prev->i_ino, ino_next);
		err = ext4_reserve_inode_write(handle, i_prev, &iloc2);
		if (err)
			goto out_brelse;
		NEXT_ORPHAN(i_prev) = ino_next;
		err = ext4_mark_iloc_dirty(handle, i_prev, &iloc2);
	}
	if (err)
		goto out_brelse;
	NEXT_ORPHAN(inode) = 0;
	err = ext4_mark_iloc_dirty(handle, inode, &iloc);

out_err:
	ext4_std_error(inode->i_sb, err);
out:
	unlock_super(inode->i_sb);
	return err;

out_brelse:
	brelse(iloc.bh);
	goto out_err;
}

static int ext4_rmdir (struct inode * dir, struct dentry *dentry)
{
	int retval;
	struct inode * inode;
	struct buffer_head * bh;
	struct ext4_dir_entry_2 * de;
	handle_t *handle;

	/* Initialize quotas before so that eventual writes go in
	 * separate transaction */
	DQUOT_INIT(dentry->d_inode);
	handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb));
	if (IS_ERR(handle))
		return PTR_ERR(handle);

	retval = -ENOENT;
	bh = ext4_find_entry (dentry, &de);
	if (!bh)
		goto end_rmdir;

	if (IS_DIRSYNC(dir))
		handle->h_sync = 1;

	inode = dentry->d_inode;

	retval = -EIO;
	if (le32_to_cpu(de->inode) != inode->i_ino)
		goto end_rmdir;

	retval = -ENOTEMPTY;
	if (!empty_dir (inode))
		goto end_rmdir;

	retval = ext4_delete_entry(handle, dir, de, bh);
	if (retval)
		goto end_rmdir;
	if (!EXT4_DIR_LINK_EMPTY(inode))
		ext4_warning (inode->i_sb, "ext4_rmdir",
			      "empty directory has too many links (%d)",
			      inode->i_nlink);
	inode->i_version++;
	clear_nlink(inode);
	/* There's no need to set i_disksize: the fact that i_nlink is
	 * zero will ensure that the right thing happens during any
	 * recovery. */
	inode->i_size = 0;
	ext4_orphan_add(handle, inode);
	inode->i_ctime = dir->i_ctime = dir->i_mtime = ext4_current_time(inode);
	ext4_mark_inode_dirty(handle, inode);
	ext4_dec_count(handle, dir);
	ext4_update_dx_flag(dir);
	ext4_mark_inode_dirty(handle, dir);

end_rmdir:
	ext4_journal_stop(handle);
	brelse (bh);
	return retval;
}

static int ext4_unlink(struct inode * dir, struct dentry *dentry)
{
	int retval;
	struct inode * inode;
	struct buffer_head * bh;
	struct ext4_dir_entry_2 * de;
	handle_t *handle;

	/* Initialize quotas before so that eventual writes go
	 * in separate transaction */
	DQUOT_INIT(dentry->d_inode);
	handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb));
	if (IS_ERR(handle))
		return PTR_ERR(handle);

	if (IS_DIRSYNC(dir))
		handle->h_sync = 1;

	retval = -ENOENT;
	bh = ext4_find_entry (dentry, &de);
	if (!bh)
		goto end_unlink;

	inode = dentry->d_inode;

	retval = -EIO;
	if (le32_to_cpu(de->inode) != inode->i_ino)
		goto end_unlink;

	if (!inode->i_nlink) {
		ext4_warning (inode->i_sb, "ext4_unlink",
			      "Deleting nonexistent file (%lu), %d",
			      inode->i_ino, inode->i_nlink);
		inode->i_nlink = 1;
	}
	retval = ext4_delete_entry(handle, dir, de, bh);
	if (retval)
		goto end_unlink;
	dir->i_ctime = dir->i_mtime = ext4_current_time(dir);
	ext4_update_dx_flag(dir);
	ext4_mark_inode_dirty(handle, dir);
	ext4_dec_count(handle, inode);
	if (!inode->i_nlink)
		ext4_orphan_add(handle, inode);
	inode->i_ctime = ext4_current_time(inode);
	ext4_mark_inode_dirty(handle, inode);
	retval = 0;

end_unlink:
	ext4_journal_stop(handle);
	brelse (bh);
	return retval;
}

static int ext4_symlink (struct inode * dir,
		struct dentry *dentry, const char * symname)
{
	handle_t *handle;
	struct inode * inode;
	int l, err, retries = 0;

	l = strlen(symname)+1;
	if (l > dir->i_sb->s_blocksize)
		return -ENAMETOOLONG;

retry:
	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 5 +
					2*EXT4_QUOTA_INIT_BLOCKS(dir->i_sb));
	if (IS_ERR(handle))
		return PTR_ERR(handle);

	if (IS_DIRSYNC(dir))
		handle->h_sync = 1;

	inode = ext4_new_inode (handle, dir, S_IFLNK|S_IRWXUGO);
	err = PTR_ERR(inode);
	if (IS_ERR(inode))
		goto out_stop;

	if (l > sizeof (EXT4_I(inode)->i_data)) {
		inode->i_op = &ext4_symlink_inode_operations;
		ext4_set_aops(inode);
		/*
		 * page_symlink() calls into ext4_prepare/commit_write.
		 * We have a transaction open.  All is sweetness.  It also sets
		 * i_size in generic_commit_write().
		 */
		err = __page_symlink(inode, symname, l,
				mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS);
		if (err) {
			ext4_dec_count(handle, inode);
			ext4_mark_inode_dirty(handle, inode);
			iput (inode);
			goto out_stop;
		}
	} else {
		inode->i_op = &ext4_fast_symlink_inode_operations;
		memcpy((char*)&EXT4_I(inode)->i_data,symname,l);
		inode->i_size = l-1;
	}
	EXT4_I(inode)->i_disksize = inode->i_size;
	err = ext4_add_nondir(handle, dentry, inode);
out_stop:
	ext4_journal_stop(handle);
	if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries))
		goto retry;
	return err;
}

static int ext4_link (struct dentry * old_dentry,
		struct inode * dir, struct dentry *dentry)
{
	handle_t *handle;
	struct inode *inode = old_dentry->d_inode;
	int err, retries = 0;

	if (EXT4_DIR_LINK_MAX(inode))
		return -EMLINK;

	/*
	 * Return -ENOENT if we've raced with unlink and i_nlink is 0.  Doing
	 * otherwise has the potential to corrupt the orphan inode list.
	 */
	if (inode->i_nlink == 0)
		return -ENOENT;

retry:
	handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) +
					EXT4_INDEX_EXTRA_TRANS_BLOCKS);
	if (IS_ERR(handle))
		return PTR_ERR(handle);

	if (IS_DIRSYNC(dir))
		handle->h_sync = 1;

	inode->i_ctime = ext4_current_time(inode);
	ext4_inc_count(handle, inode);
	atomic_inc(&inode->i_count);

	err = ext4_add_nondir(handle, dentry, inode);
	ext4_journal_stop(handle);
	if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries))
		goto retry;
	return err;
}

#define PARENT_INO(buffer) \
	((struct ext4_dir_entry_2 *) ((char *) buffer + \
	le16_to_cpu(((struct ext4_dir_entry_2 *) buffer)->rec_len)))->inode

/*
 * Anybody can rename anything with this: the permission checks are left to the
 * higher-level routines.
 */
static int ext4_rename (struct inode * old_dir, struct dentry *old_dentry,
			   struct inode * new_dir,struct dentry *new_dentry)
{
	handle_t *handle;
	struct inode * old_inode, * new_inode;
	struct buffer_head * old_bh, * new_bh, * dir_bh;
	struct ext4_dir_entry_2 * old_de, * new_de;
	int retval;

	old_bh = new_bh = dir_bh = NULL;

	/* Initialize quotas before so that eventual writes go
	 * in separate transaction */
	if (new_dentry->d_inode)
		DQUOT_INIT(new_dentry->d_inode);
	handle = ext4_journal_start(old_dir, 2 *
					EXT4_DATA_TRANS_BLOCKS(old_dir->i_sb) +
					EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2);
	if (IS_ERR(handle))
		return PTR_ERR(handle);

	if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir))
		handle->h_sync = 1;

	old_bh = ext4_find_entry (old_dentry, &old_de);
	/*
	 *  Check for inode number is _not_ due to possible IO errors.
	 *  We might rmdir the source, keep it as pwd of some process
	 *  and merrily kill the link to whatever was created under the
	 *  same name. Goodbye sticky bit ;-<
	 */
	old_inode = old_dentry->d_inode;
	retval = -ENOENT;
	if (!old_bh || le32_to_cpu(old_de->inode) != old_inode->i_ino)
		goto end_rename;

	new_inode = new_dentry->d_inode;
	new_bh = ext4_find_entry (new_dentry, &new_de);
	if (new_bh) {
		if (!new_inode) {
			brelse (new_bh);
			new_bh = NULL;
		}
	}
	if (S_ISDIR(old_inode->i_mode)) {
		if (new_inode) {
			retval = -ENOTEMPTY;
			if (!empty_dir (new_inode))
				goto end_rename;
		}
		retval = -EIO;
		dir_bh = ext4_bread (handle, old_inode, 0, 0, &retval);
		if (!dir_bh)
			goto end_rename;
		if (le32_to_cpu(PARENT_INO(dir_bh->b_data)) != old_dir->i_ino)
			goto end_rename;
		retval = -EMLINK;
		if (!new_inode && new_dir!=old_dir &&
				new_dir->i_nlink >= EXT4_LINK_MAX)
			goto end_rename;
	}
	if (!new_bh) {
		retval = ext4_add_entry (handle, new_dentry, old_inode);
		if (retval)
			goto end_rename;
	} else {
		BUFFER_TRACE(new_bh, "get write access");
		ext4_journal_get_write_access(handle, new_bh);
		new_de->inode = cpu_to_le32(old_inode->i_ino);
		if (EXT4_HAS_INCOMPAT_FEATURE(new_dir->i_sb,
					      EXT4_FEATURE_INCOMPAT_FILETYPE))
			new_de->file_type = old_de->file_type;
		new_dir->i_version++;
		BUFFER_TRACE(new_bh, "call ext4_journal_dirty_metadata");
		ext4_journal_dirty_metadata(handle, new_bh);
		brelse(new_bh);
		new_bh = NULL;
	}

	/*
	 * Like most other Unix systems, set the ctime for inodes on a
	 * rename.
	 */
	old_inode->i_ctime = ext4_current_time(old_inode);
	ext4_mark_inode_dirty(handle, old_inode);

	/*
	 * ok, that's it
	 */
	if (le32_to_cpu(old_de->inode) != old_inode->i_ino ||
	    old_de->name_len != old_dentry->