2006-10-11 08:21:03 +00:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com
|
|
|
|
* Written by Alex Tomas <alex@clusterfs.com>
|
|
|
|
*
|
|
|
|
* Architecture independence:
|
|
|
|
* Copyright (c) 2005, Bull S.A.
|
|
|
|
* Written by Pierre Peiffer <pierre.peiffer@bull.net>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public Licens
|
|
|
|
* along with this program; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Extents support for EXT4
|
|
|
|
*
|
|
|
|
* TODO:
|
|
|
|
* - ext4*_error() should be used in some situations
|
|
|
|
* - analyze all BUG()/BUG_ON(), use -EIO where appropriate
|
|
|
|
* - smart tree reduction
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/time.h>
|
2007-10-16 22:38:25 +00:00
|
|
|
#include <linux/jbd2.h>
|
2006-10-11 08:21:03 +00:00
|
|
|
#include <linux/highuid.h>
|
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/quotaops.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/slab.h>
|
2007-07-18 01:42:41 +00:00
|
|
|
#include <linux/falloc.h>
|
2006-10-11 08:21:03 +00:00
|
|
|
#include <asm/uaccess.h>
|
2008-10-07 04:46:36 +00:00
|
|
|
#include <linux/fiemap.h>
|
2008-04-29 22:13:32 +00:00
|
|
|
#include "ext4_jbd2.h"
|
|
|
|
#include "ext4_extents.h"
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2009-08-18 02:17:20 +00:00
|
|
|
static int ext4_ext_truncate_extend_restart(handle_t *handle,
|
|
|
|
struct inode *inode,
|
|
|
|
int needed)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
2009-01-07 05:06:22 +00:00
|
|
|
if (!ext4_handle_valid(handle))
|
|
|
|
return 0;
|
2006-10-11 08:21:03 +00:00
|
|
|
if (handle->h_buffer_credits > needed)
|
2008-07-11 23:27:31 +00:00
|
|
|
return 0;
|
|
|
|
err = ext4_journal_extend(handle, needed);
|
2008-08-02 00:57:54 +00:00
|
|
|
if (err <= 0)
|
2008-07-11 23:27:31 +00:00
|
|
|
return err;
|
2009-08-18 02:17:20 +00:00
|
|
|
err = ext4_truncate_restart_trans(handle, inode, needed);
|
2010-05-17 05:00:00 +00:00
|
|
|
if (err == 0)
|
|
|
|
err = -EAGAIN;
|
2009-08-18 02:17:20 +00:00
|
|
|
|
|
|
|
return err;
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* could return:
|
|
|
|
* - EROFS
|
|
|
|
* - ENOMEM
|
|
|
|
*/
|
|
|
|
static int ext4_ext_get_access(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *path)
|
|
|
|
{
|
|
|
|
if (path->p_bh) {
|
|
|
|
/* path points to block */
|
|
|
|
return ext4_journal_get_write_access(handle, path->p_bh);
|
|
|
|
}
|
|
|
|
/* path points to leaf/index in inode body */
|
|
|
|
/* we use in-core data, no need to protect them */
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* could return:
|
|
|
|
* - EROFS
|
|
|
|
* - ENOMEM
|
|
|
|
* - EIO
|
|
|
|
*/
|
|
|
|
static int ext4_ext_dirty(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *path)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
if (path->p_bh) {
|
|
|
|
/* path points to block */
|
2009-01-07 05:06:22 +00:00
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, path->p_bh);
|
2006-10-11 08:21:03 +00:00
|
|
|
} else {
|
|
|
|
/* path points to leaf/index in inode body */
|
|
|
|
err = ext4_mark_inode_dirty(handle, inode);
|
|
|
|
}
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2006-10-11 08:21:05 +00:00
|
|
|
static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_ext_path *path,
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t block)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t bg_start;
|
2008-02-15 18:43:07 +00:00
|
|
|
ext4_fsblk_t last_block;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_grpblk_t colour;
|
2009-03-12 16:18:34 +00:00
|
|
|
ext4_group_t block_group;
|
|
|
|
int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb));
|
2006-10-11 08:21:03 +00:00
|
|
|
int depth;
|
|
|
|
|
|
|
|
if (path) {
|
|
|
|
struct ext4_extent *ex;
|
|
|
|
depth = path->p_depth;
|
|
|
|
|
2011-01-10 17:12:28 +00:00
|
|
|
/*
|
|
|
|
* 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,
|
|
|
|
* especiially if the latter case turns out to be
|
|
|
|
* common.
|
|
|
|
*/
|
2006-12-07 04:41:33 +00:00
|
|
|
ex = path[depth].p_ext;
|
2011-01-10 17:12:28 +00:00
|
|
|
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);
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2006-10-11 08:21:07 +00:00
|
|
|
/* it looks like index is empty;
|
|
|
|
* try to find starting block from index itself */
|
2006-10-11 08:21:03 +00:00
|
|
|
if (path[depth].p_bh)
|
|
|
|
return path[depth].p_bh->b_blocknr;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* OK. use inode's group */
|
2009-03-12 16:18:34 +00:00
|
|
|
block_group = ei->i_block_group;
|
|
|
|
if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) {
|
|
|
|
/*
|
|
|
|
* If there are at least EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME
|
2010-05-17 11:00:00 +00:00
|
|
|
* block groups per flexgroup, reserve the first block
|
|
|
|
* group for directories and special files. Regular
|
2009-03-12 16:18:34 +00:00
|
|
|
* files will start at the second block group. This
|
2010-05-17 11:00:00 +00:00
|
|
|
* tends to speed up directory access and improves
|
2009-03-12 16:18:34 +00:00
|
|
|
* fsck times.
|
|
|
|
*/
|
|
|
|
block_group &= ~(flex_size-1);
|
|
|
|
if (S_ISREG(inode->i_mode))
|
|
|
|
block_group++;
|
|
|
|
}
|
2010-03-04 04:53:39 +00:00
|
|
|
bg_start = ext4_group_first_block_no(inode->i_sb, block_group);
|
2008-02-15 18:43:07 +00:00
|
|
|
last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1;
|
|
|
|
|
2009-03-12 16:18:34 +00:00
|
|
|
/*
|
|
|
|
* If we are doing delayed allocation, we don't need take
|
|
|
|
* colour into account.
|
|
|
|
*/
|
|
|
|
if (test_opt(inode->i_sb, DELALLOC))
|
|
|
|
return bg_start;
|
|
|
|
|
2008-02-15 18:43:07 +00:00
|
|
|
if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block)
|
|
|
|
colour = (current->pid % 16) *
|
2006-10-11 08:21:03 +00:00
|
|
|
(EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16);
|
2008-02-15 18:43:07 +00:00
|
|
|
else
|
|
|
|
colour = (current->pid % 16) * ((last_block - bg_start) / 16);
|
2006-10-11 08:21:03 +00:00
|
|
|
return bg_start + colour + block;
|
|
|
|
}
|
|
|
|
|
2008-07-11 23:27:31 +00:00
|
|
|
/*
|
|
|
|
* Allocation for a meta data block
|
|
|
|
*/
|
2006-10-11 08:21:05 +00:00
|
|
|
static ext4_fsblk_t
|
2008-07-11 23:27:31 +00:00
|
|
|
ext4_ext_new_meta_block(handle_t *handle, struct inode *inode,
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_ext_path *path,
|
|
|
|
struct ext4_extent *ex, int *err)
|
|
|
|
{
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t goal, newblock;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
|
2008-12-12 17:41:28 +00:00
|
|
|
newblock = ext4_new_meta_blocks(handle, inode, goal, NULL, err);
|
2006-10-11 08:21:03 +00:00
|
|
|
return newblock;
|
|
|
|
}
|
|
|
|
|
2009-08-28 14:40:33 +00:00
|
|
|
static inline int ext4_ext_space_block(struct inode *inode, int check)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int size;
|
|
|
|
|
|
|
|
size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
|
|
|
|
/ sizeof(struct ext4_extent);
|
2009-08-28 14:40:33 +00:00
|
|
|
if (!check) {
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
2009-08-28 14:40:33 +00:00
|
|
|
if (size > 6)
|
|
|
|
size = 6;
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
2009-08-28 14:40:33 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
2009-08-28 14:40:33 +00:00
|
|
|
static inline int ext4_ext_space_block_idx(struct inode *inode, int check)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int size;
|
|
|
|
|
|
|
|
size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
|
|
|
|
/ sizeof(struct ext4_extent_idx);
|
2009-08-28 14:40:33 +00:00
|
|
|
if (!check) {
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
2009-08-28 14:40:33 +00:00
|
|
|
if (size > 5)
|
|
|
|
size = 5;
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
2009-08-28 14:40:33 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
2009-08-28 14:40:33 +00:00
|
|
|
static inline int ext4_ext_space_root(struct inode *inode, int check)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int size;
|
|
|
|
|
|
|
|
size = sizeof(EXT4_I(inode)->i_data);
|
|
|
|
size -= sizeof(struct ext4_extent_header);
|
|
|
|
size /= sizeof(struct ext4_extent);
|
2009-08-28 14:40:33 +00:00
|
|
|
if (!check) {
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
2009-08-28 14:40:33 +00:00
|
|
|
if (size > 3)
|
|
|
|
size = 3;
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
2009-08-28 14:40:33 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
2009-08-28 14:40:33 +00:00
|
|
|
static inline int ext4_ext_space_root_idx(struct inode *inode, int check)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int size;
|
|
|
|
|
|
|
|
size = sizeof(EXT4_I(inode)->i_data);
|
|
|
|
size -= sizeof(struct ext4_extent_header);
|
|
|
|
size /= sizeof(struct ext4_extent_idx);
|
2009-08-28 14:40:33 +00:00
|
|
|
if (!check) {
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
2009-08-28 14:40:33 +00:00
|
|
|
if (size > 4)
|
|
|
|
size = 4;
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
2009-08-28 14:40:33 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
2008-07-14 21:52:37 +00:00
|
|
|
/*
|
|
|
|
* Calculate the number of metadata blocks needed
|
|
|
|
* to allocate @blocks
|
|
|
|
* Worse case is one block per extent
|
|
|
|
*/
|
2011-01-10 17:13:03 +00:00
|
|
|
int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
|
2008-07-14 21:52:37 +00:00
|
|
|
{
|
2010-01-01 07:41:30 +00:00
|
|
|
struct ext4_inode_info *ei = EXT4_I(inode);
|
|
|
|
int idxs, num = 0;
|
2008-07-14 21:52:37 +00:00
|
|
|
|
2010-01-01 07:41:30 +00:00
|
|
|
idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
|
|
|
|
/ sizeof(struct ext4_extent_idx));
|
2008-07-14 21:52:37 +00:00
|
|
|
|
|
|
|
/*
|
2010-01-01 07:41:30 +00:00
|
|
|
* 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.
|
2008-07-14 21:52:37 +00:00
|
|
|
*/
|
2010-01-01 07:41:30 +00:00
|
|
|
if (ei->i_da_metadata_calc_len &&
|
|
|
|
ei->i_da_metadata_calc_last_lblock+1 == lblock) {
|
|
|
|
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;
|
|
|
|
}
|
2008-07-14 21:52:37 +00:00
|
|
|
|
2010-01-01 07:41:30 +00:00
|
|
|
/*
|
|
|
|
* 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;
|
2008-07-14 21:52:37 +00:00
|
|
|
}
|
|
|
|
|
2007-07-18 13:19:09 +00:00
|
|
|
static int
|
|
|
|
ext4_ext_max_entries(struct inode *inode, int depth)
|
|
|
|
{
|
|
|
|
int max;
|
|
|
|
|
|
|
|
if (depth == ext_depth(inode)) {
|
|
|
|
if (depth == 0)
|
2009-08-28 14:40:33 +00:00
|
|
|
max = ext4_ext_space_root(inode, 1);
|
2007-07-18 13:19:09 +00:00
|
|
|
else
|
2009-08-28 14:40:33 +00:00
|
|
|
max = ext4_ext_space_root_idx(inode, 1);
|
2007-07-18 13:19:09 +00:00
|
|
|
} else {
|
|
|
|
if (depth == 0)
|
2009-08-28 14:40:33 +00:00
|
|
|
max = ext4_ext_space_block(inode, 1);
|
2007-07-18 13:19:09 +00:00
|
|
|
else
|
2009-08-28 14:40:33 +00:00
|
|
|
max = ext4_ext_space_block_idx(inode, 1);
|
2007-07-18 13:19:09 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return max;
|
|
|
|
}
|
|
|
|
|
2009-03-12 13:51:20 +00:00
|
|
|
static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext)
|
|
|
|
{
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_fsblk_t block = ext4_ext_pblock(ext);
|
2009-03-12 13:51:20 +00:00
|
|
|
int len = ext4_ext_get_actual_len(ext);
|
2009-04-23 00:52:25 +00:00
|
|
|
|
2009-05-17 19:38:01 +00:00
|
|
|
return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len);
|
2009-03-12 13:51:20 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_valid_extent_idx(struct inode *inode,
|
|
|
|
struct ext4_extent_idx *ext_idx)
|
|
|
|
{
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_fsblk_t block = ext4_idx_pblock(ext_idx);
|
2009-04-23 00:52:25 +00:00
|
|
|
|
2009-05-17 19:38:01 +00:00
|
|
|
return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1);
|
2009-03-12 13:51:20 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_valid_extent_entries(struct inode *inode,
|
|
|
|
struct ext4_extent_header *eh,
|
|
|
|
int depth)
|
|
|
|
{
|
|
|
|
struct ext4_extent *ext;
|
|
|
|
struct ext4_extent_idx *ext_idx;
|
|
|
|
unsigned short entries;
|
|
|
|
if (eh->eh_entries == 0)
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
entries = le16_to_cpu(eh->eh_entries);
|
|
|
|
|
|
|
|
if (depth == 0) {
|
|
|
|
/* leaf entries */
|
|
|
|
ext = EXT_FIRST_EXTENT(eh);
|
|
|
|
while (entries) {
|
|
|
|
if (!ext4_valid_extent(inode, ext))
|
|
|
|
return 0;
|
|
|
|
ext++;
|
|
|
|
entries--;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
ext_idx = EXT_FIRST_INDEX(eh);
|
|
|
|
while (entries) {
|
|
|
|
if (!ext4_valid_extent_idx(inode, ext_idx))
|
|
|
|
return 0;
|
|
|
|
ext_idx++;
|
|
|
|
entries--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2010-07-27 15:56:40 +00:00
|
|
|
static int __ext4_ext_check(const char *function, unsigned int line,
|
|
|
|
struct inode *inode, struct ext4_extent_header *eh,
|
|
|
|
int depth)
|
2007-07-18 13:19:09 +00:00
|
|
|
{
|
|
|
|
const char *error_msg;
|
|
|
|
int max = 0;
|
|
|
|
|
|
|
|
if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) {
|
|
|
|
error_msg = "invalid magic";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
|
|
|
if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) {
|
|
|
|
error_msg = "unexpected eh_depth";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
|
|
|
if (unlikely(eh->eh_max == 0)) {
|
|
|
|
error_msg = "invalid eh_max";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
|
|
|
max = ext4_ext_max_entries(inode, depth);
|
|
|
|
if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
|
|
|
|
error_msg = "too large eh_max";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
|
|
|
if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
|
|
|
|
error_msg = "invalid eh_entries";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
2009-03-12 13:51:20 +00:00
|
|
|
if (!ext4_valid_extent_entries(inode, eh, depth)) {
|
|
|
|
error_msg = "invalid extent entries";
|
|
|
|
goto corrupted;
|
|
|
|
}
|
2007-07-18 13:19:09 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
corrupted:
|
2010-07-27 15:56:40 +00:00
|
|
|
ext4_error_inode(inode, function, line, 0,
|
2010-05-17 01:00:00 +00:00
|
|
|
"bad header/extent: %s - magic %x, "
|
2007-07-18 13:19:09 +00:00
|
|
|
"entries %u, max %u(%u), depth %u(%u)",
|
2010-05-17 01:00:00 +00:00
|
|
|
error_msg, le16_to_cpu(eh->eh_magic),
|
2007-07-18 13:19:09 +00:00
|
|
|
le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
|
|
|
|
max, le16_to_cpu(eh->eh_depth), depth);
|
|
|
|
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2009-03-12 13:51:20 +00:00
|
|
|
#define ext4_ext_check(inode, eh, depth) \
|
2010-07-27 15:56:40 +00:00
|
|
|
__ext4_ext_check(__func__, __LINE__, inode, eh, depth)
|
2007-07-18 13:19:09 +00:00
|
|
|
|
2009-03-27 20:39:58 +00:00
|
|
|
int ext4_ext_check_inode(struct inode *inode)
|
|
|
|
{
|
|
|
|
return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode));
|
|
|
|
}
|
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
#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) {
|
2006-10-11 08:21:11 +00:00
|
|
|
ext_debug(" %d->%llu", le32_to_cpu(path->p_idx->ei_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_idx_pblock(path->p_idx));
|
2006-10-11 08:21:03 +00:00
|
|
|
} else if (path->p_ext) {
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug(" %d:[%d]%d:%llu ",
|
2006-10-11 08:21:03 +00:00
|
|
|
le32_to_cpu(path->p_ext->ee_block),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(path->p_ext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(path->p_ext),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(path->p_ext));
|
2006-10-11 08:21:03 +00:00
|
|
|
} 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);
|
|
|
|
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
|
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
|
|
|
|
ext4_ext_is_uninitialized(ex),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
ext_debug("\n");
|
|
|
|
}
|
|
|
|
#else
|
2008-09-09 02:25:24 +00:00
|
|
|
#define ext4_ext_show_path(inode, path)
|
|
|
|
#define ext4_ext_show_leaf(inode, path)
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
|
|
|
|
2008-02-25 21:54:37 +00:00
|
|
|
void ext4_ext_drop_refs(struct ext4_ext_path *path)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int depth = path->p_depth;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i <= depth; i++, path++)
|
|
|
|
if (path->p_bh) {
|
|
|
|
brelse(path->p_bh);
|
|
|
|
path->p_bh = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_binsearch_idx:
|
|
|
|
* binary search for the closest index of the given block
|
2007-07-18 13:19:09 +00:00
|
|
|
* the header must be checked before calling this
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
static void
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_ext_binsearch_idx(struct inode *inode,
|
|
|
|
struct ext4_ext_path *path, ext4_lblk_t block)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_header *eh = path->p_hdr;
|
|
|
|
struct ext4_extent_idx *r, *l, *m;
|
|
|
|
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("binsearch for %u(idx): ", block);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
l = EXT_FIRST_INDEX(eh) + 1;
|
2007-07-18 13:09:15 +00:00
|
|
|
r = EXT_LAST_INDEX(eh);
|
2006-10-11 08:21:03 +00:00
|
|
|
while (l <= r) {
|
|
|
|
m = l + (r - l) / 2;
|
|
|
|
if (block < le32_to_cpu(m->ei_block))
|
|
|
|
r = m - 1;
|
|
|
|
else
|
|
|
|
l = m + 1;
|
2007-07-18 12:33:37 +00:00
|
|
|
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));
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
path->p_idx = l - 1;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext_debug(" -> %d->%lld ", le32_to_cpu(path->p_idx->ei_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_idx_pblock(path->p_idx));
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
#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)) {
|
2008-09-09 03:00:52 +00:00
|
|
|
printk(KERN_DEBUG "k=%d, ix=0x%p, "
|
|
|
|
"first=0x%p\n", k,
|
|
|
|
ix, EXT_FIRST_INDEX(eh));
|
|
|
|
printk(KERN_DEBUG "%u <= %u\n",
|
2006-10-11 08:21:03 +00:00
|
|
|
le32_to_cpu(ix->ei_block),
|
|
|
|
le32_to_cpu(ix[-1].ei_block));
|
|
|
|
}
|
|
|
|
BUG_ON(k && le32_to_cpu(ix->ei_block)
|
2007-05-24 17:04:54 +00:00
|
|
|
<= le32_to_cpu(ix[-1].ei_block));
|
2006-10-11 08:21:03 +00:00
|
|
|
if (block < le32_to_cpu(ix->ei_block))
|
|
|
|
break;
|
|
|
|
chix = ix;
|
|
|
|
}
|
|
|
|
BUG_ON(chix != path->p_idx);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_binsearch:
|
|
|
|
* binary search for closest extent of the given block
|
2007-07-18 13:19:09 +00:00
|
|
|
* the header must be checked before calling this
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
static void
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_ext_binsearch(struct inode *inode,
|
|
|
|
struct ext4_ext_path *path, ext4_lblk_t block)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_header *eh = path->p_hdr;
|
|
|
|
struct ext4_extent *r, *l, *m;
|
|
|
|
|
|
|
|
if (eh->eh_entries == 0) {
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* this leaf is empty:
|
|
|
|
* we get such a leaf in split/add case
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("binsearch for %u: ", block);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
l = EXT_FIRST_EXTENT(eh) + 1;
|
2007-07-18 13:09:15 +00:00
|
|
|
r = EXT_LAST_EXTENT(eh);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
while (l <= r) {
|
|
|
|
m = l + (r - l) / 2;
|
|
|
|
if (block < le32_to_cpu(m->ee_block))
|
|
|
|
r = m - 1;
|
|
|
|
else
|
|
|
|
l = m + 1;
|
2007-07-18 12:33:37 +00:00
|
|
|
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));
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
path->p_ext = l - 1;
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug(" -> %d:%llu:[%d]%d ",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(path->p_ext->ee_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(path->p_ext),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(path->p_ext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(path->p_ext));
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
#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)
|
2007-05-24 17:04:54 +00:00
|
|
|
<= le32_to_cpu(ex[-1].ee_block));
|
2006-10-11 08:21:03 +00:00
|
|
|
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;
|
2009-08-28 14:40:33 +00:00
|
|
|
eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
|
|
|
ext4_ext_invalidate_cache(inode);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
struct ext4_ext_path *
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
|
|
|
|
struct ext4_ext_path *path)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_header *eh;
|
|
|
|
struct buffer_head *bh;
|
|
|
|
short int depth, i, ppos = 0, alloc = 0;
|
|
|
|
|
|
|
|
eh = ext_inode_hdr(inode);
|
2007-07-18 13:19:09 +00:00
|
|
|
depth = ext_depth(inode);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* account possible depth increase */
|
|
|
|
if (!path) {
|
2006-12-07 04:41:35 +00:00
|
|
|
path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 2),
|
2006-10-11 08:21:03 +00:00
|
|
|
GFP_NOFS);
|
|
|
|
if (!path)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
alloc = 1;
|
|
|
|
}
|
|
|
|
path[0].p_hdr = eh;
|
2008-07-11 23:27:31 +00:00
|
|
|
path[0].p_bh = NULL;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2007-07-18 13:19:09 +00:00
|
|
|
i = depth;
|
2006-10-11 08:21:03 +00:00
|
|
|
/* walk through the tree */
|
|
|
|
while (i) {
|
2009-03-27 20:39:58 +00:00
|
|
|
int need_to_validate = 0;
|
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
ext_debug("depth %d: num %d, max %d\n",
|
|
|
|
ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
|
2007-07-18 13:19:09 +00:00
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_binsearch_idx(inode, path + ppos, block);
|
2010-10-28 01:30:14 +00:00
|
|
|
path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx);
|
2006-10-11 08:21:03 +00:00
|
|
|
path[ppos].p_depth = i;
|
|
|
|
path[ppos].p_ext = NULL;
|
|
|
|
|
2009-03-27 20:39:58 +00:00
|
|
|
bh = sb_getblk(inode->i_sb, path[ppos].p_block);
|
|
|
|
if (unlikely(!bh))
|
2006-10-11 08:21:03 +00:00
|
|
|
goto err;
|
2009-03-27 20:39:58 +00:00
|
|
|
if (!bh_uptodate_or_lock(bh)) {
|
|
|
|
if (bh_submit_read(bh) < 0) {
|
|
|
|
put_bh(bh);
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
/* validate the extent entries */
|
|
|
|
need_to_validate = 1;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
eh = ext_block_hdr(bh);
|
|
|
|
ppos++;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(ppos > depth)) {
|
|
|
|
put_bh(bh);
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"ppos %d > depth %d", ppos, depth);
|
|
|
|
goto err;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
path[ppos].p_bh = bh;
|
|
|
|
path[ppos].p_hdr = eh;
|
|
|
|
i--;
|
|
|
|
|
2009-03-27 20:39:58 +00:00
|
|
|
if (need_to_validate && ext4_ext_check(inode, eh, i))
|
2006-10-11 08:21:03 +00:00
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
path[ppos].p_depth = i;
|
|
|
|
path[ppos].p_ext = NULL;
|
|
|
|
path[ppos].p_idx = NULL;
|
|
|
|
|
|
|
|
/* find extent */
|
|
|
|
ext4_ext_binsearch(inode, path + ppos, block);
|
2008-07-11 23:27:31 +00:00
|
|
|
/* if not an empty leaf */
|
|
|
|
if (path[ppos].p_ext)
|
2010-10-28 01:30:14 +00:00
|
|
|
path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
ext4_ext_show_path(inode, path);
|
|
|
|
|
|
|
|
return path;
|
|
|
|
|
|
|
|
err:
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
if (alloc)
|
|
|
|
kfree(path);
|
|
|
|
return ERR_PTR(-EIO);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_insert_index:
|
|
|
|
* insert new index [@logical;@ptr] into the block at @curp;
|
|
|
|
* check where to insert: before @curp or after @curp
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2010-10-28 01:30:14 +00:00
|
|
|
static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *curp,
|
|
|
|
int logical, ext4_fsblk_t ptr)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_idx *ix;
|
|
|
|
int len, err;
|
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, curp);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"logical %d == ei_block %d!",
|
|
|
|
logical, le32_to_cpu(curp->p_idx->ei_block));
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
len = EXT_MAX_INDEX(curp->p_hdr) - curp->p_idx;
|
|
|
|
if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
|
|
|
|
/* insert after */
|
|
|
|
if (curp->p_idx != EXT_LAST_INDEX(curp->p_hdr)) {
|
|
|
|
len = (len - 1) * sizeof(struct ext4_extent_idx);
|
|
|
|
len = len < 0 ? 0 : len;
|
2007-07-18 12:33:37 +00:00
|
|
|
ext_debug("insert new index %d after: %llu. "
|
2006-10-11 08:21:03 +00:00
|
|
|
"move %d from 0x%p to 0x%p\n",
|
|
|
|
logical, ptr, len,
|
|
|
|
(curp->p_idx + 1), (curp->p_idx + 2));
|
|
|
|
memmove(curp->p_idx + 2, curp->p_idx + 1, len);
|
|
|
|
}
|
|
|
|
ix = curp->p_idx + 1;
|
|
|
|
} else {
|
|
|
|
/* insert before */
|
|
|
|
len = len * sizeof(struct ext4_extent_idx);
|
|
|
|
len = len < 0 ? 0 : len;
|
2007-07-18 12:33:37 +00:00
|
|
|
ext_debug("insert new index %d before: %llu. "
|
2006-10-11 08:21:03 +00:00
|
|
|
"move %d from 0x%p to 0x%p\n",
|
|
|
|
logical, ptr, len,
|
|
|
|
curp->p_idx, (curp->p_idx + 1));
|
|
|
|
memmove(curp->p_idx + 1, curp->p_idx, len);
|
|
|
|
ix = curp->p_idx;
|
|
|
|
}
|
|
|
|
|
|
|
|
ix->ei_block = cpu_to_le32(logical);
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_idx_store_pblock(ix, ptr);
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&curp->p_hdr->eh_entries, 1);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries)
|
|
|
|
> le16_to_cpu(curp->p_hdr->eh_max))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"logical %d == ei_block %d!",
|
|
|
|
logical, le32_to_cpu(curp->p_idx->ei_block));
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) {
|
|
|
|
EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!");
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
err = ext4_ext_dirty(handle, inode, curp);
|
|
|
|
ext4_std_error(inode->i_sb, err);
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* 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
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
static int ext4_ext_split(handle_t *handle, struct inode *inode,
|
|
|
|
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;
|
|
|
|
struct ext4_extent *ex;
|
|
|
|
int i = at, k, m, a;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t newblock, oldblock;
|
2006-10-11 08:21:03 +00:00
|
|
|
__le32 border;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */
|
2006-10-11 08:21:03 +00:00
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
/* make decision: where to split? */
|
2006-10-11 08:21:07 +00:00
|
|
|
/* FIXME: now decision is simplest: at current extent */
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2006-10-11 08:21:07 +00:00
|
|
|
/* if current leaf will be split, then we should use
|
2006-10-11 08:21:03 +00:00
|
|
|
* border from split point */
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) {
|
|
|
|
EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!");
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
|
|
|
|
border = path[depth].p_ext[1].ee_block;
|
2006-10-11 08:21:07 +00:00
|
|
|
ext_debug("leaf will be split."
|
2006-10-11 08:21:03 +00:00
|
|
|
" next leaf starts at %d\n",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(border));
|
2006-10-11 08:21:03 +00:00
|
|
|
} else {
|
|
|
|
border = newext->ee_block;
|
|
|
|
ext_debug("leaf will be added."
|
|
|
|
" next leaf starts at %d\n",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(border));
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* If error occurs, then we break processing
|
|
|
|
* and mark filesystem read-only. index won't
|
2006-10-11 08:21:03 +00:00
|
|
|
* be inserted and tree will be in consistent
|
2006-10-11 08:21:07 +00:00
|
|
|
* state. Next mount will repair buffers too.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* Get array to track all allocated blocks.
|
|
|
|
* We need this to handle errors and free blocks
|
|
|
|
* upon them.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2006-12-07 04:41:35 +00:00
|
|
|
ablocks = kzalloc(sizeof(ext4_fsblk_t) * depth, GFP_NOFS);
|
2006-10-11 08:21:03 +00:00
|
|
|
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++) {
|
2008-07-11 23:27:31 +00:00
|
|
|
newblock = ext4_ext_new_meta_block(handle, inode, path,
|
|
|
|
newext, &err);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (newblock == 0)
|
|
|
|
goto cleanup;
|
|
|
|
ablocks[a] = newblock;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* initialize new leaf */
|
|
|
|
newblock = ablocks[--a];
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(newblock == 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "newblock == 0!");
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
bh = sb_getblk(inode->i_sb, newblock);
|
|
|
|
if (!bh) {
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
|
|
|
lock_buffer(bh);
|
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_journal_get_create_access(handle, bh);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
|
|
|
|
neh = ext_block_hdr(bh);
|
|
|
|
neh->eh_entries = 0;
|
2009-08-28 14:40:33 +00:00
|
|
|
neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
neh->eh_magic = EXT4_EXT_MAGIC;
|
|
|
|
neh->eh_depth = 0;
|
|
|
|
ex = EXT_FIRST_EXTENT(neh);
|
|
|
|
|
2006-10-11 08:21:07 +00:00
|
|
|
/* move remainder of path[depth] to the new leaf */
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_hdr->eh_entries !=
|
|
|
|
path[depth].p_hdr->eh_max)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!",
|
|
|
|
path[depth].p_hdr->eh_entries,
|
|
|
|
path[depth].p_hdr->eh_max);
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
/* start copy from next extent */
|
|
|
|
/* TODO: we could do it by single memmove */
|
|
|
|
m = 0;
|
|
|
|
path[depth].p_ext++;
|
|
|
|
while (path[depth].p_ext <=
|
|
|
|
EXT_MAX_EXTENT(path[depth].p_hdr)) {
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(path[depth].p_ext->ee_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(path[depth].p_ext),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(path[depth].p_ext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(path[depth].p_ext),
|
2006-10-11 08:21:03 +00:00
|
|
|
newblock);
|
|
|
|
/*memmove(ex++, path[depth].p_ext++,
|
|
|
|
sizeof(struct ext4_extent));
|
|
|
|
neh->eh_entries++;*/
|
|
|
|
path[depth].p_ext++;
|
|
|
|
m++;
|
|
|
|
}
|
|
|
|
if (m) {
|
|
|
|
memmove(ex, path[depth].p_ext-m, sizeof(struct ext4_extent)*m);
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&neh->eh_entries, m);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
unlock_buffer(bh);
|
|
|
|
|
2009-01-07 05:06:22 +00:00
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, bh);
|
2006-12-07 04:41:33 +00:00
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
brelse(bh);
|
|
|
|
bh = NULL;
|
|
|
|
|
|
|
|
/* correct old leaf */
|
|
|
|
if (m) {
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
/* create intermediate indexes */
|
|
|
|
k = depth - at - 1;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(k < 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "k %d < 0!", k);
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
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];
|
2008-01-29 04:58:27 +00:00
|
|
|
bh = sb_getblk(inode->i_sb, newblock);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (!bh) {
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
|
|
|
lock_buffer(bh);
|
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_journal_get_create_access(handle, bh);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
|
|
|
|
neh = ext_block_hdr(bh);
|
|
|
|
neh->eh_entries = cpu_to_le16(1);
|
|
|
|
neh->eh_magic = EXT4_EXT_MAGIC;
|
2009-08-28 14:40:33 +00:00
|
|
|
neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
neh->eh_depth = cpu_to_le16(depth - i);
|
|
|
|
fidx = EXT_FIRST_INDEX(neh);
|
|
|
|
fidx->ei_block = border;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_idx_store_pblock(fidx, oldblock);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("int.index at %d (block %llu): %u -> %llu\n",
|
|
|
|
i, newblock, le32_to_cpu(border), oldblock);
|
2006-10-11 08:21:03 +00:00
|
|
|
/* copy indexes */
|
|
|
|
m = 0;
|
|
|
|
path[i].p_idx++;
|
|
|
|
|
|
|
|
ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
|
|
|
|
EXT_MAX_INDEX(path[i].p_hdr));
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) !=
|
|
|
|
EXT_LAST_INDEX(path[i].p_hdr))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!",
|
|
|
|
le32_to_cpu(path[i].p_ext->ee_block));
|
|
|
|
err = -EIO;
|
|
|
|
goto cleanup;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
while (path[i].p_idx <= EXT_MAX_INDEX(path[i].p_hdr)) {
|
2007-07-18 12:33:37 +00:00
|
|
|
ext_debug("%d: move %d:%llu in new index %llu\n", i,
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(path[i].p_idx->ei_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_idx_pblock(path[i].p_idx),
|
2007-05-24 17:04:54 +00:00
|
|
|
newblock);
|
2006-10-11 08:21:03 +00:00
|
|
|
/*memmove(++fidx, path[i].p_idx++,
|
|
|
|
sizeof(struct ext4_extent_idx));
|
|
|
|
neh->eh_entries++;
|
|
|
|
BUG_ON(neh->eh_entries > neh->eh_max);*/
|
|
|
|
path[i].p_idx++;
|
|
|
|
m++;
|
|
|
|
}
|
|
|
|
if (m) {
|
|
|
|
memmove(++fidx, path[i].p_idx - m,
|
|
|
|
sizeof(struct ext4_extent_idx) * m);
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&neh->eh_entries, m);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
unlock_buffer(bh);
|
|
|
|
|
2009-01-07 05:06:22 +00:00
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, bh);
|
2006-12-07 04:41:33 +00:00
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
brelse(bh);
|
|
|
|
bh = NULL;
|
|
|
|
|
|
|
|
/* correct old index */
|
|
|
|
if (m) {
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + i);
|
|
|
|
if (err)
|
|
|
|
goto cleanup;
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
|
2006-10-11 08:21:03 +00:00
|
|
|
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;
|
2009-11-23 12:17:05 +00:00
|
|
|
ext4_free_blocks(handle, inode, 0, ablocks[i], 1,
|
|
|
|
EXT4_FREE_BLOCKS_METADATA);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
kfree(ablocks);
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* 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
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
struct ext4_extent *newext)
|
|
|
|
{
|
|
|
|
struct ext4_ext_path *curp = path;
|
|
|
|
struct ext4_extent_header *neh;
|
|
|
|
struct buffer_head *bh;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t newblock;
|
2006-10-11 08:21:03 +00:00
|
|
|
int err = 0;
|
|
|
|
|
2008-07-11 23:27:31 +00:00
|
|
|
newblock = ext4_ext_new_meta_block(handle, inode, path, newext, &err);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (newblock == 0)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
bh = sb_getblk(inode->i_sb, newblock);
|
|
|
|
if (!bh) {
|
|
|
|
err = -EIO;
|
|
|
|
ext4_std_error(inode->i_sb, err);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
lock_buffer(bh);
|
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_journal_get_create_access(handle, bh);
|
|
|
|
if (err) {
|
2006-10-11 08:21:03 +00:00
|
|
|
unlock_buffer(bh);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* move top-level index/leaf into new block */
|
|
|
|
memmove(bh->b_data, curp->p_hdr, 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))
|
2009-08-28 14:40:33 +00:00
|
|
|
neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
else
|
2009-08-28 14:40:33 +00:00
|
|
|
neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
neh->eh_magic = EXT4_EXT_MAGIC;
|
|
|
|
set_buffer_uptodate(bh);
|
|
|
|
unlock_buffer(bh);
|
|
|
|
|
2009-01-07 05:06:22 +00:00
|
|
|
err = ext4_handle_dirty_metadata(handle, inode, bh);
|
2006-12-07 04:41:33 +00:00
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* create index in new top-level index: num,max,pointer */
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, curp);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out;
|
|
|
|
|
|
|
|
curp->p_hdr->eh_magic = EXT4_EXT_MAGIC;
|
2009-08-28 14:40:33 +00:00
|
|
|
curp->p_hdr->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
curp->p_hdr->eh_entries = cpu_to_le16(1);
|
|
|
|
curp->p_idx = EXT_FIRST_INDEX(curp->p_hdr);
|
2007-07-18 13:09:15 +00:00
|
|
|
|
|
|
|
if (path[0].p_hdr->eh_depth)
|
|
|
|
curp->p_idx->ei_block =
|
|
|
|
EXT_FIRST_INDEX(path[0].p_hdr)->ei_block;
|
|
|
|
else
|
|
|
|
curp->p_idx->ei_block =
|
|
|
|
EXT_FIRST_EXTENT(path[0].p_hdr)->ee_block;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_idx_store_pblock(curp->p_idx, newblock);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
neh = ext_inode_hdr(inode);
|
2006-10-11 08:21:11 +00:00
|
|
|
ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
|
2006-10-11 08:21:03 +00:00
|
|
|
le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
|
2010-06-14 17:28:03 +00:00
|
|
|
le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
neh->eh_depth = cpu_to_le16(path->p_depth + 1);
|
|
|
|
err = ext4_ext_dirty(handle, inode, curp);
|
|
|
|
out:
|
|
|
|
brelse(bh);
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_create_new_leaf:
|
|
|
|
* finds empty index and adds new leaf.
|
|
|
|
* if no free index is found, then it requests in-depth growing.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
struct ext4_extent *newext)
|
|
|
|
{
|
|
|
|
struct ext4_ext_path *curp;
|
|
|
|
int depth, i, err = 0;
|
|
|
|
|
|
|
|
repeat:
|
|
|
|
i = depth = ext_depth(inode);
|
|
|
|
|
|
|
|
/* walk up to the tree and look for free index entry */
|
|
|
|
curp = path + depth;
|
|
|
|
while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
|
|
|
|
i--;
|
|
|
|
curp--;
|
|
|
|
}
|
|
|
|
|
2006-10-11 08:21:07 +00:00
|
|
|
/* we use already allocated block for index block,
|
|
|
|
* so subsequent data blocks should be contiguous */
|
2006-10-11 08:21:03 +00:00
|
|
|
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, path, newext, i);
|
2008-07-11 23:27:31 +00:00
|
|
|
if (err)
|
|
|
|
goto out;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* refill path */
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
path = ext4_ext_find_extent(inode,
|
2008-01-29 04:58:27 +00:00
|
|
|
(ext4_lblk_t)le32_to_cpu(newext->ee_block),
|
|
|
|
path);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (IS_ERR(path))
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
} else {
|
|
|
|
/* tree is full, time to grow in depth */
|
|
|
|
err = ext4_ext_grow_indepth(handle, inode, path, newext);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* refill path */
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
path = ext4_ext_find_extent(inode,
|
2008-01-29 04:58:27 +00:00
|
|
|
(ext4_lblk_t)le32_to_cpu(newext->ee_block),
|
|
|
|
path);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* only first (depth 0 -> 1) produces free space;
|
|
|
|
* in all other cases we have to split the grown tree
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* now we need to split */
|
2006-10-11 08:21:03 +00:00
|
|
|
goto repeat;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
/*
|
|
|
|
* 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
|
|
|
|
*/
|
2010-10-28 01:30:14 +00:00
|
|
|
static int ext4_ext_search_left(struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
ext4_lblk_t *logical, ext4_fsblk_t *phys)
|
2008-01-29 04:58:27 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_idx *ix;
|
|
|
|
struct ext4_extent *ex;
|
2008-01-29 04:58:27 +00:00
|
|
|
int depth, ee_len;
|
2008-01-29 04:58:27 +00:00
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
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;
|
2008-01-29 04:58:27 +00:00
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
if (*logical < le32_to_cpu(ex->ee_block)) {
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"EXT_FIRST_EXTENT != ex *logical %d ee_block %d!",
|
|
|
|
*logical, le32_to_cpu(ex->ee_block));
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
while (--depth >= 0) {
|
|
|
|
ix = path[depth].p_idx;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!",
|
|
|
|
ix != NULL ? ix->ei_block : 0,
|
|
|
|
EXT_FIRST_INDEX(path[depth].p_hdr) != NULL ?
|
|
|
|
EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block : 0,
|
|
|
|
depth);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"logical %d < ee_block %d + ee_len %d!",
|
|
|
|
*logical, le32_to_cpu(ex->ee_block), ee_len);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
*logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
|
2010-10-28 01:30:14 +00:00
|
|
|
*phys = ext4_ext_pblock(ex) + ee_len - 1;
|
2008-01-29 04:58:27 +00:00
|
|
|
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 smallest allocated block, the function
|
|
|
|
* returns 0 at @phys
|
|
|
|
* return value contains 0 (success) or error code
|
|
|
|
*/
|
2010-10-28 01:30:14 +00:00
|
|
|
static int ext4_ext_search_right(struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
ext4_lblk_t *logical, ext4_fsblk_t *phys)
|
2008-01-29 04:58:27 +00:00
|
|
|
{
|
|
|
|
struct buffer_head *bh = NULL;
|
|
|
|
struct ext4_extent_header *eh;
|
|
|
|
struct ext4_extent_idx *ix;
|
|
|
|
struct ext4_extent *ex;
|
|
|
|
ext4_fsblk_t block;
|
2009-03-10 22:18:47 +00:00
|
|
|
int depth; /* Note, NOT eh_depth; depth from top of tree */
|
|
|
|
int ee_len;
|
2008-01-29 04:58:27 +00:00
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
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;
|
2008-01-29 04:58:27 +00:00
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
if (*logical < le32_to_cpu(ex->ee_block)) {
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"first_extent(path[%d].p_hdr) != ex",
|
|
|
|
depth);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
while (--depth >= 0) {
|
|
|
|
ix = path[depth].p_idx;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"ix != EXT_FIRST_INDEX *logical %d!",
|
|
|
|
*logical);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
}
|
|
|
|
*logical = le32_to_cpu(ex->ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
*phys = ext4_ext_pblock(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"logical %d < ee_block %d + ee_len %d!",
|
|
|
|
*logical, le32_to_cpu(ex->ee_block), ee_len);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2008-01-29 04:58:27 +00:00
|
|
|
|
|
|
|
if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
|
|
|
|
/* next allocated block in this leaf */
|
|
|
|
ex++;
|
|
|
|
*logical = le32_to_cpu(ex->ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
*phys = ext4_ext_pblock(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* 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))
|
2008-11-25 22:24:23 +00:00
|
|
|
goto got_index;
|
2008-01-29 04:58:27 +00:00
|
|
|
}
|
|
|
|
|
2008-11-25 22:24:23 +00:00
|
|
|
/* we've gone up to the root and found no index to the right */
|
|
|
|
return 0;
|
2008-01-29 04:58:27 +00:00
|
|
|
|
2008-11-25 22:24:23 +00:00
|
|
|
got_index:
|
2008-01-29 04:58:27 +00:00
|
|
|
/* we've found index to the right, let's
|
|
|
|
* follow it and find the closest allocated
|
|
|
|
* block to the right */
|
|
|
|
ix++;
|
2010-10-28 01:30:14 +00:00
|
|
|
block = ext4_idx_pblock(ix);
|
2008-01-29 04:58:27 +00:00
|
|
|
while (++depth < path->p_depth) {
|
|
|
|
bh = sb_bread(inode->i_sb, block);
|
|
|
|
if (bh == NULL)
|
|
|
|
return -EIO;
|
|
|
|
eh = ext_block_hdr(bh);
|
2009-03-10 22:18:47 +00:00
|
|
|
/* subtract from p_depth to get proper eh_depth */
|
2009-03-12 13:51:20 +00:00
|
|
|
if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
|
2008-01-29 04:58:27 +00:00
|
|
|
put_bh(bh);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
ix = EXT_FIRST_INDEX(eh);
|
2010-10-28 01:30:14 +00:00
|
|
|
block = ext4_idx_pblock(ix);
|
2008-01-29 04:58:27 +00:00
|
|
|
put_bh(bh);
|
|
|
|
}
|
|
|
|
|
|
|
|
bh = sb_bread(inode->i_sb, block);
|
|
|
|
if (bh == NULL)
|
|
|
|
return -EIO;
|
|
|
|
eh = ext_block_hdr(bh);
|
2009-03-12 13:51:20 +00:00
|
|
|
if (ext4_ext_check(inode, eh, path->p_depth - depth)) {
|
2008-01-29 04:58:27 +00:00
|
|
|
put_bh(bh);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
ex = EXT_FIRST_EXTENT(eh);
|
|
|
|
*logical = le32_to_cpu(ex->ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
*phys = ext4_ext_pblock(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
put_bh(bh);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_next_allocated_block:
|
|
|
|
* returns allocated block in subsequent extent or EXT_MAX_BLOCK.
|
|
|
|
* NOTE: it considers block number from index entry as
|
|
|
|
* allocated block. Thus, index entries have to be consistent
|
|
|
|
* with leaves.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2008-01-29 04:58:27 +00:00
|
|
|
static ext4_lblk_t
|
2006-10-11 08:21:03 +00:00
|
|
|
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_BLOCK;
|
|
|
|
|
|
|
|
while (depth >= 0) {
|
|
|
|
if (depth == path->p_depth) {
|
|
|
|
/* leaf */
|
|
|
|
if (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_BLOCK;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_next_leaf_block:
|
2006-10-11 08:21:03 +00:00
|
|
|
* returns first allocated block from next leaf or EXT_MAX_BLOCK
|
|
|
|
*/
|
2008-01-29 04:58:27 +00:00
|
|
|
static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
|
2006-10-11 08:21:24 +00:00
|
|
|
struct ext4_ext_path *path)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int depth;
|
|
|
|
|
|
|
|
BUG_ON(path == NULL);
|
|
|
|
depth = path->p_depth;
|
|
|
|
|
|
|
|
/* zero-tree has no leaf blocks at all */
|
|
|
|
if (depth == 0)
|
|
|
|
return EXT_MAX_BLOCK;
|
|
|
|
|
|
|
|
/* go to index block */
|
|
|
|
depth--;
|
|
|
|
|
|
|
|
while (depth >= 0) {
|
|
|
|
if (path[depth].p_idx !=
|
|
|
|
EXT_LAST_INDEX(path[depth].p_hdr))
|
2008-01-29 04:58:27 +00:00
|
|
|
return (ext4_lblk_t)
|
|
|
|
le32_to_cpu(path[depth].p_idx[1].ei_block);
|
2006-10-11 08:21:03 +00:00
|
|
|
depth--;
|
|
|
|
}
|
|
|
|
|
|
|
|
return EXT_MAX_BLOCK;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_correct_indexes:
|
|
|
|
* if leaf gets modified and modified extent is first in the leaf,
|
|
|
|
* then we have to correct all indexes above.
|
2006-10-11 08:21:03 +00:00
|
|
|
* TODO: do we need to correct tree in all cases?
|
|
|
|
*/
|
2008-01-29 04:58:27 +00:00
|
|
|
static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
|
2006-10-11 08:21:03 +00:00
|
|
|
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;
|
2010-03-02 16:46:09 +00:00
|
|
|
|
|
|
|
if (unlikely(ex == NULL || eh == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode,
|
|
|
|
"ex %p == NULL or eh %p == NULL", ex, eh);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
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;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* TODO: we need correction if border is smaller than current one
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
k = depth - 1;
|
|
|
|
border = path[depth].p_ext->ee_block;
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + k);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
|
|
|
path[k].p_idx->ei_block = border;
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_dirty(handle, inode, path + k);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
|
|
|
|
|
|
|
while (k--) {
|
|
|
|
/* change all left-side indexes */
|
|
|
|
if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
|
|
|
|
break;
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + k);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
break;
|
|
|
|
path[k].p_idx->ei_block = border;
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_dirty(handle, inode, path + k);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2009-06-17 23:24:03 +00:00
|
|
|
int
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
|
|
|
|
struct ext4_extent *ex2)
|
|
|
|
{
|
2007-07-18 13:02:56 +00:00
|
|
|
unsigned short ext1_ee_len, ext2_ee_len, max_len;
|
2007-07-18 01:42:41 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Make sure that either both extents are uninitialized, or
|
|
|
|
* both are _not_.
|
|
|
|
*/
|
|
|
|
if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2))
|
|
|
|
return 0;
|
|
|
|
|
2007-07-18 13:02:56 +00:00
|
|
|
if (ext4_ext_is_uninitialized(ex1))
|
|
|
|
max_len = EXT_UNINIT_MAX_LEN;
|
|
|
|
else
|
|
|
|
max_len = EXT_INIT_MAX_LEN;
|
|
|
|
|
2007-07-18 01:42:41 +00:00
|
|
|
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 !=
|
2006-10-11 08:21:24 +00:00
|
|
|
le32_to_cpu(ex2->ee_block))
|
2006-10-11 08:21:03 +00:00
|
|
|
return 0;
|
|
|
|
|
2006-10-11 08:21:06 +00:00
|
|
|
/*
|
|
|
|
* To allow future support for preallocated extents to be added
|
|
|
|
* as an RO_COMPAT feature, refuse to merge to extents if
|
2006-10-11 08:21:07 +00:00
|
|
|
* this can result in the top bit of ee_len being set.
|
2006-10-11 08:21:06 +00:00
|
|
|
*/
|
2007-07-18 13:02:56 +00:00
|
|
|
if (ext1_ee_len + ext2_ee_len > max_len)
|
2006-10-11 08:21:06 +00:00
|
|
|
return 0;
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
2008-01-29 04:58:27 +00:00
|
|
|
if (ext1_ee_len >= 4)
|
2006-10-11 08:21:03 +00:00
|
|
|
return 0;
|
|
|
|
#endif
|
|
|
|
|
2010-10-28 01:30:14 +00:00
|
|
|
if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2))
|
2006-10-11 08:21:03 +00:00
|
|
|
return 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-07-18 01:42:38 +00:00
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2010-10-28 01:30:14 +00:00
|
|
|
static int ext4_ext_try_to_merge(struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
struct ext4_extent *ex)
|
2007-07-18 01:42:38 +00:00
|
|
|
{
|
|
|
|
struct ext4_extent_header *eh;
|
|
|
|
unsigned int depth, len;
|
|
|
|
int merge_done = 0;
|
|
|
|
int uninitialized = 0;
|
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
BUG_ON(path[depth].p_hdr == NULL);
|
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
|
|
|
|
while (ex < EXT_LAST_EXTENT(eh)) {
|
|
|
|
if (!ext4_can_extents_be_merged(inode, ex, ex + 1))
|
|
|
|
break;
|
|
|
|
/* merge with next extent! */
|
|
|
|
if (ext4_ext_is_uninitialized(ex))
|
|
|
|
uninitialized = 1;
|
|
|
|
ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
|
|
|
|
+ ext4_ext_get_actual_len(ex + 1));
|
|
|
|
if (uninitialized)
|
|
|
|
ext4_ext_mark_uninitialized(ex);
|
|
|
|
|
|
|
|
if (ex + 1 < EXT_LAST_EXTENT(eh)) {
|
|
|
|
len = (EXT_LAST_EXTENT(eh) - ex - 1)
|
|
|
|
* sizeof(struct ext4_extent);
|
|
|
|
memmove(ex + 1, ex + 2, len);
|
|
|
|
}
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&eh->eh_entries, -1);
|
2007-07-18 01:42:38 +00:00
|
|
|
merge_done = 1;
|
|
|
|
WARN_ON(eh->eh_entries == 0);
|
|
|
|
if (!eh->eh_entries)
|
2010-05-17 01:00:00 +00:00
|
|
|
EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!");
|
2007-07-18 01:42:38 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return merge_done;
|
|
|
|
}
|
|
|
|
|
2007-05-24 17:04:13 +00:00
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2010-10-28 01:30:14 +00:00
|
|
|
static unsigned int ext4_ext_check_overlap(struct inode *inode,
|
|
|
|
struct ext4_extent *newext,
|
|
|
|
struct ext4_ext_path *path)
|
2007-05-24 17:04:13 +00:00
|
|
|
{
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t b1, b2;
|
2007-05-24 17:04:13 +00:00
|
|
|
unsigned int depth, len1;
|
|
|
|
unsigned int ret = 0;
|
|
|
|
|
|
|
|
b1 = le32_to_cpu(newext->ee_block);
|
2007-07-18 01:42:41 +00:00
|
|
|
len1 = ext4_ext_get_actual_len(newext);
|
2007-05-24 17:04:13 +00:00
|
|
|
depth = ext_depth(inode);
|
|
|
|
if (!path[depth].p_ext)
|
|
|
|
goto out;
|
|
|
|
b2 = le32_to_cpu(path[depth].p_ext->ee_block);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get the next allocated block if the extent in the path
|
2008-07-26 20:15:44 +00:00
|
|
|
* is before the requested block(s)
|
2007-05-24 17:04:13 +00:00
|
|
|
*/
|
|
|
|
if (b2 < b1) {
|
|
|
|
b2 = ext4_ext_next_allocated_block(path);
|
|
|
|
if (b2 == EXT_MAX_BLOCK)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
/* check for wrap through zero on extent logical start block*/
|
2007-05-24 17:04:13 +00:00
|
|
|
if (b1 + len1 < b1) {
|
|
|
|
len1 = EXT_MAX_BLOCK - 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;
|
|
|
|
}
|
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* 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.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_ext_path *path,
|
2009-09-28 19:49:08 +00:00
|
|
|
struct ext4_extent *newext, int flag)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
2008-09-09 02:25:24 +00:00
|
|
|
struct ext4_extent_header *eh;
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_extent *ex, *fex;
|
|
|
|
struct ext4_extent *nearex; /* nearest extent */
|
|
|
|
struct ext4_ext_path *npath = NULL;
|
2008-01-29 04:58:27 +00:00
|
|
|
int depth, len, err;
|
|
|
|
ext4_lblk_t next;
|
2007-07-18 01:42:41 +00:00
|
|
|
unsigned uninitialized = 0;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(ext4_ext_get_actual_len(newext) == 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0");
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
depth = ext_depth(inode);
|
|
|
|
ex = path[depth].p_ext;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_hdr == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* try to insert block into found extent and return */
|
2010-03-04 21:14:02 +00:00
|
|
|
if (ex && !(flag & EXT4_GET_BLOCKS_PRE_IO)
|
2009-09-28 19:49:08 +00:00
|
|
|
&& ext4_can_extents_be_merged(inode, ex, newext)) {
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("append [%d]%d block to %d:[%d]%d (from %llu)\n",
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_is_uninitialized(newext),
|
|
|
|
ext4_ext_get_actual_len(newext),
|
|
|
|
le32_to_cpu(ex->ee_block),
|
|
|
|
ext4_ext_is_uninitialized(ex),
|
|
|
|
ext4_ext_get_actual_len(ex),
|
|
|
|
ext4_ext_pblock(ex));
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
2007-07-18 01:42:41 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* ext4_can_extents_be_merged should have checked that either
|
|
|
|
* both extents are uninitialized, or both aren't. Thus we
|
|
|
|
* need to check only one of them here.
|
|
|
|
*/
|
|
|
|
if (ext4_ext_is_uninitialized(ex))
|
|
|
|
uninitialized = 1;
|
|
|
|
ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex)
|
|
|
|
+ ext4_ext_get_actual_len(newext));
|
|
|
|
if (uninitialized)
|
|
|
|
ext4_ext_mark_uninitialized(ex);
|
2006-10-11 08:21:03 +00:00
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
nearex = ex;
|
|
|
|
goto merge;
|
|
|
|
}
|
|
|
|
|
|
|
|
repeat:
|
|
|
|
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 = ext4_ext_next_leaf_block(inode, path);
|
|
|
|
if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)
|
|
|
|
&& next != EXT_MAX_BLOCK) {
|
|
|
|
ext_debug("next leaf block - %d\n", next);
|
|
|
|
BUG_ON(npath != NULL);
|
|
|
|
npath = ext4_ext_find_extent(inode, next, NULL);
|
|
|
|
if (IS_ERR(npath))
|
|
|
|
return PTR_ERR(npath);
|
|
|
|
BUG_ON(npath->p_depth != path->p_depth);
|
|
|
|
eh = npath[depth].p_hdr;
|
|
|
|
if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
|
|
|
|
ext_debug("next leaf isnt full(%d)\n",
|
|
|
|
le16_to_cpu(eh->eh_entries));
|
|
|
|
path = npath;
|
|
|
|
goto repeat;
|
|
|
|
}
|
|
|
|
ext_debug("next leaf has no free space(%d,%d)\n",
|
|
|
|
le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* There is no free space in the found leaf.
|
|
|
|
* We're gonna add a new leaf in the tree.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
err = ext4_ext_create_new_leaf(handle, inode, path, newext);
|
|
|
|
if (err)
|
|
|
|
goto cleanup;
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
|
|
|
|
has_space:
|
|
|
|
nearex = path[depth].p_ext;
|
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto cleanup;
|
|
|
|
|
|
|
|
if (!nearex) {
|
|
|
|
/* there is no extent in this leaf, create first one */
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("first extent in the leaf: %d:%llu:[%d]%d\n",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(newext->ee_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(newext),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(newext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(newext));
|
2006-10-11 08:21:03 +00:00
|
|
|
path[depth].p_ext = EXT_FIRST_EXTENT(eh);
|
|
|
|
} else if (le32_to_cpu(newext->ee_block)
|
2007-05-24 17:04:54 +00:00
|
|
|
> le32_to_cpu(nearex->ee_block)) {
|
2006-10-11 08:21:03 +00:00
|
|
|
/* BUG_ON(newext->ee_block == nearex->ee_block); */
|
|
|
|
if (nearex != EXT_LAST_EXTENT(eh)) {
|
|
|
|
len = EXT_MAX_EXTENT(eh) - nearex;
|
|
|
|
len = (len - 1) * sizeof(struct ext4_extent);
|
|
|
|
len = len < 0 ? 0 : len;
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("insert %d:%llu:[%d]%d after: nearest 0x%p, "
|
2006-10-11 08:21:03 +00:00
|
|
|
"move %d from 0x%p to 0x%p\n",
|
2007-05-24 17:04:54 +00:00
|
|
|
le32_to_cpu(newext->ee_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(newext),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(newext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(newext),
|
2006-10-11 08:21:03 +00:00
|
|
|
nearex, len, nearex + 1, nearex + 2);
|
|
|
|
memmove(nearex + 2, nearex + 1, len);
|
|
|
|
}
|
|
|
|
path[depth].p_ext = nearex + 1;
|
|
|
|
} else {
|
|
|
|
BUG_ON(newext->ee_block == nearex->ee_block);
|
|
|
|
len = (EXT_MAX_EXTENT(eh) - nearex) * sizeof(struct ext4_extent);
|
|
|
|
len = len < 0 ? 0 : len;
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("insert %d:%llu:[%d]%d before: nearest 0x%p, "
|
2006-10-11 08:21:03 +00:00
|
|
|
"move %d from 0x%p to 0x%p\n",
|
|
|
|
le32_to_cpu(newext->ee_block),
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(newext),
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_is_uninitialized(newext),
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_get_actual_len(newext),
|
2006-10-11 08:21:03 +00:00
|
|
|
nearex, len, nearex + 1, nearex + 2);
|
|
|
|
memmove(nearex + 1, nearex, len);
|
|
|
|
path[depth].p_ext = nearex;
|
|
|
|
}
|
|
|
|
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&eh->eh_entries, 1);
|
2006-10-11 08:21:03 +00:00
|
|
|
nearex = path[depth].p_ext;
|
|
|
|
nearex->ee_block = newext->ee_block;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext));
|
2006-10-11 08:21:03 +00:00
|
|
|
nearex->ee_len = newext->ee_len;
|
|
|
|
|
|
|
|
merge:
|
|
|
|
/* try to merge extents to the right */
|
2010-03-04 21:14:02 +00:00
|
|
|
if (!(flag & EXT4_GET_BLOCKS_PRE_IO))
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_try_to_merge(inode, path, nearex);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* try to merge extents to the left */
|
|
|
|
|
|
|
|
/* time to correct all indexes above */
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
if (err)
|
|
|
|
goto cleanup;
|
|
|
|
|
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
|
|
|
|
cleanup:
|
|
|
|
if (npath) {
|
|
|
|
ext4_ext_drop_refs(npath);
|
|
|
|
kfree(npath);
|
|
|
|
}
|
|
|
|
ext4_ext_invalidate_cache(inode);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2010-10-28 01:30:14 +00:00
|
|
|
static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block,
|
|
|
|
ext4_lblk_t num, ext_prepare_callback func,
|
|
|
|
void *cbdata)
|
2008-10-07 04:46:36 +00:00
|
|
|
{
|
|
|
|
struct ext4_ext_path *path = NULL;
|
|
|
|
struct ext4_ext_cache cbex;
|
|
|
|
struct ext4_extent *ex;
|
|
|
|
ext4_lblk_t next, start = 0, end = 0;
|
|
|
|
ext4_lblk_t last = block + num;
|
|
|
|
int depth, exists, err = 0;
|
|
|
|
|
|
|
|
BUG_ON(func == NULL);
|
|
|
|
BUG_ON(inode == NULL);
|
|
|
|
|
|
|
|
while (block < last && block != EXT_MAX_BLOCK) {
|
|
|
|
num = last - block;
|
|
|
|
/* find extent for this block */
|
2009-12-10 02:30:02 +00:00
|
|
|
down_read(&EXT4_I(inode)->i_data_sem);
|
2008-10-07 04:46:36 +00:00
|
|
|
path = ext4_ext_find_extent(inode, block, path);
|
2009-12-10 02:30:02 +00:00
|
|
|
up_read(&EXT4_I(inode)->i_data_sem);
|
2008-10-07 04:46:36 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
path = NULL;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_hdr == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
|
|
|
|
err = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
2008-10-07 04:46:36 +00:00
|
|
|
ex = path[depth].p_ext;
|
|
|
|
next = ext4_ext_next_allocated_block(path);
|
|
|
|
|
|
|
|
exists = 0;
|
|
|
|
if (!ex) {
|
|
|
|
/* there is no extent yet, so try to allocate
|
|
|
|
* all requested space */
|
|
|
|
start = block;
|
|
|
|
end = block + num;
|
|
|
|
} else if (le32_to_cpu(ex->ee_block) > block) {
|
|
|
|
/* need to allocate space before found extent */
|
|
|
|
start = block;
|
|
|
|
end = le32_to_cpu(ex->ee_block);
|
|
|
|
if (block + num < end)
|
|
|
|
end = block + num;
|
|
|
|
} else if (block >= le32_to_cpu(ex->ee_block)
|
|
|
|
+ ext4_ext_get_actual_len(ex)) {
|
|
|
|
/* need to allocate space after found extent */
|
|
|
|
start = block;
|
|
|
|
end = block + num;
|
|
|
|
if (end >= next)
|
|
|
|
end = next;
|
|
|
|
} else if (block >= le32_to_cpu(ex->ee_block)) {
|
|
|
|
/*
|
|
|
|
* some part of requested space is covered
|
|
|
|
* by found extent
|
|
|
|
*/
|
|
|
|
start = block;
|
|
|
|
end = le32_to_cpu(ex->ee_block)
|
|
|
|
+ ext4_ext_get_actual_len(ex);
|
|
|
|
if (block + num < end)
|
|
|
|
end = block + num;
|
|
|
|
exists = 1;
|
|
|
|
} else {
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
BUG_ON(end <= start);
|
|
|
|
|
|
|
|
if (!exists) {
|
|
|
|
cbex.ec_block = start;
|
|
|
|
cbex.ec_len = end - start;
|
|
|
|
cbex.ec_start = 0;
|
|
|
|
cbex.ec_type = EXT4_EXT_CACHE_GAP;
|
|
|
|
} else {
|
|
|
|
cbex.ec_block = le32_to_cpu(ex->ee_block);
|
|
|
|
cbex.ec_len = ext4_ext_get_actual_len(ex);
|
2010-10-28 01:30:14 +00:00
|
|
|
cbex.ec_start = ext4_ext_pblock(ex);
|
2008-10-07 04:46:36 +00:00
|
|
|
cbex.ec_type = EXT4_EXT_CACHE_EXTENT;
|
|
|
|
}
|
|
|
|
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(cbex.ec_len == 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "cbex.ec_len == 0");
|
|
|
|
err = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
2008-10-07 04:46:36 +00:00
|
|
|
err = func(inode, path, &cbex, ex, cbdata);
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
|
|
|
|
if (err < 0)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (err == EXT_REPEAT)
|
|
|
|
continue;
|
|
|
|
else if (err == EXT_BREAK) {
|
|
|
|
err = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ext_depth(inode) != depth) {
|
|
|
|
/* depth was changed. we have to realloc path */
|
|
|
|
kfree(path);
|
|
|
|
path = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
block = cbex.ec_block + cbex.ec_len;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (path) {
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
kfree(path);
|
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2006-12-07 04:41:36 +00:00
|
|
|
static void
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
|
2007-07-31 07:37:46 +00:00
|
|
|
__u32 len, ext4_fsblk_t start, int type)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_ext_cache *cex;
|
|
|
|
BUG_ON(len == 0);
|
2009-05-15 13:07:28 +00:00
|
|
|
spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
|
2006-10-11 08:21:03 +00:00
|
|
|
cex = &EXT4_I(inode)->i_cached_extent;
|
|
|
|
cex->ec_type = type;
|
|
|
|
cex->ec_block = block;
|
|
|
|
cex->ec_len = len;
|
|
|
|
cex->ec_start = start;
|
2009-05-15 13:07:28 +00:00
|
|
|
spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_put_gap_in_cache:
|
|
|
|
* calculate boundaries of the gap that the requested block fits into
|
2006-10-11 08:21:03 +00:00
|
|
|
* and cache this gap
|
|
|
|
*/
|
2006-12-07 04:41:36 +00:00
|
|
|
static void
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t block)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int depth = ext_depth(inode);
|
2008-01-29 04:58:27 +00:00
|
|
|
unsigned long len;
|
|
|
|
ext4_lblk_t lblock;
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_extent *ex;
|
|
|
|
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
if (ex == NULL) {
|
|
|
|
/* there is no extent yet, so gap is [0;-] */
|
|
|
|
lblock = 0;
|
|
|
|
len = EXT_MAX_BLOCK;
|
|
|
|
ext_debug("cache gap(whole file):");
|
|
|
|
} else if (block < le32_to_cpu(ex->ee_block)) {
|
|
|
|
lblock = block;
|
|
|
|
len = le32_to_cpu(ex->ee_block) - block;
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("cache gap(before): %u [%u:%u]",
|
|
|
|
block,
|
|
|
|
le32_to_cpu(ex->ee_block),
|
|
|
|
ext4_ext_get_actual_len(ex));
|
2006-10-11 08:21:03 +00:00
|
|
|
} else if (block >= le32_to_cpu(ex->ee_block)
|
2007-07-18 01:42:41 +00:00
|
|
|
+ ext4_ext_get_actual_len(ex)) {
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t next;
|
2007-05-24 17:04:54 +00:00
|
|
|
lblock = le32_to_cpu(ex->ee_block)
|
2007-07-18 01:42:41 +00:00
|
|
|
+ ext4_ext_get_actual_len(ex);
|
2008-01-29 04:58:27 +00:00
|
|
|
|
|
|
|
next = ext4_ext_next_allocated_block(path);
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("cache gap(after): [%u:%u] %u",
|
|
|
|
le32_to_cpu(ex->ee_block),
|
|
|
|
ext4_ext_get_actual_len(ex),
|
|
|
|
block);
|
2008-01-29 04:58:27 +00:00
|
|
|
BUG_ON(next == lblock);
|
|
|
|
len = next - lblock;
|
2006-10-11 08:21:03 +00:00
|
|
|
} else {
|
|
|
|
lblock = len = 0;
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug(" -> %u:%lu\n", lblock, len);
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_put_in_cache(inode, lblock, len, 0, EXT4_EXT_CACHE_GAP);
|
|
|
|
}
|
|
|
|
|
2006-12-07 04:41:36 +00:00
|
|
|
static int
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_extent *ex)
|
|
|
|
{
|
|
|
|
struct ext4_ext_cache *cex;
|
2009-05-15 13:07:28 +00:00
|
|
|
int ret = EXT4_EXT_CACHE_NO;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2010-05-17 11:00:00 +00:00
|
|
|
/*
|
2009-05-15 13:07:28 +00:00
|
|
|
* We borrow i_block_reservation_lock to protect i_cached_extent
|
|
|
|
*/
|
|
|
|
spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
|
2006-10-11 08:21:03 +00:00
|
|
|
cex = &EXT4_I(inode)->i_cached_extent;
|
|
|
|
|
|
|
|
/* has cache valid data? */
|
|
|
|
if (cex->ec_type == EXT4_EXT_CACHE_NO)
|
2009-05-15 13:07:28 +00:00
|
|
|
goto errout;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
BUG_ON(cex->ec_type != EXT4_EXT_CACHE_GAP &&
|
|
|
|
cex->ec_type != EXT4_EXT_CACHE_EXTENT);
|
2010-03-04 04:55:01 +00:00
|
|
|
if (in_range(block, cex->ec_block, cex->ec_len)) {
|
2007-05-24 17:04:54 +00:00
|
|
|
ex->ee_block = cpu_to_le32(cex->ec_block);
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_ext_store_pblock(ex, cex->ec_start);
|
2007-05-24 17:04:54 +00:00
|
|
|
ex->ee_len = cpu_to_le16(cex->ec_len);
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("%u cached by %u:%u:%llu\n",
|
|
|
|
block,
|
|
|
|
cex->ec_block, cex->ec_len, cex->ec_start);
|
2009-05-15 13:07:28 +00:00
|
|
|
ret = cex->ec_type;
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
2009-05-15 13:07:28 +00:00
|
|
|
errout:
|
|
|
|
spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
|
|
|
|
return ret;
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_rm_idx:
|
|
|
|
* removes index from the index block.
|
|
|
|
* It's used in truncate case only, thus all requests are for
|
|
|
|
* last index in the block only.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2008-01-29 04:58:27 +00:00
|
|
|
static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_ext_path *path)
|
|
|
|
{
|
|
|
|
int err;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t leaf;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* free index block */
|
|
|
|
path--;
|
2010-10-28 01:30:14 +00:00
|
|
|
leaf = ext4_idx_pblock(path->p_idx);
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path->p_hdr->eh_entries == 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&path->p_hdr->eh_entries, -1);
|
2006-12-07 04:41:33 +00:00
|
|
|
err = ext4_ext_dirty(handle, inode, path);
|
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
2006-10-11 08:21:11 +00:00
|
|
|
ext_debug("index is empty, remove it, free block %llu\n", leaf);
|
2009-11-23 12:17:05 +00:00
|
|
|
ext4_free_blocks(handle, inode, 0, leaf, 1,
|
|
|
|
EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
|
2006-10-11 08:21:03 +00:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2008-08-20 02:16:05 +00:00
|
|
|
* 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.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2008-08-20 02:15:58 +00:00
|
|
|
int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_ext_path *path)
|
|
|
|
{
|
|
|
|
if (path) {
|
2008-08-20 02:16:05 +00:00
|
|
|
int depth = ext_depth(inode);
|
2008-08-20 02:16:03 +00:00
|
|
|
int ret = 0;
|
2008-08-20 02:16:05 +00:00
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
/* probably there is space in leaf? */
|
|
|
|
if (le16_to_cpu(path[depth].p_hdr->eh_entries)
|
2008-08-20 02:16:05 +00:00
|
|
|
< le16_to_cpu(path[depth].p_hdr->eh_max)) {
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-08-20 02:16:05 +00:00
|
|
|
/*
|
|
|
|
* There are some space in the leaf tree, no
|
|
|
|
* need to account for leaf block credit
|
|
|
|
*
|
|
|
|
* bitmaps and block group descriptor blocks
|
|
|
|
* and other metadat blocks still need to be
|
|
|
|
* accounted.
|
|
|
|
*/
|
2008-08-20 02:15:58 +00:00
|
|
|
/* 1 bitmap, 1 block group descriptor */
|
2008-08-20 02:16:05 +00:00
|
|
|
ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb);
|
2009-07-06 03:12:04 +00:00
|
|
|
return ret;
|
2008-08-20 02:16:05 +00:00
|
|
|
}
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-08-20 02:15:58 +00:00
|
|
|
return ext4_chunk_trans_blocks(inode, nrblocks);
|
2008-08-20 02:16:05 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-08-20 02:16:05 +00:00
|
|
|
/*
|
|
|
|
* How many index/leaf blocks need to change/allocate to modify nrblocks?
|
|
|
|
*
|
|
|
|
* if nrblocks are fit in a single extent (chunk flag is 1), then
|
|
|
|
* in the worse case, each tree level index/leaf need to be changed
|
|
|
|
* if the tree split due to insert a new extent, then the old tree
|
|
|
|
* index/leaf need to be updated too
|
|
|
|
*
|
|
|
|
* If the nrblocks are discontiguous, they could cause
|
|
|
|
* the whole tree split more than once, but this is really rare.
|
|
|
|
*/
|
2008-08-20 02:15:58 +00:00
|
|
|
int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
|
2008-08-20 02:16:05 +00:00
|
|
|
{
|
|
|
|
int index;
|
|
|
|
int depth = ext_depth(inode);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-08-20 02:16:05 +00:00
|
|
|
if (chunk)
|
|
|
|
index = depth * 2;
|
|
|
|
else
|
|
|
|
index = depth * 3;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-08-20 02:16:05 +00:00
|
|
|
return index;
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_extent *ex,
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t from, ext4_lblk_t to)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
2007-07-18 01:42:41 +00:00
|
|
|
unsigned short ee_len = ext4_ext_get_actual_len(ex);
|
2009-11-23 12:17:05 +00:00
|
|
|
int flags = EXT4_FREE_BLOCKS_FORGET;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-01-29 05:19:52 +00:00
|
|
|
if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
|
2009-11-23 12:17:05 +00:00
|
|
|
flags |= EXT4_FREE_BLOCKS_METADATA;
|
2006-10-11 08:21:03 +00:00
|
|
|
#ifdef EXTENTS_STATS
|
|
|
|
{
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
|
|
|
|
spin_lock(&sbi->s_ext_stats_lock);
|
|
|
|
sbi->s_ext_blocks += ee_len;
|
|
|
|
sbi->s_ext_extents++;
|
|
|
|
if (ee_len < sbi->s_ext_min)
|
|
|
|
sbi->s_ext_min = ee_len;
|
|
|
|
if (ee_len > sbi->s_ext_max)
|
|
|
|
sbi->s_ext_max = ee_len;
|
|
|
|
if (ext_depth(inode) > sbi->s_depth_max)
|
|
|
|
sbi->s_depth_max = ext_depth(inode);
|
|
|
|
spin_unlock(&sbi->s_ext_stats_lock);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
if (from >= le32_to_cpu(ex->ee_block)
|
2007-07-18 01:42:41 +00:00
|
|
|
&& to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
|
2006-10-11 08:21:03 +00:00
|
|
|
/* tail removal */
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t num;
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_fsblk_t start;
|
2008-01-29 04:58:27 +00:00
|
|
|
|
2007-07-18 01:42:41 +00:00
|
|
|
num = le32_to_cpu(ex->ee_block) + ee_len - from;
|
2010-10-28 01:30:14 +00:00
|
|
|
start = ext4_ext_pblock(ex) + ee_len - num;
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("free last %u blocks starting %llu\n", num, start);
|
2009-11-23 12:17:05 +00:00
|
|
|
ext4_free_blocks(handle, inode, 0, start, num, flags);
|
2006-10-11 08:21:03 +00:00
|
|
|
} else if (from == le32_to_cpu(ex->ee_block)
|
2007-07-18 01:42:41 +00:00
|
|
|
&& to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
|
2008-01-29 04:58:27 +00:00
|
|
|
printk(KERN_INFO "strange request: removal %u-%u from %u:%u\n",
|
2007-07-18 01:42:41 +00:00
|
|
|
from, to, le32_to_cpu(ex->ee_block), ee_len);
|
2006-10-11 08:21:03 +00:00
|
|
|
} else {
|
2008-01-29 04:58:27 +00:00
|
|
|
printk(KERN_INFO "strange request: removal(2) "
|
|
|
|
"%u-%u from %u:%u\n",
|
|
|
|
from, to, le32_to_cpu(ex->ee_block), ee_len);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
|
2008-01-29 04:58:27 +00:00
|
|
|
struct ext4_ext_path *path, ext4_lblk_t start)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
int err = 0, correct_index = 0;
|
|
|
|
int depth = ext_depth(inode), credits;
|
|
|
|
struct ext4_extent_header *eh;
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t a, b, block;
|
|
|
|
unsigned num;
|
|
|
|
ext4_lblk_t ex_ee_block;
|
2006-10-11 08:21:03 +00:00
|
|
|
unsigned short ex_ee_len;
|
2007-07-18 01:42:41 +00:00
|
|
|
unsigned uninitialized = 0;
|
2006-10-11 08:21:03 +00:00
|
|
|
struct ext4_extent *ex;
|
|
|
|
|
2007-07-18 13:19:09 +00:00
|
|
|
/* the header must be checked already in ext4_ext_remove_space() */
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("truncate since %u in leaf\n", start);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (!path[depth].p_hdr)
|
|
|
|
path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
|
|
|
|
eh = path[depth].p_hdr;
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_hdr == NULL)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
/* find where to start removing */
|
|
|
|
ex = EXT_LAST_EXTENT(eh);
|
|
|
|
|
|
|
|
ex_ee_block = le32_to_cpu(ex->ee_block);
|
2007-07-18 01:42:41 +00:00
|
|
|
ex_ee_len = ext4_ext_get_actual_len(ex);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
while (ex >= EXT_FIRST_EXTENT(eh) &&
|
|
|
|
ex_ee_block + ex_ee_len > start) {
|
2009-06-10 18:22:55 +00:00
|
|
|
|
|
|
|
if (ext4_ext_is_uninitialized(ex))
|
|
|
|
uninitialized = 1;
|
|
|
|
else
|
|
|
|
uninitialized = 0;
|
|
|
|
|
2009-09-18 17:34:55 +00:00
|
|
|
ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
|
|
|
|
uninitialized, ex_ee_len);
|
2006-10-11 08:21:03 +00:00
|
|
|
path[depth].p_ext = ex;
|
|
|
|
|
|
|
|
a = ex_ee_block > start ? ex_ee_block : start;
|
|
|
|
b = ex_ee_block + ex_ee_len - 1 < EXT_MAX_BLOCK ?
|
|
|
|
ex_ee_block + ex_ee_len - 1 : EXT_MAX_BLOCK;
|
|
|
|
|
|
|
|
ext_debug(" border %u:%u\n", a, b);
|
|
|
|
|
|
|
|
if (a != ex_ee_block && b != ex_ee_block + ex_ee_len - 1) {
|
|
|
|
block = 0;
|
|
|
|
num = 0;
|
|
|
|
BUG();
|
|
|
|
} else if (a != ex_ee_block) {
|
|
|
|
/* remove tail of the extent */
|
|
|
|
block = ex_ee_block;
|
|
|
|
num = a - block;
|
|
|
|
} else if (b != ex_ee_block + ex_ee_len - 1) {
|
|
|
|
/* remove head of the extent */
|
|
|
|
block = a;
|
|
|
|
num = b - a;
|
|
|
|
/* there is no "make a hole" API yet */
|
|
|
|
BUG();
|
|
|
|
} else {
|
|
|
|
/* remove whole extent: excellent! */
|
|
|
|
block = ex_ee_block;
|
|
|
|
num = 0;
|
|
|
|
BUG_ON(a != ex_ee_block);
|
|
|
|
BUG_ON(b != ex_ee_block + ex_ee_len - 1);
|
|
|
|
}
|
|
|
|
|
2008-08-02 01:59:19 +00:00
|
|
|
/*
|
|
|
|
* 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));
|
2006-10-11 08:21:03 +00:00
|
|
|
if (ex == EXT_FIRST_EXTENT(eh)) {
|
|
|
|
correct_index = 1;
|
|
|
|
credits += (ext_depth(inode)) + 1;
|
|
|
|
}
|
2009-12-09 03:42:15 +00:00
|
|
|
credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2009-08-18 02:17:20 +00:00
|
|
|
err = ext4_ext_truncate_extend_restart(handle, inode, credits);
|
2008-07-11 23:27:31 +00:00
|
|
|
if (err)
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out;
|
|
|
|
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
err = ext4_remove_blocks(handle, inode, ex, a, b);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
if (num == 0) {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* this extent is removed; mark slot entirely unused */
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_ext_store_pblock(ex, 0);
|
2008-04-17 14:38:59 +00:00
|
|
|
le16_add_cpu(&eh->eh_entries, -1);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
ex->ee_block = cpu_to_le32(block);
|
|
|
|
ex->ee_len = cpu_to_le16(num);
|
2007-07-18 13:02:56 +00:00
|
|
|
/*
|
|
|
|
* Do not mark uninitialized if all the blocks in the
|
|
|
|
* extent have been removed.
|
|
|
|
*/
|
|
|
|
if (uninitialized && num)
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
2006-10-11 08:21:11 +00:00
|
|
|
ext_debug("new extent: %u:%u:%llu\n", block, num,
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_pblock(ex));
|
2006-10-11 08:21:03 +00:00
|
|
|
ex--;
|
|
|
|
ex_ee_block = le32_to_cpu(ex->ee_block);
|
2007-07-18 01:42:41 +00:00
|
|
|
ex_ee_len = ext4_ext_get_actual_len(ex);
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (correct_index && eh->eh_entries)
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
|
|
|
|
/* 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;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* ext4_ext_more_to_rm:
|
|
|
|
* returns 1 if current index has to be freed (even partial)
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2006-12-07 04:41:36 +00:00
|
|
|
static int
|
2006-10-11 08:21:03 +00:00
|
|
|
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;
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* if truncate on deeper level happened, it wasn't partial,
|
2006-10-11 08:21:03 +00:00
|
|
|
* 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;
|
|
|
|
}
|
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct super_block *sb = inode->i_sb;
|
|
|
|
int depth = ext_depth(inode);
|
|
|
|
struct ext4_ext_path *path;
|
|
|
|
handle_t *handle;
|
2010-05-17 05:00:00 +00:00
|
|
|
int i, err;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
ext_debug("truncate since %u\n", start);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* probably first extent we're gonna free will be last in block */
|
|
|
|
handle = ext4_journal_start(inode, depth + 1);
|
|
|
|
if (IS_ERR(handle))
|
|
|
|
return PTR_ERR(handle);
|
|
|
|
|
2010-05-17 05:00:00 +00:00
|
|
|
again:
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_invalidate_cache(inode);
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* We start scanning from right side, freeing all the blocks
|
|
|
|
* after i_size and walking into the tree depth-wise.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
2010-05-17 05:00:00 +00:00
|
|
|
depth = ext_depth(inode);
|
2008-04-30 02:02:02 +00:00
|
|
|
path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (path == NULL) {
|
|
|
|
ext4_journal_stop(handle);
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2010-05-17 05:00:00 +00:00
|
|
|
path[0].p_depth = depth;
|
2006-10-11 08:21:03 +00:00
|
|
|
path[0].p_hdr = ext_inode_hdr(inode);
|
2009-03-12 13:51:20 +00:00
|
|
|
if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
|
2006-10-11 08:21:03 +00:00
|
|
|
err = -EIO;
|
|
|
|
goto out;
|
|
|
|
}
|
2010-05-17 05:00:00 +00:00
|
|
|
i = err = 0;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
while (i >= 0 && err == 0) {
|
|
|
|
if (i == depth) {
|
|
|
|
/* this is leaf block */
|
|
|
|
err = ext4_ext_rm_leaf(handle, inode, path, start);
|
2006-10-11 08:21:07 +00:00
|
|
|
/* root level has p_bh == NULL, brelse() eats this */
|
2006-10-11 08:21:03 +00:00
|
|
|
brelse(path[i].p_bh);
|
|
|
|
path[i].p_bh = NULL;
|
|
|
|
i--;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* this is index block */
|
|
|
|
if (!path[i].p_hdr) {
|
|
|
|
ext_debug("initialize header\n");
|
|
|
|
path[i].p_hdr = ext_block_hdr(path[i].p_bh);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!path[i].p_idx) {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* this level hasn't been touched yet */
|
2006-10-11 08:21:03 +00:00
|
|
|
path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
|
|
|
|
path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
|
|
|
|
ext_debug("init index ptr: hdr 0x%p, num %d\n",
|
|
|
|
path[i].p_hdr,
|
|
|
|
le16_to_cpu(path[i].p_hdr->eh_entries));
|
|
|
|
} else {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* we were already here, see at next index */
|
2006-10-11 08:21:03 +00:00
|
|
|
path[i].p_idx--;
|
|
|
|
}
|
|
|
|
|
|
|
|
ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
|
|
|
|
i, EXT_FIRST_INDEX(path[i].p_hdr),
|
|
|
|
path[i].p_idx);
|
|
|
|
if (ext4_ext_more_to_rm(path + i)) {
|
2007-07-18 13:19:09 +00:00
|
|
|
struct buffer_head *bh;
|
2006-10-11 08:21:03 +00:00
|
|
|
/* go to the next level */
|
2006-10-11 08:21:11 +00:00
|
|
|
ext_debug("move to level %d (block %llu)\n",
|
2010-10-28 01:30:14 +00:00
|
|
|
i + 1, ext4_idx_pblock(path[i].p_idx));
|
2006-10-11 08:21:03 +00:00
|
|
|
memset(path + i + 1, 0, sizeof(*path));
|
2010-10-28 01:30:14 +00:00
|
|
|
bh = sb_bread(sb, ext4_idx_pblock(path[i].p_idx));
|
2007-07-18 13:19:09 +00:00
|
|
|
if (!bh) {
|
2006-10-11 08:21:03 +00:00
|
|
|
/* should we reset i_size? */
|
|
|
|
err = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
2007-07-18 13:19:09 +00:00
|
|
|
if (WARN_ON(i + 1 > depth)) {
|
|
|
|
err = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
2009-03-12 13:51:20 +00:00
|
|
|
if (ext4_ext_check(inode, ext_block_hdr(bh),
|
2007-07-18 13:19:09 +00:00
|
|
|
depth - i - 1)) {
|
|
|
|
err = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
path[i + 1].p_bh = bh;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2006-10-11 08:21:07 +00:00
|
|
|
/* save actual number of indexes since this
|
|
|
|
* number is changed at the next iteration */
|
2006-10-11 08:21:03 +00:00
|
|
|
path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
|
|
|
|
i++;
|
|
|
|
} else {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* we finished processing this index, go up */
|
2006-10-11 08:21:03 +00:00
|
|
|
if (path[i].p_hdr->eh_entries == 0 && i > 0) {
|
2006-10-11 08:21:07 +00:00
|
|
|
/* index is empty, remove it;
|
2006-10-11 08:21:03 +00:00
|
|
|
* handle must be already prepared by the
|
|
|
|
* truncatei_leaf() */
|
|
|
|
err = ext4_ext_rm_idx(handle, inode, path + i);
|
|
|
|
}
|
2006-10-11 08:21:07 +00:00
|
|
|
/* root level has p_bh == NULL, brelse() eats this */
|
2006-10-11 08:21:03 +00:00
|
|
|
brelse(path[i].p_bh);
|
|
|
|
path[i].p_bh = NULL;
|
|
|
|
i--;
|
|
|
|
ext_debug("return to level %d\n", i);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* TODO: flexible tree reduction should be here */
|
|
|
|
if (path->p_hdr->eh_entries == 0) {
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* truncate to zero freed all the tree,
|
|
|
|
* so we need to correct eh_depth
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
err = ext4_ext_get_access(handle, inode, path);
|
|
|
|
if (err == 0) {
|
|
|
|
ext_inode_hdr(inode)->eh_depth = 0;
|
|
|
|
ext_inode_hdr(inode)->eh_max =
|
2009-08-28 14:40:33 +00:00
|
|
|
cpu_to_le16(ext4_ext_space_root(inode, 0));
|
2006-10-11 08:21:03 +00:00
|
|
|
err = ext4_ext_dirty(handle, inode, path);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
out:
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
kfree(path);
|
2010-05-17 05:00:00 +00:00
|
|
|
if (err == -EAGAIN)
|
|
|
|
goto again;
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_journal_stop(handle);
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* called at mount time
|
|
|
|
*/
|
|
|
|
void ext4_ext_init(struct super_block *sb)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* possible initialization would be here
|
|
|
|
*/
|
|
|
|
|
2009-01-06 19:53:16 +00:00
|
|
|
if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
|
2009-09-29 19:51:30 +00:00
|
|
|
#if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS)
|
2008-09-09 03:00:52 +00:00
|
|
|
printk(KERN_INFO "EXT4-fs: file extents enabled");
|
2007-02-17 18:20:16 +00:00
|
|
|
#ifdef AGGRESSIVE_TEST
|
|
|
|
printk(", aggressive tests");
|
2006-10-11 08:21:03 +00:00
|
|
|
#endif
|
|
|
|
#ifdef CHECK_BINSEARCH
|
|
|
|
printk(", check binsearch");
|
|
|
|
#endif
|
|
|
|
#ifdef EXTENTS_STATS
|
|
|
|
printk(", stats");
|
|
|
|
#endif
|
|
|
|
printk("\n");
|
2009-09-29 19:51:30 +00:00
|
|
|
#endif
|
2006-10-11 08:21:03 +00:00
|
|
|
#ifdef EXTENTS_STATS
|
|
|
|
spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock);
|
|
|
|
EXT4_SB(sb)->s_ext_min = 1 << 30;
|
|
|
|
EXT4_SB(sb)->s_ext_max = 0;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* called at umount time
|
|
|
|
*/
|
|
|
|
void ext4_ext_release(struct super_block *sb)
|
|
|
|
{
|
2009-01-06 19:53:16 +00:00
|
|
|
if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS))
|
2006-10-11 08:21:03 +00:00
|
|
|
return;
|
|
|
|
|
|
|
|
#ifdef EXTENTS_STATS
|
|
|
|
if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) {
|
|
|
|
struct ext4_sb_info *sbi = EXT4_SB(sb);
|
|
|
|
printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n",
|
|
|
|
sbi->s_ext_blocks, sbi->s_ext_extents,
|
|
|
|
sbi->s_ext_blocks / sbi->s_ext_extents);
|
|
|
|
printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n",
|
|
|
|
sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2008-04-29 12:11:12 +00:00
|
|
|
/* FIXME!! we need to try to merge to left or right after zero-out */
|
|
|
|
static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
|
|
|
|
{
|
2010-10-28 01:30:06 +00:00
|
|
|
ext4_fsblk_t ee_pblock;
|
|
|
|
unsigned int ee_len;
|
2010-05-12 04:00:00 +00:00
|
|
|
int ret;
|
2008-04-29 12:11:12 +00:00
|
|
|
|
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2010-10-28 01:30:14 +00:00
|
|
|
ee_pblock = ext4_ext_pblock(ex);
|
2010-05-12 04:00:00 +00:00
|
|
|
|
2010-10-28 03:44:47 +00:00
|
|
|
ret = sb_issue_zeroout(inode->i_sb, ee_pblock, ee_len, GFP_NOFS);
|
2010-10-28 01:30:06 +00:00
|
|
|
if (ret > 0)
|
|
|
|
ret = 0;
|
2008-04-29 12:11:12 +00:00
|
|
|
|
2010-10-28 01:30:06 +00:00
|
|
|
return ret;
|
2008-04-29 12:11:12 +00:00
|
|
|
}
|
|
|
|
|
2008-04-17 14:38:59 +00:00
|
|
|
#define EXT4_EXT_ZERO_LEN 7
|
2007-07-18 01:42:38 +00:00
|
|
|
/*
|
2010-05-16 23:00:00 +00:00
|
|
|
* This function is called by ext4_ext_map_blocks() if someone tries to write
|
2007-07-18 01:42:38 +00:00
|
|
|
* to an uninitialized extent. It may result in splitting the uninitialized
|
|
|
|
* extent into multiple extents (upto three - one initialized and two
|
|
|
|
* uninitialized).
|
|
|
|
* There are three possibilities:
|
|
|
|
* a> There is no split required: Entire extent should be initialized
|
|
|
|
* b> Splits in two extents: Write is happening at either end of the extent
|
|
|
|
* c> Splits in three extents: Somone is writing in middle of the extent
|
|
|
|
*/
|
2008-01-29 04:58:27 +00:00
|
|
|
static int ext4_ext_convert_to_initialized(handle_t *handle,
|
2010-05-16 23:00:00 +00:00
|
|
|
struct inode *inode,
|
|
|
|
struct ext4_map_blocks *map,
|
|
|
|
struct ext4_ext_path *path)
|
2007-07-18 01:42:38 +00:00
|
|
|
{
|
2008-04-17 14:38:59 +00:00
|
|
|
struct ext4_extent *ex, newex, orig_ex;
|
2007-07-18 01:42:38 +00:00
|
|
|
struct ext4_extent *ex1 = NULL;
|
|
|
|
struct ext4_extent *ex2 = NULL;
|
|
|
|
struct ext4_extent *ex3 = NULL;
|
|
|
|
struct ext4_extent_header *eh;
|
2010-05-16 10:00:00 +00:00
|
|
|
ext4_lblk_t ee_block, eof_block;
|
2008-01-29 04:58:27 +00:00
|
|
|
unsigned int allocated, ee_len, depth;
|
2007-07-18 01:42:38 +00:00
|
|
|
ext4_fsblk_t newblock;
|
|
|
|
int err = 0;
|
|
|
|
int ret = 0;
|
2010-05-16 10:00:00 +00:00
|
|
|
int may_zeroout;
|
|
|
|
|
|
|
|
ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
|
|
|
|
"block %llu, max_blocks %u\n", inode->i_ino,
|
2010-05-16 23:00:00 +00:00
|
|
|
(unsigned long long)map->m_lblk, map->m_len);
|
2010-05-16 10:00:00 +00:00
|
|
|
|
|
|
|
eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
|
|
|
|
inode->i_sb->s_blocksize_bits;
|
2010-05-16 23:00:00 +00:00
|
|
|
if (eof_block < map->m_lblk + map->m_len)
|
|
|
|
eof_block = map->m_lblk + map->m_len;
|
2007-07-18 01:42:38 +00:00
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
ee_block = le32_to_cpu(ex->ee_block);
|
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = ee_len - (map->m_lblk - ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
newblock = map->m_lblk - ee_block + ext4_ext_pblock(ex);
|
2010-05-16 10:00:00 +00:00
|
|
|
|
2007-07-18 01:42:38 +00:00
|
|
|
ex2 = ex;
|
2008-04-17 14:38:59 +00:00
|
|
|
orig_ex.ee_block = ex->ee_block;
|
|
|
|
orig_ex.ee_len = cpu_to_le16(ee_len);
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(&orig_ex, ext4_ext_pblock(ex));
|
2007-07-18 01:42:38 +00:00
|
|
|
|
2010-05-16 10:00:00 +00:00
|
|
|
/*
|
|
|
|
* It is safe to convert extent to initialized via explicit
|
|
|
|
* zeroout only if extent is fully insde i_size or new_size.
|
|
|
|
*/
|
|
|
|
may_zeroout = ee_block + ee_len <= eof_block;
|
|
|
|
|
2008-02-22 11:17:31 +00:00
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
2008-04-17 14:38:59 +00:00
|
|
|
/* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
|
2010-05-16 10:00:00 +00:00
|
|
|
if (ee_len <= 2*EXT4_EXT_ZERO_LEN && may_zeroout) {
|
2008-04-17 14:38:59 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
2008-04-30 02:03:59 +00:00
|
|
|
/* zeroed the full extent */
|
|
|
|
return allocated;
|
2008-04-17 14:38:59 +00:00
|
|
|
}
|
2008-02-22 11:17:31 +00:00
|
|
|
|
2010-05-16 23:00:00 +00:00
|
|
|
/* ex1: ee_block to map->m_lblk - 1 : uninitialized */
|
|
|
|
if (map->m_lblk > ee_block) {
|
2007-07-18 01:42:38 +00:00
|
|
|
ex1 = ex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block);
|
2007-07-18 01:42:38 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex1);
|
|
|
|
ex2 = &newex;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* for sanity, update the length of the ex2 extent before
|
|
|
|
* we insert ex3, if ex1 is NULL. This is to avoid temporary
|
|
|
|
* overlap of blocks.
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
if (!ex1 && allocated > map->m_len)
|
|
|
|
ex2->ee_len = cpu_to_le16(map->m_len);
|
2007-07-18 01:42:38 +00:00
|
|
|
/* ex3: to ee_block + ee_len : uninitialised */
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len) {
|
2007-07-18 01:42:38 +00:00
|
|
|
unsigned int newdepth;
|
2008-04-17 14:38:59 +00:00
|
|
|
/* If extent has less than EXT4_EXT_ZERO_LEN zerout directly */
|
2010-05-16 10:00:00 +00:00
|
|
|
if (allocated <= EXT4_EXT_ZERO_LEN && may_zeroout) {
|
2008-08-02 22:51:32 +00:00
|
|
|
/*
|
2010-05-16 23:00:00 +00:00
|
|
|
* map->m_lblk == ee_block is handled by the zerouout
|
2008-08-02 22:51:32 +00:00
|
|
|
* at the beginning.
|
|
|
|
* Mark first half uninitialized.
|
2008-04-17 14:38:59 +00:00
|
|
|
* Mark second half initialized and zero out the
|
|
|
|
* initialized extent
|
|
|
|
*/
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = cpu_to_le16(ee_len - allocated);
|
|
|
|
ext4_ext_mark_uninitialized(ex);
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
|
|
|
|
ex3 = &newex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex3->ee_block = cpu_to_le32(map->m_lblk);
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_store_pblock(ex3, newblock);
|
|
|
|
ex3->ee_len = cpu_to_le16(allocated);
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_insert_extent(handle, inode, path,
|
|
|
|
ex3, 0);
|
2008-04-17 14:38:59 +00:00
|
|
|
if (err == -ENOSPC) {
|
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex,
|
|
|
|
ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
2010-05-16 23:00:00 +00:00
|
|
|
/* blocks available from map->m_lblk */
|
2008-04-30 02:03:59 +00:00
|
|
|
return allocated;
|
2008-04-17 14:38:59 +00:00
|
|
|
|
|
|
|
} else if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
|
2008-04-30 02:03:59 +00:00
|
|
|
/*
|
|
|
|
* We need to zero out the second half because
|
|
|
|
* an fallocate request can update file size and
|
|
|
|
* converting the second half to initialized extent
|
|
|
|
* implies that we can leak some junk data to user
|
|
|
|
* space.
|
|
|
|
*/
|
|
|
|
err = ext4_ext_zeroout(inode, ex3);
|
|
|
|
if (err) {
|
|
|
|
/*
|
|
|
|
* We should actually mark the
|
|
|
|
* second half as uninit and return error
|
|
|
|
* Insert would have changed the extent
|
|
|
|
*/
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
ext4_ext_drop_refs(path);
|
2010-05-16 23:00:00 +00:00
|
|
|
path = ext4_ext_find_extent(inode, map->m_lblk,
|
|
|
|
path);
|
2008-04-30 02:03:59 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
return err;
|
|
|
|
}
|
2008-08-02 22:51:32 +00:00
|
|
|
/* get the second half extent details */
|
2008-04-30 02:03:59 +00:00
|
|
|
ex = path[depth].p_ext;
|
|
|
|
err = ext4_ext_get_access(handle, inode,
|
|
|
|
path + depth);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
ext4_ext_mark_uninitialized(ex);
|
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* zeroed the second half */
|
2008-04-17 14:38:59 +00:00
|
|
|
return allocated;
|
|
|
|
}
|
2007-07-18 01:42:38 +00:00
|
|
|
ex3 = &newex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex3->ee_block = cpu_to_le32(map->m_lblk + map->m_len);
|
|
|
|
ext4_ext_store_pblock(ex3, newblock + map->m_len);
|
|
|
|
ex3->ee_len = cpu_to_le16(allocated - map->m_len);
|
2007-07-18 01:42:38 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex3);
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_insert_extent(handle, inode, path, ex3, 0);
|
2010-05-16 10:00:00 +00:00
|
|
|
if (err == -ENOSPC && may_zeroout) {
|
2008-04-29 12:11:12 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
2008-04-17 14:38:59 +00:00
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
2008-04-30 02:03:59 +00:00
|
|
|
/* zeroed the full extent */
|
2010-05-16 23:00:00 +00:00
|
|
|
/* blocks available from map->m_lblk */
|
2008-04-30 02:03:59 +00:00
|
|
|
return allocated;
|
2008-04-29 12:11:12 +00:00
|
|
|
|
|
|
|
} else if (err)
|
|
|
|
goto fix_extent_len;
|
2007-07-18 01:42:38 +00:00
|
|
|
/*
|
|
|
|
* The depth, and hence eh & ex might change
|
|
|
|
* as part of the insert above.
|
|
|
|
*/
|
|
|
|
newdepth = ext_depth(inode);
|
2008-04-17 14:38:59 +00:00
|
|
|
/*
|
2009-01-08 02:09:16 +00:00
|
|
|
* update the extent length after successful insert of the
|
2008-04-17 14:38:59 +00:00
|
|
|
* split extent
|
|
|
|
*/
|
2010-05-16 10:00:00 +00:00
|
|
|
ee_len -= ext4_ext_get_actual_len(ex3);
|
|
|
|
orig_ex.ee_len = cpu_to_le16(ee_len);
|
|
|
|
may_zeroout = ee_block + ee_len <= eof_block;
|
|
|
|
|
2008-08-02 22:51:32 +00:00
|
|
|
depth = newdepth;
|
|
|
|
ext4_ext_drop_refs(path);
|
2010-05-16 23:00:00 +00:00
|
|
|
path = ext4_ext_find_extent(inode, map->m_lblk, path);
|
2008-08-02 22:51:32 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
goto out;
|
2007-07-18 01:42:38 +00:00
|
|
|
}
|
2008-08-02 22:51:32 +00:00
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
if (ex2 != &newex)
|
|
|
|
ex2 = ex;
|
|
|
|
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = map->m_len;
|
2008-04-17 14:38:59 +00:00
|
|
|
|
|
|
|
/* If extent has less than EXT4_EXT_ZERO_LEN and we are trying
|
|
|
|
* to insert a extent in the middle zerout directly
|
|
|
|
* otherwise give the extent a chance to merge to left
|
|
|
|
*/
|
|
|
|
if (le16_to_cpu(orig_ex.ee_len) <= EXT4_EXT_ZERO_LEN &&
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_lblk != ee_block && may_zeroout) {
|
2008-04-17 14:38:59 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
2008-04-30 02:03:59 +00:00
|
|
|
/* zero out the first half */
|
2010-05-16 23:00:00 +00:00
|
|
|
/* blocks available from map->m_lblk */
|
2008-04-30 02:03:59 +00:00
|
|
|
return allocated;
|
2008-04-17 14:38:59 +00:00
|
|
|
}
|
2007-07-18 01:42:38 +00:00
|
|
|
}
|
|
|
|
/*
|
|
|
|
* If there was a change of depth as part of the
|
|
|
|
* insertion of ex3 above, we need to update the length
|
|
|
|
* of the ex1 extent again here
|
|
|
|
*/
|
|
|
|
if (ex1 && ex1 != ex) {
|
|
|
|
ex1 = ex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block);
|
2007-07-18 01:42:38 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex1);
|
|
|
|
ex2 = &newex;
|
|
|
|
}
|
2010-05-16 23:00:00 +00:00
|
|
|
/* ex2: map->m_lblk to map->m_lblk + maxblocks-1 : initialised */
|
|
|
|
ex2->ee_block = cpu_to_le32(map->m_lblk);
|
2007-07-18 01:42:38 +00:00
|
|
|
ext4_ext_store_pblock(ex2, newblock);
|
|
|
|
ex2->ee_len = cpu_to_le16(allocated);
|
|
|
|
if (ex2 != ex)
|
|
|
|
goto insert;
|
|
|
|
/*
|
|
|
|
* New (initialized) extent starts from the first block
|
|
|
|
* in the current extent. i.e., ex2 == ex
|
|
|
|
* We have to see if it can be merged with the extent
|
|
|
|
* on the left.
|
|
|
|
*/
|
|
|
|
if (ex2 > EXT_FIRST_EXTENT(eh)) {
|
|
|
|
/*
|
|
|
|
* To merge left, pass "ex2 - 1" to try_to_merge(),
|
|
|
|
* since it merges towards right _only_.
|
|
|
|
*/
|
|
|
|
ret = ext4_ext_try_to_merge(inode, path, ex2 - 1);
|
|
|
|
if (ret) {
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
ex2--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Try to Merge towards right. This might be required
|
|
|
|
* only when the whole extent is being written to.
|
|
|
|
* i.e. ex2 == ex and ex3 == NULL.
|
|
|
|
*/
|
|
|
|
if (!ex3) {
|
|
|
|
ret = ext4_ext_try_to_merge(inode, path, ex2);
|
|
|
|
if (ret) {
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* Mark modified extent as dirty */
|
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
goto out;
|
|
|
|
insert:
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_insert_extent(handle, inode, path, &newex, 0);
|
2010-05-16 10:00:00 +00:00
|
|
|
if (err == -ENOSPC && may_zeroout) {
|
2008-04-29 12:11:12 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
2008-04-17 14:38:59 +00:00
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-17 14:38:59 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
2008-04-30 02:03:59 +00:00
|
|
|
/* zero out the first half */
|
|
|
|
return allocated;
|
2008-04-29 12:11:12 +00:00
|
|
|
} else if (err)
|
|
|
|
goto fix_extent_len;
|
2007-07-18 01:42:38 +00:00
|
|
|
out:
|
2009-09-18 17:34:55 +00:00
|
|
|
ext4_ext_show_leaf(inode, path);
|
2007-07-18 01:42:38 +00:00
|
|
|
return err ? err : allocated;
|
2008-04-29 12:11:12 +00:00
|
|
|
|
|
|
|
fix_extent_len:
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2008-04-29 12:11:12 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex);
|
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
return err;
|
2007-07-18 01:42:38 +00:00
|
|
|
}
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
/*
|
2010-05-16 23:00:00 +00:00
|
|
|
* This function is called by ext4_ext_map_blocks() from
|
2009-09-28 19:49:08 +00:00
|
|
|
* ext4_get_blocks_dio_write() when DIO to write
|
|
|
|
* to an uninitialized extent.
|
|
|
|
*
|
|
|
|
* Writing to an uninitized extent may result in splitting the uninitialized
|
|
|
|
* extent into multiple /intialized unintialized extents (up to three)
|
|
|
|
* There are three possibilities:
|
|
|
|
* a> There is no split required: Entire extent should be uninitialized
|
|
|
|
* b> Splits in two extents: Write is happening at either end of the extent
|
|
|
|
* c> Splits in three extents: Somone is writing in middle of the extent
|
|
|
|
*
|
|
|
|
* One of more index blocks maybe needed if the extent tree grow after
|
|
|
|
* the unintialized extent split. To prevent ENOSPC occur at the IO
|
|
|
|
* complete, we need to split the uninitialized extent before DIO submit
|
2010-06-11 10:17:00 +00:00
|
|
|
* the IO. The uninitialized extent called at this time will be split
|
2009-09-28 19:49:08 +00:00
|
|
|
* into three uninitialized extent(at most). After IO complete, the part
|
|
|
|
* being filled will be convert to initialized by the end_io callback function
|
|
|
|
* via ext4_convert_unwritten_extents().
|
2009-11-06 09:01:23 +00:00
|
|
|
*
|
|
|
|
* Returns the size of uninitialized extent to be written on success.
|
2009-09-28 19:49:08 +00:00
|
|
|
*/
|
|
|
|
static int ext4_split_unwritten_extents(handle_t *handle,
|
|
|
|
struct inode *inode,
|
2010-05-16 23:00:00 +00:00
|
|
|
struct ext4_map_blocks *map,
|
2009-09-28 19:49:08 +00:00
|
|
|
struct ext4_ext_path *path,
|
|
|
|
int flags)
|
|
|
|
{
|
|
|
|
struct ext4_extent *ex, newex, orig_ex;
|
|
|
|
struct ext4_extent *ex1 = NULL;
|
|
|
|
struct ext4_extent *ex2 = NULL;
|
|
|
|
struct ext4_extent *ex3 = NULL;
|
2010-05-16 10:00:00 +00:00
|
|
|
ext4_lblk_t ee_block, eof_block;
|
2009-09-28 19:49:08 +00:00
|
|
|
unsigned int allocated, ee_len, depth;
|
|
|
|
ext4_fsblk_t newblock;
|
|
|
|
int err = 0;
|
2010-05-16 10:00:00 +00:00
|
|
|
int may_zeroout;
|
|
|
|
|
|
|
|
ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
|
|
|
|
"block %llu, max_blocks %u\n", inode->i_ino,
|
2010-05-16 23:00:00 +00:00
|
|
|
(unsigned long long)map->m_lblk, map->m_len);
|
2010-05-16 10:00:00 +00:00
|
|
|
|
|
|
|
eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
|
|
|
|
inode->i_sb->s_blocksize_bits;
|
2010-05-16 23:00:00 +00:00
|
|
|
if (eof_block < map->m_lblk + map->m_len)
|
|
|
|
eof_block = map->m_lblk + map->m_len;
|
2009-09-28 19:49:08 +00:00
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
ee_block = le32_to_cpu(ex->ee_block);
|
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = ee_len - (map->m_lblk - ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
newblock = map->m_lblk - ee_block + ext4_ext_pblock(ex);
|
2010-05-16 10:00:00 +00:00
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
ex2 = ex;
|
|
|
|
orig_ex.ee_block = ex->ee_block;
|
|
|
|
orig_ex.ee_len = cpu_to_le16(ee_len);
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(&orig_ex, ext4_ext_pblock(ex));
|
2009-09-28 19:49:08 +00:00
|
|
|
|
2010-05-16 10:00:00 +00:00
|
|
|
/*
|
|
|
|
* It is safe to convert extent to initialized via explicit
|
|
|
|
* zeroout only if extent is fully insde i_size or new_size.
|
|
|
|
*/
|
|
|
|
may_zeroout = ee_block + ee_len <= eof_block;
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
/*
|
2009-11-06 09:01:23 +00:00
|
|
|
* If the uninitialized extent begins at the same logical
|
|
|
|
* block where the write begins, and the write completely
|
|
|
|
* covers the extent, then we don't need to split it.
|
2009-09-28 19:49:08 +00:00
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
if ((map->m_lblk == ee_block) && (allocated <= map->m_len))
|
2009-11-06 09:01:23 +00:00
|
|
|
return allocated;
|
2009-09-28 19:49:08 +00:00
|
|
|
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
2010-05-16 23:00:00 +00:00
|
|
|
/* ex1: ee_block to map->m_lblk - 1 : uninitialized */
|
|
|
|
if (map->m_lblk > ee_block) {
|
2009-09-28 19:49:08 +00:00
|
|
|
ex1 = ex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block);
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex1);
|
|
|
|
ex2 = &newex;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* for sanity, update the length of the ex2 extent before
|
|
|
|
* we insert ex3, if ex1 is NULL. This is to avoid temporary
|
|
|
|
* overlap of blocks.
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
if (!ex1 && allocated > map->m_len)
|
|
|
|
ex2->ee_len = cpu_to_le16(map->m_len);
|
2009-09-28 19:49:08 +00:00
|
|
|
/* ex3: to ee_block + ee_len : uninitialised */
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len) {
|
2009-09-28 19:49:08 +00:00
|
|
|
unsigned int newdepth;
|
|
|
|
ex3 = &newex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex3->ee_block = cpu_to_le32(map->m_lblk + map->m_len);
|
|
|
|
ext4_ext_store_pblock(ex3, newblock + map->m_len);
|
|
|
|
ex3->ee_len = cpu_to_le16(allocated - map->m_len);
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex3);
|
|
|
|
err = ext4_ext_insert_extent(handle, inode, path, ex3, flags);
|
2010-05-16 10:00:00 +00:00
|
|
|
if (err == -ENOSPC && may_zeroout) {
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
/* zeroed the full extent */
|
2010-05-16 23:00:00 +00:00
|
|
|
/* blocks available from map->m_lblk */
|
2009-09-28 19:49:08 +00:00
|
|
|
return allocated;
|
|
|
|
|
|
|
|
} else if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/*
|
|
|
|
* The depth, and hence eh & ex might change
|
|
|
|
* as part of the insert above.
|
|
|
|
*/
|
|
|
|
newdepth = ext_depth(inode);
|
|
|
|
/*
|
|
|
|
* update the extent length after successful insert of the
|
|
|
|
* split extent
|
|
|
|
*/
|
2010-05-16 10:00:00 +00:00
|
|
|
ee_len -= ext4_ext_get_actual_len(ex3);
|
|
|
|
orig_ex.ee_len = cpu_to_le16(ee_len);
|
|
|
|
may_zeroout = ee_block + ee_len <= eof_block;
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
depth = newdepth;
|
|
|
|
ext4_ext_drop_refs(path);
|
2010-05-16 23:00:00 +00:00
|
|
|
path = ext4_ext_find_extent(inode, map->m_lblk, path);
|
2009-09-28 19:49:08 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
if (ex2 != &newex)
|
|
|
|
ex2 = ex;
|
|
|
|
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = map->m_len;
|
2009-09-28 19:49:08 +00:00
|
|
|
}
|
|
|
|
/*
|
|
|
|
* If there was a change of depth as part of the
|
|
|
|
* insertion of ex3 above, we need to update the length
|
|
|
|
* of the ex1 extent again here
|
|
|
|
*/
|
|
|
|
if (ex1 && ex1 != ex) {
|
|
|
|
ex1 = ex;
|
2010-05-16 23:00:00 +00:00
|
|
|
ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block);
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex1);
|
|
|
|
ex2 = &newex;
|
|
|
|
}
|
|
|
|
/*
|
2010-05-16 23:00:00 +00:00
|
|
|
* ex2: map->m_lblk to map->m_lblk + map->m_len-1 : to be written
|
|
|
|
* using direct I/O, uninitialised still.
|
2009-09-28 19:49:08 +00:00
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
ex2->ee_block = cpu_to_le32(map->m_lblk);
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_store_pblock(ex2, newblock);
|
|
|
|
ex2->ee_len = cpu_to_le16(allocated);
|
|
|
|
ext4_ext_mark_uninitialized(ex2);
|
|
|
|
if (ex2 != ex)
|
|
|
|
goto insert;
|
|
|
|
/* Mark modified extent as dirty */
|
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
ext_debug("out here\n");
|
|
|
|
goto out;
|
|
|
|
insert:
|
|
|
|
err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
|
2010-05-16 10:00:00 +00:00
|
|
|
if (err == -ENOSPC && may_zeroout) {
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_zeroout(inode, &orig_ex);
|
|
|
|
if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
/* update the extent length and mark as initialized */
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
/* zero out the first half */
|
|
|
|
return allocated;
|
|
|
|
} else if (err)
|
|
|
|
goto fix_extent_len;
|
|
|
|
out:
|
|
|
|
ext4_ext_show_leaf(inode, path);
|
|
|
|
return err ? err : allocated;
|
|
|
|
|
|
|
|
fix_extent_len:
|
|
|
|
ex->ee_block = orig_ex.ee_block;
|
|
|
|
ex->ee_len = orig_ex.ee_len;
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex));
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_mark_uninitialized(ex);
|
|
|
|
ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
return err;
|
|
|
|
}
|
2010-03-02 18:28:44 +00:00
|
|
|
static int ext4_convert_unwritten_extents_endio(handle_t *handle,
|
2009-09-28 19:49:08 +00:00
|
|
|
struct inode *inode,
|
|
|
|
struct ext4_ext_path *path)
|
|
|
|
{
|
|
|
|
struct ext4_extent *ex;
|
|
|
|
struct ext4_extent_header *eh;
|
|
|
|
int depth;
|
|
|
|
int err = 0;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
|
|
|
|
err = ext4_ext_get_access(handle, inode, path + depth);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
/* first mark the extent as initialized */
|
|
|
|
ext4_ext_mark_initialized(ex);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We have to see if it can be merged with the extent
|
|
|
|
* on the left.
|
|
|
|
*/
|
|
|
|
if (ex > EXT_FIRST_EXTENT(eh)) {
|
|
|
|
/*
|
|
|
|
* To merge left, pass "ex - 1" to try_to_merge(),
|
|
|
|
* since it merges towards right _only_.
|
|
|
|
*/
|
|
|
|
ret = ext4_ext_try_to_merge(inode, path, ex - 1);
|
|
|
|
if (ret) {
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
ex--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Try to Merge towards right.
|
|
|
|
*/
|
|
|
|
ret = ext4_ext_try_to_merge(inode, path, ex);
|
|
|
|
if (ret) {
|
|
|
|
err = ext4_ext_correct_indexes(handle, inode, path);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
}
|
|
|
|
/* Mark modified extent as dirty */
|
|
|
|
err = ext4_ext_dirty(handle, inode, path + depth);
|
|
|
|
out:
|
|
|
|
ext4_ext_show_leaf(inode, path);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2009-12-30 04:39:06 +00:00
|
|
|
static void unmap_underlying_metadata_blocks(struct block_device *bdev,
|
|
|
|
sector_t block, int count)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
for (i = 0; i < count; i++)
|
|
|
|
unmap_underlying_metadata(bdev, block + i);
|
|
|
|
}
|
|
|
|
|
2010-10-28 01:23:12 +00:00
|
|
|
/*
|
|
|
|
* Handle EOFBLOCKS_FL flag, clearing it if necessary
|
|
|
|
*/
|
|
|
|
static int check_eofblocks_fl(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_map_blocks *map,
|
|
|
|
struct ext4_ext_path *path,
|
|
|
|
unsigned int len)
|
|
|
|
{
|
|
|
|
int i, depth;
|
|
|
|
struct ext4_extent_header *eh;
|
|
|
|
struct ext4_extent *ex, *last_ex;
|
|
|
|
|
|
|
|
if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
eh = path[depth].p_hdr;
|
|
|
|
ex = path[depth].p_ext;
|
|
|
|
|
|
|
|
if (unlikely(!eh->eh_entries)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "eh->eh_entries == 0 and "
|
|
|
|
"EOFBLOCKS_FL set");
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
last_ex = EXT_LAST_EXTENT(eh);
|
|
|
|
/*
|
|
|
|
* We should clear the EOFBLOCKS_FL flag if we are writing the
|
|
|
|
* last block in the last extent in the file. We test this by
|
|
|
|
* first checking to see if the caller to
|
|
|
|
* ext4_ext_get_blocks() was interested in the last block (or
|
|
|
|
* a block beyond the last block) in the current extent. If
|
|
|
|
* this turns out to be false, we can bail out from this
|
|
|
|
* function immediately.
|
|
|
|
*/
|
|
|
|
if (map->m_lblk + len < le32_to_cpu(last_ex->ee_block) +
|
|
|
|
ext4_ext_get_actual_len(last_ex))
|
|
|
|
return 0;
|
|
|
|
/*
|
|
|
|
* If the caller does appear to be planning to write at or
|
|
|
|
* beyond the end of the current extent, we then test to see
|
|
|
|
* if the current extent is the last extent in the file, by
|
|
|
|
* checking to make sure it was reached via the rightmost node
|
|
|
|
* at each level of the tree.
|
|
|
|
*/
|
|
|
|
for (i = depth-1; i >= 0; i--)
|
|
|
|
if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
|
|
|
|
return 0;
|
|
|
|
ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
|
|
|
|
return ext4_mark_inode_dirty(handle, inode);
|
|
|
|
}
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
static int
|
|
|
|
ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
|
2010-05-16 23:00:00 +00:00
|
|
|
struct ext4_map_blocks *map,
|
2009-09-28 19:49:08 +00:00
|
|
|
struct ext4_ext_path *path, int flags,
|
2010-05-16 23:00:00 +00:00
|
|
|
unsigned int allocated, ext4_fsblk_t newblock)
|
2009-09-28 19:49:08 +00:00
|
|
|
{
|
|
|
|
int ret = 0;
|
|
|
|
int err = 0;
|
2009-09-28 19:48:29 +00:00
|
|
|
ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
|
2009-09-28 19:49:08 +00:00
|
|
|
|
|
|
|
ext_debug("ext4_ext_handle_uninitialized_extents: inode %lu, logical"
|
|
|
|
"block %llu, max_blocks %u, flags %d, allocated %u",
|
2010-05-16 23:00:00 +00:00
|
|
|
inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
|
2009-09-28 19:49:08 +00:00
|
|
|
flags, allocated);
|
|
|
|
ext4_ext_show_leaf(inode, path);
|
|
|
|
|
2010-03-02 18:28:44 +00:00
|
|
|
/* get_block() before submit the IO, split the extent */
|
2010-03-04 21:14:02 +00:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
|
2010-05-16 23:00:00 +00:00
|
|
|
ret = ext4_split_unwritten_extents(handle, inode, map,
|
|
|
|
path, flags);
|
2009-11-10 15:48:04 +00:00
|
|
|
/*
|
|
|
|
* Flag the inode(non aio case) or end_io struct (aio case)
|
|
|
|
* that this IO needs to convertion to written when IO is
|
|
|
|
* completed
|
|
|
|
*/
|
2009-09-28 19:48:29 +00:00
|
|
|
if (io)
|
2010-10-28 01:30:10 +00:00
|
|
|
io->flag = EXT4_IO_END_UNWRITTEN;
|
2009-11-10 15:48:04 +00:00
|
|
|
else
|
2010-01-24 19:34:07 +00:00
|
|
|
ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
|
2010-03-04 21:14:02 +00:00
|
|
|
if (ext4_should_dioread_nolock(inode))
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_UNINIT;
|
2009-09-28 19:49:08 +00:00
|
|
|
goto out;
|
|
|
|
}
|
2010-03-02 18:28:44 +00:00
|
|
|
/* IO end_io complete, convert the filled extent to written */
|
2010-03-04 21:14:02 +00:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_CONVERT)) {
|
2010-03-02 18:28:44 +00:00
|
|
|
ret = ext4_convert_unwritten_extents_endio(handle, inode,
|
2009-09-28 19:49:08 +00:00
|
|
|
path);
|
2010-10-28 01:23:12 +00:00
|
|
|
if (ret >= 0) {
|
2009-12-09 04:51:10 +00:00
|
|
|
ext4_update_inode_fsync_trans(handle, inode, 1);
|
2010-10-28 01:23:12 +00:00
|
|
|
err = check_eofblocks_fl(handle, inode, map, path,
|
|
|
|
map->m_len);
|
|
|
|
} else
|
|
|
|
err = ret;
|
2009-09-28 19:49:08 +00:00
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
/* buffered IO case */
|
|
|
|
/*
|
|
|
|
* repeat fallocate creation request
|
|
|
|
* we already have an unwritten extent
|
|
|
|
*/
|
|
|
|
if (flags & EXT4_GET_BLOCKS_UNINIT_EXT)
|
|
|
|
goto map_out;
|
|
|
|
|
|
|
|
/* buffered READ or buffered write_begin() lookup */
|
|
|
|
if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
|
|
|
|
/*
|
|
|
|
* We have blocks reserved already. We
|
|
|
|
* return allocated blocks so that delalloc
|
|
|
|
* won't do block reservation for us. But
|
|
|
|
* the buffer head will be unmapped so that
|
|
|
|
* a read from the block returns 0s.
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_UNWRITTEN;
|
2009-09-28 19:49:08 +00:00
|
|
|
goto out1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* buffered write, writepage time, convert*/
|
2010-05-16 23:00:00 +00:00
|
|
|
ret = ext4_ext_convert_to_initialized(handle, inode, map, path);
|
2010-10-28 01:23:12 +00:00
|
|
|
if (ret >= 0) {
|
2009-12-09 04:51:10 +00:00
|
|
|
ext4_update_inode_fsync_trans(handle, inode, 1);
|
2010-10-28 01:23:12 +00:00
|
|
|
err = check_eofblocks_fl(handle, inode, map, path, map->m_len);
|
|
|
|
if (err < 0)
|
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
out:
|
|
|
|
if (ret <= 0) {
|
|
|
|
err = ret;
|
|
|
|
goto out2;
|
|
|
|
} else
|
|
|
|
allocated = ret;
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_NEW;
|
2009-12-30 04:39:06 +00:00
|
|
|
/*
|
|
|
|
* if we allocated more blocks than requested
|
|
|
|
* we need to make sure we unmap the extra block
|
|
|
|
* allocated. The actual needed block will get
|
|
|
|
* unmapped later when we find the buffer_head marked
|
|
|
|
* new.
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len) {
|
2009-12-30 04:39:06 +00:00
|
|
|
unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
|
2010-05-16 23:00:00 +00:00
|
|
|
newblock + map->m_len,
|
|
|
|
allocated - map->m_len);
|
|
|
|
allocated = map->m_len;
|
2009-12-30 04:39:06 +00:00
|
|
|
}
|
2010-01-25 09:00:31 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If we have done fallocate with the offset that is already
|
|
|
|
* delayed allocated, we would have block reservation
|
|
|
|
* and quota reservation done in the delayed write path.
|
|
|
|
* But fallocate would have already updated quota and block
|
|
|
|
* count for this offset. So cancel these reservation
|
|
|
|
*/
|
2010-01-15 06:27:59 +00:00
|
|
|
if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
|
2010-01-25 09:00:31 +00:00
|
|
|
ext4_da_update_reserve_space(inode, allocated, 0);
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
map_out:
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_MAPPED;
|
2009-09-28 19:49:08 +00:00
|
|
|
out1:
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len)
|
|
|
|
allocated = map->m_len;
|
2009-09-28 19:49:08 +00:00
|
|
|
ext4_ext_show_leaf(inode, path);
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_pblk = newblock;
|
|
|
|
map->m_len = allocated;
|
2009-09-28 19:49:08 +00:00
|
|
|
out2:
|
|
|
|
if (path) {
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
kfree(path);
|
|
|
|
}
|
|
|
|
return err ? err : allocated;
|
|
|
|
}
|
2010-10-28 01:23:12 +00:00
|
|
|
|
2008-01-29 04:58:27 +00:00
|
|
|
/*
|
2008-02-25 20:29:55 +00:00
|
|
|
* Block allocation/map/preallocation routine for extents based files
|
|
|
|
*
|
|
|
|
*
|
2008-01-29 04:58:27 +00:00
|
|
|
* Need to be called with
|
2008-01-29 04:58:26 +00:00
|
|
|
* down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
|
|
|
|
* (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
|
2008-02-25 20:29:55 +00:00
|
|
|
*
|
|
|
|
* return > 0, number of of blocks already mapped/allocated
|
|
|
|
* if create == 0 and these are pre-allocated blocks
|
|
|
|
* buffer head is unmapped
|
|
|
|
* otherwise blocks are mapped
|
|
|
|
*
|
|
|
|
* return = 0, if plain look up failed (blocks have not been allocated)
|
|
|
|
* buffer head is unmapped
|
|
|
|
*
|
|
|
|
* return < 0, error case.
|
2008-01-29 04:58:27 +00:00
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
|
|
|
|
struct ext4_map_blocks *map, int flags)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct ext4_ext_path *path = NULL;
|
2007-07-18 01:42:38 +00:00
|
|
|
struct ext4_extent_header *eh;
|
2010-10-28 01:23:12 +00:00
|
|
|
struct ext4_extent newex, *ex;
|
2008-11-05 05:14:04 +00:00
|
|
|
ext4_fsblk_t newblock;
|
2010-10-28 01:23:12 +00:00
|
|
|
int err = 0, depth, ret, cache_type;
|
2008-11-05 05:14:04 +00:00
|
|
|
unsigned int allocated = 0;
|
2008-01-29 05:19:52 +00:00
|
|
|
struct ext4_allocation_request ar;
|
2009-09-28 19:48:29 +00:00
|
|
|
ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2009-09-01 12:44:37 +00:00
|
|
|
ext_debug("blocks %u/%u requested for inode %lu\n",
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_lblk, map->m_len, inode->i_ino);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* check in cache */
|
2010-05-16 23:00:00 +00:00
|
|
|
cache_type = ext4_ext_in_cache(inode, map->m_lblk, &newex);
|
2008-11-05 05:14:04 +00:00
|
|
|
if (cache_type) {
|
|
|
|
if (cache_type == EXT4_EXT_CACHE_GAP) {
|
2009-05-14 04:58:52 +00:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
|
2007-07-18 01:42:38 +00:00
|
|
|
/*
|
|
|
|
* block isn't allocated yet and
|
|
|
|
* user doesn't want to allocate it
|
|
|
|
*/
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
/* we should allocate requested block */
|
2008-11-05 05:14:04 +00:00
|
|
|
} else if (cache_type == EXT4_EXT_CACHE_EXTENT) {
|
2006-10-11 08:21:03 +00:00
|
|
|
/* block is already allocated */
|
2010-05-16 23:00:00 +00:00
|
|
|
newblock = map->m_lblk
|
2007-05-24 17:04:54 +00:00
|
|
|
- le32_to_cpu(newex.ee_block)
|
2010-10-28 01:30:14 +00:00
|
|
|
+ ext4_ext_pblock(&newex);
|
2006-10-11 08:21:07 +00:00
|
|
|
/* number of remaining blocks in the extent */
|
2008-01-29 04:58:27 +00:00
|
|
|
allocated = ext4_ext_get_actual_len(&newex) -
|
2010-05-16 23:00:00 +00:00
|
|
|
(map->m_lblk - le32_to_cpu(newex.ee_block));
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out;
|
|
|
|
} else {
|
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* find extent for this block */
|
2010-05-16 23:00:00 +00:00
|
|
|
path = ext4_ext_find_extent(inode, map->m_lblk, NULL);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (IS_ERR(path)) {
|
|
|
|
err = PTR_ERR(path);
|
|
|
|
path = NULL;
|
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
|
|
|
|
depth = ext_depth(inode);
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* consistent leaf must not be empty;
|
|
|
|
* this situation is possible, though, _during_ tree modification;
|
2006-10-11 08:21:03 +00:00
|
|
|
* this is why assert can't be put in ext4_ext_find_extent()
|
|
|
|
*/
|
2010-03-02 16:46:09 +00:00
|
|
|
if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
|
|
|
|
EXT4_ERROR_INODE(inode, "bad extent address "
|
2010-05-17 03:00:00 +00:00
|
|
|
"lblock: %lu, depth: %d pblock %lld",
|
|
|
|
(unsigned long) map->m_lblk, depth,
|
|
|
|
path[depth].p_block);
|
2009-12-14 14:53:52 +00:00
|
|
|
err = -EIO;
|
|
|
|
goto out2;
|
|
|
|
}
|
2007-07-18 01:42:38 +00:00
|
|
|
eh = path[depth].p_hdr;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2006-12-07 04:41:33 +00:00
|
|
|
ex = path[depth].p_ext;
|
|
|
|
if (ex) {
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_fsblk_t ee_start = ext4_ext_pblock(ex);
|
2007-07-18 01:42:41 +00:00
|
|
|
unsigned short ee_len;
|
2006-10-11 08:21:06 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Uninitialized extents are treated as holes, except that
|
2007-07-18 01:42:38 +00:00
|
|
|
* we split out initialized portions during a write.
|
2006-10-11 08:21:06 +00:00
|
|
|
*/
|
2007-07-18 01:42:41 +00:00
|
|
|
ee_len = ext4_ext_get_actual_len(ex);
|
2006-10-11 08:21:07 +00:00
|
|
|
/* if found extent covers block, simply return it */
|
2010-05-16 23:00:00 +00:00
|
|
|
if (in_range(map->m_lblk, ee_block, ee_len)) {
|
|
|
|
newblock = map->m_lblk - ee_block + ee_start;
|
2006-10-11 08:21:07 +00:00
|
|
|
/* number of remaining blocks in the extent */
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = ee_len - (map->m_lblk - ee_block);
|
|
|
|
ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
|
|
|
|
ee_block, ee_len, newblock);
|
2007-07-18 01:42:38 +00:00
|
|
|
|
2007-07-18 01:42:41 +00:00
|
|
|
/* Do not put uninitialized extent in the cache */
|
2007-07-18 01:42:38 +00:00
|
|
|
if (!ext4_ext_is_uninitialized(ex)) {
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_put_in_cache(inode, ee_block,
|
|
|
|
ee_len, ee_start,
|
|
|
|
EXT4_EXT_CACHE_EXTENT);
|
2007-07-18 01:42:38 +00:00
|
|
|
goto out;
|
|
|
|
}
|
2009-09-28 19:49:08 +00:00
|
|
|
ret = ext4_ext_handle_uninitialized_extents(handle,
|
2010-05-16 23:00:00 +00:00
|
|
|
inode, map, path, flags, allocated,
|
|
|
|
newblock);
|
2009-09-28 19:49:08 +00:00
|
|
|
return ret;
|
2006-10-11 08:21:03 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* requested block isn't allocated yet;
|
2006-10-11 08:21:03 +00:00
|
|
|
* we couldn't try to create block if create flag is zero
|
|
|
|
*/
|
2009-05-14 04:58:52 +00:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) {
|
2007-07-18 01:42:38 +00:00
|
|
|
/*
|
|
|
|
* put just found gap into cache to speed up
|
|
|
|
* subsequent requests
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
ext4_ext_put_gap_in_cache(inode, path, map->m_lblk);
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
/*
|
2008-10-10 13:40:52 +00:00
|
|
|
* Okay, we need to do block allocation.
|
2006-10-11 08:21:24 +00:00
|
|
|
*/
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-01-29 05:19:52 +00:00
|
|
|
/* find neighbour allocated blocks */
|
2010-05-16 23:00:00 +00:00
|
|
|
ar.lleft = map->m_lblk;
|
2008-01-29 05:19:52 +00:00
|
|
|
err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
|
|
|
|
if (err)
|
|
|
|
goto out2;
|
2010-05-16 23:00:00 +00:00
|
|
|
ar.lright = map->m_lblk;
|
2008-01-29 05:19:52 +00:00
|
|
|
err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright);
|
|
|
|
if (err)
|
|
|
|
goto out2;
|
2007-05-24 17:04:13 +00:00
|
|
|
|
2007-07-18 13:02:56 +00:00
|
|
|
/*
|
|
|
|
* See if request is beyond maximum number of blocks we can have in
|
|
|
|
* a single extent. For an initialized extent this limit is
|
|
|
|
* EXT_INIT_MAX_LEN and for an uninitialized extent this limit is
|
|
|
|
* EXT_UNINIT_MAX_LEN.
|
|
|
|
*/
|
2010-05-16 23:00:00 +00:00
|
|
|
if (map->m_len > EXT_INIT_MAX_LEN &&
|
2009-05-14 04:58:52 +00:00
|
|
|
!(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_len = EXT_INIT_MAX_LEN;
|
|
|
|
else if (map->m_len > EXT_UNINIT_MAX_LEN &&
|
2009-05-14 04:58:52 +00:00
|
|
|
(flags & EXT4_GET_BLOCKS_UNINIT_EXT))
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_len = EXT_UNINIT_MAX_LEN;
|
2007-07-18 13:02:56 +00:00
|
|
|
|
2010-05-16 23:00:00 +00:00
|
|
|
/* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */
|
|
|
|
newex.ee_block = cpu_to_le32(map->m_lblk);
|
|
|
|
newex.ee_len = cpu_to_le16(map->m_len);
|
2007-05-24 17:04:13 +00:00
|
|
|
err = ext4_ext_check_overlap(inode, &newex, path);
|
|
|
|
if (err)
|
2008-01-29 04:58:27 +00:00
|
|
|
allocated = ext4_ext_get_actual_len(&newex);
|
2007-05-24 17:04:13 +00:00
|
|
|
else
|
2010-05-16 23:00:00 +00:00
|
|
|
allocated = map->m_len;
|
2008-01-29 05:19:52 +00:00
|
|
|
|
|
|
|
/* allocate new block */
|
|
|
|
ar.inode = inode;
|
2010-05-16 23:00:00 +00:00
|
|
|
ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk);
|
|
|
|
ar.logical = map->m_lblk;
|
2008-01-29 05:19:52 +00:00
|
|
|
ar.len = allocated;
|
|
|
|
if (S_ISREG(inode->i_mode))
|
|
|
|
ar.flags = EXT4_MB_HINT_DATA;
|
|
|
|
else
|
|
|
|
/* disable in-core preallocation for non-regular files */
|
|
|
|
ar.flags = 0;
|
|
|
|
newblock = ext4_mb_new_blocks(handle, &ar, &err);
|
2006-10-11 08:21:03 +00:00
|
|
|
if (!newblock)
|
|
|
|
goto out2;
|
2009-09-01 12:44:37 +00:00
|
|
|
ext_debug("allocate new block: goal %llu, found %llu/%u\n",
|
2008-11-05 05:14:04 +00:00
|
|
|
ar.goal, newblock, allocated);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* try to insert new extent into found leaf and return */
|
2006-10-11 08:21:05 +00:00
|
|
|
ext4_ext_store_pblock(&newex, newblock);
|
2008-01-29 05:19:52 +00:00
|
|
|
newex.ee_len = cpu_to_le16(ar.len);
|
2009-09-28 19:48:29 +00:00
|
|
|
/* Mark uninitialized */
|
|
|
|
if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_ext_mark_uninitialized(&newex);
|
2009-09-28 19:48:29 +00:00
|
|
|
/*
|
2010-03-04 21:14:02 +00:00
|
|
|
* io_end structure was created for every IO write to an
|
|
|
|
* uninitialized extent. To avoid unecessary conversion,
|
|
|
|
* here we flag the IO that really needs the conversion.
|
2009-11-10 15:48:04 +00:00
|
|
|
* For non asycn direct IO case, flag the inode state
|
|
|
|
* that we need to perform convertion when IO is done.
|
2009-09-28 19:48:29 +00:00
|
|
|
*/
|
2010-03-04 21:14:02 +00:00
|
|
|
if ((flags & EXT4_GET_BLOCKS_PRE_IO)) {
|
2009-11-10 15:48:04 +00:00
|
|
|
if (io)
|
2010-10-28 01:30:10 +00:00
|
|
|
io->flag = EXT4_IO_END_UNWRITTEN;
|
2009-11-10 15:48:04 +00:00
|
|
|
else
|
2010-01-24 19:34:07 +00:00
|
|
|
ext4_set_inode_state(inode,
|
|
|
|
EXT4_STATE_DIO_UNWRITTEN);
|
2009-11-10 15:48:04 +00:00
|
|
|
}
|
2010-03-04 21:14:02 +00:00
|
|
|
if (ext4_should_dioread_nolock(inode))
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_UNINIT;
|
2009-09-28 19:48:29 +00:00
|
|
|
}
|
2010-02-24 14:52:53 +00:00
|
|
|
|
2010-10-28 01:23:12 +00:00
|
|
|
err = check_eofblocks_fl(handle, inode, map, path, ar.len);
|
|
|
|
if (err)
|
|
|
|
goto out2;
|
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
|
2007-05-24 17:04:25 +00:00
|
|
|
if (err) {
|
|
|
|
/* free data blocks we just allocated */
|
2008-01-29 05:19:52 +00:00
|
|
|
/* not a good idea to call discard here directly,
|
|
|
|
* but otherwise we'd need to call it every free() */
|
2008-10-10 13:40:52 +00:00
|
|
|
ext4_discard_preallocations(inode);
|
2010-10-28 01:30:14 +00:00
|
|
|
ext4_free_blocks(handle, inode, 0, ext4_ext_pblock(&newex),
|
2009-11-23 12:17:05 +00:00
|
|
|
ext4_ext_get_actual_len(&newex), 0);
|
2006-10-11 08:21:03 +00:00
|
|
|
goto out2;
|
2007-05-24 17:04:25 +00:00
|
|
|
}
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
/* previous routine could use block we allocated */
|
2010-10-28 01:30:14 +00:00
|
|
|
newblock = ext4_ext_pblock(&newex);
|
2008-01-29 04:58:27 +00:00
|
|
|
allocated = ext4_ext_get_actual_len(&newex);
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len)
|
|
|
|
allocated = map->m_len;
|
|
|
|
map->m_flags |= EXT4_MAP_NEW;
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2010-01-25 09:00:31 +00:00
|
|
|
/*
|
|
|
|
* Update reserved blocks/metadata blocks after successful
|
|
|
|
* block allocation which had been deferred till now.
|
|
|
|
*/
|
2010-01-15 06:27:59 +00:00
|
|
|
if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
|
2010-01-25 09:00:31 +00:00
|
|
|
ext4_da_update_reserve_space(inode, allocated, 1);
|
|
|
|
|
2009-12-09 04:51:10 +00:00
|
|
|
/*
|
|
|
|
* Cache the extent and update transaction to commit on fdatasync only
|
|
|
|
* when it is _not_ an uninitialized extent.
|
|
|
|
*/
|
|
|
|
if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) {
|
2010-05-16 23:00:00 +00:00
|
|
|
ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock,
|
2007-07-18 01:42:41 +00:00
|
|
|
EXT4_EXT_CACHE_EXTENT);
|
2009-12-09 04:51:10 +00:00
|
|
|
ext4_update_inode_fsync_trans(handle, inode, 1);
|
|
|
|
} else
|
|
|
|
ext4_update_inode_fsync_trans(handle, inode, 0);
|
2006-10-11 08:21:03 +00:00
|
|
|
out:
|
2010-05-16 23:00:00 +00:00
|
|
|
if (allocated > map->m_len)
|
|
|
|
allocated = map->m_len;
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_show_leaf(inode, path);
|
2010-05-16 23:00:00 +00:00
|
|
|
map->m_flags |= EXT4_MAP_MAPPED;
|
|
|
|
map->m_pblk = newblock;
|
|
|
|
map->m_len = allocated;
|
2006-10-11 08:21:03 +00:00
|
|
|
out2:
|
|
|
|
if (path) {
|
|
|
|
ext4_ext_drop_refs(path);
|
|
|
|
kfree(path);
|
|
|
|
}
|
|
|
|
return err ? err : allocated;
|
|
|
|
}
|
|
|
|
|
2008-07-11 23:27:31 +00:00
|
|
|
void ext4_ext_truncate(struct inode *inode)
|
2006-10-11 08:21:03 +00:00
|
|
|
{
|
|
|
|
struct address_space *mapping = inode->i_mapping;
|
|
|
|
struct super_block *sb = inode->i_sb;
|
2008-01-29 04:58:27 +00:00
|
|
|
ext4_lblk_t last_block;
|
2006-10-11 08:21:03 +00:00
|
|
|
handle_t *handle;
|
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* probably first extent we're gonna free will be last in block
|
|
|
|
*/
|
2008-08-20 02:16:03 +00:00
|
|
|
err = ext4_writepage_trans_blocks(inode);
|
2006-10-11 08:21:03 +00:00
|
|
|
handle = ext4_journal_start(inode, err);
|
2008-07-11 23:27:31 +00:00
|
|
|
if (IS_ERR(handle))
|
2006-10-11 08:21:03 +00:00
|
|
|
return;
|
|
|
|
|
2008-07-11 23:27:31 +00:00
|
|
|
if (inode->i_size & (sb->s_blocksize - 1))
|
|
|
|
ext4_block_truncate_page(handle, mapping, inode->i_size);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
2008-07-11 23:27:31 +00:00
|
|
|
if (ext4_orphan_add(handle, inode))
|
|
|
|
goto out_stop;
|
|
|
|
|
2008-01-29 04:58:26 +00:00
|
|
|
down_write(&EXT4_I(inode)->i_data_sem);
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_ext_invalidate_cache(inode);
|
|
|
|
|
2008-10-10 13:40:52 +00:00
|
|
|
ext4_discard_preallocations(inode);
|
2008-01-29 05:19:52 +00:00
|
|
|
|
2006-10-11 08:21:03 +00:00
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* TODO: optimization is possible here.
|
|
|
|
* Probably we need not scan at all,
|
|
|
|
* because page truncation is enough.
|
2006-10-11 08:21:03 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
/* we have to know where to truncate from in crash case */
|
|
|
|
EXT4_I(inode)->i_disksize = inode->i_size;
|
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
|
|
|
|
|
|
|
last_block = (inode->i_size + sb->s_blocksize - 1)
|
|
|
|
>> EXT4_BLOCK_SIZE_BITS(sb);
|
|
|
|
err = ext4_ext_remove_space(inode, last_block);
|
|
|
|
|
|
|
|
/* In a multi-transaction truncate, we only make the final
|
2007-07-18 01:42:38 +00:00
|
|
|
* transaction synchronous.
|
|
|
|
*/
|
2006-10-11 08:21:03 +00:00
|
|
|
if (IS_SYNC(inode))
|
2009-01-07 05:06:22 +00:00
|
|
|
ext4_handle_sync(handle);
|
2006-10-11 08:21:03 +00:00
|
|
|
|
|
|
|
out_stop:
|
2008-07-11 23:27:31 +00:00
|
|
|
up_write(&EXT4_I(inode)->i_data_sem);
|
2006-10-11 08:21:03 +00:00
|
|
|
/*
|
2006-10-11 08:21:07 +00:00
|
|
|
* If this was a simple ftruncate() and the file will remain alive,
|
2006-10-11 08:21:03 +00:00
|
|
|
* then we need to clear up the orphan record which we created above.
|
|
|
|
* However, if this was a real unlink then we were called by
|
|
|
|
* ext4_delete_inode(), and we allow that function to clean up the
|
|
|
|
* orphan info for us.
|
|
|
|
*/
|
|
|
|
if (inode->i_nlink)
|
|
|
|
ext4_orphan_del(handle, inode);
|
|
|
|
|
2008-04-30 02:00:41 +00:00
|
|
|
inode->i_mtime = inode->i_ctime = ext4_current_time(inode);
|
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
2006-10-11 08:21:03 +00:00
|
|
|
ext4_journal_stop(handle);
|
|
|
|
}
|
|
|
|
|
2008-04-29 12:11:12 +00:00
|
|
|
static void ext4_falloc_update_inode(struct inode *inode,
|
|
|
|
int mode, loff_t new_size, int update_ctime)
|
|
|
|
{
|
|
|
|
struct timespec now;
|
|
|
|
|
|
|
|
if (update_ctime) {
|
|
|
|
now = current_fs_time(inode->i_sb);
|
|
|
|
if (!timespec_equal(&inode->i_ctime, &now))
|
|
|
|
inode->i_ctime = now;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Update only when preallocation was requested beyond
|
|
|
|
* the file size.
|
|
|
|
*/
|
2008-09-13 17:06:18 +00:00
|
|
|
if (!(mode & FALLOC_FL_KEEP_SIZE)) {
|
|
|
|
if (new_size > i_size_read(inode))
|
|
|
|
i_size_write(inode, new_size);
|
|
|
|
if (new_size > EXT4_I(inode)->i_disksize)
|
|
|
|
ext4_update_i_disksize(inode, new_size);
|
2010-02-24 14:52:53 +00:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Mark that we allocate beyond EOF so the subsequent truncate
|
|
|
|
* can proceed even if the new size is the same as i_size.
|
|
|
|
*/
|
|
|
|
if (new_size > i_size_read(inode))
|
2010-05-17 02:00:00 +00:00
|
|
|
ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
|
2008-04-29 12:11:12 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2007-07-18 01:42:41 +00:00
|
|
|
/*
|
|
|
|
* preallocate space for a file. This implements ext4's fallocate inode
|
|
|
|
* operation, which gets called from sys_fallocate system call.
|
|
|
|
* For block-mapped files, posix_fallocate should fall back to the method
|
|
|
|
* of writing zeroes to the required new blocks (the same behavior which is
|
|
|
|
* expected for file systems which do not support fallocate() system call).
|
|
|
|
*/
|
|
|
|
long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
|
|
|
|
{
|
|
|
|
handle_t *handle;
|
2008-04-29 12:11:12 +00:00
|
|
|
loff_t new_size;
|
2008-11-05 05:14:04 +00:00
|
|
|
unsigned int max_blocks;
|
2007-07-18 01:42:41 +00:00
|
|
|
int ret = 0;
|
|
|
|
int ret2 = 0;
|
|
|
|
int retries = 0;
|
2010-05-17 00:00:00 +00:00
|
|
|
struct ext4_map_blocks map;
|
2007-07-18 01:42:41 +00:00
|
|
|
unsigned int credits, blkbits = inode->i_blkbits;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* currently supporting (pre)allocate mode for extent-based
|
|
|
|
* files _only_
|
|
|
|
*/
|
2010-05-17 02:00:00 +00:00
|
|
|
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
|
2007-07-18 01:42:41 +00:00
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
|
|
|
/* preallocation to directories is currently not supported */
|
|
|
|
if (S_ISDIR(inode->i_mode))
|
|
|
|
return -ENODEV;
|
|
|
|
|
2010-05-17 00:00:00 +00:00
|
|
|
map.m_lblk = offset >> blkbits;
|
2008-04-29 12:11:12 +00:00
|
|
|
/*
|
|
|
|
* We can't just convert len to max_blocks because
|
|
|
|
* If blocksize = 4096 offset = 3072 and len = 2048
|
|
|
|
*/
|
2007-07-18 01:42:41 +00:00
|
|
|
max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
|
2010-05-17 00:00:00 +00:00
|
|
|
- map.m_lblk;
|
2007-07-18 01:42:41 +00:00
|
|
|
/*
|
2008-08-20 02:16:03 +00:00
|
|
|
* credits to insert 1 extent into extent tree
|
2007-07-18 01:42:41 +00:00
|
|
|
*/
|
2008-08-20 02:16:03 +00:00
|
|
|
credits = ext4_chunk_trans_blocks(inode, max_blocks);
|
2008-02-15 17:47:21 +00:00
|
|
|
mutex_lock(&inode->i_mutex);
|
2010-05-16 18:00:00 +00:00
|
|
|
ret = inode_newsize_ok(inode, (len + offset));
|
|
|
|
if (ret) {
|
|
|
|
mutex_unlock(&inode->i_mutex);
|
|
|
|
return ret;
|
|
|
|
}
|
2007-07-18 01:42:41 +00:00
|
|
|
retry:
|
|
|
|
while (ret >= 0 && ret < max_blocks) {
|
2010-05-17 00:00:00 +00:00
|
|
|
map.m_lblk = map.m_lblk + ret;
|
|
|
|
map.m_len = max_blocks = max_blocks - ret;
|
2007-07-18 01:42:41 +00:00
|
|
|
handle = ext4_journal_start(inode, credits);
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
|
|
|
break;
|
|
|
|
}
|
2010-05-17 00:00:00 +00:00
|
|
|
ret = ext4_map_blocks(handle, inode, &map,
|
2009-05-14 04:58:52 +00:00
|
|
|
EXT4_GET_BLOCKS_CREATE_UNINIT_EXT);
|
2008-01-29 04:58:27 +00:00
|
|
|
if (ret <= 0) {
|
2008-02-25 20:41:35 +00:00
|
|
|
#ifdef EXT4FS_DEBUG
|
|
|
|
WARN_ON(ret <= 0);
|
2010-05-16 23:00:00 +00:00
|
|
|
printk(KERN_ERR "%s: ext4_ext_map_blocks "
|
2008-02-25 20:41:35 +00:00
|
|
|
"returned error inode#%lu, block=%u, "
|
2009-01-27 00:26:26 +00:00
|
|
|
"max_blocks=%u", __func__,
|
2010-10-28 01:30:15 +00:00
|
|
|
inode->i_ino, map.m_lblk, max_blocks);
|
2008-02-25 20:41:35 +00:00
|
|
|
#endif
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
|
|
|
ret2 = ext4_journal_stop(handle);
|
|
|
|
break;
|
|
|
|
}
|
2010-05-17 00:00:00 +00:00
|
|
|
if ((map.m_lblk + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
|
2008-04-29 12:11:12 +00:00
|
|
|
blkbits) >> blkbits))
|
|
|
|
new_size = offset + len;
|
|
|
|
else
|
2010-05-17 00:00:00 +00:00
|
|
|
new_size = (map.m_lblk + ret) << blkbits;
|
2007-07-18 01:42:41 +00:00
|
|
|
|
2008-04-29 12:11:12 +00:00
|
|
|
ext4_falloc_update_inode(inode, mode, new_size,
|
2010-05-17 00:00:00 +00:00
|
|
|
(map.m_flags & EXT4_MAP_NEW));
|
2007-07-18 01:42:41 +00:00
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
|
|
|
ret2 = ext4_journal_stop(handle);
|
|
|
|
if (ret2)
|
|
|
|
break;
|
|
|
|
}
|
2008-04-29 12:11:12 +00:00
|
|
|
if (ret == -ENOSPC &&
|
|
|
|
ext4_should_retry_alloc(inode->i_sb, &retries)) {
|
|
|
|
ret = 0;
|
2007-07-18 01:42:41 +00:00
|
|
|
goto retry;
|
|
|
|
}
|
2008-02-15 17:47:21 +00:00
|
|
|
mutex_unlock(&inode->i_mutex);
|
2007-07-18 01:42:41 +00:00
|
|
|
return ret > 0 ? ret2 : ret;
|
|
|
|
}
|
2008-10-07 04:46:36 +00:00
|
|
|
|
2009-09-28 19:49:08 +00:00
|
|
|
/*
|
|
|
|
* This function convert a range of blocks to written extents
|
|
|
|
* The caller of this function will pass the start offset and the size.
|
|
|
|
* all unwritten extents within this range will be converted to
|
|
|
|
* written extents.
|
|
|
|
*
|
|
|
|
* This function is called from the direct IO end io call back
|
|
|
|
* function, to convert the fallocated extents after IO is completed.
|
2009-11-10 15:48:08 +00:00
|
|
|
* Returns 0 on success.
|
2009-09-28 19:49:08 +00:00
|
|
|
*/
|
|
|
|
int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
|
2010-02-05 04:58:38 +00:00
|
|
|
ssize_t len)
|
2009-09-28 19:49:08 +00:00
|
|
|
{
|
|
|
|
handle_t *handle;
|
|
|
|
unsigned int max_blocks;
|
|
|
|
int ret = 0;
|
|
|
|
int ret2 = 0;
|
2010-05-17 00:00:00 +00:00
|
|
|
struct ext4_map_blocks map;
|
2009-09-28 19:49:08 +00:00
|
|
|
unsigned int credits, blkbits = inode->i_blkbits;
|
|
|
|
|
2010-05-17 00:00:00 +00:00
|
|
|
map.m_lblk = offset >> blkbits;
|
2009-09-28 19:49:08 +00:00
|
|
|
/*
|
|
|
|
* We can't just convert len to max_blocks because
|
|
|
|
* If blocksize = 4096 offset = 3072 and len = 2048
|
|
|
|
*/
|
2010-05-17 00:00:00 +00:00
|
|
|
max_blocks = ((EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits) -
|
|
|
|
map.m_lblk);
|
2009-09-28 19:49:08 +00:00
|
|
|
/*
|
|
|
|
* credits to insert 1 extent into extent tree
|
|
|
|
*/
|
|
|
|
credits = ext4_chunk_trans_blocks(inode, max_blocks);
|
|
|
|
while (ret >= 0 && ret < max_blocks) {
|
2010-05-17 00:00:00 +00:00
|
|
|
map.m_lblk += ret;
|
|
|
|
map.m_len = (max_blocks -= ret);
|
2009-09-28 19:49:08 +00:00
|
|
|
handle = ext4_journal_start(inode, credits);
|
|
|
|
if (IS_ERR(handle)) {
|
|
|
|
ret = PTR_ERR(handle);
|
|
|
|
break;
|
|
|
|
}
|
2010-05-17 00:00:00 +00:00
|
|
|
ret = ext4_map_blocks(handle, inode, &map,
|
2010-03-02 18:28:44 +00:00
|
|
|
EXT4_GET_BLOCKS_IO_CONVERT_EXT);
|
2009-09-28 19:49:08 +00:00
|
|
|
if (ret <= 0) {
|
|
|
|
WARN_ON(ret <= 0);
|
2010-05-16 23:00:00 +00:00
|
|
|
printk(KERN_ERR "%s: ext4_ext_map_blocks "
|
2009-09-28 19:49:08 +00:00
|
|
|
"returned error inode#%lu, block=%u, "
|
|
|
|
"max_blocks=%u", __func__,
|
2010-05-17 00:00:00 +00:00
|
|
|
inode->i_ino, map.m_lblk, map.m_len);
|
2009-09-28 19:49:08 +00:00
|
|
|
}
|
|
|
|
ext4_mark_inode_dirty(handle, inode);
|
|
|
|
ret2 = ext4_journal_stop(handle);
|
|
|
|
if (ret <= 0 || ret2 )
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return ret > 0 ? ret2 : ret;
|
|
|
|
}
|
2008-10-07 04:46:36 +00:00
|
|
|
/*
|
|
|
|
* Callback function called for each extent to gather FIEMAP information.
|
|
|
|
*/
|
2008-11-22 20:04:59 +00:00
|
|
|
static int ext4_ext_fiemap_cb(struct inode *inode, struct ext4_ext_path *path,
|
2008-10-07 04:46:36 +00:00
|
|
|
struct ext4_ext_cache *newex, struct ext4_extent *ex,
|
|
|
|
void *data)
|
|
|
|
{
|
|
|
|
struct fiemap_extent_info *fieinfo = data;
|
2009-05-02 03:32:06 +00:00
|
|
|
unsigned char blksize_bits = inode->i_sb->s_blocksize_bits;
|
2008-10-07 04:46:36 +00:00
|
|
|
__u64 logical;
|
|
|
|
__u64 physical;
|
|
|
|
__u64 length;
|
|
|
|
__u32 flags = 0;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
logical = (__u64)newex->ec_block << blksize_bits;
|
|
|
|
|
|
|
|
if (newex->ec_type == EXT4_EXT_CACHE_GAP) {
|
|
|
|
pgoff_t offset;
|
|
|
|
struct page *page;
|
|
|
|
struct buffer_head *bh = NULL;
|
|
|
|
|
|
|
|
offset = logical >> PAGE_SHIFT;
|
|
|
|
page = find_get_page(inode->i_mapping, offset);
|
|
|
|
if (!page || !page_has_buffers(page))
|
|
|
|
return EXT_CONTINUE;
|
|
|
|
|
|
|
|
bh = page_buffers(page);
|
|
|
|
|
|
|
|
if (!bh)
|
|
|
|
return EXT_CONTINUE;
|
|
|
|
|
|
|
|
if (buffer_delay(bh)) {
|
|
|
|
flags |= FIEMAP_EXTENT_DELALLOC;
|
|
|
|
page_cache_release(page);
|
|
|
|
} else {
|
|
|
|
page_cache_release(page);
|
|
|
|
return EXT_CONTINUE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
physical = (__u64)newex->ec_start << blksize_bits;
|
|
|
|
length = (__u64)newex->ec_len << blksize_bits;
|
|
|
|
|
|
|
|
if (ex && ext4_ext_is_uninitialized(ex))
|
|
|
|
flags |= FIEMAP_EXTENT_UNWRITTEN;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If this extent reaches EXT_MAX_BLOCK, it must be last.
|
|
|
|
*
|
|
|
|
* Or if ext4_ext_next_allocated_block is EXT_MAX_BLOCK,
|
|
|
|
* this also indicates no more allocated blocks.
|
|
|
|
*
|
|
|
|
* XXX this might miss a single-block extent at EXT_MAX_BLOCK
|
|
|
|
*/
|
2009-05-02 03:32:06 +00:00
|
|
|
if (ext4_ext_next_allocated_block(path) == EXT_MAX_BLOCK ||
|
2009-05-02 23:05:37 +00:00
|
|
|
newex->ec_block + newex->ec_len - 1 == EXT_MAX_BLOCK) {
|
|
|
|
loff_t size = i_size_read(inode);
|
|
|
|
loff_t bs = EXT4_BLOCK_SIZE(inode->i_sb);
|
|
|
|
|
2008-10-07 04:46:36 +00:00
|
|
|
flags |= FIEMAP_EXTENT_LAST;
|
2009-05-02 23:05:37 +00:00
|
|
|
if ((flags & FIEMAP_EXTENT_DELALLOC) &&
|
|
|
|
logical+length > size)
|
|
|
|
length = (size - logical + bs - 1) & ~(bs-1);
|
|
|
|
}
|
2008-10-07 04:46:36 +00:00
|
|
|
|
|
|
|
error = fiemap_fill_next_extent(fieinfo, logical, physical,
|
|
|
|
length, flags);
|
|
|
|
if (error < 0)
|
|
|
|
return error;
|
|
|
|
if (error == 1)
|
|
|
|
return EXT_BREAK;
|
|
|
|
|
|
|
|
return EXT_CONTINUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* fiemap flags we can handle specified here */
|
|
|
|
#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR)
|
|
|
|
|
2008-11-22 20:04:59 +00:00
|
|
|
static int ext4_xattr_fiemap(struct inode *inode,
|
|
|
|
struct fiemap_extent_info *fieinfo)
|
2008-10-07 04:46:36 +00:00
|
|
|
{
|
|
|
|
__u64 physical = 0;
|
|
|
|
__u64 length;
|
|
|
|
__u32 flags = FIEMAP_EXTENT_LAST;
|
|
|
|
int blockbits = inode->i_sb->s_blocksize_bits;
|
|
|
|
int error = 0;
|
|
|
|
|
|
|
|
/* in-inode? */
|
2010-01-24 19:34:07 +00:00
|
|
|
if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
|
2008-10-07 04:46:36 +00:00
|
|
|
struct ext4_iloc iloc;
|
|
|
|
int offset; /* offset of xattr in inode */
|
|
|
|
|
|
|
|
error = ext4_get_inode_loc(inode, &iloc);
|
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
physical = iloc.bh->b_blocknr << blockbits;
|
|
|
|
offset = EXT4_GOOD_OLD_INODE_SIZE +
|
|
|
|
EXT4_I(inode)->i_extra_isize;
|
|
|
|
physical += offset;
|
|
|
|
length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
|
|
|
|
flags |= FIEMAP_EXTENT_DATA_INLINE;
|
2010-04-03 21:44:16 +00:00
|
|
|
brelse(iloc.bh);
|
2008-10-07 04:46:36 +00:00
|
|
|
} else { /* external block */
|
|
|
|
physical = EXT4_I(inode)->i_file_acl << blockbits;
|
|
|
|
length = inode->i_sb->s_blocksize;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (physical)
|
|
|
|
error = fiemap_fill_next_extent(fieinfo, 0, physical,
|
|
|
|
length, flags);
|
|
|
|
return (error < 0 ? error : 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
|
|
|
|
__u64 start, __u64 len)
|
|
|
|
{
|
|
|
|
ext4_lblk_t start_blk;
|
|
|
|
int error = 0;
|
|
|
|
|
|
|
|
/* fallback to generic here if not in extents fmt */
|
2010-05-17 02:00:00 +00:00
|
|
|
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
|
2008-10-07 04:46:36 +00:00
|
|
|
return generic_block_fiemap(inode, fieinfo, start, len,
|
|
|
|
ext4_get_block);
|
|
|
|
|
|
|
|
if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
|
|
|
|
return -EBADR;
|
|
|
|
|
|
|
|
if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
|
|
|
|
error = ext4_xattr_fiemap(inode, fieinfo);
|
|
|
|
} else {
|
2010-03-04 22:07:28 +00:00
|
|
|
ext4_lblk_t len_blks;
|
|
|
|
__u64 last_blk;
|
|
|
|
|
2008-10-07 04:46:36 +00:00
|
|
|
start_blk = start >> inode->i_sb->s_blocksize_bits;
|
2010-03-04 22:07:28 +00:00
|
|
|
last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
|
|
|
|
if (last_blk >= EXT_MAX_BLOCK)
|
|
|
|
last_blk = EXT_MAX_BLOCK-1;
|
|
|
|
len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
|
2008-10-07 04:46:36 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Walk the extent tree gathering extent information.
|
|
|
|
* ext4_ext_fiemap_cb will push extents back to user.
|
|
|
|
*/
|
|
|
|
error = ext4_ext_walk_space(inode, start_blk, len_blks,
|
|
|
|
ext4_ext_fiemap_cb, fieinfo);
|
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|