#error "not redhat kernel"
#endif
],[
- RHEL_KENEL="yes"
RHEL_KERNEL="yes"
AC_MSG_RESULT([yes])
],[
# don't need to do this if only configuring for make dist
AC_MSG_CHECKING([which ldiskfs series to use])
case $LINUXRELEASE in
-2.6.5*) LDISKFS_SERIES="2.6-suse.series" ;;
-2.6.9*) LDISKFS_SERIES="2.6-rhel4.series" ;;
-2.6.10-ac*) LDISKFS_SERIES="2.6-fc3.series" ;;
-2.6.10*) LDISKFS_SERIES="2.6-rhel4.series" ;;
-2.6.12*) LDISKFS_SERIES="2.6.12-vanilla.series" ;;
-2.6.15*) LDISKFS_SERIES="2.6-fc5.series";;
-2.6.16*) LDISKFS_SERIES="2.6-sles10.series";;
2.6.18*)
- if test x$RHEL_KENEL = xyes; then
- if test x$enable_ext4 = xyes; then
- LDISKFS_SERIES="2.6-rhel5-ext4.series"
- else
- LDISKFS_SERIES="2.6-rhel5.series"
- fi
- else
- LDISKFS_SERIES="2.6.18-vanilla.series"
+ if test x$RHEL_KERNEL = xyes; then
+ LDISKFS_SERIES="2.6-rhel5-ext4.series"
+ fi
+ ;;
+2.6.32*)
+ if test x$RHEL_KERNEL = xyes; then
+ LDISKFS_SERIES="2.6-rhel6.series"
fi
;;
-2.6.22*) LDISKFS_SERIES="2.6.22-vanilla.series";;
-2.6.27*) LDISKFS_SERIES="2.6-sles11.series";;
-2.6.32*) LDISKFS_SERIES="2.6-rhel6.series";;
*) AC_MSG_WARN([Unknown kernel version $LINUXRELEASE, fix ldiskfs/configure.ac])
esac
AC_MSG_RESULT([$LDISKFS_SERIES])
+++ /dev/null
-diff -Nrpu /tmp/linux-stage/fs/ext3/ialloc.c linux-stage/fs/ext3/ialloc.c
---- /tmp/linux-stage/fs/ext3/ialloc.c 2009-05-09 06:44:02.000000000 +0400
-+++ linux-stage/fs/ext3/ialloc.c 2009-05-09 06:45:33.000000000 +0400
-@@ -822,6 +822,36 @@ fail_drop:
- return ERR_PTR(err);
- }
-
-+unsigned long ext3_find_reverse(struct super_block *sb)
-+{
-+ struct ext3_group_desc *desc;
-+ struct buffer_head *bitmap_bh = NULL;
-+ int group;
-+ unsigned long ino, offset;
-+
-+ for (offset = (EXT3_INODES_PER_GROUP(sb) >> 1); offset >= 0;
-+ offset >>= 1) {
-+ for (group = EXT3_SB(sb)->s_groups_count - 1; group >= 0;
-+ --group) {
-+ desc = ext3_get_group_desc(sb, group, NULL);
-+ if (desc->bg_free_inodes_count == 0)
-+ continue;
-+
-+ bitmap_bh = read_inode_bitmap(sb, group);
-+ if (!bitmap_bh)
-+ continue;
-+
-+ ino = ext3_find_next_zero_bit((unsigned long *)
-+ bitmap_bh->b_data,
-+ EXT3_INODES_PER_GROUP(sb), offset);
-+ if (ino < EXT3_INODES_PER_GROUP(sb))
-+ return(group * EXT3_INODES_PER_GROUP(sb) +
-+ ino + 1);
-+ }
-+ }
-+ return 0;
-+}
-+
- /* Verify that we are loading a valid orphan from disk */
- struct inode *ext3_orphan_get(struct super_block *sb, unsigned long ino)
- {
-diff -Nrpu /tmp/linux-stage/fs/ext3/namei.c linux-stage/fs/ext3/namei.c
---- /tmp/linux-stage/fs/ext3/namei.c 2009-05-09 06:44:02.000000000 +0400
-+++ linux-stage/fs/ext3/namei.c 2009-05-09 06:45:33.000000000 +0400
-@@ -145,14 +145,25 @@ struct dx_map_entry
- u32 offs;
- };
-
-+/*
-+ * dentry_param used by ext3_new_inode_wantedi()
-+ */
- #define LVFS_DENTRY_PARAM_MAGIC 20070216UL
- struct lvfs_dentry_params
- {
-- unsigned long p_inum;
-- void *p_ptr;
-- u32 magic;
-+ unsigned long ldp_inum;
-+ long ldp_flags;
-+ u32 ldp_magic;
- };
-
-+/* Only use the least 3 bits of ldp_flags for goal policy */
-+typedef enum {
-+ DP_GOAL_POLICY = 0,
-+ DP_LASTGROUP_REVERSE = 1,
-+} dp_policy_t;
-+
-+#define LDP_FLAGS_RANGE 0x07
-+
- #ifdef CONFIG_EXT3_INDEX
- static inline unsigned dx_get_block (struct dx_entry *entry);
- static void dx_set_block (struct dx_entry *entry, unsigned value);
-@@ -1718,8 +1727,13 @@ static struct inode * ext3_new_inode_wan
- if (dentry->d_fsdata != NULL) {
- struct lvfs_dentry_params *param = dentry->d_fsdata;
-
-- if (param->magic == LVFS_DENTRY_PARAM_MAGIC)
-- inum = param->p_inum;
-+ if (param->ldp_magic == LVFS_DENTRY_PARAM_MAGIC) {
-+ if ((dp_policy_t)(param->ldp_flags & LDP_FLAGS_RANGE) ==
-+ DP_LASTGROUP_REVERSE)
-+ inum = ext3_find_reverse(dir->i_sb);
-+ else /* DP_GOAL_POLICY */
-+ inum = param->ldp_inum;
-+ }
- }
- return ext3_new_inode(handle, dir, mode, inum);
- }
-diff -Nrpu /tmp/linux-stage/include/linux/ext3_fs.h linux-stage/include/linux/ext3_fs.h
---- /tmp/linux-stage/include/linux/ext3_fs.h 2009-05-09 06:44:02.000000000 +0400
-+++ linux-stage/include/linux/ext3_fs.h 2009-05-09 06:45:33.000000000 +0400
-@@ -973,6 +973,7 @@ extern int ext3fs_dirhash(const char *na
- /* ialloc.c */
- extern struct inode * ext3_new_inode (handle_t *, struct inode *, int,
- unsigned long);
-+extern unsigned long ext3_find_reverse(struct super_block *);
- extern void ext3_free_inode (handle_t *, struct inode *);
- extern struct inode * ext3_orphan_get (struct super_block *, unsigned long);
- extern unsigned long ext3_count_free_inodes (struct super_block *);
+++ /dev/null
-Index: linux-2.6.9-42.0.10.EL_lustre.1.4.10/fs/ext3/super.c
-===================================================================
---- linux-2.6.9-42.0.10.EL_lustre.1.4.10.orig/fs/ext3/super.c 2007-05-16 08:46:24.000000000 +0200
-+++ linux-2.6.9-42.0.10.EL_lustre.1.4.10/fs/ext3/super.c 2007-05-16 08:48:58.000000000 +0200
-@@ -123,6 +123,8 @@ void ext3_journal_abort_handle(const cha
- journal_abort_handle(handle);
- }
-
-+EXPORT_SYMBOL(ext3_journal_abort_handle);
-+
- /* Deal with the reporting of failure conditions on a filesystem such as
- * inconsistencies detected or read IO failures.
- *
-@@ -2064,6 +2066,8 @@ int ext3_force_commit(struct super_block
- return ret;
- }
-
-+EXPORT_SYMBOL(ext3_force_commit);
-+
- /*
- * Ext3 always journals updates to the superblock itself, so we don't
- * have to propagate any other updates to the superblock on disk at this
-@@ -2586,6 +2590,12 @@ int ext3_map_inode_page(struct inode *in
- unsigned long *blocks, int *created, int create);
- EXPORT_SYMBOL(ext3_map_inode_page);
-
-+EXPORT_SYMBOL(ext3_xattr_get);
-+EXPORT_SYMBOL(ext3_xattr_set_handle);
-+EXPORT_SYMBOL(ext3_bread);
-+EXPORT_SYMBOL(ext3_journal_start_sb);
-+EXPORT_SYMBOL(__ext3_journal_stop);
-+
- MODULE_AUTHOR("Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others");
- MODULE_DESCRIPTION("Second Extended Filesystem with journaling extensions");
- MODULE_LICENSE("GPL");
+++ /dev/null
-Index: linux-2.6.18-92.1.6/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-92.1.6.orig/fs/ext3/super.c
-+++ linux-2.6.18-92.1.6/fs/ext3/super.c
-@@ -71,6 +71,8 @@ static void ext3_unlockfs(struct super_b
- static void ext3_write_super (struct super_block * sb);
- static void ext3_write_super_lockfs(struct super_block *sb);
-
-+static int bigendian_extents;
-+
- /*
- * Wrappers for journal_start/end.
- *
-@@ -706,7 +708,7 @@ enum {
- Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
- Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
- Opt_grpquota,
-- Opt_extents, Opt_noextents, Opt_extdebug,
-+ Opt_extents, Opt_noextents, Opt_bigendian_extents, Opt_extdebug,
- Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_maxdirsize
- };
-
-@@ -766,6 +768,7 @@ static match_table_t tokens = {
- {Opt_barrier, "barrier=%u"},
- {Opt_extents, "extents"},
- {Opt_noextents, "noextents"},
-+ {Opt_bigendian_extents, "bigendian_extents"},
- {Opt_extdebug, "extdebug"},
- {Opt_mballoc, "mballoc"},
- {Opt_nomballoc, "nomballoc"},
-@@ -1129,6 +1132,9 @@ clear_qf_name:
- case Opt_noextents:
- clear_opt (sbi->s_mount_opt, EXTENTS);
- break;
-+ case Opt_bigendian_extents:
-+ bigendian_extents = 1;
-+ break;
- case Opt_extdebug:
- set_opt (sbi->s_mount_opt, EXTDEBUG);
- break;
-@@ -2224,6 +2230,15 @@ static int ext3_fill_super (struct super
- goto failed_mount;
- }
-
-+#ifdef __BIG_ENDIAN
-+ if (bigendian_extents == 0) {
-+ printk(KERN_ERR "EXT3-fs: extents feature is not guaranteed to "
-+ "work on big-endian systems. Use \"bigendian_extents\" "
-+ "mount option to override.\n");
-+ goto failed_mount;
-+ }
-+#endif
-+
- bgl_lock_init(&sbi->s_blockgroup_lock);
-
- sbi->s_last_alloc_group = -1;
+++ /dev/null
- fs/ext3/balloc.c | 99 ++++++++++++++++++++++++++++++++++++++++++++----------
- 1 files changed, 81 insertions(+), 18 deletions(-)
-diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c
-index ff3428e..a9140ea 100644
-Index: linux-stage/fs/ext3/balloc.c
-===================================================================
---- linux-stage.orig/fs/ext3/balloc.c
-+++ linux-stage/fs/ext3/balloc.c
-@@ -143,9 +143,96 @@ unsigned ext3_init_block_bitmap(struct s
- return free_blocks - sbi->s_itb_per_group - 2;
- }
-
--/*
-- * Read the bitmap for a given block_group, reading into the specified
-- * slot in the superblock's bitmap cache.
-+/**
-+* bh_uptodate_or_lock: Test whether the buffer is uptodate
-+* @bh: struct buffer_head
-+*
-+* Return true if the buffer is up-to-date and false,
-+* with the buffer locked, if not.
-+*/
-+int bh_uptodate_or_lock(struct buffer_head *bh)
-+{
-+ if (!buffer_uptodate(bh)) {
-+ lock_buffer(bh);
-+ if (!buffer_uptodate(bh))
-+ return 0;
-+ unlock_buffer(bh);
-+ }
-+ return 1;
-+}
-+
-+/**
-+* bh_submit_read: Submit a locked buffer for reading
-+* @bh: struct buffer_head
-+*
-+* Returns a negative error
-+*/
-+int bh_submit_read(struct buffer_head *bh)
-+{
-+ if (!buffer_locked(bh))
-+ lock_buffer(bh);
-+ if (buffer_uptodate(bh))
-+ return 0;
-+ get_bh(bh);
-+ bh->b_end_io = end_buffer_read_sync;
-+ submit_bh(READ, bh);
-+ wait_on_buffer(bh);
-+ if (buffer_uptodate(bh))
-+ return 0;
-+ return -EIO;
-+}
-+
-+static int ext3_valid_block_bitmap(struct super_block *sb,
-+ struct ext3_group_desc *desc,
-+ unsigned int block_group,
-+ struct buffer_head *bh)
-+{
-+ ext3_grpblk_t offset;
-+ ext3_grpblk_t next_zero_bit;
-+ ext3_fsblk_t bitmap_blk;
-+ ext3_fsblk_t group_first_block;
-+
-+ group_first_block = ext3_group_first_block_no(sb, block_group);
-+
-+ /* check whether block bitmap block number is set */
-+ bitmap_blk = le32_to_cpu(desc->bg_block_bitmap);
-+ offset = bitmap_blk - group_first_block;
-+ if (!ext3_test_bit(offset, bh->b_data))
-+ /* bad block bitmap */
-+ goto err_out;
-+
-+ /* check whether the inode bitmap block number is set */
-+ bitmap_blk = le32_to_cpu(desc->bg_inode_bitmap);
-+ offset = bitmap_blk - group_first_block;
-+ if (!ext3_test_bit(offset, bh->b_data))
-+ /* bad block bitmap */
-+ goto err_out;
-+
-+ /* check whether the inode table block number is set */
-+ bitmap_blk = le32_to_cpu(desc->bg_inode_table);
-+ offset = bitmap_blk - group_first_block;
-+ next_zero_bit = ext3_find_next_zero_bit(bh->b_data,
-+ offset + EXT3_SB(sb)->s_itb_per_group,
-+ offset);
-+ if (next_zero_bit >= offset + EXT3_SB(sb)->s_itb_per_group)
-+ /* good bitmap for inode tables */
-+ return 1;
-+
-+err_out:
-+ ext3_error(sb, __FUNCTION__,
-+ "Invalid block bitmap - "
-+ "block_group = %d, block = %lu",
-+ (int)block_group, bitmap_blk);
-+ return 0;
-+}
-+
-+/**
-+ * read_block_bitmap()
-+ * @sb: super block
-+ * @block_group: given block group
-+ *
-+ * Read the bitmap for a given block_group,and validate the
-+ * bits for block/inode/inode tables are set in the bitmaps.
- *
- * Return buffer_head on success or NULL in case of failure.
- */
-@@ -154,29 +241,42 @@ read_block_bitmap(struct super_block *sb
- {
- struct ext3_group_desc * desc;
- struct buffer_head * bh = NULL;
-+ ext3_fsblk_t bitmap_blk;
-
- desc = ext3_get_group_desc (sb, block_group, NULL);
- if (!desc)
-- goto error_out;
-+ return NULL;
-+ bitmap_blk = desc->bg_block_bitmap;
-+ bh = sb_getblk(sb, bitmap_blk);
-+ if (unlikely(!bh)) {
-+ ext3_error(sb, __FUNCTION__,
-+ "Can not read block bitmap - "
-+ "block group = %d, block_bitmap = %lu",
-+ (int)block_group, bitmap_blk);
-+ return NULL;
-+ }
-+ if (bh_uptodate_or_lock(bh))
-+ return bh;
-+
- if (desc->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-- bh = sb_getblk(sb, le32_to_cpu(desc->bg_block_bitmap));
-- if (!buffer_uptodate(bh)) {
-- lock_buffer(bh);
-- if (!buffer_uptodate(bh)) {
-- ext3_init_block_bitmap(sb, bh,block_group,desc);
-- set_buffer_uptodate(bh);
-- }
-- unlock_buffer(bh);
-- }
-- } else {
-- bh = sb_bread(sb, le32_to_cpu(desc->bg_block_bitmap));
-+ ext3_init_block_bitmap(sb, bh, block_group, desc);
-+ set_buffer_uptodate(bh);
-+ unlock_buffer(bh);
-+ return bh;
-+ }
-+ if (bh_submit_read(bh) < 0) {
-+ brelse(bh);
-+ ext3_error(sb, __FUNCTION__,
-+ "Cannot read block bitmap - "
-+ "block group = %d block_bitmap = %lu",
-+ (int)block_group, bitmap_blk);
-+ return NULL;
-+ }
-+ if (!ext3_valid_block_bitmap(sb, desc, block_group, bh)) {
-+ brelse(bh);
-+ return NULL;
- }
-- if (!bh)
-- ext3_error (sb, "read_block_bitmap",
-- "Cannot read block bitmap - "
-- "block_group = %d, block_bitmap = %u",
-- block_group, le32_to_cpu(desc->bg_block_bitmap));
--error_out:
-+
- return bh;
- }
- /*
+++ /dev/null
---- linux-2.6.18-128.7.1.orig/include/linux/ext3_fs.h 2006-09-19 23:42:06.000000000 -0400
-+++ linux-2.6.18-128.7.1/include/linux/ext3_fs.h 2009-10-12 19:37:54.000000000 -0400
-@@ -809,6 +809,7 @@ extern void ext3_discard_reservation (st
- extern void ext3_dirty_inode(struct inode *);
- extern int ext3_change_inode_journal_flag(struct inode *, int);
- extern int ext3_get_inode_loc(struct inode *, struct ext3_iloc *);
-+extern int ext3_can_truncate(struct inode *inode);
- extern void ext3_truncate (struct inode *);
- extern void ext3_set_inode_flags(struct inode *);
- extern void ext3_set_aops(struct inode *inode);
---- linux-2.6.18-128.7.1.orig/fs/ext3/inode.c 2009-09-15 10:38:31.000000000 -0400
-+++ linux-2.6.18-128.7.1/fs/ext3/inode.c 2009-10-12 18:49:01.000000000 -0400
-@@ -2194,6 +2194,19 @@ static void ext3_free_branches(handle_t
- }
- }
-
-+int ext3_can_truncate(struct inode *inode)
-+{
-+ if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
-+ return 0;
-+ if (S_ISREG(inode->i_mode))
-+ return 1;
-+ if (S_ISDIR(inode->i_mode))
-+ return 1;
-+ if (S_ISLNK(inode->i_mode))
-+ return !ext3_inode_is_fast_symlink(inode);
-+ return 0;
-+}
-+
- /*
- * ext3_truncate()
- *
-@@ -2238,12 +2251,7 @@ void ext3_truncate(struct inode *inode)
- unsigned blocksize = inode->i_sb->s_blocksize;
- struct page *page;
-
-- if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) ||
-- S_ISLNK(inode->i_mode)))
-- goto out_notrans;
-- if (ext3_inode_is_fast_symlink(inode))
-- goto out_notrans;
-- if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
-+ if (!ext3_can_truncate(inode))
- goto out_notrans;
-
- /*
---- linux-2.6.18-128.7.1.orig/fs/ext3/ialloc.c 2009-09-15 10:24:17.000000000 -0400
-+++ linux-2.6.18-128.7.1/fs/ext3/ialloc.c 2009-10-14 15:36:27.000000000 -0400
-@@ -645,54 +645,71 @@ struct inode *ext3_orphan_get(struct sup
- unsigned long max_ino = le32_to_cpu(EXT3_SB(sb)->s_es->s_inodes_count);
- unsigned long block_group;
- int bit;
-- struct buffer_head *bitmap_bh = NULL;
-+ struct buffer_head *bitmap_bh;
- struct inode *inode = NULL;
-
- /* Error cases - e2fsck has already cleaned up for us */
- if (ino > max_ino) {
-- ext3_warning(sb, __FUNCTION__,
-+ ext3_warning(sb, __func__,
- "bad orphan ino %lu! e2fsck was run?", ino);
-- goto out;
-+ goto error;
- }
-
- block_group = (ino - 1) / EXT3_INODES_PER_GROUP(sb);
- bit = (ino - 1) % EXT3_INODES_PER_GROUP(sb);
- bitmap_bh = read_inode_bitmap(sb, block_group);
- if (!bitmap_bh) {
-- ext3_warning(sb, __FUNCTION__,
-+ ext3_warning(sb, __func__,
- "inode bitmap error for orphan %lu", ino);
-- goto out;
-+ goto error;
- }
-
- /* Having the inode bit set should be a 100% indicator that this
- * is a valid orphan (no e2fsck run on fs). Orphans also include
- * inodes that were being truncated, so we can't check i_nlink==0.
- */
-- if (!ext3_test_bit(bit, bitmap_bh->b_data) ||
-- !(inode = iget(sb, ino)) || is_bad_inode(inode) ||
-- NEXT_ORPHAN(inode) > max_ino) {
-- ext3_warning(sb, __FUNCTION__,
-- "bad orphan inode %lu! e2fsck was run?", ino);
-- printk(KERN_NOTICE "ext3_test_bit(bit=%d, block=%llu) = %d\n",
-- bit, (unsigned long long)bitmap_bh->b_blocknr,
-- ext3_test_bit(bit, bitmap_bh->b_data));
-- printk(KERN_NOTICE "inode=%p\n", inode);
-- if (inode) {
-- printk(KERN_NOTICE "is_bad_inode(inode)=%d\n",
-- is_bad_inode(inode));
-- printk(KERN_NOTICE "NEXT_ORPHAN(inode)=%u\n",
-- NEXT_ORPHAN(inode));
-- printk(KERN_NOTICE "max_ino=%lu\n", max_ino);
-- }
-+ if (!ext3_test_bit(bit, bitmap_bh->b_data))
-+ goto bad_orphan;
-+
-+ inode = iget(sb, ino);
-+ if ((inode == NULL) || is_bad_inode(inode))
-+ goto bad_orphan;
-+
-+ /*
-+ * If the orphan has i_nlinks > 0 then it should be able to be
-+ * truncated, otherwise it won't be removed from the orphan list
-+ * during processing and an infinite loop will result.
-+ */
-+ if (inode->i_nlink && !ext3_can_truncate(inode))
-+ goto bad_orphan;
-+
-+ if (NEXT_ORPHAN(inode) > max_ino)
-+ goto bad_orphan;
-+ brelse(bitmap_bh);
-+ return inode;
-+
-+bad_orphan:
-+ ext3_warning(sb, __func__,
-+ "bad orphan inode %lu! e2fsck was run?", ino);
-+ printk(KERN_NOTICE "ext3_test_bit(bit=%d, block=%llu) = %d\n",
-+ bit, (unsigned long long)bitmap_bh->b_blocknr,
-+ ext3_test_bit(bit, bitmap_bh->b_data));
-+ printk(KERN_NOTICE "inode=%p\n", inode);
-+ if (inode) {
-+ printk(KERN_NOTICE "is_bad_inode(inode)=%d\n",
-+ is_bad_inode(inode));
-+ printk(KERN_NOTICE "NEXT_ORPHAN(inode)=%u\n",
-+ NEXT_ORPHAN(inode));
-+ printk(KERN_NOTICE "max_ino=%lu\n", max_ino);
-+ printk(KERN_NOTICE "i_nlink=%u\n", inode->i_nlink);
- /* Avoid freeing blocks if we got a bad deleted inode */
-- if (inode && inode->i_nlink == 0)
-+ if (inode->i_nlink == 0)
- inode->i_blocks = 0;
- iput(inode);
-- inode = NULL;
- }
--out:
- brelse(bitmap_bh);
-- return inode;
-+error:
-+ return NULL;
- }
-
- unsigned long ext3_count_free_inodes (struct super_block * sb)
+++ /dev/null
-Index: linux-stage/fs/ext3/xattr.c
-===================================================================
---- linux-stage.orig/fs/ext3/xattr.c
-+++ linux-stage/fs/ext3/xattr.c
-@@ -93,7 +93,8 @@
- # define ea_bdebug(f...)
- #endif
-
--static void ext3_xattr_cache_insert(struct buffer_head *);
-+static void ext3_xattr_cache_insert(struct super_block *,
-+ struct buffer_head *);
- static struct buffer_head *ext3_xattr_cache_find(struct inode *,
- struct ext3_xattr_header *,
- struct mb_cache_entry **);
-@@ -238,7 +239,7 @@ bad_block: ext3_error(inode->i_sb, __FUN
- error = -EIO;
- goto cleanup;
- }
-- ext3_xattr_cache_insert(bh);
-+ ext3_xattr_cache_insert(inode->i_sb, bh);
- entry = BFIRST(bh);
- error = ext3_xattr_find_entry(&entry, name_index, name, bh->b_size, 1);
- if (error == -EIO)
-@@ -380,7 +381,7 @@ ext3_xattr_block_list(struct inode *inod
- error = -EIO;
- goto cleanup;
- }
-- ext3_xattr_cache_insert(bh);
-+ ext3_xattr_cache_insert(inode->i_sb, bh);
- error = ext3_xattr_list_entries(inode, BFIRST(bh), buffer, buffer_size);
-
- cleanup:
-@@ -479,7 +480,9 @@ ext3_xattr_release_block(handle_t *handl
- {
- struct mb_cache_entry *ce = NULL;
-
-- ce = mb_cache_entry_get(ext3_xattr_cache, bh->b_bdev, bh->b_blocknr);
-+ if (!test_opt(inode->i_sb, NO_MBCACHE))
-+ ce = mb_cache_entry_get(ext3_xattr_cache, bh->b_bdev,
-+ bh->b_blocknr);
- if (BHDR(bh)->h_refcount == cpu_to_le32(1)) {
- ea_bdebug(bh, "refcount now=0; freeing");
- if (ce)
-@@ -699,8 +702,10 @@ ext3_xattr_block_set(handle_t *handle, s
- if (i->value && i->value_len > sb->s_blocksize)
- return -ENOSPC;
- if (s->base) {
-- ce = mb_cache_entry_get(ext3_xattr_cache, bs->bh->b_bdev,
-- bs->bh->b_blocknr);
-+ if (!test_opt(inode->i_sb, NO_MBCACHE))
-+ ce = mb_cache_entry_get(ext3_xattr_cache,
-+ bs->bh->b_bdev,
-+ bs->bh->b_blocknr);
- if (header(s->base)->h_refcount == cpu_to_le32(1)) {
- if (ce) {
- mb_cache_entry_free(ce);
-@@ -716,7 +721,7 @@ ext3_xattr_block_set(handle_t *handle, s
- if (!IS_LAST_ENTRY(s->first))
- ext3_xattr_rehash(header(s->base),
- s->here);
-- ext3_xattr_cache_insert(bs->bh);
-+ ext3_xattr_cache_insert(sb, bs->bh);
- }
- unlock_buffer(bs->bh);
- if (error == -EIO)
-@@ -797,7 +802,8 @@ inserted:
- if (error)
- goto cleanup_dquot;
- }
-- mb_cache_entry_release(ce);
-+ if (ce)
-+ mb_cache_entry_release(ce);
- ce = NULL;
- } else if (bs->bh && s->base == bs->bh->b_data) {
- /* We were modifying this block in-place. */
-@@ -832,7 +838,7 @@ getblk_failed:
- memcpy(new_bh->b_data, s->base, new_bh->b_size);
- set_buffer_uptodate(new_bh);
- unlock_buffer(new_bh);
-- ext3_xattr_cache_insert(new_bh);
-+ ext3_xattr_cache_insert(sb, new_bh);
- error = ext3_journal_dirty_metadata(handle, new_bh);
- if (error)
- goto cleanup;
-@@ -1387,12 +1393,15 @@ ext3_xattr_put_super(struct super_block
- * Returns 0, or a negative error number on failure.
- */
- static void
--ext3_xattr_cache_insert(struct buffer_head *bh)
-+ext3_xattr_cache_insert(struct super_block *sb, struct buffer_head *bh)
- {
- __u32 hash = le32_to_cpu(BHDR(bh)->h_hash);
- struct mb_cache_entry *ce;
- int error;
-
-+ if (test_opt(sb, NO_MBCACHE))
-+ return;
-+
- ce = mb_cache_entry_alloc(ext3_xattr_cache);
- if (!ce) {
- ea_bdebug(bh, "out of memory");
-@@ -1466,6 +1475,8 @@ ext3_xattr_cache_find(struct inode *inod
- __u32 hash = le32_to_cpu(header->h_hash);
- struct mb_cache_entry *ce;
-
-+ if (test_opt(inode->i_sb, NO_MBCACHE))
-+ return NULL;
- if (!header->h_hash)
- return NULL; /* never share */
- ea_idebug(inode, "looking for cached blocks [%x]", (int)hash);
-Index: linux-stage/fs/ext3/super.c
-===================================================================
---- linux-stage.orig/fs/ext3/super.c
-+++ linux-stage/fs/ext3/super.c
-@@ -722,6 +722,7 @@ enum {
- Opt_grpquota,
- Opt_extents, Opt_noextents, Opt_bigendian_extents, Opt_extdebug,
- Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_maxdirsize, Opt_force_over_8tb,
-+ Opt_no_mbcache,
- };
-
- static match_table_t tokens = {
-@@ -788,6 +789,7 @@ static match_table_t tokens = {
- {Opt_force_over_8tb, "force_over_8tb"},
- {Opt_resize, "resize"},
- {Opt_maxdirsize, "maxdirsize=%u"},
-+ {Opt_no_mbcache, "no_mbcache"},
- {Opt_err, NULL}
- };
-
-@@ -1175,6 +1177,9 @@ clear_qf_name:
- case Opt_force_over_8tb:
- force_over_8tb = 1;
- break;
-+ case Opt_no_mbcache:
-+ set_opt(sbi->s_mount_opt, NO_MBCACHE);
-+ break;
- default:
- printk (KERN_ERR
- "EXT3-fs: Unrecognized mount option \"%s\" "
-Index: linux-stage/include/linux/ext3_fs.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs.h
-+++ linux-stage/include/linux/ext3_fs.h
-@@ -483,6 +483,8 @@ do { \
- #define EXT3_MOUNT_JOURNAL_ASYNC_COMMIT 0x20000000 /* Journal Async Commit */
- #endif
-
-+#define EXT3_MOUNT_NO_MBCACHE 0x40000000 /* Disable mbcache */
-+
- /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
- #ifndef clear_opt
- #define clear_opt(o, opt) o &= ~EXT3_MOUNT_##opt
+++ /dev/null
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/Makefile linux-2.6.18-128.1.6_2/fs/ext3/Makefile
---- linux-2.6.18-128.1.6_1/fs/ext3/Makefile 2009-08-13 19:19:54.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/Makefile 2009-08-13 19:20:30.000000000 +0530
-@@ -5,7 +5,8 @@
- obj-$(CONFIG_EXT3_FS) += ext3.o
-
- ext3-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o \
-- ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o mballoc.o
-+ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \
-+ mballoc.o dynlocks.o
-
- ext3-$(CONFIG_EXT3_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o
- ext3-$(CONFIG_EXT3_FS_POSIX_ACL) += acl.o
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/super.c linux-2.6.18-128.1.6_2/fs/ext3/super.c
---- linux-2.6.18-128.1.6_1/fs/ext3/super.c 2009-08-13 19:19:54.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/super.c 2009-08-13 19:23:23.000000000 +0530
-@@ -3529,6 +3530,7 @@ static int __init init_ext3_fs(void)
- err = init_inodecache();
- if (err)
- goto out1;
-+ dynlock_cache_init();
- err = register_filesystem(&ext3_fs_type);
- if (err)
- goto out;
-@@ -3546,6 +3548,7 @@ out1:
- static void __exit exit_ext3_fs(void)
- {
- unregister_filesystem(&ext3_fs_type);
-+ dynlock_cache_exit();
- destroy_inodecache();
- exit_ext3_xattr();
- exit_ext3_proc();
-
+++ /dev/null
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/dynlocks.c linux-2.6.18-128.1.6_2/fs/ext3/dynlocks.c
---- linux-2.6.18-128.1.6_1/fs/ext3/dynlocks.c 1970-01-01 05:30:00.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/dynlocks.c 2009-08-13 20:42:59.000000000 +0530
-@@ -0,0 +1,236 @@
-+/*
-+ * Dynamic Locks
-+ *
-+ * struct dynlock is lockspace
-+ * one may request lock (exclusive or shared) for some value
-+ * in that lockspace
-+ *
-+ */
-+
-+#include <linux/dynlocks.h>
-+#include <linux/module.h>
-+#include <linux/slab.h>
-+#include <linux/sched.h>
-+
-+#define DYNLOCK_HANDLE_MAGIC 0xd19a10c
-+#define DYNLOCK_HANDLE_DEAD 0xd1956ee
-+#define DYNLOCK_LIST_MAGIC 0x11ee91e6
-+
-+static kmem_cache_t * dynlock_cachep = NULL;
-+
-+struct dynlock_handle {
-+ unsigned dh_magic;
-+ struct list_head dh_list;
-+ unsigned long dh_value; /* lock value */
-+ int dh_refcount; /* number of users */
-+ int dh_readers;
-+ int dh_writers;
-+ int dh_pid; /* holder of the lock */
-+ wait_queue_head_t dh_wait;
-+};
-+
-+int __init dynlock_cache_init(void)
-+{
-+ int rc = 0;
-+
-+ printk(KERN_INFO "init dynlocks cache\n");
-+ dynlock_cachep = kmem_cache_create("dynlock_cache",
-+ sizeof(struct dynlock_handle),
-+ 0,
-+ SLAB_HWCACHE_ALIGN,
-+ NULL, NULL);
-+ if (dynlock_cachep == NULL) {
-+ printk(KERN_ERR "Not able to create dynlock cache");
-+ rc = -ENOMEM;
-+ }
-+ return rc;
-+}
-+
-+void __exit dynlock_cache_exit(void)
-+{
-+ printk(KERN_INFO "exit dynlocks cache\n");
-+ kmem_cache_destroy(dynlock_cachep);
-+}
-+
-+/*
-+ * dynlock_init
-+ *
-+ * initialize lockspace
-+ *
-+ */
-+void dynlock_init(struct dynlock *dl)
-+{
-+ spin_lock_init(&dl->dl_list_lock);
-+ INIT_LIST_HEAD(&dl->dl_list);
-+ dl->dl_magic = DYNLOCK_LIST_MAGIC;
-+}
-+EXPORT_SYMBOL(dynlock_init);
-+
-+/*
-+ * dynlock_lock
-+ *
-+ * acquires lock (exclusive or shared) in specified lockspace
-+ * each lock in lockspace is allocated separately, so user have
-+ * to specify GFP flags.
-+ * routine returns pointer to lock. this pointer is intended to
-+ * be passed to dynlock_unlock
-+ *
-+ */
-+struct dynlock_handle *dynlock_lock(struct dynlock *dl, unsigned long value,
-+ enum dynlock_type lt, gfp_t gfp)
-+{
-+ struct dynlock_handle *nhl = NULL;
-+ struct dynlock_handle *hl;
-+
-+ BUG_ON(dl == NULL);
-+ BUG_ON(dl->dl_magic != DYNLOCK_LIST_MAGIC);
-+
-+repeat:
-+ /* find requested lock in lockspace */
-+ spin_lock(&dl->dl_list_lock);
-+ BUG_ON(dl->dl_list.next == NULL);
-+ BUG_ON(dl->dl_list.prev == NULL);
-+ list_for_each_entry(hl, &dl->dl_list, dh_list) {
-+ BUG_ON(hl->dh_list.next == NULL);
-+ BUG_ON(hl->dh_list.prev == NULL);
-+ BUG_ON(hl->dh_magic != DYNLOCK_HANDLE_MAGIC);
-+ if (hl->dh_value == value) {
-+ /* lock is found */
-+ if (nhl) {
-+ /* someone else just allocated
-+ * lock we didn't find and just created
-+ * so, we drop our lock
-+ */
-+ kmem_cache_free(dynlock_cachep, nhl);
-+ nhl = NULL;
-+ }
-+ hl->dh_refcount++;
-+ goto found;
-+ }
-+ }
-+ /* lock not found */
-+ if (nhl) {
-+ /* we already have allocated lock. use it */
-+ hl = nhl;
-+ nhl = NULL;
-+ list_add(&hl->dh_list, &dl->dl_list);
-+ goto found;
-+ }
-+ spin_unlock(&dl->dl_list_lock);
-+
-+ /* lock not found and we haven't allocated lock yet. allocate it */
-+ nhl = kmem_cache_alloc(dynlock_cachep, gfp);
-+ if (nhl == NULL)
-+ return NULL;
-+ nhl->dh_refcount = 1;
-+ nhl->dh_value = value;
-+ nhl->dh_readers = 0;
-+ nhl->dh_writers = 0;
-+ nhl->dh_magic = DYNLOCK_HANDLE_MAGIC;
-+ init_waitqueue_head(&nhl->dh_wait);
-+
-+ /* while lock is being allocated, someone else may allocate it
-+ * and put onto to list. check this situation
-+ */
-+ goto repeat;
-+
-+found:
-+ if (lt == DLT_WRITE) {
-+ /* exclusive lock: user don't want to share lock at all
-+ * NOTE: one process may take the same lock several times
-+ * this functionaly is useful for rename operations */
-+ while ((hl->dh_writers && hl->dh_pid != current->pid) ||
-+ hl->dh_readers) {
-+ spin_unlock(&dl->dl_list_lock);
-+ wait_event(hl->dh_wait,
-+ hl->dh_writers == 0 && hl->dh_readers == 0);
-+ spin_lock(&dl->dl_list_lock);
-+ }
-+ hl->dh_writers++;
-+ } else {
-+ /* shared lock: user do not want to share lock with writer */
-+ while (hl->dh_writers) {
-+ spin_unlock(&dl->dl_list_lock);
-+ wait_event(hl->dh_wait, hl->dh_writers == 0);
-+ spin_lock(&dl->dl_list_lock);
-+ }
-+ hl->dh_readers++;
-+ }
-+ hl->dh_pid = current->pid;
-+ spin_unlock(&dl->dl_list_lock);
-+
-+ return hl;
-+}
-+EXPORT_SYMBOL(dynlock_lock);
-+
-+
-+/*
-+ * dynlock_unlock
-+ *
-+ * user have to specify lockspace (dl) and pointer to lock structure
-+ * returned by dynlock_lock()
-+ *
-+ */
-+void dynlock_unlock(struct dynlock *dl, struct dynlock_handle *hl)
-+{
-+ int wakeup = 0;
-+
-+ BUG_ON(dl == NULL);
-+ BUG_ON(hl == NULL);
-+ BUG_ON(dl->dl_magic != DYNLOCK_LIST_MAGIC);
-+
-+ if (hl->dh_magic != DYNLOCK_HANDLE_MAGIC)
-+ printk(KERN_EMERG "wrong lock magic: %#x\n", hl->dh_magic);
-+
-+ BUG_ON(hl->dh_magic != DYNLOCK_HANDLE_MAGIC);
-+ BUG_ON(hl->dh_writers != 0 && current->pid != hl->dh_pid);
-+
-+ spin_lock(&dl->dl_list_lock);
-+ if (hl->dh_writers) {
-+ BUG_ON(hl->dh_readers != 0);
-+ hl->dh_writers--;
-+ if (hl->dh_writers == 0)
-+ wakeup = 1;
-+ } else if (hl->dh_readers) {
-+ hl->dh_readers--;
-+ if (hl->dh_readers == 0)
-+ wakeup = 1;
-+ } else {
-+ BUG();
-+ }
-+ if (wakeup) {
-+ hl->dh_pid = 0;
-+ wake_up(&hl->dh_wait);
-+ }
-+ if (--(hl->dh_refcount) == 0) {
-+ hl->dh_magic = DYNLOCK_HANDLE_DEAD;
-+ list_del(&hl->dh_list);
-+ kmem_cache_free(dynlock_cachep, hl);
-+ }
-+ spin_unlock(&dl->dl_list_lock);
-+}
-+EXPORT_SYMBOL(dynlock_unlock);
-+
-+int dynlock_is_locked(struct dynlock *dl, unsigned long value)
-+{
-+ struct dynlock_handle *hl;
-+ int result = 0;
-+
-+ /* find requested lock in lockspace */
-+ spin_lock(&dl->dl_list_lock);
-+ BUG_ON(dl->dl_list.next == NULL);
-+ BUG_ON(dl->dl_list.prev == NULL);
-+ list_for_each_entry(hl, &dl->dl_list, dh_list) {
-+ BUG_ON(hl->dh_list.next == NULL);
-+ BUG_ON(hl->dh_list.prev == NULL);
-+ BUG_ON(hl->dh_magic != DYNLOCK_HANDLE_MAGIC);
-+ if (hl->dh_value == value && hl->dh_pid == current->pid) {
-+ /* lock is found */
-+ result = 1;
-+ break;
-+ }
-+ }
-+ spin_unlock(&dl->dl_list_lock);
-+ return result;
-+}
-+EXPORT_SYMBOL(dynlock_is_locked);
-diff -rupN linux-2.6.18-128.1.6_1/include/linux/dynlocks.h linux-2.6.18-128.1.6_2/include/linux/dynlocks.h
---- linux-2.6.18-128.1.6_1/include/linux/dynlocks.h 1970-01-01 05:30:00.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/include/linux/dynlocks.h 2009-08-13 20:43:18.000000000 +0530
-@@ -0,0 +1,34 @@
-+#ifndef _LINUX_DYNLOCKS_H
-+#define _LINUX_DYNLOCKS_H
-+
-+#include <linux/list.h>
-+#include <linux/wait.h>
-+
-+struct dynlock_handle;
-+
-+/*
-+ * lock's namespace:
-+ * - list of locks
-+ * - lock to protect this list
-+ */
-+struct dynlock {
-+ unsigned dl_magic;
-+ struct list_head dl_list;
-+ spinlock_t dl_list_lock;
-+};
-+
-+enum dynlock_type {
-+ DLT_WRITE,
-+ DLT_READ
-+};
-+
-+int dynlock_cache_init(void);
-+void dynlock_cache_exit(void);
-+void dynlock_init(struct dynlock *dl);
-+struct dynlock_handle *dynlock_lock(struct dynlock *dl, unsigned long value,
-+ enum dynlock_type lt, gfp_t gfp);
-+void dynlock_unlock(struct dynlock *dl, struct dynlock_handle *lock);
-+int dynlock_is_locked(struct dynlock *dl, unsigned long value);
-+
-+#endif
-+
+++ /dev/null
-Date: Mon, 12 May 2008 11:24:40 +0800
-From: Tiger Yang <tiger.yang@oracle.com>
-Subject: [PATCH] ext3/4: fix uninitialized bs in ext3/4_xattr_set_handle()
-To: linux-ext4@vger.kernel.org
-Cc: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
-
-This fix the uninitialized bs when we try to replace a xattr entry in ibody
-with the new value which require more than free space.
-
-Signed-off-by: Tiger Yang <tiger.yang@oracle.com>
-
-
-diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c
---- a/fs/ext3/xattr.c
-+++ b/fs/ext3/xattr.c
-@@ -1000,6 +1000,11 @@ ext3_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index,
- i.value = NULL;
- error = ext3_xattr_block_set(handle, inode, &i, &bs);
- } else if (error == -ENOSPC) {
-+ if (EXT3_I(inode)->i_file_acl && !bs.s.base) {
-+ error = ext3_xattr_block_find(inode, &i, &bs);
-+ if (error)
-+ goto cleanup;
-+ }
- error = ext3_xattr_block_set(handle, inode, &i, &bs);
- if (error)
- goto cleanup;
+++ /dev/null
-Index: linux-stage/fs/ext3/dir.c
-===================================================================
---- linux-stage.orig/fs/ext3/dir.c 2011-04-19 01:39:47.000000000 +0800
-+++ linux-stage/fs/ext3/dir.c 2011-04-19 01:44:19.000000000 +0800
-@@ -237,22 +237,50 @@
- }
-
- #ifdef CONFIG_EXT3_INDEX
-+static inline int is_32bit_api(void)
-+{
-+#ifdef HAVE_IS_COMPAT_TASK
-+ return is_compat_task();
-+#else
-+ return (BITS_PER_LONG == 32);
-+#endif
-+}
-+
- /*
- * These functions convert from the major/minor hash to an f_pos
- * value.
-- *
-- * Currently we only use major hash numer. This is unfortunate, but
-- * on 32-bit machines, the same VFS interface is used for lseek and
-- * llseek, so if we use the 64 bit offset, then the 32-bit versions of
-- * lseek/telldir/seekdir will blow out spectacularly, and from within
-- * the ext2 low-level routine, we don't know if we're being called by
-- * a 64-bit version of the system call or the 32-bit version of the
-- * system call. Worse yet, NFSv2 only allows for a 32-bit readdir
-- * cookie. Sigh.
-+ *
-+ * Up layer (OSD) should specify O_32BITHASH or O_64BITHASH explicitly.
-+ * On the other hand, we allow ldiskfs to be mounted directly on both 32-bit
-+ * and 64-bit nodes, under such case, neither O_32BITHASH nor O_64BITHASH is
-+ * specified.
- */
--#define hash2pos(major, minor) (major >> 1)
--#define pos2maj_hash(pos) ((pos << 1) & 0xffffffff)
--#define pos2min_hash(pos) (0)
-+static inline loff_t hash2pos(struct file *filp, __u32 major, __u32 minor)
-+{
-+ if ((filp->f_flags & O_32BITHASH) ||
-+ (!(filp->f_flags & O_64BITHASH) && is_32bit_api()))
-+ return (major >> 1);
-+ else
-+ return (((__u64)(major >> 1) << 32) | (__u64)minor);
-+}
-+
-+static inline __u32 pos2maj_hash(struct file *filp, loff_t pos)
-+{
-+ if ((filp->f_flags & O_32BITHASH) ||
-+ (!(filp->f_flags & O_64BITHASH) && is_32bit_api()))
-+ return ((pos << 1) & 0xffffffff);
-+ else
-+ return (((pos >> 32) << 1) & 0xffffffff);
-+}
-+
-+static inline __u32 pos2min_hash(struct file *filp, loff_t pos)
-+{
-+ if ((filp->f_flags & O_32BITHASH) ||
-+ (!(filp->f_flags & O_64BITHASH) && is_32bit_api()))
-+ return (0);
-+ else
-+ return (pos & 0xffffffff);
-+}
-
- /*
- * This structure holds the nodes of the red-black tree used to store
-@@ -314,7 +342,7 @@
- }
-
-
--static struct dir_private_info *create_dir_info(loff_t pos)
-+static struct dir_private_info *create_dir_info(struct file* filp, loff_t pos)
- {
- struct dir_private_info *p;
-
-@@ -325,8 +353,8 @@
- p->curr_node = NULL;
- p->extra_fname = NULL;
- p->last_pos = 0;
-- p->curr_hash = pos2maj_hash(pos);
-- p->curr_minor_hash = pos2min_hash(pos);
-+ p->curr_hash = pos2maj_hash(filp, pos);
-+ p->curr_minor_hash = pos2min_hash(filp, pos);
- p->next_hash = 0;
- return p;
- }
-@@ -422,7 +450,7 @@
- printk("call_filldir: called with null fname?!?\n");
- return 0;
- }
-- curr_pos = hash2pos(fname->hash, fname->minor_hash);
-+ curr_pos = hash2pos(filp, fname->hash, fname->minor_hash);
- while (fname) {
- error = filldir(dirent, fname->name,
- fname->name_len, curr_pos,
-@@ -447,7 +475,7 @@
- int ret;
-
- if (!info) {
-- info = create_dir_info(filp->f_pos);
-+ info = create_dir_info(filp, filp->f_pos);
- if (!info)
- return -ENOMEM;
- filp->private_data = info;
-@@ -461,8 +489,8 @@
- free_rb_tree_fname(&info->root);
- info->curr_node = NULL;
- info->extra_fname = NULL;
-- info->curr_hash = pos2maj_hash(filp->f_pos);
-- info->curr_minor_hash = pos2min_hash(filp->f_pos);
-+ info->curr_hash = pos2maj_hash(filp, filp->f_pos);
-+ info->curr_minor_hash = pos2min_hash(filp, filp->f_pos);
- }
-
- /*
-Index: linux-stage/include/linux/ext3_fs.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs.h 2011-04-19 01:39:47.000000000 +0800
-+++ linux-stage/include/linux/ext3_fs.h 2011-04-19 01:45:21.000000000 +0800
-@@ -54,6 +54,14 @@
- #define ext3_debug(f, a...) do {} while (0)
- #endif
-
-+#ifndef O_32BITHASH
-+# define O_32BITHASH 0x10000000
-+#endif
-+
-+#ifndef O_64BITHASH
-+# define O_64BITHASH 0x20000000
-+#endif
-+
- #define EXT3_MULTIBLOCK_ALLOCATOR 1
-
- #define EXT3_MB_HINT_MERGE 1 /* prefer goal again. length */
+++ /dev/null
-Index: linux-2.6.18.8/fs/ext3/dir.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/dir.c 2007-02-24 00:52:30.000000000 +0100
-+++ linux-2.6.18.8/fs/ext3/dir.c 2007-07-17 09:18:14.000000000 +0200
-@@ -131,8 +131,7 @@ static int ext3_readdir(struct file * fi
- struct buffer_head *bh = NULL;
-
- map_bh.b_state = 0;
-- err = ext3_get_blocks_handle(NULL, inode, blk, 1,
-- &map_bh, 0, 0);
-+ err = ext3_get_blocks_wrap(NULL, inode, blk, 1, &map_bh, 0, 0);
- if (err > 0) {
- page_cache_readahead(sb->s_bdev->bd_inode->i_mapping,
- &filp->f_ra,
-Index: linux-2.6.18.8/fs/ext3/extents.c
-===================================================================
---- /dev/null 1970-01-01 00:00:00.000000000 +0000
-+++ linux-2.6.18.8/fs/ext3/extents.c 2007-07-17 11:08:59.000000000 +0200
-@@ -0,0 +1,2276 @@
-+/*
-+ * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
-+ * 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 EXT3
-+ *
-+ * TODO:
-+ * - ext3*_error() should be used in some situations
-+ * - analyze all BUG()/BUG_ON(), use -EIO where appropriate
-+ * - smart tree reduction
-+ */
-+
-+#include <linux/module.h>
-+#include <linux/fs.h>
-+#include <linux/time.h>
-+#include <linux/ext3_jbd.h>
-+#include <linux/jbd.h>
-+#include <linux/smp_lock.h>
-+#include <linux/highuid.h>
-+#include <linux/pagemap.h>
-+#include <linux/quotaops.h>
-+#include <linux/string.h>
-+#include <linux/slab.h>
-+#include <linux/ext3_extents.h>
-+#include <asm/uaccess.h>
-+
-+
-+static handle_t *ext3_ext_journal_restart(handle_t *handle, int needed)
-+{
-+ int err;
-+
-+ if (handle->h_buffer_credits > needed)
-+ return handle;
-+ if (!ext3_journal_extend(handle, needed))
-+ return handle;
-+ err = ext3_journal_restart(handle, needed);
-+
-+ return handle;
-+}
-+
-+/*
-+ * could return:
-+ * - EROFS
-+ * - ENOMEM
-+ */
-+static int ext3_ext_get_access(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ if (path->p_bh) {
-+ /* path points to block */
-+ return ext3_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 ext3_ext_dirty(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ int err;
-+ if (path->p_bh) {
-+ /* path points to block */
-+ err = ext3_journal_dirty_metadata(handle, path->p_bh);
-+ } else {
-+ /* path points to leaf/index in inode body */
-+ err = ext3_mark_inode_dirty(handle, inode);
-+ }
-+ return err;
-+}
-+
-+static int ext3_ext_find_goal(struct inode *inode,
-+ struct ext3_ext_path *path,
-+ unsigned long block)
-+{
-+ struct ext3_inode_info *ei = EXT3_I(inode);
-+ unsigned long bg_start;
-+ unsigned long colour;
-+ int depth;
-+
-+ if (path) {
-+ struct ext3_extent *ex;
-+ depth = path->p_depth;
-+
-+ /* try to predict block placement */
-+ if ((ex = path[depth].p_ext))
-+ return le32_to_cpu(ex->ee_start)
-+ + (block - le32_to_cpu(ex->ee_block));
-+
-+ /* it looks index is empty
-+ * try to find starting from index itself */
-+ if (path[depth].p_bh)
-+ return path[depth].p_bh->b_blocknr;
-+ }
-+
-+ /* OK. use inode's group */
-+ bg_start = (ei->i_block_group * EXT3_BLOCKS_PER_GROUP(inode->i_sb)) +
-+ le32_to_cpu(EXT3_SB(inode->i_sb)->s_es->s_first_data_block);
-+ colour = (current->pid % 16) *
-+ (EXT3_BLOCKS_PER_GROUP(inode->i_sb) / 16);
-+ return bg_start + colour + block;
-+}
-+
-+static int
-+ext3_ext_new_block(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path,
-+ struct ext3_extent *ex, int *err)
-+{
-+ int goal, newblock;
-+
-+ goal = ext3_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block));
-+ newblock = ext3_new_block(handle, inode, goal, err);
-+ return newblock;
-+}
-+
-+static inline int ext3_ext_space_block(struct inode *inode)
-+{
-+ int size;
-+
-+ size = (inode->i_sb->s_blocksize - sizeof(struct ext3_extent_header))
-+ / sizeof(struct ext3_extent);
-+#ifdef AGRESSIVE_TEST
-+ if (size > 6)
-+ size = 6;
-+#endif
-+ return size;
-+}
-+
-+static inline int ext3_ext_space_block_idx(struct inode *inode)
-+{
-+ int size;
-+
-+ size = (inode->i_sb->s_blocksize - sizeof(struct ext3_extent_header))
-+ / sizeof(struct ext3_extent_idx);
-+#ifdef AGRESSIVE_TEST
-+ if (size > 5)
-+ size = 5;
-+#endif
-+ return size;
-+}
-+
-+static inline int ext3_ext_space_root(struct inode *inode)
-+{
-+ int size;
-+
-+ size = sizeof(EXT3_I(inode)->i_data);
-+ size -= sizeof(struct ext3_extent_header);
-+ size /= sizeof(struct ext3_extent);
-+#ifdef AGRESSIVE_TEST
-+ if (size > 3)
-+ size = 3;
-+#endif
-+ return size;
-+}
-+
-+static inline int ext3_ext_space_root_idx(struct inode *inode)
-+{
-+ int size;
-+
-+ size = sizeof(EXT3_I(inode)->i_data);
-+ size -= sizeof(struct ext3_extent_header);
-+ size /= sizeof(struct ext3_extent_idx);
-+#ifdef AGRESSIVE_TEST
-+ if (size > 4)
-+ size = 4;
-+#endif
-+ return size;
-+}
-+
-+static inline int
-+ext3_ext_max_entries(struct inode *inode, int depth)
-+{
-+ int max;
-+
-+ if (depth == ext_depth(inode)) {
-+ if (depth == 0)
-+ max = ext3_ext_space_root(inode);
-+ else
-+ max = ext3_ext_space_root_idx(inode);
-+ } else {
-+ if (depth == 0)
-+ max = ext3_ext_space_block(inode);
-+ else
-+ max = ext3_ext_space_block_idx(inode);
-+ }
-+
-+ return max;
-+}
-+
-+static int __ext3_ext_check_header(const char *function, int line, struct inode *inode,
-+ struct ext3_extent_header *eh,
-+ int depth)
-+{
-+ const char *error_msg = NULL;
-+ int max = 0;
-+
-+ if (unlikely(eh->eh_magic != cpu_to_le16(EXT3_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 = ext3_ext_max_entries(inode, depth);
-+#ifdef AGRESSIVE_TEST
-+ if (eh->eh_max > 3) {
-+ /* inode probably got extent without defining AGRESSIVE_TEST */
-+ max = eh->eh_max;
-+ }
-+#endif
-+ if (unlikely(le16_to_cpu(eh->eh_max) > max)) {
-+ error_msg = "too large eh_max";
-+ goto corrupted;
-+ }
-+ if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) {
-+ error_msg = "invalid eh_entries";
-+ goto corrupted;
-+ }
-+ if (unlikely((eh->eh_entries == 0) && (eh->eh_depth != 0))) {
-+ error_msg = "invalid index, eh_entries=0 && eh_depth != 0";
-+ goto corrupted;
-+ }
-+ return 0;
-+
-+corrupted:
-+ ext3_error(inode->i_sb, function,
-+ ":%d: bad header in inode #%lu: %s - magic %x, "
-+ "entries %u, max %u(%u), depth %u(%u)", line,
-+ inode->i_ino, error_msg, le16_to_cpu(eh->eh_magic),
-+ le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
-+ max, le16_to_cpu(eh->eh_depth), depth);
-+
-+ return -EIO;
-+}
-+
-+#define ext3_ext_check_header(inode,eh,depth) \
-+ __ext3_ext_check_header(__FUNCTION__,__LINE__,inode,eh,depth)
-+
-+#ifdef EXT_DEBUG
-+static void ext3_ext_show_path(struct inode *inode, struct ext3_ext_path *path)
-+{
-+ int k, l = path->p_depth;
-+
-+ ext_debug(inode, "path:");
-+ for (k = 0; k <= l; k++, path++) {
-+ if (path->p_idx) {
-+ ext_debug(inode, " %d->%d", le32_to_cpu(path->p_idx->ei_block),
-+ le32_to_cpu(path->p_idx->ei_leaf));
-+ } else if (path->p_ext) {
-+ ext_debug(inode, " %d:%d:%d",
-+ le32_to_cpu(path->p_ext->ee_block),
-+ le16_to_cpu(path->p_ext->ee_len),
-+ le32_to_cpu(path->p_ext->ee_start));
-+ } else
-+ ext_debug(inode, " []");
-+ }
-+ ext_debug(inode, "\n");
-+}
-+
-+static void ext3_ext_show_leaf(struct inode *inode, struct ext3_ext_path *path)
-+{
-+ int depth = ext_depth(inode);
-+ struct ext3_extent_header *eh;
-+ struct ext3_extent *ex;
-+ int i;
-+
-+ if (!path)
-+ return;
-+
-+ eh = path[depth].p_hdr;
-+ ex = EXT_FIRST_EXTENT(eh);
-+
-+ for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
-+ ext_debug(inode, "%d:%d:%d ", le32_to_cpu(ex->ee_block),
-+ le16_to_cpu(ex->ee_len),
-+ le32_to_cpu(ex->ee_start));
-+ }
-+ ext_debug(inode, "\n");
-+}
-+#else
-+#define ext3_ext_show_path(inode,path)
-+#define ext3_ext_show_leaf(inode,path)
-+#endif
-+
-+static void ext3_ext_drop_refs(struct ext3_ext_path *path)
-+{
-+ int depth = path->p_depth;
-+ int i;
-+
-+ for (i = 0; i <= depth; i++, path++)
-+ if (path->p_bh) {
-+ brelse(path->p_bh);
-+ path->p_bh = NULL;
-+ }
-+}
-+
-+/*
-+ * binary search for closest index by given block
-+ * the header must be checked before calling this
-+ */
-+static void
-+ext3_ext_binsearch_idx(struct inode *inode, struct ext3_ext_path *path, int block)
-+{
-+ struct ext3_extent_header *eh = path->p_hdr;
-+ struct ext3_extent_idx *r, *l, *m;
-+
-+ ext_debug(inode, "binsearch for %d(idx): ", block);
-+
-+ l = EXT_FIRST_INDEX(eh) + 1;
-+ r = EXT_FIRST_INDEX(eh) + le16_to_cpu(eh->eh_entries) - 1;
-+ while (l <= r) {
-+ m = l + (r - l) / 2;
-+ if (block < le32_to_cpu(m->ei_block))
-+ r = m - 1;
-+ else
-+ l = m + 1;
-+ ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l, l->ei_block,
-+ m, m->ei_block, r, r->ei_block);
-+ }
-+
-+ path->p_idx = l - 1;
-+ ext_debug(inode, " -> %d->%d ", le32_to_cpu(path->p_idx->ei_block),
-+ le32_to_cpu(path->p_idx->ei_leaf));
-+
-+#ifdef CHECK_BINSEARCH
-+ {
-+ struct ext3_extent_idx *chix, *ix;
-+ int k;
-+
-+ chix = ix = EXT_FIRST_INDEX(eh);
-+ for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
-+ if (k != 0 &&
-+ le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
-+ printk("k=%d, ix=0x%p, first=0x%p\n", k,
-+ ix, EXT_FIRST_INDEX(eh));
-+ printk("%u <= %u\n",
-+ le32_to_cpu(ix->ei_block),
-+ le32_to_cpu(ix[-1].ei_block));
-+ }
-+ BUG_ON(k && le32_to_cpu(ix->ei_block)
-+ <= le32_to_cpu(ix[-1].ei_block));
-+ if (block < le32_to_cpu(ix->ei_block))
-+ break;
-+ chix = ix;
-+ }
-+ BUG_ON(chix != path->p_idx);
-+ }
-+#endif
-+
-+}
-+
-+/*
-+ * binary search for closest extent by given block
-+ * the header must be checked before calling this
-+ */
-+static void
-+ext3_ext_binsearch(struct inode *inode, struct ext3_ext_path *path, int block)
-+{
-+ struct ext3_extent_header *eh = path->p_hdr;
-+ struct ext3_extent *r, *l, *m;
-+
-+ if (eh->eh_entries == 0) {
-+ /*
-+ * this leaf is empty yet:
-+ * we get such a leaf in split/add case
-+ */
-+ return;
-+ }
-+
-+ ext_debug(inode, "binsearch for %d: ", block);
-+
-+ l = EXT_FIRST_EXTENT(eh) + 1;
-+ r = EXT_FIRST_EXTENT(eh) + le16_to_cpu(eh->eh_entries) - 1;
-+
-+ while (l <= r) {
-+ m = l + (r - l) / 2;
-+ if (block < le32_to_cpu(m->ee_block))
-+ r = m - 1;
-+ else
-+ l = m + 1;
-+ ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l, l->ee_block,
-+ m, m->ee_block, r, r->ee_block);
-+ }
-+
-+ path->p_ext = l - 1;
-+ ext_debug(inode, " -> %d:%d:%d ",
-+ le32_to_cpu(path->p_ext->ee_block),
-+ le32_to_cpu(path->p_ext->ee_start),
-+ le16_to_cpu(path->p_ext->ee_len));
-+
-+#ifdef CHECK_BINSEARCH
-+ {
-+ struct ext3_extent *chex, *ex;
-+ int k;
-+
-+ chex = ex = EXT_FIRST_EXTENT(eh);
-+ for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) {
-+ BUG_ON(k && le32_to_cpu(ex->ee_block)
-+ <= le32_to_cpu(ex[-1].ee_block));
-+ if (block < le32_to_cpu(ex->ee_block))
-+ break;
-+ chex = ex;
-+ }
-+ BUG_ON(chex != path->p_ext);
-+ }
-+#endif
-+
-+}
-+
-+int ext3_ext_tree_init(handle_t *handle, struct inode *inode)
-+{
-+ struct ext3_extent_header *eh;
-+
-+ eh = ext_inode_hdr(inode);
-+ eh->eh_depth = 0;
-+ eh->eh_entries = 0;
-+ eh->eh_magic = cpu_to_le16(EXT3_EXT_MAGIC);
-+ eh->eh_max = cpu_to_le16(ext3_ext_space_root(inode));
-+ ext3_mark_inode_dirty(handle, inode);
-+ ext3_ext_invalidate_cache(inode);
-+ return 0;
-+}
-+
-+struct ext3_ext_path *
-+ext3_ext_find_extent(struct inode *inode, int block, struct ext3_ext_path *path)
-+{
-+ struct ext3_extent_header *eh;
-+ struct buffer_head *bh;
-+ short int depth, i, ppos = 0, alloc = 0;
-+
-+ eh = ext_inode_hdr(inode);
-+ i = depth = ext_depth(inode);
-+ if (ext3_ext_check_header(inode, eh, depth))
-+ return ERR_PTR(-EIO);
-+
-+ /* account possible depth increase */
-+ if (!path) {
-+ path = kmalloc(sizeof(struct ext3_ext_path) * (depth + 2),
-+ GFP_NOFS);
-+ if (!path)
-+ return ERR_PTR(-ENOMEM);
-+ alloc = 1;
-+ }
-+ memset(path, 0, sizeof(struct ext3_ext_path) * (depth + 1));
-+ path[0].p_hdr = eh;
-+
-+ /* walk through the tree */
-+ while (i) {
-+ ext_debug(inode, "depth %d: num %d, max %d\n",
-+ ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
-+
-+ ext3_ext_binsearch_idx(inode, path + ppos, block);
-+ path[ppos].p_block = le32_to_cpu(path[ppos].p_idx->ei_leaf);
-+ path[ppos].p_depth = i;
-+ path[ppos].p_ext = NULL;
-+
-+ bh = sb_bread(inode->i_sb, path[ppos].p_block);
-+ if (!bh)
-+ goto err;
-+
-+ eh = ext_block_hdr(bh);
-+ ppos++;
-+ BUG_ON(ppos > depth);
-+ path[ppos].p_bh = bh;
-+ path[ppos].p_hdr = eh;
-+ i--;
-+
-+ if (ext3_ext_check_header(inode, eh, i))
-+ goto err;
-+ }
-+
-+ path[ppos].p_depth = i;
-+ path[ppos].p_hdr = eh;
-+ path[ppos].p_ext = NULL;
-+ path[ppos].p_idx = NULL;
-+
-+ /* find extent */
-+ ext3_ext_binsearch(inode, path + ppos, block);
-+
-+ ext3_ext_show_path(inode, path);
-+
-+ return path;
-+
-+err:
-+ ext3_ext_drop_refs(path);
-+ if (alloc)
-+ kfree(path);
-+ return ERR_PTR(-EIO);
-+}
-+
-+/*
-+ * insert new index [logical;ptr] into the block at cupr
-+ * it check where to insert: before curp or after curp
-+ */
-+static int ext3_ext_insert_index(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *curp,
-+ int logical, int ptr)
-+{
-+ struct ext3_extent_idx *ix;
-+ int len, err;
-+
-+ if ((err = ext3_ext_get_access(handle, inode, curp)))
-+ return err;
-+
-+ BUG_ON(logical == le32_to_cpu(curp->p_idx->ei_block));
-+ 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 ext3_extent_idx);
-+ len = len < 0 ? 0 : len;
-+ ext_debug(inode, "insert new index %d after: %d. "
-+ "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 ext3_extent_idx);
-+ len = len < 0 ? 0 : len;
-+ ext_debug(inode, "insert new index %d before: %d. "
-+ "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);
-+ ix->ei_leaf = cpu_to_le32(ptr);
-+ ix->ei_leaf_hi = ix->ei_unused = 0;
-+ curp->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(curp->p_hdr->eh_entries)+1);
-+
-+ BUG_ON(le16_to_cpu(curp->p_hdr->eh_entries)
-+ > le16_to_cpu(curp->p_hdr->eh_max));
-+ BUG_ON(ix > EXT_LAST_INDEX(curp->p_hdr));
-+
-+ err = ext3_ext_dirty(handle, inode, curp);
-+ ext3_std_error(inode->i_sb, err);
-+
-+ return err;
-+}
-+
-+/*
-+ * routine 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 extens and index entries (right to the split point)
-+ * into the newly allocated blocks
-+ * - initialize subtree
-+ */
-+static int ext3_ext_split(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path,
-+ struct ext3_extent *newext, int at)
-+{
-+ struct buffer_head *bh = NULL;
-+ int depth = ext_depth(inode);
-+ struct ext3_extent_header *neh;
-+ struct ext3_extent_idx *fidx;
-+ struct ext3_extent *ex;
-+ int i = at, k, m, a;
-+ unsigned long newblock, oldblock;
-+ __le32 border;
-+ int *ablocks = NULL; /* array of allocated blocks */
-+ int err = 0;
-+
-+ /* make decision: where to split? */
-+ /* FIXME: now desicion is simplest: at current extent */
-+
-+ /* if current leaf will be splitted, then we should use
-+ * border from split point */
-+ BUG_ON(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr));
-+ if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
-+ border = path[depth].p_ext[1].ee_block;
-+ ext_debug(inode, "leaf will be splitted."
-+ " next leaf starts at %d\n",
-+ le32_to_cpu(border));
-+ } else {
-+ border = newext->ee_block;
-+ ext_debug(inode, "leaf will be added."
-+ " next leaf starts at %d\n",
-+ le32_to_cpu(border));
-+ }
-+
-+ /*
-+ * if error occurs, then we break processing
-+ * and turn filesystem read-only. so, index won't
-+ * be inserted and tree will be in consistent
-+ * state. next mount will repair buffers too
-+ */
-+
-+ /*
-+ * get array to track all allocated blocks
-+ * we need this to handle errors and free blocks
-+ * upon them
-+ */
-+ ablocks = kmalloc(sizeof(unsigned long) * depth, GFP_NOFS);
-+ if (!ablocks)
-+ return -ENOMEM;
-+ memset(ablocks, 0, sizeof(unsigned long) * depth);
-+
-+ /* allocate all needed blocks */
-+ ext_debug(inode, "allocate %d blocks for indexes/leaf\n", depth - at);
-+ for (a = 0; a < depth - at; a++) {
-+ newblock = ext3_ext_new_block(handle, inode, path, newext, &err);
-+ if (newblock == 0)
-+ goto cleanup;
-+ ablocks[a] = newblock;
-+ }
-+
-+ /* initialize new leaf */
-+ newblock = ablocks[--a];
-+ BUG_ON(newblock == 0);
-+ bh = sb_getblk(inode->i_sb, newblock);
-+ if (!bh) {
-+ err = -EIO;
-+ goto cleanup;
-+ }
-+ lock_buffer(bh);
-+
-+ if ((err = ext3_journal_get_create_access(handle, bh)))
-+ goto cleanup;
-+
-+ neh = ext_block_hdr(bh);
-+ neh->eh_entries = 0;
-+ neh->eh_max = cpu_to_le16(ext3_ext_space_block(inode));
-+ neh->eh_magic = cpu_to_le16(EXT3_EXT_MAGIC);
-+ neh->eh_depth = 0;
-+ ex = EXT_FIRST_EXTENT(neh);
-+
-+ /* move remain of path[depth] to the new leaf */
-+ BUG_ON(path[depth].p_hdr->eh_entries != path[depth].p_hdr->eh_max);
-+ /* 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)) {
-+ ext_debug(inode, "move %d:%d:%d in new leaf %lu\n",
-+ le32_to_cpu(path[depth].p_ext->ee_block),
-+ le32_to_cpu(path[depth].p_ext->ee_start),
-+ le16_to_cpu(path[depth].p_ext->ee_len),
-+ newblock);
-+ /*memmove(ex++, path[depth].p_ext++,
-+ sizeof(struct ext3_extent));
-+ neh->eh_entries++;*/
-+ path[depth].p_ext++;
-+ m++;
-+ }
-+ if (m) {
-+ memmove(ex, path[depth].p_ext-m, sizeof(struct ext3_extent)*m);
-+ neh->eh_entries = cpu_to_le16(le16_to_cpu(neh->eh_entries)+m);
-+ }
-+
-+ set_buffer_uptodate(bh);
-+ unlock_buffer(bh);
-+
-+ if ((err = ext3_journal_dirty_metadata(handle, bh)))
-+ goto cleanup;
-+ brelse(bh);
-+ bh = NULL;
-+
-+ /* correct old leaf */
-+ if (m) {
-+ if ((err = ext3_ext_get_access(handle, inode, path + depth)))
-+ goto cleanup;
-+ path[depth].p_hdr->eh_entries =
-+ cpu_to_le16(le16_to_cpu(path[depth].p_hdr->eh_entries)-m);
-+ if ((err = ext3_ext_dirty(handle, inode, path + depth)))
-+ goto cleanup;
-+
-+ }
-+
-+ /* create intermediate indexes */
-+ k = depth - at - 1;
-+ BUG_ON(k < 0);
-+ if (k)
-+ ext_debug(inode, "create %d intermediate indices\n", k);
-+ /* insert new index into current index block */
-+ /* current depth stored in i var */
-+ i = depth - 1;
-+ while (k--) {
-+ oldblock = newblock;
-+ newblock = ablocks[--a];
-+ bh = sb_getblk(inode->i_sb, newblock);
-+ if (!bh) {
-+ err = -EIO;
-+ goto cleanup;
-+ }
-+ lock_buffer(bh);
-+
-+ if ((err = ext3_journal_get_create_access(handle, bh)))
-+ goto cleanup;
-+
-+ neh = ext_block_hdr(bh);
-+ neh->eh_entries = cpu_to_le16(1);
-+ neh->eh_magic = cpu_to_le16(EXT3_EXT_MAGIC);
-+ neh->eh_max = cpu_to_le16(ext3_ext_space_block_idx(inode));
-+ neh->eh_depth = cpu_to_le16(depth - i);
-+ fidx = EXT_FIRST_INDEX(neh);
-+ fidx->ei_block = border;
-+ fidx->ei_leaf = cpu_to_le32(oldblock);
-+ fidx->ei_leaf_hi = fidx->ei_unused = 0;
-+
-+ ext_debug(inode, "int.index at %d (block %lu): %lu -> %lu\n", i,
-+ newblock, (unsigned long) le32_to_cpu(border),
-+ oldblock);
-+ /* copy indexes */
-+ m = 0;
-+ path[i].p_idx++;
-+
-+ ext_debug(inode, "cur 0x%p, last 0x%p\n", path[i].p_idx,
-+ EXT_MAX_INDEX(path[i].p_hdr));
-+ BUG_ON(EXT_MAX_INDEX(path[i].p_hdr) !=
-+ EXT_LAST_INDEX(path[i].p_hdr));
-+ while (path[i].p_idx <= EXT_MAX_INDEX(path[i].p_hdr)) {
-+ ext_debug(inode, "%d: move %d:%d in new index %lu\n", i,
-+ le32_to_cpu(path[i].p_idx->ei_block),
-+ le32_to_cpu(path[i].p_idx->ei_leaf),
-+ newblock);
-+ /*memmove(++fidx, path[i].p_idx++,
-+ sizeof(struct ext3_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 ext3_extent_idx) * m);
-+ neh->eh_entries =
-+ cpu_to_le16(le16_to_cpu(neh->eh_entries) + m);
-+ }
-+ set_buffer_uptodate(bh);
-+ unlock_buffer(bh);
-+
-+ if ((err = ext3_journal_dirty_metadata(handle, bh)))
-+ goto cleanup;
-+ brelse(bh);
-+ bh = NULL;
-+
-+ /* correct old index */
-+ if (m) {
-+ err = ext3_ext_get_access(handle, inode, path + i);
-+ if (err)
-+ goto cleanup;
-+ path[i].p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path[i].p_hdr->eh_entries)-m);
-+ err = ext3_ext_dirty(handle, inode, path + i);
-+ if (err)
-+ goto cleanup;
-+ }
-+
-+ i--;
-+ }
-+
-+ /* insert new index */
-+ if (err)
-+ goto cleanup;
-+
-+ err = ext3_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;
-+ ext3_free_blocks(handle, inode, ablocks[i], 1);
-+ }
-+ }
-+ kfree(ablocks);
-+
-+ return err;
-+}
-+
-+/*
-+ * routine implements tree growing procedure:
-+ * - allocates new block
-+ * - moves top-level data (index block or leaf) into the new block
-+ * - initialize new top-level, creating index that points to the
-+ * just created block
-+ */
-+static int ext3_ext_grow_indepth(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path,
-+ struct ext3_extent *newext)
-+{
-+ struct ext3_ext_path *curp = path;
-+ struct ext3_extent_header *neh;
-+ struct ext3_extent_idx *fidx;
-+ struct buffer_head *bh;
-+ unsigned long newblock;
-+ int err = 0;
-+
-+ newblock = ext3_ext_new_block(handle, inode, path, newext, &err);
-+ if (newblock == 0)
-+ return err;
-+
-+ bh = sb_getblk(inode->i_sb, newblock);
-+ if (!bh) {
-+ err = -EIO;
-+ ext3_std_error(inode->i_sb, err);
-+ return err;
-+ }
-+ lock_buffer(bh);
-+
-+ if ((err = ext3_journal_get_create_access(handle, bh))) {
-+ unlock_buffer(bh);
-+ goto out;
-+ }
-+
-+ /* move top-level index/leaf into new block */
-+ memmove(bh->b_data, curp->p_hdr, sizeof(EXT3_I(inode)->i_data));
-+
-+ /* set size of new block */
-+ neh = ext_block_hdr(bh);
-+ /* old root could have indexes or leaves
-+ * so calculate e_max right way */
-+ if (ext_depth(inode))
-+ neh->eh_max = cpu_to_le16(ext3_ext_space_block_idx(inode));
-+ else
-+ neh->eh_max = cpu_to_le16(ext3_ext_space_block(inode));
-+ neh->eh_magic = cpu_to_le16(EXT3_EXT_MAGIC);
-+ set_buffer_uptodate(bh);
-+ unlock_buffer(bh);
-+
-+ if ((err = ext3_journal_dirty_metadata(handle, bh)))
-+ goto out;
-+
-+ /* create index in new top-level index: num,max,pointer */
-+ if ((err = ext3_ext_get_access(handle, inode, curp)))
-+ goto out;
-+
-+ curp->p_hdr->eh_magic = cpu_to_le16(EXT3_EXT_MAGIC);
-+ curp->p_hdr->eh_max = cpu_to_le16(ext3_ext_space_root_idx(inode));
-+ curp->p_hdr->eh_entries = cpu_to_le16(1);
-+ curp->p_idx = EXT_FIRST_INDEX(curp->p_hdr);
-+ /* FIXME: it works, but actually path[0] can be index */
-+ curp->p_idx->ei_block = EXT_FIRST_EXTENT(path[0].p_hdr)->ee_block;
-+ curp->p_idx->ei_leaf = cpu_to_le32(newblock);
-+ curp->p_idx->ei_leaf_hi = curp->p_idx->ei_unused = 0;
-+
-+ neh = ext_inode_hdr(inode);
-+ fidx = EXT_FIRST_INDEX(neh);
-+ ext_debug(inode, "new root: num %d(%d), lblock %d, ptr %d\n",
-+ le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
-+ le32_to_cpu(fidx->ei_block), le32_to_cpu(fidx->ei_leaf));
-+
-+ neh->eh_depth = cpu_to_le16(path->p_depth + 1);
-+ err = ext3_ext_dirty(handle, inode, curp);
-+out:
-+ brelse(bh);
-+
-+ return err;
-+}
-+
-+/*
-+ * routine finds empty index and adds new leaf. if no free index found
-+ * then it requests in-depth growing
-+ */
-+static int ext3_ext_create_new_leaf(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path,
-+ struct ext3_extent *newext)
-+{
-+ struct ext3_ext_path *curp;
-+ int depth, i, err = 0;
-+
-+repeat:
-+ i = depth = ext_depth(inode);
-+
-+ /* walk up to the tree and look for free index entry */
-+ curp = path + depth;
-+ while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) {
-+ i--;
-+ curp--;
-+ }
-+
-+ /* we use already allocated block for index block
-+ * so, subsequent data blocks should be contigoues */
-+ 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 = ext3_ext_split(handle, inode, path, newext, i);
-+ if (err)
-+ goto out;
-+
-+ /* refill path */
-+ ext3_ext_drop_refs(path);
-+ path = ext3_ext_find_extent(inode,
-+ le32_to_cpu(newext->ee_block),
-+ path);
-+ if (IS_ERR(path))
-+ err = PTR_ERR(path);
-+ } else {
-+ /* tree is full, time to grow in depth */
-+ err = ext3_ext_grow_indepth(handle, inode, path, newext);
-+ if (err)
-+ goto out;
-+
-+ /* refill path */
-+ ext3_ext_drop_refs(path);
-+ path = ext3_ext_find_extent(inode,
-+ le32_to_cpu(newext->ee_block),
-+ path);
-+ if (IS_ERR(path)) {
-+ err = PTR_ERR(path);
-+ goto out;
-+ }
-+
-+ /*
-+ * only first (depth 0 -> 1) produces free space
-+ * in all other cases we have to split growed tree
-+ */
-+ depth = ext_depth(inode);
-+ if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) {
-+ /* now we need split */
-+ goto repeat;
-+ }
-+ }
-+
-+out:
-+ return err;
-+}
-+
-+/*
-+ * search the closest allocated block to the left for *logical
-+ * and returns it at @logical + it's physical address at @phys
-+ * if *logical is the smallest allocated block, the function
-+ * returns 0 at @phys
-+ * return value contains 0 (success) or error code
-+ */
-+int
-+ext3_ext_search_left(struct inode *inode, struct ext3_ext_path *path,
-+ unsigned long *logical, unsigned long *phys)
-+{
-+ struct ext3_extent_idx *ix;
-+ struct ext3_extent *ex;
-+ int depth;
-+
-+ BUG_ON(path == NULL);
-+ 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;
-+ if (*logical < le32_to_cpu(ex->ee_block)) {
-+ BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
-+ while (--depth >= 0) {
-+ ix = path[depth].p_idx;
-+ BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
-+ }
-+ return 0;
-+ }
-+
-+ BUG_ON(*logical < le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len));
-+
-+ *logical = le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - 1;
-+ *phys = le32_to_cpu(ex->ee_start) + le16_to_cpu(ex->ee_len) - 1;
-+ return 0;
-+}
-+EXPORT_SYMBOL(ext3_ext_search_left);
-+
-+/*
-+ * 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
-+ */
-+int
-+ext3_ext_search_right(struct inode *inode, struct ext3_ext_path *path,
-+ unsigned long *logical, unsigned long *phys)
-+{
-+ struct buffer_head *bh = NULL;
-+ struct ext3_extent_header *eh;
-+ struct ext3_extent_idx *ix = NULL;
-+ struct ext3_extent *ex;
-+ unsigned long block;
-+ int depth;
-+
-+ BUG_ON(path == NULL);
-+ 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;
-+ if (*logical < le32_to_cpu(ex->ee_block)) {
-+ BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
-+ while (--depth >= 0) {
-+ ix = path[depth].p_idx;
-+ BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
-+ }
-+ *logical = le32_to_cpu(ex->ee_block);
-+ *phys = le32_to_cpu(ex->ee_start);
-+ return 0;
-+ }
-+
-+ BUG_ON(*logical < le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len));
-+
-+ if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
-+ /* next allocated block in this leaf */
-+ ex++;
-+ *logical = le32_to_cpu(ex->ee_block);
-+ *phys = le32_to_cpu(ex->ee_start);
-+ 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))
-+ break;
-+ }
-+
-+ if (depth < 0) {
-+ /* we've gone up to the root and
-+ * found no index to the right */
-+ return 0;
-+ }
-+
-+ /* we've found index to the right, let's
-+ * follow it and find the closest allocated
-+ * block to the right */
-+ ix++;
-+ block = le32_to_cpu(ix->ei_leaf);
-+ while (++depth < path->p_depth) {
-+ bh = sb_bread(inode->i_sb, block);
-+ if (bh == NULL)
-+ return -EIO;
-+ eh = ext_block_hdr(bh);
-+ if (ext3_ext_check_header(inode, eh, path->p_depth - depth)) {
-+ brelse(bh);
-+ return -EIO;
-+ }
-+ ix = EXT_FIRST_INDEX(eh);
-+ block = le32_to_cpu(ix->ei_leaf);
-+ brelse(bh);
-+ }
-+
-+ bh = sb_bread(inode->i_sb, block);
-+ if (bh == NULL)
-+ return -EIO;
-+ eh = ext_block_hdr(bh);
-+ if (ext3_ext_check_header(inode, eh, 0)) {
-+ brelse(bh);
-+ return -EIO;
-+ }
-+ ex = EXT_FIRST_EXTENT(eh);
-+ *logical = le32_to_cpu(ex->ee_block);
-+ *phys = le32_to_cpu(ex->ee_start);
-+ brelse(bh);
-+ return 0;
-+
-+}
-+EXPORT_SYMBOL(ext3_ext_search_right);
-+
-+
-+
-+/*
-+ * returns allocated block in subsequent extent or EXT_MAX_BLOCK
-+ * NOTE: it consider block number from index entry as
-+ * allocated block. thus, index entries have to be consistent
-+ * with leafs
-+ */
-+static unsigned long
-+ext3_ext_next_allocated_block(struct ext3_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;
-+}
-+
-+/*
-+ * returns first allocated block from next leaf or EXT_UNSET_BLOCK
-+ */
-+static unsigned ext3_ext_next_leaf_block(struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ int depth;
-+
-+ BUG_ON(path == NULL);
-+ depth = path->p_depth;
-+
-+ /* zero-tree has no leaf blocks at all */
-+ if (depth == 0)
-+ return EXT_UNSET_BLOCK;
-+
-+ /* go to index block */
-+ depth--;
-+
-+ while (depth >= 0) {
-+ 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_UNSET_BLOCK;
-+}
-+
-+/*
-+ * if leaf gets modified and modified extent is first in the leaf
-+ * then we have to correct all indexes above
-+ * TODO: do we need to correct tree in all cases?
-+ */
-+int ext3_ext_correct_indexes(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ struct ext3_extent_header *eh;
-+ int depth = ext_depth(inode);
-+ struct ext3_extent *ex;
-+ __le32 border;
-+ int k, err = 0;
-+
-+ eh = path[depth].p_hdr;
-+ ex = path[depth].p_ext;
-+ BUG_ON(ex == NULL);
-+ BUG_ON(eh == NULL);
-+
-+ if (depth == 0) {
-+ /* there is no tree at all */
-+ return 0;
-+ }
-+
-+ if (ex != EXT_FIRST_EXTENT(eh)) {
-+ /* we correct tree if first leaf got modified only */
-+ return 0;
-+ }
-+
-+ /*
-+ * TODO: we need correction if border is smaller then current one
-+ */
-+ k = depth - 1;
-+ border = path[depth].p_ext->ee_block;
-+ if ((err = ext3_ext_get_access(handle, inode, path + k)))
-+ return err;
-+ path[k].p_idx->ei_block = border;
-+ if ((err = ext3_ext_dirty(handle, inode, path + k)))
-+ return err;
-+
-+ while (k--) {
-+ /* change all left-side indexes */
-+ if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr))
-+ break;
-+ if ((err = ext3_ext_get_access(handle, inode, path + k)))
-+ break;
-+ path[k].p_idx->ei_block = border;
-+ if ((err = ext3_ext_dirty(handle, inode, path + k)))
-+ break;
-+ }
-+
-+ return err;
-+}
-+
-+static int inline
-+ext3_can_extents_be_merged(struct inode *inode, struct ext3_extent *ex1,
-+ struct ext3_extent *ex2)
-+{
-+ /* FIXME: 48bit support */
-+ if (le32_to_cpu(ex1->ee_block) + le16_to_cpu(ex1->ee_len) !=
-+ le32_to_cpu(ex2->ee_block))
-+ return 0;
-+
-+#ifdef AGRESSIVE_TEST
-+ if (le16_to_cpu(ex1->ee_len) >= 4)
-+ return 0;
-+#endif
-+
-+ if (le32_to_cpu(ex1->ee_start) + le16_to_cpu(ex1->ee_len) ==
-+ le32_to_cpu(ex2->ee_start))
-+ return 1;
-+ return 0;
-+}
-+
-+/*
-+ * this routine tries to merge requsted extent into the existing
-+ * extent or inserts requested extent as new one into the tree,
-+ * creating new leaf in no-space case
-+ */
-+int ext3_ext_insert_extent(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path,
-+ struct ext3_extent *newext)
-+{
-+ struct ext3_extent_header * eh;
-+ struct ext3_extent *ex, *fex;
-+ struct ext3_extent *nearex; /* nearest extent */
-+ struct ext3_ext_path *npath = NULL;
-+ int depth, len, err, next;
-+
-+ BUG_ON(newext->ee_len == 0);
-+ depth = ext_depth(inode);
-+ ex = path[depth].p_ext;
-+ BUG_ON(path[depth].p_hdr == NULL);
-+
-+ /* try to insert block into found extent and return */
-+ if (ex && ext3_can_extents_be_merged(inode, ex, newext)) {
-+ ext_debug(inode, "append %d block to %d:%d (from %d)\n",
-+ le16_to_cpu(newext->ee_len),
-+ le32_to_cpu(ex->ee_block),
-+ le16_to_cpu(ex->ee_len),
-+ le32_to_cpu(ex->ee_start));
-+ if ((err = ext3_ext_get_access(handle, inode, path + depth)))
-+ return err;
-+ ex->ee_len = cpu_to_le16(le16_to_cpu(ex->ee_len)
-+ + le16_to_cpu(newext->ee_len));
-+ 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 = ext3_ext_next_leaf_block(inode, path);
-+ if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)
-+ && next != EXT_UNSET_BLOCK) {
-+ ext_debug(inode, "next leaf block - %d\n", next);
-+ BUG_ON(npath != NULL);
-+ npath = ext3_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(inode, "next leaf isnt full(%d)\n",
-+ le16_to_cpu(eh->eh_entries));
-+ path = npath;
-+ goto repeat;
-+ }
-+ ext_debug(inode, "next leaf has no free space(%d,%d)\n",
-+ le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
-+ }
-+
-+ /*
-+ * there is no free space in found leaf
-+ * we're gonna add new leaf in the tree
-+ */
-+ err = ext3_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;
-+
-+ if ((err = ext3_ext_get_access(handle, inode, path + depth)))
-+ goto cleanup;
-+
-+ if (!nearex) {
-+ /* there is no extent in this leaf, create first one */
-+ ext_debug(inode, "first extent in the leaf: %d:%d:%d\n",
-+ le32_to_cpu(newext->ee_block),
-+ le32_to_cpu(newext->ee_start),
-+ le16_to_cpu(newext->ee_len));
-+ path[depth].p_ext = EXT_FIRST_EXTENT(eh);
-+ } else if (le32_to_cpu(newext->ee_block)
-+ > le32_to_cpu(nearex->ee_block)) {
-+ /* 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 ext3_extent);
-+ len = len < 0 ? 0 : len;
-+ ext_debug(inode, "insert %d:%d:%d after: nearest 0x%p, "
-+ "move %d from 0x%p to 0x%p\n",
-+ le32_to_cpu(newext->ee_block),
-+ le32_to_cpu(newext->ee_start),
-+ le16_to_cpu(newext->ee_len),
-+ 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 ext3_extent);
-+ len = len < 0 ? 0 : len;
-+ ext_debug(inode, "insert %d:%d:%d before: nearest 0x%p, "
-+ "move %d from 0x%p to 0x%p\n",
-+ le32_to_cpu(newext->ee_block),
-+ le32_to_cpu(newext->ee_start),
-+ le16_to_cpu(newext->ee_len),
-+ nearex, len, nearex + 1, nearex + 2);
-+ memmove(nearex + 1, nearex, len);
-+ path[depth].p_ext = nearex;
-+ }
-+
-+ eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)+1);
-+ nearex = path[depth].p_ext;
-+ nearex->ee_block = newext->ee_block;
-+ nearex->ee_start = newext->ee_start;
-+ nearex->ee_len = newext->ee_len;
-+ /* FIXME: support for large fs */
-+ nearex->ee_start_hi = 0;
-+
-+merge:
-+ /* try to merge extents to the right */
-+ while (nearex < EXT_LAST_EXTENT(eh)) {
-+ if (!ext3_can_extents_be_merged(inode, nearex, nearex + 1))
-+ break;
-+ /* merge with next extent! */
-+ nearex->ee_len = cpu_to_le16(le16_to_cpu(nearex->ee_len)
-+ + le16_to_cpu(nearex[1].ee_len));
-+ if (nearex + 1 < EXT_LAST_EXTENT(eh)) {
-+ len = (EXT_LAST_EXTENT(eh) - nearex - 1)
-+ * sizeof(struct ext3_extent);
-+ memmove(nearex + 1, nearex + 2, len);
-+ }
-+ eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)-1);
-+ BUG_ON(eh->eh_entries == 0);
-+ }
-+
-+ /* try to merge extents to the left */
-+
-+ /* time to correct all indexes above */
-+ err = ext3_ext_correct_indexes(handle, inode, path);
-+ if (err)
-+ goto cleanup;
-+
-+ err = ext3_ext_dirty(handle, inode, path + depth);
-+
-+cleanup:
-+ if (npath) {
-+ ext3_ext_drop_refs(npath);
-+ kfree(npath);
-+ }
-+ ext3_ext_tree_changed(inode);
-+ ext3_ext_invalidate_cache(inode);
-+ return err;
-+}
-+
-+int ext3_ext_walk_space(struct inode *inode, unsigned long block,
-+ unsigned long num, ext_prepare_callback func,
-+ void *cbdata)
-+{
-+ struct ext3_ext_path *path = NULL;
-+ struct ext3_ext_cache cbex;
-+ struct ext3_extent *ex;
-+ unsigned long next, start = 0, end = 0;
-+ unsigned long 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 */
-+ path = ext3_ext_find_extent(inode, block, path);
-+ if (IS_ERR(path)) {
-+ err = PTR_ERR(path);
-+ path = NULL;
-+ break;
-+ }
-+
-+ depth = ext_depth(inode);
-+ BUG_ON(path[depth].p_hdr == NULL);
-+ ex = path[depth].p_ext;
-+ next = ext3_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) + le16_to_cpu(ex->ee_len)) {
-+ /* 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) + le16_to_cpu(ex->ee_len);
-+ 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 = EXT3_EXT_CACHE_GAP;
-+ } else {
-+ cbex.ec_block = le32_to_cpu(ex->ee_block);
-+ cbex.ec_len = le16_to_cpu(ex->ee_len);
-+ cbex.ec_start = le32_to_cpu(ex->ee_start);
-+ cbex.ec_type = EXT3_EXT_CACHE_EXTENT;
-+ }
-+
-+ BUG_ON(cbex.ec_len == 0);
-+ err = func(inode, path, &cbex, cbdata);
-+ ext3_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) {
-+ ext3_ext_drop_refs(path);
-+ kfree(path);
-+ }
-+
-+ return err;
-+}
-+
-+static inline void
-+ext3_ext_put_in_cache(struct inode *inode, __u32 block,
-+ __u32 len, __u32 start, int type)
-+{
-+ struct ext3_ext_cache *cex;
-+ BUG_ON(len == 0);
-+ cex = &EXT3_I(inode)->i_cached_extent;
-+ cex->ec_type = type;
-+ cex->ec_block = block;
-+ cex->ec_len = len;
-+ cex->ec_start = start;
-+}
-+
-+/*
-+ * this routine calculate boundaries of the gap requested block fits into
-+ * and cache this gap
-+ */
-+static inline void
-+ext3_ext_put_gap_in_cache(struct inode *inode, struct ext3_ext_path *path,
-+ unsigned long block)
-+{
-+ int depth = ext_depth(inode);
-+ unsigned long lblock, len;
-+ struct ext3_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(inode, "cache gap(whole file):");
-+ } else if (block < le32_to_cpu(ex->ee_block)) {
-+ lblock = block;
-+ len = le32_to_cpu(ex->ee_block) - block;
-+ ext_debug(inode, "cache gap(before): %lu [%lu:%lu]",
-+ (unsigned long) block,
-+ (unsigned long) le32_to_cpu(ex->ee_block),
-+ (unsigned long) le16_to_cpu(ex->ee_len));
-+ } else if (block >= le32_to_cpu(ex->ee_block)
-+ + le16_to_cpu(ex->ee_len)) {
-+ lblock = le32_to_cpu(ex->ee_block)
-+ + le16_to_cpu(ex->ee_len);
-+ len = ext3_ext_next_allocated_block(path);
-+ ext_debug(inode, "cache gap(after): [%lu:%lu] %lu",
-+ (unsigned long) le32_to_cpu(ex->ee_block),
-+ (unsigned long) le16_to_cpu(ex->ee_len),
-+ (unsigned long) block);
-+ BUG_ON(len == lblock);
-+ len = len - lblock;
-+ } else {
-+ lblock = len = 0;
-+ BUG();
-+ }
-+
-+ ext_debug(inode, " -> %lu:%lu\n", (unsigned long) lblock, len);
-+ ext3_ext_put_in_cache(inode, lblock, len, 0, EXT3_EXT_CACHE_GAP);
-+}
-+
-+static inline int
-+ext3_ext_in_cache(struct inode *inode, unsigned long block,
-+ struct ext3_extent *ex)
-+{
-+ struct ext3_ext_cache *cex;
-+
-+ cex = &EXT3_I(inode)->i_cached_extent;
-+
-+ /* has cache valid data? */
-+ if (cex->ec_type == EXT3_EXT_CACHE_NO)
-+ return EXT3_EXT_CACHE_NO;
-+
-+ BUG_ON(cex->ec_type != EXT3_EXT_CACHE_GAP &&
-+ cex->ec_type != EXT3_EXT_CACHE_EXTENT);
-+ if (block >= cex->ec_block && block < cex->ec_block + cex->ec_len) {
-+ ex->ee_block = cpu_to_le32(cex->ec_block);
-+ ex->ee_start = cpu_to_le32(cex->ec_start);
-+ ex->ee_start_hi = 0;
-+ ex->ee_len = cpu_to_le16(cex->ec_len);
-+ ext_debug(inode, "%lu cached by %lu:%lu:%lu\n",
-+ (unsigned long) block,
-+ (unsigned long) cex->ec_block,
-+ (unsigned long) cex->ec_len,
-+ (unsigned long) cex->ec_start);
-+ return cex->ec_type;
-+ }
-+
-+ /* not in cache */
-+ return EXT3_EXT_CACHE_NO;
-+}
-+
-+/*
-+ * routine removes index from the index block
-+ * it's used in truncate case only. thus all requests are for
-+ * last index in the block only
-+ */
-+int ext3_ext_rm_idx(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ struct buffer_head *bh;
-+ int err;
-+ unsigned long leaf;
-+
-+ /* free index block */
-+ path--;
-+ leaf = le32_to_cpu(path->p_idx->ei_leaf);
-+ BUG_ON(path->p_hdr->eh_entries == 0);
-+ if ((err = ext3_ext_get_access(handle, inode, path)))
-+ return err;
-+ path->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path->p_hdr->eh_entries)-1);
-+ if ((err = ext3_ext_dirty(handle, inode, path)))
-+ return err;
-+ ext_debug(inode, "index is empty, remove it, free block %lu\n", leaf);
-+ bh = sb_find_get_block(inode->i_sb, leaf);
-+ ext3_forget(handle, 1, inode, bh, leaf);
-+ ext3_free_blocks(handle, inode, leaf, 1);
-+ return err;
-+}
-+
-+/*
-+ * This routine returns max. credits extent tree can consume.
-+ * It should be OK for low-performance paths like ->writepage()
-+ * To allow many writing process to fit a single transaction,
-+ * caller should calculate credits under truncate_mutex and
-+ * pass actual path.
-+ */
-+int inline ext3_ext_calc_credits_for_insert(struct inode *inode,
-+ struct ext3_ext_path *path)
-+{
-+ int depth, needed;
-+
-+ if (path) {
-+ /* probably there is space in leaf? */
-+ depth = ext_depth(inode);
-+ if (le16_to_cpu(path[depth].p_hdr->eh_entries)
-+ < le16_to_cpu(path[depth].p_hdr->eh_max))
-+ return 1;
-+ }
-+
-+ /*
-+ * given 32bit logical block (4294967296 blocks), max. tree
-+ * can be 4 levels in depth -- 4 * 340^4 == 53453440000.
-+ * let's also add one more level for imbalance.
-+ */
-+ depth = 5;
-+
-+ /* allocation of new data block(s) */
-+ needed = 2;
-+
-+ /*
-+ * tree can be full, so it'd need to grow in depth:
-+ * we need one credit to modify old root, credits for
-+ * new root will be added in split accounting
-+ */
-+ needed += 1;
-+
-+ /*
-+ * Index split can happen, we'd need:
-+ * allocate intermediate indexes (bitmap + group)
-+ * + change two blocks at each level, but root (already included)
-+ */
-+ needed += (depth * 2) + (depth * 2);
-+
-+ /* any allocation modifies superblock */
-+ needed += 1;
-+
-+ return needed;
-+}
-+
-+static int ext3_remove_blocks(handle_t *handle, struct inode *inode,
-+ struct ext3_extent *ex,
-+ unsigned long from, unsigned long to)
-+{
-+ struct buffer_head *bh;
-+ int i;
-+
-+#ifdef EXTENTS_STATS
-+ {
-+ struct ext3_sb_info *sbi = EXT3_SB(inode->i_sb);
-+ unsigned short ee_len = le16_to_cpu(ex->ee_len);
-+ spin_lock(&sbi->s_ext_stats_lock);
-+ sbi->s_ext_blocks += ee_len;
-+ sbi->s_ext_extents++;
-+ if (ee_len < sbi->s_ext_min)
-+ sbi->s_ext_min = ee_len;
-+ if (ee_len > sbi->s_ext_max)
-+ sbi->s_ext_max = ee_len;
-+ if (ext_depth(inode) > sbi->s_depth_max)
-+ sbi->s_depth_max = ext_depth(inode);
-+ spin_unlock(&sbi->s_ext_stats_lock);
-+ }
-+#endif
-+ if (from >= le32_to_cpu(ex->ee_block)
-+ && to == le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - 1) {
-+ /* tail removal */
-+ unsigned long num, start;
-+ num = le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - from;
-+ start = le32_to_cpu(ex->ee_start) + le16_to_cpu(ex->ee_len) - num;
-+ ext_debug(inode, "free last %lu blocks starting %lu\n", num, start);
-+ for (i = 0; i < num; i++) {
-+ bh = sb_find_get_block(inode->i_sb, start + i);
-+ ext3_forget(handle, 0, inode, bh, start + i);
-+ }
-+ ext3_free_blocks(handle, inode, start, num);
-+ } else if (from == le32_to_cpu(ex->ee_block)
-+ && to <= le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - 1) {
-+ printk("strange request: removal %lu-%lu from %u:%u\n",
-+ from, to, le32_to_cpu(ex->ee_block), le16_to_cpu(ex->ee_len));
-+ } else {
-+ printk("strange request: removal(2) %lu-%lu from %u:%u\n",
-+ from, to, le32_to_cpu(ex->ee_block), le16_to_cpu(ex->ee_len));
-+ }
-+ return 0;
-+}
-+
-+static int
-+ext3_ext_rm_leaf(handle_t *handle, struct inode *inode,
-+ struct ext3_ext_path *path, unsigned long start)
-+{
-+ int err = 0, correct_index = 0;
-+ int depth = ext_depth(inode), credits;
-+ struct ext3_extent_header *eh;
-+ unsigned a, b, block, num;
-+ unsigned long ex_ee_block;
-+ unsigned short ex_ee_len;
-+ struct ext3_extent *ex;
-+
-+ /* the header must be checked already in ext3_ext_remove_space() */
-+ ext_debug(inode, "truncate since %lu in leaf\n", start);
-+ if (!path[depth].p_hdr)
-+ path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
-+ eh = path[depth].p_hdr;
-+ BUG_ON(eh == NULL);
-+
-+ /* find where to start removing */
-+ ex = EXT_LAST_EXTENT(eh);
-+
-+ ex_ee_block = le32_to_cpu(ex->ee_block);
-+ ex_ee_len = le16_to_cpu(ex->ee_len);
-+
-+ while (ex >= EXT_FIRST_EXTENT(eh) &&
-+ ex_ee_block + ex_ee_len > start) {
-+ ext_debug(inode, "remove ext %lu:%u\n", ex_ee_block, ex_ee_len);
-+ path[depth].p_ext = ex;
-+
-+ a = ex_ee_block > start ? ex_ee_block : start;
-+ b = (unsigned long long)ex_ee_block + ex_ee_len - 1 <
-+ EXT_MAX_BLOCK ? ex_ee_block + ex_ee_len - 1 : EXT_MAX_BLOCK;
-+
-+ ext_debug(inode, " 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);
-+ }
-+
-+ /* at present, extent can't cross block group */
-+ /* leaf + bitmap + group desc + sb + inode */
-+ credits = 5;
-+ if (ex == EXT_FIRST_EXTENT(eh)) {
-+ correct_index = 1;
-+ credits += (ext_depth(inode)) + 1;
-+ }
-+#ifdef CONFIG_QUOTA
-+ credits += 2 * EXT3_QUOTA_TRANS_BLOCKS(inode->i_sb);
-+#endif
-+
-+ handle = ext3_ext_journal_restart(handle, credits);
-+ if (IS_ERR(handle)) {
-+ err = PTR_ERR(handle);
-+ goto out;
-+ }
-+
-+ err = ext3_ext_get_access(handle, inode, path + depth);
-+ if (err)
-+ goto out;
-+
-+ err = ext3_remove_blocks(handle, inode, ex, a, b);
-+ if (err)
-+ goto out;
-+
-+ if (num == 0) {
-+ /* this extent is removed entirely mark slot unused */
-+ ex->ee_start = ex->ee_start_hi = 0;
-+ eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)-1);
-+ }
-+
-+ ex->ee_block = cpu_to_le32(block);
-+ ex->ee_len = cpu_to_le16(num);
-+
-+ err = ext3_ext_dirty(handle, inode, path + depth);
-+ if (err)
-+ goto out;
-+
-+ ext_debug(inode, "new extent: %u:%u:%u\n", block, num,
-+ le32_to_cpu(ex->ee_start));
-+ ex--;
-+ ex_ee_block = le32_to_cpu(ex->ee_block);
-+ ex_ee_len = le16_to_cpu(ex->ee_len);
-+ }
-+
-+ if (correct_index && eh->eh_entries)
-+ err = ext3_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 = ext3_ext_rm_idx(handle, inode, path + depth);
-+
-+out:
-+ return err;
-+}
-+
-+/*
-+ * returns 1 if current index have to be freed (even partial)
-+ */
-+static int inline
-+ext3_ext_more_to_rm(struct ext3_ext_path *path)
-+{
-+ BUG_ON(path->p_idx == NULL);
-+
-+ if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr))
-+ return 0;
-+
-+ /*
-+ * if truncate on deeper level happened it it wasn't partial
-+ * so we have to consider current index for truncation
-+ */
-+ if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block)
-+ return 0;
-+ return 1;
-+}
-+
-+int ext3_ext_remove_space(struct inode *inode, unsigned long start)
-+{
-+ struct super_block *sb = inode->i_sb;
-+ int depth = ext_depth(inode);
-+ struct ext3_ext_path *path;
-+ handle_t *handle;
-+ int i = 0, err = 0;
-+
-+ ext_debug(inode, "truncate since %lu\n", start);
-+
-+ /* probably first extent we're gonna free will be last in block */
-+ handle = ext3_journal_start(inode, depth + 1);
-+ if (IS_ERR(handle))
-+ return PTR_ERR(handle);
-+
-+ ext3_ext_invalidate_cache(inode);
-+
-+ /*
-+ * we start scanning from right side freeing all the blocks
-+ * after i_size and walking into the deep
-+ */
-+ path = kmalloc(sizeof(struct ext3_ext_path) * (depth + 1), GFP_KERNEL);
-+ if (path == NULL) {
-+ ext3_journal_stop(handle);
-+ return -ENOMEM;
-+ }
-+ memset(path, 0, sizeof(struct ext3_ext_path) * (depth + 1));
-+ path[0].p_hdr = ext_inode_hdr(inode);
-+ if (ext3_ext_check_header(inode, path[0].p_hdr, depth)) {
-+ err = -EIO;
-+ goto out;
-+ }
-+ path[0].p_depth = depth;
-+
-+ while (i >= 0 && err == 0) {
-+ if (i == depth) {
-+ /* this is leaf block */
-+ err = ext3_ext_rm_leaf(handle, inode, path, start);
-+ /* root level have p_bh == NULL, brelse() eats this */
-+ brelse(path[i].p_bh);
-+ path[i].p_bh = NULL;
-+ i--;
-+ continue;
-+ }
-+
-+ /* this is index block */
-+ if (!path[i].p_hdr) {
-+ ext_debug(inode, "initialize header\n");
-+ path[i].p_hdr = ext_block_hdr(path[i].p_bh);
-+ }
-+
-+ if (!path[i].p_idx) {
-+ /* this level hasn't touched yet */
-+ path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
-+ path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
-+ ext_debug(inode, "init index ptr: hdr 0x%p, num %d\n",
-+ path[i].p_hdr,
-+ le16_to_cpu(path[i].p_hdr->eh_entries));
-+ } else {
-+ /* we've already was here, see at next index */
-+ path[i].p_idx--;
-+ }
-+
-+ ext_debug(inode, "level %d - index, first 0x%p, cur 0x%p\n",
-+ i, EXT_FIRST_INDEX(path[i].p_hdr),
-+ path[i].p_idx);
-+ if (ext3_ext_more_to_rm(path + i)) {
-+ struct buffer_head *bh;
-+ /* go to the next level */
-+ ext_debug(inode, "move to level %d (block %d)\n",
-+ i + 1, le32_to_cpu(path[i].p_idx->ei_leaf));
-+ memset(path + i + 1, 0, sizeof(*path));
-+ bh = sb_bread(sb, le32_to_cpu(path[i].p_idx->ei_leaf));
-+ if (!bh) {
-+ /* should we reset i_size? */
-+ err = -EIO;
-+ break;
-+ }
-+ BUG_ON(i + 1 > depth);
-+ if (ext3_ext_check_header(inode, ext_block_hdr(bh),
-+ depth - i - 1)) {
-+ err = -EIO;
-+ break;
-+ }
-+ path[i+1].p_bh = bh;
-+
-+ /* put actual number of indexes to know is this
-+ * number got changed at the next iteration */
-+ path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries);
-+ i++;
-+ } else {
-+ /* we finish processing this index, go up */
-+ if (path[i].p_hdr->eh_entries == 0 && i > 0) {
-+ /* index is empty, remove it
-+ * handle must be already prepared by the
-+ * truncatei_leaf() */
-+ err = ext3_ext_rm_idx(handle, inode, path + i);
-+ }
-+ /* root level have p_bh == NULL, brelse() eats this */
-+ brelse(path[i].p_bh);
-+ path[i].p_bh = NULL;
-+ i--;
-+ ext_debug(inode, "return to level %d\n", i);
-+ }
-+ }
-+
-+ /* TODO: flexible tree reduction should be here */
-+ if (path->p_hdr->eh_entries == 0) {
-+ /*
-+ * truncate to zero freed all the tree
-+ * so, we need to correct eh_depth
-+ */
-+ err = ext3_ext_get_access(handle, inode, path);
-+ if (err == 0) {
-+ ext_inode_hdr(inode)->eh_depth = 0;
-+ ext_inode_hdr(inode)->eh_max =
-+ cpu_to_le16(ext3_ext_space_root(inode));
-+ err = ext3_ext_dirty(handle, inode, path);
-+ }
-+ }
-+out:
-+ ext3_ext_tree_changed(inode);
-+ ext3_ext_drop_refs(path);
-+ kfree(path);
-+ ext3_journal_stop(handle);
-+
-+ return err;
-+}
-+
-+/*
-+ * called at mount time
-+ */
-+void ext3_ext_init(struct super_block *sb)
-+{
-+ /*
-+ * possible initialization would be here
-+ */
-+
-+ if (test_opt(sb, EXTENTS)) {
-+ printk("EXT3-fs: file extents enabled");
-+#ifdef AGRESSIVE_TEST
-+ printk(", agressive tests");
-+#endif
-+#ifdef CHECK_BINSEARCH
-+ printk(", check binsearch");
-+#endif
-+#ifdef EXTENTS_STATS
-+ printk(", stats");
-+#endif
-+ printk("\n");
-+#ifdef EXTENTS_STATS
-+ spin_lock_init(&EXT3_SB(sb)->s_ext_stats_lock);
-+ EXT3_SB(sb)->s_ext_min = 1 << 30;
-+ EXT3_SB(sb)->s_ext_max = 0;
-+#endif
-+ }
-+}
-+
-+/*
-+ * called at umount time
-+ */
-+void ext3_ext_release(struct super_block *sb)
-+{
-+ if (!test_opt(sb, EXTENTS))
-+ return;
-+
-+#ifdef EXTENTS_STATS
-+ if (EXT3_SB(sb)->s_ext_blocks && EXT3_SB(sb)->s_ext_extents) {
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ printk(KERN_ERR "EXT3-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 "EXT3-fs: extents: %lu min, %lu max, max depth %lu\n",
-+ sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max);
-+ }
-+#endif
-+}
-+
-+int ext3_ext_get_blocks(handle_t *handle, struct inode *inode, sector_t iblock,
-+ unsigned long max_blocks, struct buffer_head *bh_result,
-+ int create, int extend_disksize)
-+{
-+ struct ext3_ext_path *path = NULL;
-+ struct ext3_extent newex, *ex;
-+ int goal, newblock, err = 0, depth;
-+ unsigned long allocated = 0;
-+ unsigned long next;
-+
-+ __clear_bit(BH_New, &bh_result->b_state);
-+ ext_debug(inode, "blocks %d/%lu requested for inode %u\n", (int) iblock,
-+ max_blocks, (unsigned) inode->i_ino);
-+ mutex_lock(&EXT3_I(inode)->truncate_mutex);
-+
-+ /* check in cache */
-+ if ((goal = ext3_ext_in_cache(inode, iblock, &newex))) {
-+ if (goal == EXT3_EXT_CACHE_GAP) {
-+ if (!create) {
-+ /* block isn't allocated yet and
-+ * user don't want to allocate it */
-+ goto out2;
-+ }
-+ /* we should allocate requested block */
-+ } else if (goal == EXT3_EXT_CACHE_EXTENT) {
-+ /* block is already allocated */
-+ newblock = iblock
-+ - le32_to_cpu(newex.ee_block)
-+ + le32_to_cpu(newex.ee_start);
-+ /* number of remain blocks in the extent */
-+ BUG_ON(iblock < le32_to_cpu(newex.ee_block));
-+ allocated = le16_to_cpu(newex.ee_len) -
-+ (iblock - le32_to_cpu(newex.ee_block));
-+ goto out;
-+ } else {
-+ BUG();
-+ }
-+ }
-+
-+ /* find extent for this block */
-+ path = ext3_ext_find_extent(inode, iblock, NULL);
-+ if (IS_ERR(path)) {
-+ err = PTR_ERR(path);
-+ path = NULL;
-+ goto out2;
-+ }
-+
-+ depth = ext_depth(inode);
-+
-+ /*
-+ * consistent leaf must not be empty
-+ * this situations is possible, though, _during_ tree modification
-+ * this is why assert can't be put in ext3_ext_find_extent()
-+ */
-+ BUG_ON(path[depth].p_ext == NULL && depth != 0);
-+
-+ if ((ex = path[depth].p_ext)) {
-+ unsigned long ee_block = le32_to_cpu(ex->ee_block);
-+ unsigned long ee_start = le32_to_cpu(ex->ee_start);
-+ unsigned short ee_len = le16_to_cpu(ex->ee_len);
-+ /* if found exent covers block, simple return it */
-+ if (iblock >= ee_block && iblock < ee_block + ee_len) {
-+ newblock = iblock - ee_block + ee_start;
-+ /* number of remain blocks in the extent */
-+ allocated = ee_len - (iblock - ee_block);
-+ ext_debug(inode, "%d fit into %lu:%d -> %d\n", (int) iblock,
-+ ee_block, ee_len, newblock);
-+ ext3_ext_put_in_cache(inode, ee_block, ee_len,
-+ ee_start, EXT3_EXT_CACHE_EXTENT);
-+ goto out;
-+ }
-+ }
-+
-+ /*
-+ * requested block isn't allocated yet
-+ * we couldn't try to create block if create flag is zero
-+ */
-+ if (!create) {
-+ /* put just found gap into cache to speedup subsequest reqs */
-+ ext3_ext_put_gap_in_cache(inode, path, iblock);
-+ goto out2;
-+ }
-+
-+ /*
-+ * Okay, we need to do block allocation. Lazily initialize the block
-+ * allocation info here if necessary
-+ */
-+ if (S_ISREG(inode->i_mode) && (!EXT3_I(inode)->i_block_alloc_info))
-+ ext3_init_block_alloc_info(inode);
-+
-+ /* find next allocated block so that we know how many
-+ * blocks we can allocate without ovelapping next extent */
-+ BUG_ON(iblock < le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len));
-+ next = ext3_ext_next_allocated_block(path);
-+ BUG_ON(next <= iblock);
-+ allocated = next - iblock;
-+ if (allocated > max_blocks)
-+ allocated = max_blocks;
-+
-+ /* allocate new block */
-+ goal = ext3_ext_find_goal(inode, path, iblock);
-+ newblock = ext3_new_blocks(handle, inode, goal, &allocated, &err);
-+ if (!newblock)
-+ goto out2;
-+ ext_debug(inode, "allocate new block: goal %d, found %d/%lu\n",
-+ goal, newblock, allocated);
-+
-+ /* try to insert new extent into found leaf and return */
-+ newex.ee_block = cpu_to_le32(iblock);
-+ newex.ee_start = cpu_to_le32(newblock);
-+ newex.ee_start_hi = 0;
-+ newex.ee_len = cpu_to_le16(allocated);
-+ err = ext3_ext_insert_extent(handle, inode, path, &newex);
-+ if (err) {
-+ /* free data blocks we just allocated */
-+ ext3_free_blocks(handle, inode, le32_to_cpu(newex.ee_start),
-+ le16_to_cpu(newex.ee_len));
-+ goto out2;
-+ }
-+
-+ if (extend_disksize && inode->i_size > EXT3_I(inode)->i_disksize)
-+ EXT3_I(inode)->i_disksize = inode->i_size;
-+
-+ /* previous routine could use block we allocated */
-+ newblock = le32_to_cpu(newex.ee_start);
-+ __set_bit(BH_New, &bh_result->b_state);
-+
-+ ext3_ext_put_in_cache(inode, iblock, allocated, newblock,
-+ EXT3_EXT_CACHE_EXTENT);
-+out:
-+ if (allocated > max_blocks)
-+ allocated = max_blocks;
-+ ext3_ext_show_leaf(inode, path);
-+ __set_bit(BH_Mapped, &bh_result->b_state);
-+ bh_result->b_bdev = inode->i_sb->s_bdev;
-+ bh_result->b_blocknr = newblock;
-+ bh_result->b_size = (allocated << inode->i_blkbits);
-+out2:
-+ if (path) {
-+ ext3_ext_drop_refs(path);
-+ kfree(path);
-+ }
-+ mutex_unlock(&EXT3_I(inode)->truncate_mutex);
-+
-+ return err ? err : allocated;
-+}
-+
-+void ext3_ext_truncate(struct inode * inode, struct page *page)
-+{
-+ struct address_space *mapping = inode->i_mapping;
-+ struct super_block *sb = inode->i_sb;
-+ unsigned long last_block;
-+ handle_t *handle;
-+ int err = 0;
-+
-+ /*
-+ * probably first extent we're gonna free will be last in block
-+ */
-+ err = ext3_writepage_trans_blocks(inode) + 3;
-+ handle = ext3_journal_start(inode, err);
-+ if (IS_ERR(handle)) {
-+ if (page) {
-+ clear_highpage(page);
-+ flush_dcache_page(page);
-+ unlock_page(page);
-+ page_cache_release(page);
-+ }
-+ return;
-+ }
-+
-+ if (page)
-+ ext3_block_truncate_page(handle, page, mapping, inode->i_size);
-+
-+ mutex_lock(&EXT3_I(inode)->truncate_mutex);
-+ ext3_ext_invalidate_cache(inode);
-+
-+ /*
-+ * TODO: optimization is possible here
-+ * probably we need not scaning at all,
-+ * because page truncation is enough
-+ */
-+ if (ext3_orphan_add(handle, inode))
-+ goto out_stop;
-+
-+ /* we have to know where to truncate from in crash case */
-+ EXT3_I(inode)->i_disksize = inode->i_size;
-+ ext3_mark_inode_dirty(handle, inode);
-+
-+ last_block = (inode->i_size + sb->s_blocksize - 1)
-+ >> EXT3_BLOCK_SIZE_BITS(sb);
-+ err = ext3_ext_remove_space(inode, last_block);
-+
-+ /* In a multi-transaction truncate, we only make the final
-+ * transaction synchronous */
-+ if (IS_SYNC(inode))
-+ handle->h_sync = 1;
-+
-+out_stop:
-+ /*
-+ * If this was a simple ftruncate(), and the file will remain alive
-+ * 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
-+ * ext3_delete_inode(), and we allow that function to clean up the
-+ * orphan info for us.
-+ */
-+ if (inode->i_nlink)
-+ ext3_orphan_del(handle, inode);
-+
-+ mutex_unlock(&EXT3_I(inode)->truncate_mutex);
-+ ext3_journal_stop(handle);
-+}
-+
-+/*
-+ * this routine calculate max number of blocks we could modify
-+ * in order to allocate new block for an inode
-+ */
-+int ext3_ext_writepage_trans_blocks(struct inode *inode, int num)
-+{
-+ int needed;
-+
-+ needed = ext3_ext_calc_credits_for_insert(inode, NULL);
-+
-+ /* caller want to allocate num blocks, but note it includes sb */
-+ needed = needed * num - (num - 1);
-+
-+#ifdef CONFIG_QUOTA
-+ needed += 2 * EXT3_QUOTA_TRANS_BLOCKS(inode->i_sb);
-+#endif
-+
-+ return needed;
-+}
-+
-+EXPORT_SYMBOL(ext3_mark_inode_dirty);
-+EXPORT_SYMBOL(ext3_ext_invalidate_cache);
-+EXPORT_SYMBOL(ext3_ext_insert_extent);
-+EXPORT_SYMBOL(ext3_ext_walk_space);
-+EXPORT_SYMBOL(ext3_ext_find_goal);
-+EXPORT_SYMBOL(ext3_ext_calc_credits_for_insert);
-Index: linux-2.6.18.8/fs/ext3/ialloc.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/ialloc.c 2007-07-17 09:18:09.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/ialloc.c 2007-07-17 11:08:09.000000000 +0200
-@@ -652,6 +652,17 @@ got:
- ext3_std_error(sb, err);
- goto fail_free_drop;
- }
-+ if (test_opt(sb, EXTENTS) && S_ISREG(inode->i_mode)) {
-+ EXT3_I(inode)->i_flags |= EXT3_EXTENTS_FL;
-+ ext3_ext_tree_init(handle, inode);
-+ if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_EXTENTS)) {
-+ err = ext3_journal_get_write_access(handle, EXT3_SB(sb)->s_sbh);
-+ if (err) goto fail;
-+ EXT3_SET_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_EXTENTS);
-+ BUFFER_TRACE(EXT3_SB(sb)->s_sbh, "call ext3_journal_dirty_metadata");
-+ err = ext3_journal_dirty_metadata(handle, EXT3_SB(sb)->s_sbh);
-+ }
-+ }
-
- ext3_debug("allocating inode %lu\n", inode->i_ino);
- goto really_out;
-Index: linux-2.6.18.8/fs/ext3/inode.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/inode.c 2007-07-17 09:18:12.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/inode.c 2007-07-17 11:08:11.000000000 +0200
-@@ -40,8 +40,6 @@
- #include "xattr.h"
- #include "acl.h"
-
--static int ext3_writepage_trans_blocks(struct inode *inode);
--
- /*
- * Test whether an inode is a fast symlink.
- */
-@@ -804,6 +802,7 @@ int ext3_get_blocks_handle(handle_t *han
- ext3_fsblk_t first_block = 0;
-
-
-+ J_ASSERT(!(EXT3_I(inode)->i_flags & EXT3_EXTENTS_FL));
- J_ASSERT(handle != NULL || create == 0);
- depth = ext3_block_to_path(inode,iblock,offsets,&blocks_to_boundary);
-
-@@ -984,12 +983,10 @@ static int ext3_get_block(struct inode *
- started = 1;
- }
-
-- ret = ext3_get_blocks_handle(handle, inode, iblock,
-+ ret = ext3_get_blocks_wrap(handle, inode, iblock,
- max_blocks, bh_result, create, 0);
-- if (ret > 0) {
-- bh_result->b_size = (ret << inode->i_blkbits);
-+ if (ret > 0)
- ret = 0;
-- }
- if (started)
- ext3_journal_stop(handle);
- out:
-@@ -1008,7 +1005,7 @@ struct buffer_head *ext3_getblk(handle_t
- dummy.b_state = 0;
- dummy.b_blocknr = -1000;
- buffer_trace_init(&dummy.b_history);
-- err = ext3_get_blocks_handle(handle, inode, block, 1,
-+ err = ext3_get_blocks_wrap(handle, inode, block, 1,
- &dummy, create, 1);
- /*
- * ext3_get_blocks_handle() returns number of blocks
-@@ -1759,7 +1756,7 @@ void ext3_set_aops(struct inode *inode)
- * This required during truncate. We need to physically zero the tail end
- * of that block so it doesn't yield old data if the file is later grown.
- */
--static int ext3_block_truncate_page(handle_t *handle, struct page *page,
-+int ext3_block_truncate_page(handle_t *handle, struct page *page,
- struct address_space *mapping, loff_t from)
- {
- ext3_fsblk_t index = from >> PAGE_CACHE_SHIFT;
-@@ -2263,6 +2260,9 @@ void ext3_truncate(struct inode *inode)
- return;
- }
-
-+ if (EXT3_I(inode)->i_flags & EXT3_EXTENTS_FL)
-+ return ext3_ext_truncate(inode, page);
-+
- handle = start_transaction(inode);
- if (IS_ERR(handle)) {
- if (page) {
-@@ -3008,12 +3008,15 @@ err_out:
- * block and work out the exact number of indirects which are touched. Pah.
- */
-
--static int ext3_writepage_trans_blocks(struct inode *inode)
-+int ext3_writepage_trans_blocks(struct inode *inode)
- {
- int bpp = ext3_journal_blocks_per_page(inode);
- int indirects = (EXT3_NDIR_BLOCKS % bpp) ? 5 : 3;
- int ret;
-
-+ if (EXT3_I(inode)->i_flags & EXT3_EXTENTS_FL)
-+ return ext3_ext_writepage_trans_blocks(inode, bpp);
-+
- if (ext3_should_journal_data(inode))
- ret = 3 * (bpp + indirects) + 2;
- else
-@@ -3260,7 +3263,7 @@ int ext3_map_inode_page(struct inode *in
- if (blocks[i] != 0)
- continue;
-
-- rc = ext3_get_blocks_handle(handle, inode, iblock, 1, &dummy, 1, 1);
-+ rc = ext3_get_blocks_wrap(handle, inode, iblock, 1, &dummy, 1, 1);
- if (rc < 0) {
- printk(KERN_INFO "ext3_map_inode_page: error reading "
- "block %ld\n", iblock);
-Index: linux-2.6.18.8/fs/ext3/Makefile
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/Makefile 2007-07-17 09:18:11.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/Makefile 2007-07-17 11:08:11.000000000 +0200
-@@ -5,7 +5,7 @@
- obj-$(CONFIG_EXT3_FS) += ext3.o
-
- ext3-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o \
-- ioctl.o namei.o super.o symlink.o hash.o resize.o
-+ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o
-
- ext3-$(CONFIG_EXT3_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o
- ext3-$(CONFIG_EXT3_FS_POSIX_ACL) += acl.o
-Index: linux-2.6.18.8/fs/ext3/super.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/super.c 2007-07-17 09:18:12.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/super.c 2007-07-17 11:08:12.000000000 +0200
-@@ -391,6 +391,7 @@ static void ext3_put_super (struct super
- struct ext3_super_block *es = sbi->s_es;
- int i;
-
-+ ext3_ext_release(sb);
- ext3_xattr_put_super(sb);
- journal_destroy(sbi->s_journal);
- if (!(sb->s_flags & MS_RDONLY)) {
-@@ -455,6 +456,8 @@ static struct inode *ext3_alloc_inode(st
- #endif
- ei->i_block_alloc_info = NULL;
- ei->vfs_inode.i_version = 1;
-+
-+ memset(&ei->i_cached_extent, 0, sizeof(ei->i_cached_extent));
- return &ei->vfs_inode;
- }
-
-@@ -680,7 +683,8 @@ enum {
- Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
- Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
- Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
-- Opt_grpquota
-+ Opt_grpquota,
-+ Opt_extents, Opt_noextents, Opt_extdebug,
- };
-
- static match_table_t tokens = {
-@@ -733,6 +737,9 @@ static match_table_t tokens = {
- {Opt_noiopen, "noiopen"},
- {Opt_iopen_nopriv, "iopen_nopriv"},
- {Opt_barrier, "barrier=%u"},
-+ {Opt_extents, "extents"},
-+ {Opt_noextents, "noextents"},
-+ {Opt_extdebug, "extdebug"},
- {Opt_err, NULL},
- {Opt_resize, "resize"},
- };
-@@ -1077,6 +1084,15 @@ clear_qf_name:
- case Opt_bh:
- clear_opt(sbi->s_mount_opt, NOBH);
- break;
-+ case Opt_extents:
-+ set_opt (sbi->s_mount_opt, EXTENTS);
-+ break;
-+ case Opt_noextents:
-+ clear_opt (sbi->s_mount_opt, EXTENTS);
-+ break;
-+ case Opt_extdebug:
-+ set_opt (sbi->s_mount_opt, EXTDEBUG);
-+ break;
- default:
- printk (KERN_ERR
- "EXT3-fs: Unrecognized mount option \"%s\" "
-@@ -1806,6 +1822,8 @@ static int ext3_fill_super (struct super
- test_opt(sb,DATA_FLAGS) == EXT3_MOUNT_ORDERED_DATA ? "ordered":
- "writeback");
-
-+ ext3_ext_init(sb);
-+
- lock_kernel();
- return 0;
-
-Index: linux-2.6.18.8/include/linux/ext3_extents.h
-===================================================================
---- /dev/null 1970-01-01 00:00:00.000000000 +0000
-+++ linux-2.6.18.8/include/linux/ext3_extents.h 2007-07-17 09:18:14.000000000 +0200
-@@ -0,0 +1,232 @@
-+/*
-+ * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
-+ * Written by Alex Tomas <alex@clusterfs.com>
-+ *
-+ * 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-
-+ */
-+
-+#ifndef _LINUX_EXT3_EXTENTS
-+#define _LINUX_EXT3_EXTENTS
-+
-+#include <linux/ext3_fs.h>
-+
-+/*
-+ * with AGRESSIVE_TEST defined capacity of index/leaf blocks
-+ * become very little, so index split, in-depth growing and
-+ * other hard changes happens much more often
-+ * this is for debug purposes only
-+ */
-+#define AGRESSIVE_TEST_
-+
-+/*
-+ * with EXTENTS_STATS defined number of blocks and extents
-+ * are collected in truncate path. they'll be showed at
-+ * umount time
-+ */
-+#define EXTENTS_STATS__
-+
-+/*
-+ * if CHECK_BINSEARCH defined, then results of binary search
-+ * will be checked by linear search
-+ */
-+#define CHECK_BINSEARCH__
-+
-+/*
-+ * if EXT_DEBUG is defined you can use 'extdebug' mount option
-+ * to get lots of info what's going on
-+ */
-+#define EXT_DEBUG_
-+#ifdef EXT_DEBUG
-+#define ext_debug(inode,fmt,a...) \
-+do { \
-+ if (test_opt(inode->i_sb, EXTDEBUG)) \
-+ printk(fmt, ##a); \
-+} while (0)
-+#else
-+#define ext_debug(inode,fmt,a...)
-+#endif
-+
-+
-+/*
-+ * if EXT_STATS is defined then stats numbers are collected
-+ * these number will be displayed at umount time
-+ */
-+#define EXT_STATS_
-+
-+/*
-+ * define EXT3_ALLOC_NEEDED to 0 since block bitmap, group desc. and sb
-+ * are now accounted in ext3_ext_calc_credits_for_insert()
-+ */
-+#define EXT3_ALLOC_NEEDED 0
-+
-+/*
-+ * ext3_inode has i_block array (60 bytes total)
-+ * first 12 bytes store ext3_extent_header
-+ * the remain stores array of ext3_extent
-+ */
-+
-+/*
-+ * this is extent on-disk structure
-+ * it's used at the bottom of the tree
-+ */
-+struct ext3_extent {
-+ __le32 ee_block; /* first logical block extent covers */
-+ __le16 ee_len; /* number of blocks covered by extent */
-+ __le16 ee_start_hi; /* high 16 bits of physical block */
-+ __le32 ee_start; /* low 32 bigs of physical block */
-+};
-+
-+/*
-+ * this is index on-disk structure
-+ * it's used at all the levels, but the bottom
-+ */
-+struct ext3_extent_idx {
-+ __le32 ei_block; /* index covers logical blocks from 'block' */
-+ __le32 ei_leaf; /* pointer to the physical block of the next *
-+ * level. leaf or next index could bet here */
-+ __le16 ei_leaf_hi; /* high 16 bits of physical block */
-+ __u16 ei_unused;
-+};
-+
-+/*
-+ * each block (leaves and indexes), even inode-stored has header
-+ */
-+struct ext3_extent_header {
-+ __le16 eh_magic; /* probably will support different formats */
-+ __le16 eh_entries; /* number of valid entries */
-+ __le16 eh_max; /* capacity of store in entries */
-+ __le16 eh_depth; /* has tree real underlaying blocks? */
-+ __le32 eh_generation; /* flags(8 bits) | generation of the tree */
-+};
-+
-+#define EXT3_EXT_MAGIC 0xf30a
-+
-+/*
-+ * array of ext3_ext_path contains path to some extent
-+ * creation/lookup routines use it for traversal/splitting/etc
-+ * truncate uses it to simulate recursive walking
-+ */
-+struct ext3_ext_path {
-+ __u32 p_block;
-+ __u16 p_depth;
-+ struct ext3_extent *p_ext;
-+ struct ext3_extent_idx *p_idx;
-+ struct ext3_extent_header *p_hdr;
-+ struct buffer_head *p_bh;
-+};
-+
-+/*
-+ * structure for external API
-+ */
-+
-+#define EXT3_EXT_CACHE_NO 0
-+#define EXT3_EXT_CACHE_GAP 1
-+#define EXT3_EXT_CACHE_EXTENT 2
-+#define EXT3_EXT_HAS_NO_TREE /* ext3_extents_tree struct is not used*/
-+
-+/*
-+ * to be called by ext3_ext_walk_space()
-+ * negative retcode - error
-+ * positive retcode - signal for ext3_ext_walk_space(), see below
-+ * callback must return valid extent (passed or newly created)
-+ */
-+typedef int (*ext_prepare_callback)(struct inode *, struct ext3_ext_path *,
-+ struct ext3_ext_cache *,
-+ void *);
-+
-+#define EXT_CONTINUE 0
-+#define EXT_BREAK 1
-+#define EXT_REPEAT 2
-+
-+
-+#define EXT_MAX_BLOCK 0xffffffff
-+#define EXT_UNSET_BLOCK 1
-+
-+#define EXT_FLAGS_CLR_UNKNOWN 0x7 /* Flags cleared on modification */
-+#define EXT_HDR_GEN_BITS 24
-+#define EXT_HDR_GEN_MASK ((1 << EXT_HDR_GEN_BITS) - 1)
-+
-+#define EXT_FIRST_EXTENT(__hdr__) \
-+ ((struct ext3_extent *) (((char *) (__hdr__)) + \
-+ sizeof(struct ext3_extent_header)))
-+#define EXT_FIRST_INDEX(__hdr__) \
-+ ((struct ext3_extent_idx *) (((char *) (__hdr__)) + \
-+ sizeof(struct ext3_extent_header)))
-+#define EXT_HAS_FREE_INDEX(__path__) \
-+ (le16_to_cpu((__path__)->p_hdr->eh_entries) < \
-+ le16_to_cpu((__path__)->p_hdr->eh_max))
-+#define EXT_LAST_EXTENT(__hdr__) \
-+ (EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1)
-+#define EXT_LAST_INDEX(__hdr__) \
-+ (EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1)
-+#define EXT_MAX_EXTENT(__hdr__) \
-+ (EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)
-+#define EXT_MAX_INDEX(__hdr__) \
-+ (EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)
-+
-+
-+static inline struct ext3_extent_header *ext_inode_hdr(struct inode *inode)
-+{
-+ return (struct ext3_extent_header *) EXT3_I(inode)->i_data;
-+}
-+
-+static inline struct ext3_extent_header *ext_block_hdr(struct buffer_head *bh)
-+{
-+ return (struct ext3_extent_header *) bh->b_data;
-+}
-+
-+static inline unsigned short ext_depth(struct inode *inode)
-+{
-+ return le16_to_cpu(ext_inode_hdr(inode)->eh_depth);
-+}
-+
-+static inline unsigned short ext_flags(struct ext3_extent_header *neh)
-+{
-+ return le16_to_cpu(neh->eh_generation) >> EXT_HDR_GEN_BITS;
-+}
-+
-+static inline unsigned short ext_hdr_gen(struct ext3_extent_header *neh)
-+{
-+ return le16_to_cpu(neh->eh_generation) & EXT_HDR_GEN_MASK;
-+}
-+
-+static inline unsigned short ext_generation(struct inode *inode)
-+{
-+ return ext_hdr_gen(ext_inode_hdr(inode));
-+}
-+
-+static inline void ext3_ext_tree_changed(struct inode *inode)
-+{
-+ struct ext3_extent_header *neh = ext_inode_hdr(inode);
-+ neh->eh_generation = cpu_to_le32(
-+ ((ext_flags(neh) & ~EXT_FLAGS_CLR_UNKNOWN) << EXT_HDR_GEN_BITS)
-+ | ((ext_hdr_gen(neh) + 1) & EXT_HDR_GEN_MASK));
-+}
-+
-+static inline void
-+ext3_ext_invalidate_cache(struct inode *inode)
-+{
-+ EXT3_I(inode)->i_cached_extent.ec_type = EXT3_EXT_CACHE_NO;
-+}
-+
-+extern int ext3_ext_search_left(struct inode *, struct ext3_ext_path *, unsigned long *, unsigned long *);
-+extern int ext3_ext_search_right(struct inode *, struct ext3_ext_path *, unsigned long *, unsigned long *);
-+extern int ext3_extent_tree_init(handle_t *, struct inode *);
-+extern int ext3_ext_calc_credits_for_insert(struct inode *, struct ext3_ext_path *);
-+extern int ext3_ext_insert_extent(handle_t *, struct inode *, struct ext3_ext_path *, struct ext3_extent *);
-+extern int ext3_ext_walk_space(struct inode *, unsigned long, unsigned long, ext_prepare_callback, void *);
-+extern struct ext3_ext_path * ext3_ext_find_extent(struct inode *, int, struct ext3_ext_path *);
-+
-+#endif /* _LINUX_EXT3_EXTENTS */
-+
-Index: linux-2.6.18.8/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs.h 2007-07-17 09:18:13.000000000 +0200
-+++ linux-2.6.18.8/include/linux/ext3_fs.h 2007-07-17 11:08:12.000000000 +0200
-@@ -182,8 +182,9 @@ struct ext3_group_desc
- #define EXT3_DIRSYNC_FL 0x00010000 /* dirsync behaviour (directories only) */
- #define EXT3_TOPDIR_FL 0x00020000 /* Top of directory hierarchies*/
- #define EXT3_RESERVED_FL 0x80000000 /* reserved for ext3 lib */
-+#define EXT3_EXTENTS_FL 0x00080000 /* Inode uses extents */
-
--#define EXT3_FL_USER_VISIBLE 0x0003DFFF /* User visible flags */
-+#define EXT3_FL_USER_VISIBLE 0x000BDFFF /* User visible flags */
- #define EXT3_FL_USER_MODIFIABLE 0x000380FF /* User modifiable flags */
-
- /*
-@@ -373,6 +374,8 @@ struct ext3_inode {
- #define EXT3_MOUNT_QUOTA 0x80000 /* Some quota option set */
- #define EXT3_MOUNT_USRQUOTA 0x100000 /* "old" user quota */
- #define EXT3_MOUNT_GRPQUOTA 0x200000 /* "old" group quota */
-+#define EXT3_MOUNT_EXTENTS 0x2000000/* Extents support */
-+#define EXT3_MOUNT_EXTDEBUG 0x4000000/* Extents debug */
-
- /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
- #ifndef clear_opt
-@@ -572,11 +575,13 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_INCOMPAT_RECOVER 0x0004 /* Needs recovery */
- #define EXT3_FEATURE_INCOMPAT_JOURNAL_DEV 0x0008 /* Journal device */
- #define EXT3_FEATURE_INCOMPAT_META_BG 0x0010
-+#define EXT3_FEATURE_INCOMPAT_EXTENTS 0x0040 /* extents support */
-
- #define EXT3_FEATURE_COMPAT_SUPP EXT2_FEATURE_COMPAT_EXT_ATTR
- #define EXT3_FEATURE_INCOMPAT_SUPP (EXT3_FEATURE_INCOMPAT_FILETYPE| \
- EXT3_FEATURE_INCOMPAT_RECOVER| \
-- EXT3_FEATURE_INCOMPAT_META_BG)
-+ EXT3_FEATURE_INCOMPAT_META_BG| \
-+ EXT3_FEATURE_INCOMPAT_EXTENTS)
- #define EXT3_FEATURE_RO_COMPAT_SUPP (EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER| \
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
- EXT3_FEATURE_RO_COMPAT_BTREE_DIR)
-@@ -816,6 +821,9 @@ extern int ext3_get_inode_loc(struct ino
- extern void ext3_set_aops(struct inode *inode);
- extern int ext3_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
- u64 start, u64 len);
-+extern int ext3_writepage_trans_blocks(struct inode *);
-+extern int ext3_block_truncate_page(handle_t *handle, struct page *page,
-+ struct address_space *mapping, loff_t from);
-
- /* ioctl.c */
- extern int ext3_ioctl (struct inode *, struct file *, unsigned int,
-@@ -869,6 +877,30 @@ extern struct inode_operations ext3_spec
- extern struct inode_operations ext3_symlink_inode_operations;
- extern struct inode_operations ext3_fast_symlink_inode_operations;
-
-+/* extents.c */
-+extern int ext3_ext_tree_init(handle_t *handle, struct inode *);
-+extern int ext3_ext_writepage_trans_blocks(struct inode *, int);
-+extern int ext3_ext_get_blocks(handle_t *, struct inode *, sector_t,
-+ unsigned long, struct buffer_head *, int, int);
-+extern void ext3_ext_truncate(struct inode *, struct page *);
-+extern void ext3_ext_init(struct super_block *);
-+extern void ext3_ext_release(struct super_block *);
-+static inline int
-+ext3_get_blocks_wrap(handle_t *handle, struct inode *inode, sector_t block,
-+ unsigned long max_blocks, struct buffer_head *bh,
-+ int create, int extend_disksize)
-+{
-+ int ret;
-+ if (EXT3_I(inode)->i_flags & EXT3_EXTENTS_FL)
-+ return ext3_ext_get_blocks(handle, inode, block, max_blocks,
-+ bh, create, extend_disksize);
-+ ret = ext3_get_blocks_handle(handle, inode, block, max_blocks, bh, create,
-+ extend_disksize);
-+ if (ret > 0)
-+ bh->b_size = (ret << inode->i_blkbits);
-+ return ret;
-+}
-+
-
- #endif /* __KERNEL__ */
-
-Index: linux-2.6.18.8/include/linux/ext3_fs_i.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_i.h 2007-02-24 00:52:30.000000000 +0100
-+++ linux-2.6.18.8/include/linux/ext3_fs_i.h 2007-07-17 11:08:11.000000000 +0200
-@@ -65,6 +65,16 @@ struct ext3_block_alloc_info {
- #define rsv_end rsv_window._rsv_end
-
- /*
-+ * storage for cached extent
-+ */
-+struct ext3_ext_cache {
-+ __u32 ec_start;
-+ __u32 ec_block;
-+ __u32 ec_len; /* must be 32bit to return holes */
-+ __u32 ec_type;
-+};
-+
-+/*
- * third extended file system inode data in memory
- */
- struct ext3_inode_info {
-@@ -142,6 +152,8 @@ struct ext3_inode_info {
- */
- struct mutex truncate_mutex;
- struct inode vfs_inode;
-+
-+ struct ext3_ext_cache i_cached_extent;
- };
-
- #endif /* _LINUX_EXT3_FS_I */
-Index: linux-2.6.18.8/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_sb.h 2007-02-24 00:52:30.000000000 +0100
-+++ linux-2.6.18.8/include/linux/ext3_fs_sb.h 2007-07-17 11:08:12.000000000 +0200
-@@ -78,6 +78,16 @@ struct ext3_sb_info {
- char *s_qf_names[MAXQUOTAS]; /* Names of quota files with journalled quota */
- int s_jquota_fmt; /* Format of quota to use */
- #endif
-+
-+#ifdef EXTENTS_STATS
-+ /* ext3 extents stats */
-+ unsigned long s_ext_min;
-+ unsigned long s_ext_max;
-+ unsigned long s_depth_max;
-+ spinlock_t s_ext_stats_lock;
-+ unsigned long s_ext_blocks;
-+ unsigned long s_ext_extents;
-+#endif
- };
-
- #endif /* _LINUX_EXT3_FS_SB */
-Index: linux-2.6.18.8/include/linux/ext3_jbd.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_jbd.h 2007-02-24 00:52:30.000000000 +0100
-+++ linux-2.6.18.8/include/linux/ext3_jbd.h 2007-07-17 09:18:14.000000000 +0200
-@@ -23,12 +23,20 @@
-
- /* Define the number of blocks we need to account to a transaction to
- * modify one block of data.
-- *
-+ *
- * We may have to touch one inode, one bitmap buffer, up to three
- * indirection blocks, the group and superblock summaries, and the data
-- * block to complete the transaction. */
-+ * block to complete the transaction.
-+ *
-+ * For extents-enabled fs we may have to allocate and modify upto
-+ * 5 levels of tree + root which is stored in inode. */
-+
-+#define EXT3_SINGLEDATA_TRANS_BLOCKS(sb) \
-+ (EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_EXTENTS) \
-+ || test_opt(sb, EXTENTS) ? 27U : 8U)
-
--#define EXT3_SINGLEDATA_TRANS_BLOCKS 8U
-+/* Indicate that EXT3_SINGLEDATA_TRANS_BLOCKS takes the sb as argument */
-+#define EXT3_SINGLEDATA_TRANS_BLOCKS_HAS_SB
-
- /* Extended attribute operations touch at most two data buffers,
- * two bitmap buffers, and two group summaries, in addition to the inode
-@@ -42,7 +50,7 @@
- * superblock only gets updated once, of course, so don't bother
- * counting that again for the quota updates. */
-
--#define EXT3_DATA_TRANS_BLOCKS(sb) (EXT3_SINGLEDATA_TRANS_BLOCKS + \
-+#define EXT3_DATA_TRANS_BLOCKS(sb) (EXT3_SINGLEDATA_TRANS_BLOCKS(sb) + \
- EXT3_XATTR_TRANS_BLOCKS - 2 + \
- 2*EXT3_QUOTA_TRANS_BLOCKS(sb))
-
-@@ -78,9 +86,9 @@
- /* Amount of blocks needed for quota insert/delete - we do some block writes
- * but inode, sb and group updates are done only once */
- #define EXT3_QUOTA_INIT_BLOCKS(sb) (test_opt(sb, QUOTA) ? (DQUOT_INIT_ALLOC*\
-- (EXT3_SINGLEDATA_TRANS_BLOCKS-3)+3+DQUOT_INIT_REWRITE) : 0)
-+ (EXT3_SINGLEDATA_TRANS_BLOCKS(sb)-3)+3+DQUOT_INIT_REWRITE) : 0)
- #define EXT3_QUOTA_DEL_BLOCKS(sb) (test_opt(sb, QUOTA) ? (DQUOT_DEL_ALLOC*\
-- (EXT3_SINGLEDATA_TRANS_BLOCKS-3)+3+DQUOT_DEL_REWRITE) : 0)
-+ (EXT3_SINGLEDATA_TRANS_BLOCKS(sb)-3)+3+DQUOT_DEL_REWRITE) : 0)
- #else
- #define EXT3_QUOTA_TRANS_BLOCKS(sb) 0
- #define EXT3_QUOTA_INIT_BLOCKS(sb) 0
+++ /dev/null
-diff -rupN linux-2.6.18-164.6.1_1/fs/ext3/ioctl.c linux-2.6.18-164.6.1_2/fs/ext3/ioctl.c
---- linux-2.6.18-164.6.1_1/fs/ext3/ioctl.c 2009-12-22 12:37:55.000000000 +0530
-+++ linux-2.6.18-164.6.1_2/fs/ext3/ioctl.c 2009-12-22 12:40:44.000000000 +0530
-@@ -16,6 +16,70 @@
- #include <asm/uaccess.h>
- #include <linux/namei.h>
-
-+/* So that the fiemap access checks can't overflow on 32 bit machines. */
-+#define FIEMAP_MAX_EXTENTS (UINT_MAX / sizeof(struct fiemap_extent))
-+
-+static int fiemap_check_ranges(struct super_block *sb,
-+ u64 start, u64 len, u64 *new_len)
-+{
-+ *new_len = len;
-+
-+ if (len == 0)
-+ return -EINVAL;
-+
-+ if (start > sb->s_maxbytes)
-+ return -EFBIG;
-+
-+ /*
-+ * Shrink request scope to what the fs can actually handle.
-+ */
-+ if ((len > sb->s_maxbytes) ||
-+ (sb->s_maxbytes - len) < start)
-+ *new_len = sb->s_maxbytes - start;
-+
-+ return 0;
-+}
-+
-+int ioctl_fiemap(struct inode *inode, struct file *filp, unsigned long arg)
-+{
-+ struct fiemap fiemap;
-+ u64 len;
-+ struct fiemap_extent_info fieinfo = {0, };
-+ struct super_block *sb = inode->i_sb;
-+ int error = 0;
-+
-+ if (copy_from_user(&fiemap, (struct fiemap __user *) arg,
-+ sizeof(struct fiemap)))
-+ return -EFAULT;
-+
-+ if (fiemap.fm_extent_count > FIEMAP_MAX_EXTENTS)
-+ return -EINVAL;
-+
-+ error = fiemap_check_ranges(sb, fiemap.fm_start, fiemap.fm_length,
-+ &len);
-+ if (error)
-+ return error;
-+
-+ fieinfo.fi_flags = fiemap.fm_flags;
-+ fieinfo.fi_extents_max = fiemap.fm_extent_count;
-+ fieinfo.fi_extents_start = (struct fiemap_extent *)(arg + sizeof(fiemap));
-+
-+ if (fiemap.fm_extent_count != 0 &&
-+ !access_ok(VERIFY_WRITE, (void *)arg,
-+ offsetof(typeof(fiemap), fm_extents[fiemap.fm_extent_count])))
-+ return -EFAULT;
-+
-+ if (fieinfo.fi_flags & FIEMAP_FLAG_SYNC)
-+ filemap_write_and_wait(inode->i_mapping);
-+
-+ error = ext3_fiemap(inode, &fieinfo, fiemap.fm_start, len);
-+ fiemap.fm_flags = fieinfo.fi_flags;
-+ fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;
-+ if (copy_to_user((char *)arg, &fiemap, sizeof(fiemap)))
-+ error = -EFAULT;
-+
-+ return error;
-+}
-
- int ext3_ioctl (struct inode * inode, struct file * filp, unsigned int cmd,
- unsigned long arg)
-@@ -276,7 +340,9 @@ flags_err:
-
- return err;
- }
--
-+ case EXT3_IOC_FIEMAP: {
-+ return ioctl_fiemap(inode, filp, arg);
-+ }
-
- default:
- return -ENOTTY;
-diff -rupN linux-2.6.18-164.6.1_1/include/linux/ext3_fs.h linux-2.6.18-164.6.1_2/include/linux/ext3_fs.h
---- linux-2.6.18-164.6.1_1/include/linux/ext3_fs.h 2009-12-22 12:37:55.000000000 +0530
-+++ linux-2.6.18-164.6.1_2/include/linux/ext3_fs.h 2009-12-22 12:43:02.000000000 +0530
-@@ -260,6 +260,7 @@ struct ext3_new_group_data {
- #define EXT3_IOC_SETVERSION _IOW('f', 4, long)
- #define EXT3_IOC_GROUP_EXTEND _IOW('f', 7, unsigned long)
- #define EXT3_IOC_GROUP_ADD _IOW('f', 8,struct ext3_new_group_input)
-+#define EXT3_IOC_FIEMAP _IOWR('f', 11, struct fiemap)
- #define EXT3_IOC_GETVERSION_OLD _IOR('v', 1, long)
- #define EXT3_IOC_SETVERSION_OLD _IOW('v', 2, long)
- #ifdef CONFIG_JBD_DEBUG
+++ /dev/null
-Index: linux-2.6.18-128.1.6/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext3/super.c
-+++ linux-2.6.18-128.1.6/fs/ext3/super.c
-@@ -51,6 +51,8 @@
- #define EXT3_DEFAULT_MAX_DIR_SIZE 0
- #define EXT3_MAX_DIR_SIZE_NAME "max_dir_size"
-
-+static int force_over_8tb;
-+
- static int ext3_load_journal(struct super_block *, struct ext3_super_block *,
- unsigned long journal_devnum);
- static int ext3_create_journal(struct super_block *, struct ext3_super_block *,
-@@ -715,7 +717,7 @@ enum {
- Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
- Opt_grpquota,
- Opt_extents, Opt_noextents, Opt_bigendian_extents, Opt_extdebug,
-- Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_maxdirsize
-+ Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_maxdirsize, Opt_force_over_8tb,
- };
-
- static match_table_t tokens = {
-@@ -779,5 +781,6 @@ static match_table_t tokens = {
- {Opt_mballoc, "mballoc"},
- {Opt_nomballoc, "nomballoc"},
- {Opt_stripe, "stripe=%u"},
-+ {Opt_force_over_8tb, "force_over_8tb"},
- {Opt_resize, "resize"},
- };
-@@ -1157,6 +1160,9 @@ clear_qf_name:
- return 0;
- sbi->s_max_dir_size = option;
- break;
-+ case Opt_force_over_8tb:
-+ force_over_8tb = 1;
-+ break;
- default:
- printk (KERN_ERR
- "EXT3-fs: Unrecognized mount option \"%s\" "
-@@ -2248,6 +2254,17 @@ static int ext3_fill_super (struct super
- goto failed_mount;
- }
-
-+ if (le32_to_cpu(es->s_blocks_count) >
-+ ((0x80000000000ULL >> sb->s_blocksize_bits) - 1)) {
-+ if (force_over_8tb == 0) {
-+ printk(KERN_ERR "EXT3-fs does not support filesystems "
-+ "greater than 8TB and can cause data corruption."
-+ "Use \"force_over_8tb\" mount option to override."
-+ "\n");
-+ goto failed_mount;
-+ }
-+ }
-+
- if (EXT3_BLOCKS_PER_GROUP(sb) == 0)
- goto cantfind_ext3;
- sbi->s_groups_count = ((le32_to_cpu(es->s_blocks_count) -
+++ /dev/null
-Index: linux-2.6.18-53.1.14/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/super.c
-+++ linux-2.6.18-53.1.14/fs/ext3/super.c
-@@ -1900,6 +1900,34 @@ static int ext3_max_dir_size_write(struc
- return count;
- }
-
-+/*
-+ * ext3_get_stripe_size: Get the stripe size.
-+ * @sbi: In memory super block info
-+ *
-+ * If we have specified it via mount option, then
-+ * use the mount option value. If the value specified at mount time is
-+ * greater than the blocks per group use the super block value.
-+ * If the super block value is greater than blocks per group return 0.
-+ * Allocator needs it be less than blocks per group.
-+ */
-+static unsigned long ext3_get_stripe_size(struct ext3_sb_info *sbi)
-+{
-+ unsigned long stride = le16_to_cpu(sbi->s_es->s_raid_stride);
-+ unsigned long stripe_width =
-+ le32_to_cpu(sbi->s_es->s_raid_stripe_width);
-+
-+ if (sbi->s_stripe && sbi->s_stripe <= sbi->s_blocks_per_group)
-+ return sbi->s_stripe;
-+
-+ if (stripe_width <= sbi->s_blocks_per_group)
-+ return stripe_width;
-+
-+ if (stride <= sbi->s_blocks_per_group)
-+ return stride;
-+
-+ return 0;
-+}
-+
- static int ext3_fill_super (struct super_block *sb, void *data, int silent)
- {
- struct buffer_head * bh;
-@@ -2221,6 +2249,8 @@ static int ext3_fill_super (struct super
- sbi->s_rsv_window_head.rsv_goal_size = 0;
- ext3_rsv_window_add(sb, &sbi->s_rsv_window_head);
-
-+ sbi->s_stripe = ext3_get_stripe_size(sbi);
-+
- /*
- * set up enough so that it can read an inode
- */
-
+++ /dev/null
-Index: linux-stage/fs/ext3/namei.c
-===================================================================
---- linux-stage.orig/fs/ext3/namei.c 2009-08-10 22:31:03.000000000 +0800
-+++ linux-stage/fs/ext3/namei.c 2009-08-10 22:33:38.000000000 +0800
-@@ -1471,6 +1471,72 @@
- }
- #endif
-
-+/* update ".." for hash-indexed directory, split the item "." if necessary */
-+static int ext3_update_dotdot(handle_t *handle, struct dentry *dentry,
-+ struct inode *inode)
-+{
-+ struct inode * dir = dentry->d_parent->d_inode;
-+ struct buffer_head * dir_block;
-+ struct ext3_dir_entry_2 * de;
-+ int len, journal = 0, err = 0;
-+
-+ if (IS_ERR(handle))
-+ return PTR_ERR(handle);
-+
-+ if (IS_DIRSYNC(dir))
-+ handle->h_sync = 1;
-+
-+ dir_block = ext3_bread(handle, dir, 0, 0, &err);
-+ if (!dir_block)
-+ goto out;
-+
-+ de = (struct ext3_dir_entry_2 *)dir_block->b_data;
-+ /* the first item must be "." */
-+ assert(de->name_len == 1 && de->name[0] == '.');
-+ len = le16_to_cpu(de->rec_len);
-+ assert(len >= EXT3_DIR_REC_LEN(1));
-+ if (len > EXT3_DIR_REC_LEN(1)) {
-+ BUFFER_TRACE(dir_block, "get_write_access");
-+ err = ext3_journal_get_write_access(handle, dir_block);
-+ if (err)
-+ goto out_journal;
-+
-+ journal = 1;
-+ de->rec_len = cpu_to_le16(EXT3_DIR_REC_LEN(1));
-+ }
-+
-+ len -= EXT3_DIR_REC_LEN(1);
-+ assert(len == 0 || len >= EXT3_DIR_REC_LEN(2));
-+ de = (struct ext3_dir_entry_2 *)
-+ ((char *) de + le16_to_cpu(de->rec_len));
-+ if (!journal) {
-+ BUFFER_TRACE(dir_block, "get_write_access");
-+ err = ext3_journal_get_write_access(handle, dir_block);
-+ if (err)
-+ goto out_journal;
-+ }
-+
-+ de->inode = cpu_to_le32(inode->i_ino);
-+ if (len > 0)
-+ de->rec_len = cpu_to_le16(len);
-+ else
-+ assert(le16_to_cpu(de->rec_len) >= EXT3_DIR_REC_LEN(2));
-+ de->name_len = 2;
-+ strcpy (de->name, "..");
-+ ext3_set_de_type(dir->i_sb, de, S_IFDIR);
-+
-+out_journal:
-+ if (journal) {
-+ BUFFER_TRACE(dir_block, "call ext3_journal_dirty_metadata");
-+ err = ext3_journal_dirty_metadata(handle, dir_block);
-+ ext3_mark_inode_dirty(handle, dir);
-+ }
-+ brelse (dir_block);
-+
-+out:
-+ return err;
-+}
-+
- /*
- * ext3_add_entry()
- *
-@@ -1502,6 +1568,9 @@
- return -EINVAL;
- #ifdef CONFIG_EXT3_INDEX
- if (is_dx(dir)) {
-+ if (dentry->d_name.len == 2 &&
-+ memcmp(dentry->d_name.name, "..", 2) == 0)
-+ return ext3_update_dotdot(handle, dentry, inode);
- retval = ext3_dx_add_entry(handle, dentry, inode);
- if (!retval || (retval != ERR_BAD_DX_DIR))
- return retval;
+++ /dev/null
-Index: linux-stage/include/linux/ext3_fs.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs.h 2005-02-25 14:53:56.424908168 +0200
-+++ linux-stage/include/linux/ext3_fs.h 2005-02-25 14:53:59.376459464 +0200
-@@ -361,12 +361,13 @@
- #define EXT3_MOUNT_IOPEN_NOPRIV 0x100000/* Make iopen world-readable */
-
- /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
--#ifndef _LINUX_EXT2_FS_H
-+#ifndef clear_opt
- #define clear_opt(o, opt) o &= ~EXT3_MOUNT_##opt
- #define set_opt(o, opt) o |= EXT3_MOUNT_##opt
- #define test_opt(sb, opt) (EXT3_SB(sb)->s_mount_opt & \
- EXT3_MOUNT_##opt)
--#else
-+#endif
-+#ifndef EXT2_MOUNT_NOLOAD
- #define EXT2_MOUNT_NOLOAD EXT3_MOUNT_NOLOAD
- #define EXT2_MOUNT_ABORT EXT3_MOUNT_ABORT
- #define EXT2_MOUNT_DATA_FLAGS EXT3_MOUNT_DATA_FLAGS
-Index: linux-2.6.18-53.1.21/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-53.1.21.orig/fs/ext3/super.c
-+++ linux-2.6.18-53.1.21/fs/ext3/super.c
-@@ -677,6 +677,7 @@ enum {
- Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
- Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
- Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
-+ Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
- Opt_grpquota
- };
-
-@@ -726,6 +727,9 @@ static match_table_t tokens = {
- {Opt_noquota, "noquota"},
- {Opt_quota, "quota"},
- {Opt_usrquota, "usrquota"},
-+ {Opt_iopen, "iopen"},
-+ {Opt_noiopen, "noiopen"},
-+ {Opt_iopen_nopriv, "iopen_nopriv"},
- {Opt_barrier, "barrier=%u"},
- {Opt_err, NULL},
- {Opt_resize, "resize"},
-@@ -1041,6 +1045,10 @@ clear_qf_name:
- else
- clear_opt(sbi->s_mount_opt, BARRIER);
- break;
-+ case Opt_iopen:
-+ case Opt_noiopen:
-+ case Opt_iopen_nopriv:
-+ break;
- case Opt_ignore:
- break;
- case Opt_resize:
+++ /dev/null
-Index: linux-2.6.18/fs/ext3/inode.c
-===================================================================
---- linux-2.6.18.orig/fs/ext3/inode.c
-+++ linux-2.6.18/fs/ext3/inode.c
-@@ -2690,6 +2690,13 @@ void ext3_read_inode(struct inode * inod
- EXT3_INODE_GET_XTIME(i_atime, inode, raw_inode);
- EXT3_EINODE_GET_XTIME(i_crtime, ei, raw_inode);
-
-+ ei->i_fs_version = le32_to_cpu(raw_inode->i_disk_version);
-+ if (EXT3_INODE_SIZE(inode->i_sb) > EXT3_GOOD_OLD_INODE_SIZE) {
-+ if (EXT3_FITS_IN_INODE(raw_inode, ei, i_version_hi))
-+ ei->i_fs_version |= (__u64)(le32_to_cpu(raw_inode->i_version_hi))
-+ << 32;
-+ }
-+
- if (S_ISREG(inode->i_mode)) {
- inode->i_op = &ext3_file_inode_operations;
- inode->i_fop = &ext3_file_operations;
-@@ -2828,8 +2835,14 @@ static int ext3_do_update_inode(handle_t
- } else for (block = 0; block < EXT3_N_BLOCKS; block++)
- raw_inode->i_block[block] = ei->i_data[block];
-
-- if (ei->i_extra_isize)
-+ raw_inode->i_disk_version = cpu_to_le32(ei->i_fs_version);
-+ if (ei->i_extra_isize) {
-+ if (EXT3_FITS_IN_INODE(raw_inode, ei, i_version_hi)) {
-+ raw_inode->i_version_hi = cpu_to_le32(ei->i_fs_version
-+ >> 32);
-+ }
- raw_inode->i_extra_isize = cpu_to_le16(ei->i_extra_isize);
-+ }
-
- BUFFER_TRACE(bh, "call ext3_journal_dirty_metadata");
- rc = ext3_journal_dirty_metadata(handle, bh);
-@@ -3103,10 +3116,32 @@ ext3_reserve_inode_write(handle_t *handl
- int ext3_mark_inode_dirty(handle_t *handle, struct inode *inode)
- {
- struct ext3_iloc iloc;
-- int err;
-+ int err, ret;
-+ static int expand_message;
-
- might_sleep();
- err = ext3_reserve_inode_write(handle, inode, &iloc);
-+ if (EXT3_I(inode)->i_extra_isize <
-+ EXT3_SB(inode->i_sb)->s_want_extra_isize &&
-+ !(EXT3_I(inode)->i_state & EXT3_STATE_NO_EXPAND)) {
-+ /* We need extra buffer credits since we may write into EA block
-+ * with this same handle */
-+ if ((ext3_journal_extend(handle,
-+ EXT3_DATA_TRANS_BLOCKS(inode->i_sb))) == 0) {
-+ ret = ext3_expand_extra_isize(inode,
-+ EXT3_SB(inode->i_sb)->s_want_extra_isize,
-+ iloc, handle);
-+ if (ret) {
-+ EXT3_I(inode)->i_state |= EXT3_STATE_NO_EXPAND;
-+ if (!expand_message) {
-+ ext3_warning(inode->i_sb, __FUNCTION__,
-+ "Unable to expand inode %lu. Delete some"
-+ " EAs or run e2fsck.", inode->i_ino);
-+ expand_message = 1;
-+ }
-+ }
-+ }
-+ }
- if (!err)
- err = ext3_mark_iloc_dirty(handle, inode, &iloc);
- return err;
-Index: linux-2.6.18/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18.orig/include/linux/ext3_fs.h
-+++ linux-2.6.18/include/linux/ext3_fs.h
-@@ -224,6 +224,7 @@ struct ext3_group_desc
- #define EXT3_STATE_JDATA 0x00000001 /* journaled data exists */
- #define EXT3_STATE_NEW 0x00000002 /* inode is newly created */
- #define EXT3_STATE_XATTR 0x00000004 /* has in-inode xattrs */
-+#define EXT3_STATE_NO_EXPAND 0x00000008 /* No space for expansion */
-
- /* Used to pass group descriptor data when online resize is done */
- struct ext3_new_group_input {
-@@ -297,7 +298,7 @@ struct ext3_inode {
- __le32 i_flags; /* File flags */
- union {
- struct {
-- __u32 l_i_reserved1;
-+ __u32 l_i_version;
- } linux1;
- struct {
- __u32 h_i_translator;
-@@ -342,6 +343,7 @@ struct ext3_inode {
- __le32 i_atime_extra; /* extra Access time (nsec << 2 | epoch) */
- __le32 i_crtime; /* File Creation time */
- __le32 i_crtime_extra; /* extra File Creation time (nsec << 2 | epoch) */
-+ __le32 i_version_hi; /* high 32 bits for 64-bit version */
- };
-
- #define i_size_high i_dir_acl
-@@ -404,6 +406,8 @@ do { \
- raw_inode->xtime ## _extra); \
- } while (0)
-
-+#define i_disk_version osd1.linux1.l_i_version
-+
- #if defined(__KERNEL__) || defined(__linux__)
- #define i_reserved1 osd1.linux1.l_i_reserved1
- #define i_frag osd2.linux2.l_i_frag
-Index: linux-2.6.18/include/linux/ext3_fs_i.h
-===================================================================
---- linux-2.6.18.orig/include/linux/ext3_fs_i.h
-+++ linux-2.6.18/include/linux/ext3_fs_i.h
-@@ -21,6 +21,8 @@
- #include <linux/seqlock.h>
- #include <linux/mutex.h>
-
-+#define HAVE_DISK_INODE_VERSION
-+
- /* data type for block offset of block group */
- typedef int ext3_grpblk_t;
-
-@@ -162,6 +164,8 @@ struct ext3_inode_info {
- /* mballoc */
- struct list_head i_prealloc_list;
- spinlock_t i_prealloc_lock;
-+
-+ __u64 i_fs_version;
- };
-
- #endif /* _LINUX_EXT3_FS_I */
-Index: linux-2.6.18/fs/ext3/xattr.c
-===================================================================
---- linux-2.6.18.orig/fs/ext3/xattr.c
-+++ linux-2.6.18/fs/ext3/xattr.c
-@@ -505,6 +505,20 @@ ext3_xattr_release_block(handle_t *handl
- }
- }
-
-+static inline size_t ext3_xattr_free_space(struct ext3_xattr_entry *last,
-+ size_t *min_offs, void *base, int *total)
-+{
-+ for (; !IS_LAST_ENTRY(last); last = EXT3_XATTR_NEXT(last)) {
-+ *total += EXT3_XATTR_LEN(last->e_name_len);
-+ if (!last->e_value_block && last->e_value_size) {
-+ size_t offs = le16_to_cpu(last->e_value_offs);
-+ if (offs < *min_offs)
-+ *min_offs = offs;
-+ }
-+ }
-+ return (*min_offs - ((void *)last - base) - sizeof(__u32));
-+}
-+
- struct ext3_xattr_info {
- int name_index;
- const char *name;
-@@ -945,13 +959,18 @@ ext3_xattr_set_handle(handle_t *handle,
- struct ext3_xattr_block_find bs = {
- .s = { .not_found = -ENODATA, },
- };
-+ unsigned long no_expand;
- int error;
-
- if (!name)
- return -EINVAL;
- if (strlen(name) > 255)
- return -ERANGE;
-+
- down_write(&EXT3_I(inode)->xattr_sem);
-+ no_expand = EXT3_I(inode)->i_state & EXT3_STATE_NO_EXPAND;
-+ EXT3_I(inode)->i_state |= EXT3_STATE_NO_EXPAND;
-+
- error = ext3_get_inode_loc(inode, &is.iloc);
- if (error)
- goto cleanup;
-@@ -1009,6 +1028,8 @@ ext3_xattr_set_handle(handle_t *handle,
- ext3_xattr_update_super_block(handle, inode->i_sb);
- inode->i_ctime = ext3_current_time(inode);
- error = ext3_mark_iloc_dirty(handle, inode, &is.iloc);
-+ if (!value)
-+ EXT3_I(inode)->i_state &= ~EXT3_STATE_NO_EXPAND;
- /*
- * The bh is consumed by ext3_mark_iloc_dirty, even with
- * error != 0.
-@@ -1021,6 +1042,8 @@ ext3_xattr_set_handle(handle_t *handle,
- cleanup:
- brelse(is.iloc.bh);
- brelse(bs.bh);
-+ if (no_expand == 0)
-+ EXT3_I(inode)->i_state &= ~EXT3_STATE_NO_EXPAND;
- up_write(&EXT3_I(inode)->xattr_sem);
- return error;
- }
-@@ -1060,6 +1083,249 @@ retry:
- return error;
- }
-
-+static void ext3_xattr_shift_entries(struct ext3_xattr_entry *entry,
-+ int value_offs_shift, void *to,
-+ void *from, size_t n, int blocksize)
-+{
-+ struct ext3_xattr_entry *last = entry;
-+ int new_offs;
-+
-+ /* Adjust the value offsets of the entries */
-+ for (; !IS_LAST_ENTRY(last); last = EXT3_XATTR_NEXT(last)) {
-+ if (!last->e_value_block && last->e_value_size) {
-+ new_offs = le16_to_cpu(last->e_value_offs) +
-+ value_offs_shift;
-+ BUG_ON(new_offs + le32_to_cpu(last->e_value_size) >
-+ blocksize);
-+ last->e_value_offs = cpu_to_le16(new_offs);
-+ }
-+ }
-+ /* Shift the entries by n bytes */
-+ memmove(to, from, n);
-+}
-+
-+/* Expand an inode by new_extra_isize bytes.
-+ * Returns 0 on success or negative error number on failure.
-+ */
-+int ext3_expand_extra_isize(struct inode *inode, int new_extra_isize,
-+ struct ext3_iloc iloc, handle_t *handle)
-+{
-+ struct ext3_inode *raw_inode;
-+ struct ext3_xattr_ibody_header *header;
-+ struct ext3_xattr_entry *entry, *last, *first;
-+ struct buffer_head *bh = NULL;
-+ struct ext3_xattr_ibody_find *is = NULL;
-+ struct ext3_xattr_block_find *bs = NULL;
-+ char *buffer = NULL, *b_entry_name = NULL;
-+ size_t min_offs, free;
-+ int total_ino, total_blk;
-+ void *base, *start, *end;
-+ int extra_isize = 0, error = 0, tried_min_extra_isize = 0;
-+ int s_min_extra_isize = EXT3_SB(inode->i_sb)->s_es->s_min_extra_isize;
-+
-+ down_write(&EXT3_I(inode)->xattr_sem);
-+
-+retry:
-+ if (EXT3_I(inode)->i_extra_isize >= new_extra_isize) {
-+ up_write(&EXT3_I(inode)->xattr_sem);
-+ return 0;
-+ }
-+
-+ raw_inode = ext3_raw_inode(&iloc);
-+
-+ header = IHDR(inode, raw_inode);
-+ entry = IFIRST(header);
-+
-+ /* No extended attributes present */
-+ if (!(EXT3_I(inode)->i_state & EXT3_STATE_XATTR) ||
-+ header->h_magic != cpu_to_le32(EXT3_XATTR_MAGIC)) {
-+ memset((void *)raw_inode + EXT3_GOOD_OLD_INODE_SIZE, 0,
-+ new_extra_isize);
-+ EXT3_I(inode)->i_extra_isize = new_extra_isize;
-+ goto cleanup;
-+ }
-+
-+ /*
-+ * Check if enough free space is available in the inode to shift the
-+ * entries ahead by new_extra_isize.
-+ */
-+
-+ base = start = entry;
-+ end = (void *)raw_inode + EXT3_SB(inode->i_sb)->s_inode_size;
-+ min_offs = end - base;
-+ last = entry;
-+ total_ino = sizeof(struct ext3_xattr_ibody_header);
-+
-+ free = ext3_xattr_free_space(last, &min_offs, base, &total_ino);
-+ if (free >= new_extra_isize) {
-+ entry = IFIRST(header);
-+ ext3_xattr_shift_entries(entry, EXT3_I(inode)->i_extra_isize -
-+ new_extra_isize, (void *)raw_inode +
-+ EXT3_GOOD_OLD_INODE_SIZE + new_extra_isize,
-+ (void *)header, total_ino,
-+ inode->i_sb->s_blocksize);
-+ EXT3_I(inode)->i_extra_isize = new_extra_isize;
-+ error = 0;
-+ goto cleanup;
-+ }
-+
-+ /*
-+ * Enough free space isn't available in the inode, check if
-+ * EA block can hold new_extra_isize bytes.
-+ */
-+ if (EXT3_I(inode)->i_file_acl) {
-+ bh = sb_bread(inode->i_sb, EXT3_I(inode)->i_file_acl);
-+ error = -EIO;
-+ if (!bh)
-+ goto cleanup;
-+ if (ext3_xattr_check_block(bh)) {
-+ ext3_error(inode->i_sb, __FUNCTION__,
-+ "inode %lu: bad block "E3FSBLK, inode->i_ino,
-+ EXT3_I(inode)->i_file_acl);
-+ error = -EIO;
-+ goto cleanup;
-+ }
-+ base = BHDR(bh);
-+ first = BFIRST(bh);
-+ end = bh->b_data + bh->b_size;
-+ min_offs = end - base;
-+ free = ext3_xattr_free_space(first, &min_offs, base,
-+ &total_blk);
-+ if (free < new_extra_isize) {
-+ if (!tried_min_extra_isize && s_min_extra_isize) {
-+ tried_min_extra_isize++;
-+ new_extra_isize = s_min_extra_isize;
-+ goto retry;
-+ }
-+ error = -1;
-+ goto cleanup;
-+ }
-+ } else {
-+ free = inode->i_sb->s_blocksize;
-+ }
-+
-+ while (new_extra_isize > 0) {
-+ size_t offs, size, entry_size;
-+ struct ext3_xattr_entry *small_entry = NULL;
-+ struct ext3_xattr_info i = {
-+ .value = NULL,
-+ .value_len = 0,
-+ };
-+ unsigned int total_size, shift_bytes, temp = ~0U;
-+
-+ is = (struct ext3_xattr_ibody_find *) kmalloc(sizeof(struct
-+ ext3_xattr_ibody_find), GFP_KERNEL);
-+ bs = (struct ext3_xattr_block_find *) kmalloc(sizeof(struct
-+ ext3_xattr_block_find), GFP_KERNEL);
-+ memset((void *)is, 0, sizeof(struct ext3_xattr_ibody_find));
-+ memset((void *)bs, 0, sizeof(struct ext3_xattr_block_find));
-+
-+ is->s.not_found = bs->s.not_found = -ENODATA;
-+ is->iloc.bh = NULL;
-+ bs->bh = NULL;
-+
-+ last = IFIRST(header);
-+ /* Find the entry best suited to be pushed into EA block */
-+ entry = NULL;
-+ for (; !IS_LAST_ENTRY(last); last = EXT3_XATTR_NEXT(last)) {
-+ total_size = EXT3_XATTR_SIZE(le32_to_cpu(last->e_value_size)) +
-+ EXT3_XATTR_LEN(last->e_name_len);
-+ if (total_size <= free && total_size < temp) {
-+ if (total_size < new_extra_isize) {
-+ small_entry = last;
-+ } else {
-+ entry = last;
-+ temp = total_size;
-+ }
-+ }
-+ }
-+
-+ if (entry == NULL) {
-+ if (small_entry) {
-+ entry = small_entry;
-+ } else {
-+ if (!tried_min_extra_isize &&
-+ s_min_extra_isize) {
-+ tried_min_extra_isize++;
-+ new_extra_isize = s_min_extra_isize;
-+ goto retry;
-+ }
-+ error = -1;
-+ goto cleanup;
-+ }
-+ }
-+ offs = le16_to_cpu(entry->e_value_offs);
-+ size = le32_to_cpu(entry->e_value_size);
-+ entry_size = EXT3_XATTR_LEN(entry->e_name_len);
-+ i.name_index = entry->e_name_index,
-+ buffer = kmalloc(EXT3_XATTR_SIZE(size), GFP_KERNEL);
-+ b_entry_name = kmalloc(entry->e_name_len + 1, GFP_KERNEL);
-+ /* Save the entry name and the entry value */
-+ memcpy((void *)buffer, (void *)IFIRST(header) + offs,
-+ EXT3_XATTR_SIZE(size));
-+ memcpy((void *)b_entry_name, (void *)entry->e_name,
-+ entry->e_name_len);
-+ b_entry_name[entry->e_name_len] = '\0';
-+ i.name = b_entry_name;
-+
-+ error = ext3_get_inode_loc(inode, &is->iloc);
-+ if (error)
-+ goto cleanup;
-+
-+ error = ext3_xattr_ibody_find(inode, &i, is);
-+ if (error)
-+ goto cleanup;
-+
-+ /* Remove the chosen entry from the inode */
-+ error = ext3_xattr_ibody_set(handle, inode, &i, is);
-+
-+ entry = IFIRST(header);
-+ if (entry_size + EXT3_XATTR_SIZE(size) >= new_extra_isize)
-+ shift_bytes = new_extra_isize;
-+ else
-+ shift_bytes = entry_size + size;
-+ /* Adjust the offsets and shift the remaining entries ahead */
-+ ext3_xattr_shift_entries(entry, EXT3_I(inode)->i_extra_isize -
-+ shift_bytes, (void *)raw_inode +
-+ EXT3_GOOD_OLD_INODE_SIZE + extra_isize + shift_bytes,
-+ (void *)header, total_ino - entry_size,
-+ inode->i_sb->s_blocksize);
-+
-+ extra_isize += shift_bytes;
-+ new_extra_isize -= shift_bytes;
-+ EXT3_I(inode)->i_extra_isize = extra_isize;
-+
-+ i.name = b_entry_name;
-+ i.value = buffer;
-+ i.value_len = cpu_to_le32(size);
-+ error = ext3_xattr_block_find(inode, &i, bs);
-+ if (error)
-+ goto cleanup;
-+
-+ /* Add entry which was removed from the inode into the block */
-+ error = ext3_xattr_block_set(handle, inode, &i, bs);
-+ if (error)
-+ goto cleanup;
-+ }
-+
-+cleanup:
-+ if (b_entry_name)
-+ kfree(b_entry_name);
-+ if (buffer)
-+ kfree(buffer);
-+ if (is) {
-+ brelse(is->iloc.bh);
-+ kfree(is);
-+ }
-+ if (bs)
-+ kfree(bs);
-+ brelse(bh);
-+ up_write(&EXT3_I(inode)->xattr_sem);
-+ return error;
-+}
-+
-+
-+
- /*
- * ext3_xattr_delete_inode()
- *
-Index: linux-2.6.18/fs/ext3/xattr.h
-===================================================================
---- linux-2.6.18.orig/fs/ext3/xattr.h
-+++ linux-2.6.18/fs/ext3/xattr.h
-@@ -74,6 +74,9 @@ extern int ext3_xattr_set_handle(handle_
- extern void ext3_xattr_delete_inode(handle_t *, struct inode *);
- extern void ext3_xattr_put_super(struct super_block *);
-
-+int ext3_expand_extra_isize(struct inode *inode, int new_extra_isize,
-+ struct ext3_iloc iloc, handle_t *handle);
-+
- extern int init_ext3_xattr(void);
- extern void exit_ext3_xattr(void);
-
-Index: linux-2.6.18/fs/ext3/ialloc.c
-===================================================================
---- linux-2.6.18.orig/fs/ext3/ialloc.c
-+++ linux-2.6.18/fs/ext3/ialloc.c
-@@ -751,6 +751,7 @@ got:
- ei->i_dtime = 0;
- ei->i_block_alloc_info = NULL;
- ei->i_block_group = group;
-+ ei->i_fs_version = 0;
-
- ext3_set_inode_flags(inode);
- if (IS_DIRSYNC(inode))
+++ /dev/null
-Index: linux-2.6.18/fs/ext3/super.c
-===================================================================
---- linux-2.6.18.orig/fs/ext3/super.c
-+++ linux-2.6.18/fs/ext3/super.c
-@@ -688,6 +688,9 @@ enum {
- Opt_user_xattr, Opt_nouser_xattr, Opt_acl, Opt_noacl,
- Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh,
- Opt_commit, Opt_journal_update, Opt_journal_inum, Opt_journal_dev,
-+#ifdef JFS_FEATURE_COMPAT_CHECKSUM
-+ Opt_journal_checksum, Opt_journal_async_commit,
-+#endif
- Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback,
- Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
- Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
-@@ -730,6 +733,10 @@ static match_table_t tokens = {
- {Opt_journal_update, "journal=update"},
- {Opt_journal_inum, "journal=%u"},
- {Opt_journal_dev, "journal_dev=%u"},
-+#ifdef JFS_FEATURE_COMPAT_CHECKSUM
-+ {Opt_journal_checksum, "journal_checksum"},
-+ {Opt_journal_async_commit, "journal_async_commit"},
-+#endif
- {Opt_abort, "abort"},
- {Opt_data_journal, "data=journal"},
- {Opt_data_ordered, "data=ordered"},
-@@ -924,6 +931,15 @@ static int parse_options (char *options,
- return 0;
- *journal_devnum = option;
- break;
-+#ifdef JFS_FEATURE_COMPAT_CHECKSUM
-+ case Opt_journal_checksum:
-+ set_opt (sbi->s_mount_opt, JOURNAL_CHECKSUM);
-+ break;
-+ case Opt_journal_async_commit:
-+ set_opt (sbi->s_mount_opt, JOURNAL_ASYNC_COMMIT);
-+ set_opt (sbi->s_mount_opt, JOURNAL_CHECKSUM);
-+ break;
-+#endif
- case Opt_noload:
- set_opt (sbi->s_mount_opt, NOLOAD);
- break;
-@@ -2180,6 +2196,23 @@ static int ext3_fill_super (struct super
- goto failed_mount3;
- }
-
-+#ifdef JFS_FEATURE_COMPAT_CHECKSUM
-+ if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) {
-+ journal_set_features(sbi->s_journal,
-+ JFS_FEATURE_COMPAT_CHECKSUM, 0,
-+ JFS_FEATURE_INCOMPAT_ASYNC_COMMIT);
-+ } else if (test_opt(sb, JOURNAL_CHECKSUM)) {
-+ journal_set_features(sbi->s_journal,
-+ JFS_FEATURE_COMPAT_CHECKSUM, 0, 0);
-+ journal_clear_features(sbi->s_journal, 0, 0,
-+ JFS_FEATURE_INCOMPAT_ASYNC_COMMIT);
-+ } else {
-+ journal_clear_features(sbi->s_journal,
-+ JFS_FEATURE_COMPAT_CHECKSUM, 0,
-+ JFS_FEATURE_INCOMPAT_ASYNC_COMMIT);
-+ }
-+#endif
-+
- /* We have now updated the journal if required, so we can
- * validate the data journaling mode. */
- switch (test_opt(sb, DATA_FLAGS)) {
-Index: linux-2.6.18/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18.orig/include/linux/ext3_fs.h
-+++ linux-2.6.18/include/linux/ext3_fs.h
-@@ -476,6 +476,11 @@ do { \
- #define EXT3_MOUNT_EXTDEBUG 0x4000000/* Extents debug */
- #define EXT3_MOUNT_MBALLOC 0x8000000/* Buddy allocation support */
-
-+#ifdef JFS_FEATURE_COMPAT_CHECKSUM
-+#define EXT3_MOUNT_JOURNAL_CHECKSUM 0x10000000 /* Journal checksums */
-+#define EXT3_MOUNT_JOURNAL_ASYNC_COMMIT 0x20000000 /* Journal Async Commit */
-+#endif
-+
- /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
- #ifndef clear_opt
- #define clear_opt(o, opt) o &= ~EXT3_MOUNT_##opt
+++ /dev/null
-Index: b/fs/ext3/namei.c
-===================================================================
---- a/fs/ext3/namei.c
-+++ b/fs/ext3/namei.c
-@@ -115,22 +115,13 @@ struct dx_entry
- * hash version mod 4 should never be 0. Sincerely, the paranoia department.
- */
-
--struct dx_root
-+struct dx_root_info
- {
-- struct fake_dirent dot;
-- char dot_name[4];
-- struct fake_dirent dotdot;
-- char dotdot_name[4];
-- struct dx_root_info
-- {
-- __le32 reserved_zero;
-- u8 hash_version;
-- u8 info_length; /* 8 */
-- u8 indirect_levels;
-- u8 unused_flags;
-- }
-- info;
-- struct dx_entry entries[0];
-+ __le32 reserved_zero;
-+ u8 hash_version;
-+ u8 info_length; /* 8 */
-+ u8 indirect_levels;
-+ u8 unused_flags;
- };
-
- struct dx_node
-@@ -174,6 +165,7 @@ typedef enum {
- #define LDP_FLAGS_RANGE 0x07
-
- #ifdef CONFIG_EXT3_INDEX
-+struct dx_root_info * dx_get_dx_info(struct ext3_dir_entry_2 *de);
- static inline unsigned dx_get_block (struct dx_entry *entry);
- static void dx_set_block (struct dx_entry *entry, unsigned value);
- static inline unsigned dx_get_hash (struct dx_entry *entry);
-@@ -210,6 +202,16 @@ static int ext3_dx_add_entry(handle_t *h
- * Future: use high four bits of block for coalesce-on-delete flags
- * Mask them off for now.
- */
-+struct dx_root_info * dx_get_dx_info(struct ext3_dir_entry_2 *de)
-+{
-+ /* get dotdot first */
-+ de = (struct ext3_dir_entry_2 *)((char *)de + le16_to_cpu(de->rec_len));
-+
-+ /* dx root info is after dotdot entry */
-+ de = (struct ext3_dir_entry_2 *)((char *)de + EXT3_DIR_REC_LEN(2));
-+
-+ return (struct dx_root_info *) de;
-+}
-
- static inline unsigned dx_get_block (struct dx_entry *entry)
- {
-@@ -362,7 +364,7 @@ dx_probe(struct dentry *dentry, struct i
- {
- unsigned count, indirect;
- struct dx_entry *at, *entries, *p, *q, *m;
-- struct dx_root *root;
-+ struct dx_root_info * info;
- struct buffer_head *bh;
- struct dx_frame *frame = frame_in;
- u32 hash;
-@@ -372,46 +374,45 @@ dx_probe(struct dentry *dentry, struct i
- dir = dentry->d_parent->d_inode;
- if (!(bh = ext3_bread (NULL,dir, 0, 0, err)))
- goto fail;
-- root = (struct dx_root *) bh->b_data;
-- if (root->info.hash_version != DX_HASH_TEA &&
-- root->info.hash_version != DX_HASH_HALF_MD4 &&
-- root->info.hash_version != DX_HASH_LEGACY) {
-+
-+ info = dx_get_dx_info((struct ext3_dir_entry_2*)bh->b_data);
-+ if (info->hash_version != DX_HASH_TEA &&
-+ info->hash_version != DX_HASH_HALF_MD4 &&
-+ info->hash_version != DX_HASH_LEGACY) {
- ext3_warning(dir->i_sb, __FUNCTION__,
- "Unrecognised inode hash code %d for directory "
-- "#%lu", root->info.hash_version, dir->i_ino);
-+ "#%lu", info->hash_version, dir->i_ino);
- brelse(bh);
- *err = ERR_BAD_DX_DIR;
- goto fail;
- }
-- hinfo->hash_version = root->info.hash_version;
-+ hinfo->hash_version = info->hash_version;
- hinfo->seed = EXT3_SB(dir->i_sb)->s_hash_seed;
- if (dentry)
- ext3fs_dirhash(dentry->d_name.name, dentry->d_name.len, hinfo);
- hash = hinfo->hash;
-
-- if (root->info.unused_flags & 1) {
-+ if (info->unused_flags & 1) {
- ext3_warning(dir->i_sb, __FUNCTION__,
- "Unimplemented inode hash flags: %#06x",
-- root->info.unused_flags);
-+ info->unused_flags);
- brelse(bh);
- *err = ERR_BAD_DX_DIR;
- goto fail;
- }
-
-- if ((indirect = root->info.indirect_levels) > 1) {
-+ if ((indirect = info->indirect_levels) > 1) {
- ext3_warning(dir->i_sb, __FUNCTION__,
- "Unimplemented inode hash depth: %#06x",
-- root->info.indirect_levels);
-+ info->indirect_levels);
- brelse(bh);
- *err = ERR_BAD_DX_DIR;
- goto fail;
- }
-
-- entries = (struct dx_entry *) (((char *)&root->info) +
-- root->info.info_length);
-+ entries = (struct dx_entry *) (((char *)info) + info->info_length);
-
-- if (dx_get_limit(entries) != dx_root_limit(dir,
-- root->info.info_length)) {
-+ if (dx_get_limit(entries) != dx_root_limit(dir, info->info_length)) {
- ext3_warning(dir->i_sb, __FUNCTION__,
- "dx entry: limit != root limit");
- brelse(bh);
-@@ -467,10 +469,12 @@ fail:
-
- static void dx_release (struct dx_frame *frames)
- {
-+ struct dx_root_info *info;
-+
- if (frames[0].bh == NULL)
- return;
--
-- if (((struct dx_root *) frames[0].bh->b_data)->info.indirect_levels)
-+ info = dx_get_dx_info((struct ext3_dir_entry_2*)frames[0].bh->b_data);
-+ if (info->indirect_levels)
- brelse(frames[1].bh);
- brelse(frames[0].bh);
- }
-@@ -1419,17 +1423,16 @@ static int make_indexed_dir(handle_t *ha
- const char *name = dentry->d_name.name;
- int namelen = dentry->d_name.len;
- struct buffer_head *bh2;
-- struct dx_root *root;
- struct dx_frame frames[2], *frame;
- struct dx_entry *entries;
-- struct ext3_dir_entry_2 *de, *de2;
-+ struct ext3_dir_entry_2 *de, *de2, *dot_de, *dotdot_de;
- char *data1, *top;
- unsigned len;
- int retval;
- unsigned blocksize;
- struct dx_hash_info hinfo;
- u32 block;
-- struct fake_dirent *fde;
-+ struct dx_root_info *dx_info;
-
- blocksize = dir->i_sb->s_blocksize;
- dxtrace(printk("Creating index\n"));
-@@ -1439,7 +1442,7 @@ static int make_indexed_dir(handle_t *ha
- brelse(bh);
- return retval;
- }
-- root = (struct dx_root *) bh->b_data;
-+ dot_de = (struct ext3_dir_entry_2 *) bh->b_data;
-
- bh2 = ext3_append (handle, dir, &block, &retval);
- if (!(bh2)) {
-@@ -1450,9 +1453,9 @@ static int make_indexed_dir(handle_t *ha
- data1 = bh2->b_data;
-
- /* The 0th block becomes the root, move the dirents out */
-- fde = &root->dotdot;
-- de = (struct ext3_dir_entry_2 *)((char *)fde + le16_to_cpu(fde->rec_len));
-- len = ((char *) root) + blocksize - (char *) de;
-+ dotdot_de = (struct ext3_dir_entry_2 *)((char *)dot_de + le16_to_cpu(dot_de->rec_len));
-+ de = (struct ext3_dir_entry_2 *)((char *)dotdot_de + le16_to_cpu(dotdot_de->rec_len));
-+ len = ((char *) dot_de) + blocksize - (char *) de;
- memcpy (data1, de, len);
- de = (struct ext3_dir_entry_2 *) data1;
- top = data1 + len;
-@@ -1460,18 +1463,21 @@ static int make_indexed_dir(handle_t *ha
- de = de2;
- de->rec_len = cpu_to_le16(data1 + blocksize - (char *) de);
- /* Initialize the root; the dot dirents already exist */
-- de = (struct ext3_dir_entry_2 *) (&root->dotdot);
-- de->rec_len = cpu_to_le16(blocksize - EXT3_DIR_REC_LEN(2));
-- memset (&root->info, 0, sizeof(root->info));
-- root->info.info_length = sizeof(root->info);
-- root->info.hash_version = EXT3_SB(dir->i_sb)->s_def_hash_version;
-- entries = root->entries;
-+ dotdot_de->rec_len = cpu_to_le16(blocksize - le16_to_cpu(dot_de->rec_len));
-+
-+ /* initialize hashing info */
-+ dx_info = dx_get_dx_info(dot_de);
-+ memset (dx_info, 0, sizeof(*dx_info));
-+ dx_info->info_length = sizeof(*dx_info);
-+ dx_info->hash_version = EXT3_SB(dir->i_sb)->s_def_hash_version;
-+
-+ entries = (void *)dx_info + sizeof(*dx_info);
- dx_set_block (entries, 1);
- dx_set_count (entries, 1);
-- dx_set_limit (entries, dx_root_limit(dir, sizeof(root->info)));
-+ dx_set_limit (entries, dx_root_limit(dir, sizeof(*dx_info)));
-
- /* Initialize as for dx_probe */
-- hinfo.hash_version = root->info.hash_version;
-+ hinfo.hash_version = dx_info->hash_version;
- hinfo.seed = EXT3_SB(dir->i_sb)->s_hash_seed;
- ext3fs_dirhash(name, namelen, &hinfo);
- frame = frames;
-@@ -1720,6 +1726,7 @@ static int ext3_dx_add_entry(handle_t *h
- goto journal_error;
- brelse (bh2);
- } else {
-+ struct dx_root_info *info;
- dxtrace(printk("Creating second level index...\n"));
- memcpy((char *) entries2, (char *) entries,
- icount * sizeof(struct dx_entry));
-@@ -1728,7 +1735,8 @@ static int ext3_dx_add_entry(handle_t *h
- /* Set up root */
- dx_set_count(entries, 1);
- dx_set_block(entries + 0, newblock);
-- ((struct dx_root *) frames[0].bh->b_data)->info.indirect_levels = 1;
-+ info = dx_get_dx_info((struct ext3_dir_entry_2*)frames[0].bh->b_data);
-+ info->indirect_levels = 1;
-
- /* Add new access path frame */
- frame = frames + 1;
+++ /dev/null
-Index: linux-2.6.9-full/fs/ext3/namei.c
-===================================================================
---- linux-2.6.9-full.orig/fs/ext3/namei.c 2006-05-06 01:21:10.000000000 +0400
-+++ linux-2.6.9-full/fs/ext3/namei.c 2006-05-06 01:29:30.000000000 +0400
-@@ -1003,6 +1003,38 @@ static struct dentry *ext3_lookup(struct
-
- return ERR_PTR(-EACCES);
- }
-+ /* ".." shouldn't go into dcache to preserve dcache hierarchy
-+ * otherwise we'll get parent being a child of actual child.
-+ * see bug 10458 for details -bzzz */
-+ if (inode && (dentry->d_name.name[0] == '.' && (dentry->d_name.len == 1 ||
-+ (dentry->d_name.len == 2 && dentry->d_name.name[1] == '.')))) {
-+ struct dentry *tmp, *goal = NULL;
-+ struct list_head *lp;
-+
-+ /* first, look for an existing dentry - any one is good */
-+ spin_lock(&dcache_lock);
-+ list_for_each(lp, &inode->i_dentry) {
-+ tmp = list_entry(lp, struct dentry, d_alias);
-+ goal = tmp;
-+ dget_locked(goal);
-+ break;
-+ }
-+ if (goal == NULL) {
-+ /* there is no alias, we need to make current dentry:
-+ * a) inaccessible for __d_lookup()
-+ * b) inaccessible for iopen */
-+ J_ASSERT(list_empty(&dentry->d_alias));
-+ dentry->d_flags |= DCACHE_NFSFS_RENAMED;
-+ /* this is d_instantiate() ... */
-+ list_add(&dentry->d_alias, &inode->i_dentry);
-+ dentry->d_inode = inode;
-+ }
-+ spin_unlock(&dcache_lock);
-+ if (goal)
-+ iput(inode);
-+ return goal;
-+ }
-+
- return d_splice_alias(inode, dentry);
- }
-
+++ /dev/null
-Index: linux-2.6.18.8/fs/ext3/inode.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/inode.c 2007-06-20 18:10:33.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/inode.c 2007-06-20 18:53:48.000000000 +0200
-@@ -3222,3 +3222,66 @@ int ext3_change_inode_journal_flag(struc
-
- return err;
- }
-+
-+int ext3_map_inode_page(struct inode *inode, struct page *page,
-+ unsigned long *blocks, int *created, int create)
-+{
-+ unsigned int blocksize, blocks_per_page;
-+ unsigned long iblock;
-+ struct buffer_head dummy;
-+ void *handle;
-+ int i, rc = 0, failed = 0, needed_blocks;
-+
-+ blocksize = inode->i_sb->s_blocksize;
-+ blocks_per_page = PAGE_SIZE >> inode->i_sb->s_blocksize_bits;
-+ iblock = page->index * blocks_per_page;
-+
-+ for (i = 0; i < blocks_per_page; i++, iblock++) {
-+ blocks[i] = ext3_bmap(inode->i_mapping, iblock);
-+ if (blocks[i] == 0) {
-+ failed++;
-+ if (created)
-+ created[i] = -1;
-+ } else if (created) {
-+ created[i] = 0;
-+ }
-+ }
-+
-+ if (failed == 0 || create == 0)
-+ return 0;
-+
-+ needed_blocks = ext3_writepage_trans_blocks(inode);
-+ handle = ext3_journal_start(inode, needed_blocks);
-+ if (IS_ERR(handle))
-+ return PTR_ERR(handle);
-+
-+ iblock = page->index * blocks_per_page;
-+ for (i = 0; i < blocks_per_page; i++, iblock++) {
-+ if (blocks[i] != 0)
-+ continue;
-+
-+ rc = ext3_get_blocks_handle(handle, inode, iblock, 1, &dummy, 1, 1);
-+ if (rc < 0) {
-+ printk(KERN_INFO "ext3_map_inode_page: error reading "
-+ "block %ld\n", iblock);
-+ goto out;
-+ } else {
-+ if (rc > 1)
-+ WARN_ON(1);
-+ rc = 0;
-+ }
-+ /* Unmap any metadata buffers from the block mapping, to avoid
-+ * data corruption due to direct-write from Lustre being
-+ * clobbered by a later flush of the blockdev metadata buffer.*/
-+ if (buffer_new(&dummy))
-+ unmap_underlying_metadata(dummy.b_bdev,
-+ dummy.b_blocknr);
-+ blocks[i] = dummy.b_blocknr;
-+ if (created)
-+ created[i] = 1;
-+ }
-+
-+out:
-+ ext3_journal_stop(handle);
-+ return rc;
-+}
-Index: linux-2.6.18.8/fs/ext3/super.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/super.c 2007-06-20 18:10:33.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/super.c 2007-06-20 18:51:12.000000000 +0200
-@@ -2765,6 +2765,10 @@ static void __exit exit_ext3_fs(void)
- exit_ext3_xattr();
- }
-
-+int ext3_map_inode_page(struct inode *inode, struct page *page,
-+ unsigned long *blocks, int *created, int create);
-+EXPORT_SYMBOL(ext3_map_inode_page);
-+
- MODULE_AUTHOR("Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others");
- MODULE_DESCRIPTION("Second Extended Filesystem with journaling extensions");
- MODULE_LICENSE("GPL");
+++ /dev/null
-Index: linux-stage/fs/ext3/ialloc.c
-===================================================================
---- linux-stage.orig/fs/ext3/ialloc.c
-+++ linux-stage/fs/ext3/ialloc.c
-@@ -521,12 +521,15 @@ struct inode *ext3_new_inode(handle_t
- return ERR_PTR(-EPERM);
-
- sb = dir->i_sb;
-+ sbi = EXT3_SB(sb);
-+ if (sbi->s_max_dir_size > 0 && i_size_read(dir) >= sbi->s_max_dir_size)
-+ return ERR_PTR(-EFBIG);
-+
- inode = new_inode(sb);
- if (!inode)
- return ERR_PTR(-ENOMEM);
- ei = EXT3_I(inode);
-
-- sbi = EXT3_SB(sb);
- es = sbi->s_es;
- if (goal) {
- group = (goal - 1) / EXT3_INODES_PER_GROUP(sb);
-Index: linux-stage/fs/ext3/super.c
-===================================================================
---- linux-stage.orig/fs/ext3/super.c
-+++ linux-stage/fs/ext3/super.c
-@@ -45,6 +45,12 @@
- #include "namei.h"
- #include "group.h"
-
-+/*
-+ * max directory size tunable
-+ */
-+#define EXT3_DEFAULT_MAX_DIR_SIZE 0
-+#define EXT3_MAX_DIR_SIZE_NAME "max_dir_size"
-+
- static int ext3_load_journal(struct super_block *, struct ext3_super_block *,
- unsigned long journal_devnum);
- static int ext3_create_journal(struct super_block *, struct ext3_super_block *,
-@@ -444,6 +450,7 @@ static void ext3_put_super (struct su
- }
- if (sbi->s_mmp_tsk)
- kthread_stop(sbi->s_mmp_tsk);
-+ remove_proc_entry(EXT3_MAX_DIR_SIZE_NAME, sbi->s_dev_proc);
- if (sbi->s_dev_proc) {
- remove_proc_entry(sbi->s_dev_proc->name, proc_root_ext3);
- sbi->s_dev_proc = NULL;
-@@ -703,7 +710,7 @@ enum {
- Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
- Opt_grpquota,
- Opt_extents, Opt_noextents, Opt_extdebug,
-- Opt_mballoc, Opt_nomballoc, Opt_stripe,
-+ Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_maxdirsize
- };
-
- static match_table_t tokens = {
-@@ -762,8 +769,9 @@ static match_table_t tokens = {
- {Opt_mballoc, "mballoc"},
- {Opt_nomballoc, "nomballoc"},
- {Opt_stripe, "stripe=%u"},
-- {Opt_err, NULL},
- {Opt_resize, "resize"},
-+ {Opt_maxdirsize, "maxdirsize=%u"},
-+ {Opt_err, NULL}
- };
-
- static ext3_fsblk_t get_sb_block(void **data)
-@@ -1128,6 +1136,13 @@ clear_qf_name:
- return 0;
- sbi->s_stripe = option;
- break;
-+ case Opt_maxdirsize:
-+ if (match_int(&args[0], &option))
-+ return 0;
-+ if (option <= 0)
-+ return 0;
-+ sbi->s_max_dir_size = option;
-+ break;
- default:
- printk (KERN_ERR
- "EXT3-fs: Unrecognized mount option \"%s\" "
-@@ -1875,6 +1890,45 @@ failed:
- return 1;
- }
-
-+static int ext3_max_dir_size_read(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
-+{
-+ struct ext3_sb_info *sbi = data;
-+ int len;
-+
-+ *eof = 1;
-+ if (off != 0)
-+ return 0;
-+
-+ len = sprintf(page, "%lu\n", sbi->s_max_dir_size);
-+ *start = page;
-+ return len;
-+}
-+
-+static int ext3_max_dir_size_write(struct file *file, const char *buffer,
-+ unsigned long count, void *data)
-+{
-+ struct ext3_sb_info *sbi = data;
-+ char str[32];
-+ unsigned long value;
-+ char *end;
-+
-+ if (count >= sizeof(str)) {
-+ printk(KERN_ERR "EXT3-fs: %s string too long, max %u bytes\n",
-+ EXT3_MAX_DIR_SIZE_NAME, (int)sizeof(str));
-+ return -EOVERFLOW;
-+ }
-+
-+ if (copy_from_user(str, buffer, count))
-+ return -EFAULT;
-+
-+ value = simple_strtol(str, &end, 0);
-+ if (value < 0)
-+ return -ERANGE;
-+
-+ sbi->s_max_dir_size = value;
-+ return count;
-+}
-
- static int ext3_fill_super (struct super_block *sb, void *data, int silent)
- {
-@@ -1896,6 +1950,7 @@ static int ext3_fill_super (struct su
- int i;
- int needs_recovery;
- __le32 features;
-+ struct proc_dir_entry *proc;
-
- sbi = kmalloc(sizeof(*sbi), GFP_KERNEL);
- if (!sbi)
-@@ -1924,6 +1979,23 @@ static int ext3_fill_super (struct su
-
- unlock_kernel();
-
-+ sbi->s_max_dir_size = EXT3_DEFAULT_MAX_DIR_SIZE;
-+ proc = create_proc_entry(EXT3_MAX_DIR_SIZE_NAME,
-+ S_IFREG | S_IRUGO | S_IWUSR, sbi->s_dev_proc);
-+ if (proc == NULL) {
-+ printk(KERN_ERR "EXT3-fs: unable to create %s\n",
-+ EXT3_MAX_DIR_SIZE_NAME);
-+ remove_proc_entry(EXT3_MAX_DIR_SIZE_NAME, sbi->s_dev_proc);
-+ remove_proc_entry(sb->s_id, proc_root_ext3);
-+ sbi->s_dev_proc = NULL;
-+ sb->s_fs_info = NULL;
-+ kfree(sbi);
-+ return -ENOMEM;
-+ }
-+ proc->data = sbi;
-+ proc->read_proc = ext3_max_dir_size_read;
-+ proc->write_proc = ext3_max_dir_size_write;
-+
- blocksize = sb_min_blocksize(sb, EXT3_MIN_BLOCK_SIZE);
- if (!blocksize) {
- printk(KERN_ERR "EXT3-fs: unable to set blocksize\n");
-@@ -2361,6 +2433,7 @@ failed_mount:
- ext3_blkdev_remove(sbi);
- brelse(bh);
- out_fail:
-+ remove_proc_entry(EXT3_MAX_DIR_SIZE_NAME, sbi->s_dev_proc);
- if (sbi->s_dev_proc) {
- remove_proc_entry(sbi->s_dev_proc->name, proc_root_ext3);
- sbi->s_dev_proc = NULL;
-Index: linux-stage/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs_sb.h
-+++ linux-stage/include/linux/ext3_fs_sb.h
-@@ -132,6 +132,8 @@ struct ext3_sb_info {
- unsigned long s_mb_last_group;
- unsigned long s_mb_last_start;
-
-+ unsigned long s_max_dir_size;
-+
- /* history to debug policy */
- struct ext3_mb_history *s_mb_history;
- int s_mb_history_cur;
+++ /dev/null
-Index: linux-stage/fs/ext3/mballoc.c
-===================================================================
---- linux-stage.orig/fs/ext3/mballoc.c 2010-01-26 23:36:25.000000000 +0800
-+++ linux-stage/fs/ext3/mballoc.c 2010-01-26 23:42:57.000000000 +0800
-@@ -35,6 +35,7 @@
- #include <linux/pagemap.h>
- #include <linux/seq_file.h>
- #include <linux/version.h>
-+#include <linux/genhd.h>
-
- #include "group.h"
-
-@@ -360,6 +361,7 @@
- unsigned short pa_free; /* how many blocks are free */
- unsigned short pa_linear; /* consumed in one direction
- * strictly, for group prealloc */
-+ unsigned short pa_error;
- spinlock_t *pa_obj_lock;
- struct inode *pa_inode; /* hack, for history only */
- };
-@@ -3558,6 +3560,7 @@
- spin_lock_init(&pa->pa_lock);
- pa->pa_deleted = 0;
- pa->pa_linear = 0;
-+ pa->pa_error = 0;
-
- mb_debug("new inode pa %p: %lu/%lu for %lu\n", pa,
- pa->pa_pstart, pa->pa_len, pa->pa_lstart);
-@@ -3615,6 +3618,7 @@
- spin_lock_init(&pa->pa_lock);
- pa->pa_deleted = 0;
- pa->pa_linear = 1;
-+ pa->pa_error = 0;
-
- mb_debug("new group pa %p: %lu/%lu for %lu\n", pa,
- pa->pa_pstart, pa->pa_len, pa->pa_lstart);
-@@ -3671,7 +3675,10 @@
- sector_t start;
- int err = 0, free = 0;
-
-+ BUG_ON(!ext3_is_group_locked(sb, e3b->bd_group));
- BUG_ON(pa->pa_deleted == 0);
-+ BUG_ON(pa->pa_linear != 0);
-+ BUG_ON(pa->pa_inode == NULL);
- ext3_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit);
- BUG_ON(group != e3b->bd_group && pa->pa_len != 0);
- end = bit + pa->pa_len;
-@@ -3704,14 +3711,19 @@
- mb_free_blocks(pa->pa_inode, e3b, bit, next - bit);
- bit = next + 1;
- }
-- if (free != pa->pa_free) {
-- printk("pa %p: logic %lu, phys. %lu, len %lu\n",
-- pa, (unsigned long) pa->pa_lstart,
-- (unsigned long) pa->pa_pstart,
-- (unsigned long) pa->pa_len);
-- printk("free %u, pa_free %u\n", free, pa->pa_free);
-- }
-- BUG_ON(free != pa->pa_free);
-+
-+ /* "free < pa->pa_free" means we maybe double alloc the same blocks,
-+ * otherwise maybe leave some free blocks unavailable, no need to BUG.*/
-+ if ((free > pa->pa_free && !pa->pa_error) || (free < pa->pa_free))
-+ ext3_error(sb, __FUNCTION__, "pa free mismatch: [pa %p] "
-+ "[phy %lu] [logic %lu] [len %u] [free %u] "
-+ "[error %u] [inode %lu] [freed %u]", pa,
-+ (unsigned long)pa->pa_pstart,
-+ (unsigned long)pa->pa_lstart,
-+ (unsigned)pa->pa_len, (unsigned)pa->pa_free,
-+ (unsigned)pa->pa_error, pa->pa_inode->i_ino,
-+ free);
-+ BUG_ON(pa->pa_free != free);
- atomic_add(free, &sbi->s_mb_discarded);
-
- return err;
-@@ -4189,6 +4201,19 @@
- *errp = -EDQUOT;
- return 0;
- }
-+
-+ if (dev_check_rdonly(sb->s_bdev)) {
-+ struct block_device *bdev = sb->s_bdev;
-+
-+ printk(KERN_WARNING "Alloc from readonly device %s (%#x): "
-+ "[inode %lu] [logic %lu] [goal %lu] [ll %lu] [pl %lu] "
-+ "[lr %lu] [pr %lu] [len %lu] [flags %lu]\n",
-+ bdev->bd_disk ? bdev->bd_disk->disk_name : "",
-+ bdev->bd_dev, ar->inode->i_ino, ar->logical, ar->goal,
-+ ar->lleft, ar->pleft, ar->lright, ar->pright, ar->len,
-+ ar->flags);
-+ }
-+
- inquota = ar->len;
-
- ext3_mb_poll_new_transaction(sb, handle);
-@@ -4217,10 +4242,34 @@
- }
-
- if (likely(ac.ac_status == AC_STATUS_FOUND)) {
-- ext3_mb_mark_diskspace_used(&ac, handle);
-- *errp = 0;
-- block = ext3_grp_offs_to_block(sb, &ac.ac_b_ex);
-- ar->len = ac.ac_b_ex.fe_len;
-+ *errp = ext3_mb_mark_diskspace_used(&ac, handle);
-+ if (!*errp) {
-+ block = ext3_grp_offs_to_block(sb, &ac.ac_b_ex);
-+ ar->len = ac.ac_b_ex.fe_len;
-+ } else {
-+ ac.ac_b_ex.fe_len = 0;
-+ ar->len = 0;
-+ ext3_mb_show_ac(&ac);
-+ if (ac.ac_pa) {
-+ struct ext3_prealloc_space *pa = ac.ac_pa;
-+
-+ /* We can not make sure whether the bitmap has
-+ * been updated or not when fail case. So can
-+ * not revert pa_free back, just mark pa_error*/
-+ pa->pa_error++;
-+ ext3_error(sb, __FUNCTION__,
-+ "Updating bitmap error: [err %d] "
-+ "[pa %p] [phy %lu] [logic %lu] "
-+ "[len %u] [free %u] [error %u] "
-+ "[inode %lu]", *errp, pa,
-+ (unsigned long)pa->pa_pstart,
-+ (unsigned long)pa->pa_lstart,
-+ (unsigned)pa->pa_len,
-+ (unsigned)pa->pa_free,
-+ (unsigned)pa->pa_error,
-+ pa->pa_inode ? pa->pa_inode->i_ino : 0);
-+ }
-+ }
- } else {
- freed = ext3_mb_discard_preallocations(sb, ac.ac_o_ex.fe_len);
- if (freed)
-@@ -4388,6 +4437,15 @@
- goto error_return;
- }
-
-+ if (dev_check_rdonly(sb->s_bdev)) {
-+ struct block_device *bdev = sb->s_bdev;
-+
-+ printk(KERN_WARNING "Release to readonly device %s (%#x): "
-+ "[inode %lu] [block %lu] [count %lu] [is_meta %d]\n",
-+ bdev->bd_disk ? bdev->bd_disk->disk_name : "",
-+ bdev->bd_dev, inode->i_ino, block, count, metadata);
-+ }
-+
- ext3_debug("freeing block %lu\n", block);
-
- ac.ac_op = EXT3_MB_HISTORY_FREE;
+++ /dev/null
-Index: linux-2.6.18.8/include/linux/ext3_fs_i.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_i.h
-+++ linux-2.6.18.8/include/linux/ext3_fs_i.h
-@@ -154,6 +154,10 @@ struct ext3_inode_info {
- struct inode vfs_inode;
-
- struct ext3_ext_cache i_cached_extent;
-+
-+ /* mballoc */
-+ struct list_head i_prealloc_list;
-+ spinlock_t i_prealloc_lock;
- };
-
- #endif /* _LINUX_EXT3_FS_I */
-Index: linux-2.6.18.8/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_sb.h
-+++ linux-2.6.18.8/include/linux/ext3_fs_sb.h
-@@ -21,8 +21,15 @@
- #include <linux/wait.h>
- #include <linux/blockgroup_lock.h>
- #include <linux/percpu_counter.h>
-+#include <linux/list.h>
- #endif
- #include <linux/rbtree.h>
-+#include <linux/proc_fs.h>
-+
-+struct ext3_buddy_group_blocks;
-+struct ext3_locality_group;
-+struct ext3_mb_history;
-+#define EXT3_BB_MAX_BLOCKS
-
- /*
- * third extended-fs super-block data in memory
-Index: linux-2.6.18.8/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs.h
-+++ linux-2.6.18.8/include/linux/ext3_fs.h
-@@ -17,6 +17,7 @@
- #define _LINUX_EXT3_FS_H
-
- #include <linux/types.h>
-+#include <linux/ext3_fs_i.h>
-
- /*
- * The second extended filesystem constants/structures
-@@ -67,12 +68,12 @@
-
- struct ext3_allocation_request {
- struct inode *inode; /* target inode for block we're allocating */
-- unsigned long logical; /* logical block in target inode */
-- unsigned long goal; /* phys. target (a hint) */
-- unsigned long lleft; /* the closest logical allocated block to the left */
-- unsigned long pleft; /* phys. block for ^^^ */
-- unsigned long lright; /* the closest logical allocated block to the right */
-- unsigned long pright; /* phys. block for ^^^ */
-+ ext3_fsblk_t logical; /* logical block in target inode */
-+ ext3_fsblk_t goal; /* phys. target (a hint) */
-+ ext3_fsblk_t lleft; /* the closest logical allocated block to the left */
-+ ext3_fsblk_t pleft; /* phys. block for ^^^ */
-+ ext3_fsblk_t lright; /* the closest logical allocated block to the right */
-+ ext3_fsblk_t pright; /* phys. block for ^^^ */
- unsigned long len; /* how many blocks we want to allocate */
- unsigned long flags; /* flags. see above EXT3_MB_HINT_* */
- };
-@@ -400,6 +401,7 @@ struct ext3_inode {
- #define EXT3_MOUNT_IOPEN_NOPRIV 0x800000/* Make iopen world-readable */
- #define EXT3_MOUNT_EXTENTS 0x2000000/* Extents support */
- #define EXT3_MOUNT_EXTDEBUG 0x4000000/* Extents debug */
-+#define EXT3_MOUNT_MBALLOC 0x8000000/* Buddy allocation support */
-
- /* Compatibility, for having both ext2_fs.h and ext3_fs.h included at once */
- #ifndef clear_opt
-@@ -787,12 +789,12 @@ ext3_group_first_block_no(struct super_b
- /* balloc.c */
- extern int ext3_bg_has_super(struct super_block *sb, int group);
- extern unsigned long ext3_bg_num_gdb(struct super_block *sb, int group);
--extern ext3_fsblk_t ext3_new_block (handle_t *handle, struct inode *inode,
-+extern ext3_fsblk_t ext3_new_block_old (handle_t *handle, struct inode *inode,
- ext3_fsblk_t goal, int *errp);
--extern ext3_fsblk_t ext3_new_blocks (handle_t *handle, struct inode *inode,
-+extern ext3_fsblk_t ext3_new_blocks_old (handle_t *handle, struct inode *inode,
- ext3_fsblk_t goal, unsigned long *count, int *errp);
- extern void ext3_free_blocks (handle_t *handle, struct inode *inode,
-- ext3_fsblk_t block, unsigned long count);
-+ ext3_fsblk_t block, unsigned long count, int metadata);
- extern void ext3_free_blocks_sb (handle_t *handle, struct super_block *sb,
- ext3_fsblk_t block, unsigned long count,
- unsigned long *pdquot_freed_blocks);
-@@ -836,15 +838,45 @@ extern long ext3_mb_stats;
- extern long ext3_mb_max_to_scan;
- extern int ext3_mb_init(struct super_block *, int);
- extern int ext3_mb_release(struct super_block *);
--extern unsigned long ext3_mb_new_blocks(handle_t *, struct ext3_allocation_request *, int *);
-+extern ext3_fsblk_t ext3_new_block(handle_t *handle, struct inode *inode,
-+ ext3_fsblk_t goal, int *errp);
-+extern ext3_fsblk_t ext3_mb_new_blocks(handle_t *,
-+ struct ext3_allocation_request *, int *);
- extern int ext3_mb_reserve_blocks(struct super_block *, int);
- extern void ext3_mb_release_blocks(struct super_block *, int);
- extern void ext3_mb_release_blocks(struct super_block *, int);
- extern void ext3_mb_discard_inode_preallocations(struct inode *);
- extern int __init init_ext3_mb_proc(void);
- extern void exit_ext3_mb_proc(void);
--extern void ext3_mb_free_blocks(handle_t *, struct inode *, unsigned long, unsigned long, int, int *);
-+extern void ext3_mb_free_blocks(handle_t *, struct inode *, unsigned long,
-+ unsigned long, int, unsigned long *);
-+
-+static inline ext3_fsblk_t ext3_new_blocks(handle_t *handle,
-+ struct inode *inode,
-+ ext3_fsblk_t goal,
-+ unsigned long *count, int *errp)
-+{
-+ struct ext3_allocation_request ar;
-+ ext3_fsblk_t ret;
-
-+ if (!test_opt(inode->i_sb, MBALLOC)) {
-+ ret = ext3_new_blocks_old(handle, inode, goal, count, errp);
-+ return ret;
-+ }
-+
-+ ar.inode = inode;
-+ ar.goal = goal;
-+ ar.len = *count;
-+ ar.logical = 0;
-+ ar.lleft = 0;
-+ ar.pleft = 0;
-+ ar.lright = 0;
-+ ar.pright = 0;
-+ ar.flags = 0;
-+ ret = ext3_mb_new_blocks(handle, &ar, errp);
-+ *count = ar.len;
-+ return ret;
-+}
-
- /* inode.c */
- int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode,
-Index: linux-2.6.18.8/fs/ext3/super.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/super.c
-+++ linux-2.6.18.8/fs/ext3/super.c
-@@ -688,6 +688,7 @@ enum {
- Opt_iopen, Opt_noiopen, Opt_iopen_nopriv,
- Opt_grpquota,
- Opt_extents, Opt_noextents, Opt_extdebug,
-+ Opt_mballoc, Opt_nomballoc, Opt_stripe,
- };
-
- static match_table_t tokens = {
-@@ -743,6 +744,9 @@ static match_table_t tokens = {
- {Opt_extents, "extents"},
- {Opt_noextents, "noextents"},
- {Opt_extdebug, "extdebug"},
-+ {Opt_mballoc, "mballoc"},
-+ {Opt_nomballoc, "nomballoc"},
-+ {Opt_stripe, "stripe=%u"},
- {Opt_err, NULL},
- {Opt_resize, "resize"},
- };
-@@ -1096,6 +1100,19 @@ clear_qf_name:
- case Opt_extdebug:
- set_opt (sbi->s_mount_opt, EXTDEBUG);
- break;
-+ case Opt_mballoc:
-+ set_opt(sbi->s_mount_opt, MBALLOC);
-+ break;
-+ case Opt_nomballoc:
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ break;
-+ case Opt_stripe:
-+ if (match_int(&args[0], &option))
-+ return 0;
-+ if (option < 0)
-+ return 0;
-+ sbi->s_stripe = option;
-+ break;
- default:
- printk (KERN_ERR
- "EXT3-fs: Unrecognized mount option \"%s\" "
-@@ -1826,6 +1843,7 @@ static int ext3_fill_super (struct super
- "writeback");
-
- ext3_ext_init(sb);
-+ ext3_mb_init(sb, needs_recovery);
-
- lock_kernel();
- return 0;
-Index: linux-2.6.18.8/fs/ext3/extents.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/extents.c
-+++ linux-2.6.18.8/fs/ext3/extents.c
-@@ -801,7 +801,7 @@ cleanup:
- for (i = 0; i < depth; i++) {
- if (!ablocks[i])
- continue;
-- ext3_free_blocks(handle, inode, ablocks[i], 1);
-+ ext3_free_blocks(handle, inode, ablocks[i], 1, 1);
- }
- }
- kfree(ablocks);
-@@ -1619,7 +1619,7 @@ int ext3_ext_rm_idx(handle_t *handle, st
- ext_debug(inode, "index is empty, remove it, free block %lu\n", leaf);
- bh = sb_find_get_block(inode->i_sb, leaf);
- ext3_forget(handle, 1, inode, bh, leaf);
-- ext3_free_blocks(handle, inode, leaf, 1);
-+ ext3_free_blocks(handle, inode, leaf, 1, 1);
- return err;
- }
-
-@@ -1678,7 +1678,7 @@ static int ext3_remove_blocks(handle_t *
- unsigned long from, unsigned long to)
- {
- struct buffer_head *bh;
-- int i;
-+ int i, metadata = 0;
-
- #ifdef EXTENTS_STATS
- {
-@@ -1696,6 +1696,8 @@ static int ext3_remove_blocks(handle_t *
- spin_unlock(&sbi->s_ext_stats_lock);
- }
- #endif
-+ if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
-+ metadata = 1;
- if (from >= le32_to_cpu(ex->ee_block)
- && to == le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - 1) {
- /* tail removal */
-@@ -1707,7 +1709,7 @@ static int ext3_remove_blocks(handle_t *
- bh = sb_find_get_block(inode->i_sb, start + i);
- ext3_forget(handle, 0, inode, bh, start + i);
- }
-- ext3_free_blocks(handle, inode, start, num);
-+ ext3_free_blocks(handle, inode, start, num, metadata);
- } else if (from == le32_to_cpu(ex->ee_block)
- && to <= le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len) - 1) {
- printk("strange request: removal %lu-%lu from %u:%u\n",
-@@ -2040,7 +2042,7 @@ int ext3_ext_get_blocks(handle_t *handle
- struct ext3_extent newex, *ex;
- int goal, newblock, err = 0, depth;
- unsigned long allocated = 0;
-- unsigned long next;
-+ struct ext3_allocation_request ar;
-
- __clear_bit(BH_New, &bh_result->b_state);
- ext_debug(inode, "blocks %d/%lu requested for inode %u\n", (int) iblock,
-@@ -2122,18 +2124,36 @@ int ext3_ext_get_blocks(handle_t *handle
- if (S_ISREG(inode->i_mode) && (!EXT3_I(inode)->i_block_alloc_info))
- ext3_init_block_alloc_info(inode);
-
-+ /* find neighbour allocated blocks */
-+ ar.lleft = iblock;
-+ err = ext3_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
-+ if (err)
-+ goto out2;
-+ ar.lright = iblock;
-+ err = ext3_ext_search_right(inode, path, &ar.lright, &ar.pright);
-+ if (err)
-+ goto out2;
-+
- /* find next allocated block so that we know how many
- * blocks we can allocate without ovelapping next extent */
-- BUG_ON(iblock < le32_to_cpu(ex->ee_block) + le16_to_cpu(ex->ee_len));
-- next = ext3_ext_next_allocated_block(path);
-- BUG_ON(next <= iblock);
-- allocated = next - iblock;
-+ BUG_ON(ar.pright != 0 && ar.lright <= iblock);
-+ if (ar.pright == 0)
-+ allocated = EXT_MAX_BLOCK - iblock;
-+ else
-+ allocated = ar.lright - iblock;
- if (allocated > max_blocks)
- allocated = max_blocks;
-
- /* allocate new block */
-- goal = ext3_ext_find_goal(inode, path, iblock);
-- newblock = ext3_new_blocks(handle, inode, goal, &allocated, &err);
-+ ar.inode = inode;
-+ ar.goal = ext3_ext_find_goal(inode, path, iblock);
-+ ar.logical = iblock;
-+ ar.len = allocated;
-+ if (S_ISREG(inode->i_mode))
-+ ar.flags = EXT3_MB_HINT_DATA;
-+ else
-+ ar.flags = 0;
-+ newblock = ext3_mb_new_blocks(handle, &ar, &err);
- if (!newblock)
- goto out2;
- ext_debug(inode, "allocate new block: goal %d, found %d/%lu\n",
-@@ -2143,12 +2163,16 @@ int ext3_ext_get_blocks(handle_t *handle
- newex.ee_block = cpu_to_le32(iblock);
- newex.ee_start = cpu_to_le32(newblock);
- newex.ee_start_hi = 0;
-- newex.ee_len = cpu_to_le16(allocated);
-+ newex.ee_len = cpu_to_le16(ar.len);
- err = ext3_ext_insert_extent(handle, inode, path, &newex);
- if (err) {
- /* free data blocks we just allocated */
-- ext3_free_blocks(handle, inode, le32_to_cpu(newex.ee_start),
-- le16_to_cpu(newex.ee_len));
-+ /* not a good idea to call discard here directly,
-+ * but otherwise we'd need to call it every free() */
-+ ext3_mb_discard_inode_preallocations(inode);
-+ ext3_free_blocks(handle, inode, newex.ee_start,
-+ newex.ee_len, 0);
-+
- goto out2;
- }
-
-@@ -2157,6 +2181,7 @@ int ext3_ext_get_blocks(handle_t *handle
-
- /* previous routine could use block we allocated */
- newblock = le32_to_cpu(newex.ee_start);
-+ allocated = le16_to_cpu(newex.ee_len);
- __set_bit(BH_New, &bh_result->b_state);
-
- ext3_ext_put_in_cache(inode, iblock, allocated, newblock,
-@@ -2208,6 +2233,9 @@ void ext3_ext_truncate(struct inode * in
- mutex_lock(&EXT3_I(inode)->truncate_mutex);
- ext3_ext_invalidate_cache(inode);
-
-+ /* it's important to discard preallocations under truncate_mutex */
-+ ext3_mb_discard_inode_preallocations(inode);
-+
- /*
- * TODO: optimization is possible here
- * probably we need not scaning at all,
-Index: linux-2.6.18.8/fs/ext3/Makefile
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/Makefile
-+++ linux-2.6.18.8/fs/ext3/Makefile
-@@ -5,7 +5,7 @@
- obj-$(CONFIG_EXT3_FS) += ext3.o
-
- ext3-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o \
-- ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o
-+ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o mballoc.o
-
- ext3-$(CONFIG_EXT3_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o
- ext3-$(CONFIG_EXT3_FS_POSIX_ACL) += acl.o
-Index: linux-2.6.18.8/fs/ext3/xattr.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/xattr.c
-+++ linux-2.6.18.8/fs/ext3/xattr.c
-@@ -484,7 +484,7 @@ ext3_xattr_release_block(handle_t *handl
- ea_bdebug(bh, "refcount now=0; freeing");
- if (ce)
- mb_cache_entry_free(ce);
-- ext3_free_blocks(handle, inode, bh->b_blocknr, 1);
-+ ext3_free_blocks(handle, inode, bh->b_blocknr, 1, 1);
- get_bh(bh);
- ext3_forget(handle, 1, inode, bh, bh->b_blocknr);
- } else {
-@@ -805,7 +805,7 @@ inserted:
- new_bh = sb_getblk(sb, block);
- if (!new_bh) {
- getblk_failed:
-- ext3_free_blocks(handle, inode, block, 1);
-+ ext3_free_blocks(handle, inode, block, 1, 1);
- error = -EIO;
- goto cleanup;
- }
-Index: linux-2.6.18.8/fs/ext3/balloc.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/balloc.c
-+++ linux-2.6.18.8/fs/ext3/balloc.c
-@@ -79,7 +79,7 @@ struct ext3_group_desc * ext3_get_group_
- *
- * Return buffer_head on success or NULL in case of failure.
- */
--static struct buffer_head *
-+struct buffer_head *
- read_block_bitmap(struct super_block *sb, unsigned int block_group)
- {
- struct ext3_group_desc * desc;
-@@ -294,6 +294,8 @@ void ext3_discard_reservation(struct ino
- struct ext3_reserve_window_node *rsv;
- spinlock_t *rsv_lock = &EXT3_SB(inode->i_sb)->s_rsv_window_lock;
-
-+ ext3_mb_discard_inode_preallocations(inode);
-+
- if (!block_i)
- return;
-
-@@ -490,21 +492,25 @@ error_return:
- return;
- }
-
--/* Free given blocks, update quota and i_blocks field */
- void ext3_free_blocks(handle_t *handle, struct inode *inode,
-- ext3_fsblk_t block, unsigned long count)
-+ ext3_fsblk_t block, unsigned long count, int metadata)
- {
-- struct super_block * sb;
-- unsigned long dquot_freed_blocks;
-+ struct super_block *sb;
-+ unsigned long freed;
-+
-+ /* this isn't the right place to decide whether block is metadata
-+ * inode.c/extents.c knows better, but for safety ... */
-+ if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode) ||
-+ ext3_should_journal_data(inode))
-+ metadata = 1;
-
- sb = inode->i_sb;
-- if (!sb) {
-- printk ("ext3_free_blocks: nonexistent device");
-- return;
-- }
-- ext3_free_blocks_sb(handle, sb, block, count, &dquot_freed_blocks);
-- if (dquot_freed_blocks)
-- DQUOT_FREE_BLOCK(inode, dquot_freed_blocks);
-+ if (!test_opt(sb, MBALLOC) || !EXT3_SB(sb)->s_group_info)
-+ ext3_free_blocks_sb(handle, sb, block, count, &freed);
-+ else
-+ ext3_mb_free_blocks(handle, inode, block, count, metadata, &freed);
-+ if (freed)
-+ DQUOT_FREE_BLOCK(inode, freed);
- return;
- }
-
-@@ -1199,7 +1205,7 @@ int ext3_should_retry_alloc(struct super
- * bitmap, and then for any free bit if that fails.
- * This function also updates quota and i_blocks field.
- */
--ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode,
-+ext3_fsblk_t ext3_new_blocks_old(handle_t *handle, struct inode *inode,
- ext3_fsblk_t goal, unsigned long *count, int *errp)
- {
- struct buffer_head *bitmap_bh = NULL;
-@@ -1463,7 +1469,7 @@ out:
- return 0;
- }
-
--ext3_fsblk_t ext3_new_block(handle_t *handle, struct inode *inode,
-+ext3_fsblk_t ext3_new_block_old(handle_t *handle, struct inode *inode,
- ext3_fsblk_t goal, int *errp)
- {
- unsigned long count = 1;
-Index: linux-2.6.18.8/fs/ext3/inode.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/inode.c
-+++ linux-2.6.18.8/fs/ext3/inode.c
-@@ -560,7 +560,7 @@ static int ext3_alloc_blocks(handle_t *h
- return ret;
- failed_out:
- for (i = 0; i <index; i++)
-- ext3_free_blocks(handle, inode, new_blocks[i], 1);
-+ ext3_free_blocks(handle, inode, new_blocks[i], 1, 0);
- return ret;
- }
-
-@@ -659,9 +659,9 @@ failed:
- ext3_journal_forget(handle, branch[i].bh);
- }
- for (i = 0; i <indirect_blks; i++)
-- ext3_free_blocks(handle, inode, new_blocks[i], 1);
-+ ext3_free_blocks(handle, inode, new_blocks[i], 1, 0);
-
-- ext3_free_blocks(handle, inode, new_blocks[i], num);
-+ ext3_free_blocks(handle, inode, new_blocks[i], num, 0);
-
- return err;
- }
-@@ -758,9 +758,10 @@ err_out:
- for (i = 1; i <= num; i++) {
- BUFFER_TRACE(where[i].bh, "call journal_forget");
- ext3_journal_forget(handle, where[i].bh);
-- ext3_free_blocks(handle,inode,le32_to_cpu(where[i-1].key),1);
-+ ext3_free_blocks(handle,inode,le32_to_cpu(where[i-1].key), 1,
-+ 0);
- }
-- ext3_free_blocks(handle, inode, le32_to_cpu(where[num].key), blks);
-+ ext3_free_blocks(handle, inode, le32_to_cpu(where[num].key), blks, 0);
-
- return err;
- }
-@@ -1996,7 +1997,7 @@ static void ext3_clear_blocks(handle_t *
- }
- }
-
-- ext3_free_blocks(handle, inode, block_to_free, count);
-+ ext3_free_blocks(handle, inode, block_to_free, count, 0);
- }
-
- /**
-@@ -2169,7 +2170,7 @@ static void ext3_free_branches(handle_t
- ext3_journal_test_restart(handle, inode);
- }
-
-- ext3_free_blocks(handle, inode, nr, 1);
-+ ext3_free_blocks(handle, inode, nr, 1, 1);
-
- if (parent_bh) {
- /*
-Index: linux-2.6.18.8/fs/ext3/mballoc.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/mballoc.c
-+++ linux-2.6.18.8/fs/ext3/mballoc.c
-@@ -350,8 +350,8 @@ struct ext3_prealloc_space {
- spinlock_t pa_lock;
- atomic_t pa_count;
- unsigned pa_deleted;
-- unsigned long pa_pstart; /* phys. block */
-- unsigned long pa_lstart; /* log. block */
-+ ext3_fsblk_t pa_pstart; /* phys. block */
-+ ext3_fsblk_t pa_lstart; /* log. block */
- unsigned short pa_len; /* len of preallocated chunk */
- unsigned short pa_free; /* how many blocks are free */
- unsigned short pa_linear; /* consumed in one direction
-@@ -460,8 +460,8 @@ static struct proc_dir_entry *proc_root_
-
- int ext3_create (struct inode *, struct dentry *, int, struct nameidata *);
- struct buffer_head * read_block_bitmap(struct super_block *, unsigned int);
--unsigned long ext3_new_blocks_old(handle_t *handle, struct inode *inode,
-- unsigned long goal, unsigned long *count, int *errp);
-+ext3_fsblk_t ext3_new_blocks_old(handle_t *handle, struct inode *inode,
-+ ext3_fsblk_t goal, unsigned long *count, int *errp);
- void ext3_mb_release_blocks(struct super_block *, int);
- void ext3_mb_poll_new_transaction(struct super_block *, handle_t *);
- void ext3_mb_free_committed_blocks(struct super_block *);
-@@ -479,7 +479,7 @@ int ext3_mb_destroy_per_dev_proc(struct
- * Calculate the block group number and offset, given a block number
- */
- static void ext3_get_group_no_and_offset(struct super_block *sb,
-- unsigned long blocknr,
-+ ext3_fsblk_t blocknr,
- unsigned long *blockgrpp,
- unsigned long *offsetp)
- {
-@@ -517,12 +517,12 @@ ext3_is_group_locked(struct super_block
- &EXT3_GROUP_INFO(sb, group)->bb_state);
- }
-
--unsigned long ext3_grp_offs_to_block(struct super_block *sb,
-+ext3_fsblk_t ext3_grp_offs_to_block(struct super_block *sb,
- struct ext3_free_extent *fex)
- {
-- unsigned long block;
-+ ext3_fsblk_t block;
-
-- block = (unsigned long) fex->fe_group * EXT3_BLOCKS_PER_GROUP(sb)
-+ block = (ext3_fsblk_t) fex->fe_group * EXT3_BLOCKS_PER_GROUP(sb)
- + fex->fe_start
- + le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block);
- return block;
-@@ -3202,7 +3202,7 @@ void ext3_mb_collect_stats(struct ext3_a
- void ext3_mb_use_inode_pa(struct ext3_allocation_context *ac,
- struct ext3_prealloc_space *pa)
- {
-- unsigned long start, len;
-+ ext3_fsblk_t start, len;
-
- /* found preallocated blocks, use them */
- start = pa->pa_pstart + (ac->ac_o_ex.fe_logical - pa->pa_lstart);
-@@ -4053,13 +4053,13 @@ int ext3_mb_discard_preallocations(struc
- * it tries to use preallocation first, then falls back
- * to usual allocation
- */
--unsigned long ext3_mb_new_blocks(handle_t *handle,
-+ext3_fsblk_t ext3_mb_new_blocks(handle_t *handle,
- struct ext3_allocation_request *ar, int *errp)
- {
- struct ext3_allocation_context ac;
- struct ext3_sb_info *sbi;
- struct super_block *sb;
-- unsigned long block = 0;
-+ ext3_fsblk_t block = 0;
- int freed, inquota;
-
- sb = ar->inode->i_sb;
-@@ -4070,8 +4070,8 @@ unsigned long ext3_mb_new_blocks(handle_
- if (ext3_mballoc_warning++ == 0)
- printk(KERN_ERR "EXT3-fs: multiblock request with "
- "mballoc disabled!\n");
-- ar->len = 1;
-- block = ext3_new_block_old(handle, ar->inode, ar->goal, errp);
-+ block = ext3_new_blocks_old(handle, ar->inode, ar->goal,
-+ &ar->len, errp);
- return block;
- }
-
-@@ -4135,11 +4135,11 @@ out:
- }
- EXPORT_SYMBOL(ext3_mb_new_blocks);
-
--int ext3_new_block(handle_t *handle, struct inode *inode,
-- unsigned long goal, int *errp)
-+ext3_fsblk_t ext3_new_block(handle_t *handle, struct inode *inode,
-+ ext3_fsblk_t goal, int *errp)
- {
- struct ext3_allocation_request ar;
-- unsigned long ret;
-+ ext3_fsblk_t ret;
-
- if (!test_opt(inode->i_sb, MBALLOC)) {
- ret = ext3_new_block_old(handle, inode, goal, errp);
-@@ -4254,8 +4254,8 @@ int ext3_mb_free_metadata(handle_t *hand
- * Main entry point into mballoc to free blocks
- */
- void ext3_mb_free_blocks(handle_t *handle, struct inode *inode,
-- unsigned long block, unsigned long count,
-- int metadata, int *freed)
-+ ext3_fsblk_t block, unsigned long count,
-+ int metadata, unsigned long *freed)
- {
- struct buffer_head *bitmap_bh = NULL;
- struct super_block *sb = inode->i_sb;
+++ /dev/null
-Index: linux-stage/include/linux/ext3_fs.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs.h
-+++ linux-stage/include/linux/ext3_fs.h
-@@ -53,6 +53,31 @@
- #define ext3_debug(f, a...) do {} while (0)
- #endif
-
-+#define EXT3_MULTIBLOCK_ALLOCATOR 1
-+
-+#define EXT3_MB_HINT_MERGE 1 /* prefer goal again. length */
-+#define EXT3_MB_HINT_RESERVED 2 /* blocks already reserved */
-+#define EXT3_MB_HINT_METADATA 4 /* metadata is being allocated */
-+#define EXT3_MB_HINT_FIRST 8 /* first blocks in the file */
-+#define EXT3_MB_HINT_BEST 16 /* search for the best chunk */
-+#define EXT3_MB_HINT_DATA 32 /* data is being allocated */
-+#define EXT3_MB_HINT_NOPREALLOC 64 /* don't preallocate (for tails) */
-+#define EXT3_MB_HINT_GROUP_ALLOC 128 /* allocate for locality group */
-+#define EXT3_MB_HINT_GOAL_ONLY 256 /* allocate goal blocks or none */
-+#define EXT3_MB_HINT_TRY_GOAL 512 /* goal is meaningful */
-+
-+struct ext3_allocation_request {
-+ struct inode *inode; /* target inode for block we're allocating */
-+ unsigned long logical; /* logical block in target inode */
-+ unsigned long goal; /* phys. target (a hint) */
-+ unsigned long lleft; /* the closest logical allocated block to the left */
-+ unsigned long pleft; /* phys. block for ^^^ */
-+ unsigned long lright; /* the closest logical allocated block to the right */
-+ unsigned long pright; /* phys. block for ^^^ */
-+ unsigned long len; /* how many blocks we want to allocate */
-+ unsigned long flags; /* flags. see above EXT3_MB_HINT_* */
-+};
-+
- /*
- * Special inodes numbers
- */
-@@ -398,6 +423,14 @@ struct ext3_inode {
- #define ext3_find_first_zero_bit ext2_find_first_zero_bit
- #define ext3_find_next_zero_bit ext2_find_next_zero_bit
-
-+#ifndef ext2_find_next_le_bit
-+#ifdef __LITTLE_ENDIAN
-+#define ext2_find_next_le_bit(addr, size, off) find_next_bit((addr), (size), (off))
-+#else
-+#error "mballoc needs a patch for big-endian systems - CFS bug 10634"
-+#endif /* __LITTLE_ENDIAN */
-+#endif /* !ext2_find_next_le_bit */
-+
- /*
- * Maximal mount counts between two filesystem checks
- */
-@@ -799,6 +832,20 @@ extern unsigned long ext3_count_dirs (st
- extern void ext3_check_inodes_bitmap (struct super_block *);
- extern unsigned long ext3_count_free (struct buffer_head *, unsigned);
-
-+/* mballoc.c */
-+extern long ext3_mb_stats;
-+extern long ext3_mb_max_to_scan;
-+extern int ext3_mb_init(struct super_block *, int);
-+extern int ext3_mb_release(struct super_block *);
-+extern unsigned long ext3_mb_new_blocks(handle_t *, struct ext3_allocation_request *, int *);
-+extern int ext3_mb_reserve_blocks(struct super_block *, int);
-+extern void ext3_mb_release_blocks(struct super_block *, int);
-+extern void ext3_mb_release_blocks(struct super_block *, int);
-+extern void ext3_mb_discard_inode_preallocations(struct inode *);
-+extern int __init init_ext3_mb_proc(void);
-+extern void exit_ext3_mb_proc(void);
-+extern void ext3_mb_free_blocks(handle_t *, struct inode *, unsigned long, unsigned long, int, int *);
-+
-
- /* inode.c */
- int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode,
-@@ -845,6 +892,10 @@ extern int ext3_group_extend(struct supe
- ext3_fsblk_t n_blocks_count);
-
- /* super.c */
-+extern struct proc_dir_entry *proc_root_ext3;
-+extern int __init init_ext3_proc(void);
-+extern void exit_ext3_proc(void);
-+
- extern void ext3_error (struct super_block *, const char *, const char *, ...)
- __attribute__ ((format (printf, 3, 4)));
- extern void __ext3_std_error (struct super_block *, const char *, int);
-Index: linux-stage/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs_sb.h
-+++ linux-stage/include/linux/ext3_fs_sb.h
-@@ -89,6 +89,68 @@ struct ext3_sb_info {
- unsigned long s_ext_blocks;
- unsigned long s_ext_extents;
- #endif
-+
-+ /* for buddy allocator */
-+ struct ext3_group_info ***s_group_info;
-+ struct inode *s_buddy_cache;
-+ long s_blocks_reserved;
-+ spinlock_t s_reserve_lock;
-+ struct list_head s_active_transaction;
-+ struct list_head s_closed_transaction;
-+ struct list_head s_committed_transaction;
-+ spinlock_t s_md_lock;
-+ tid_t s_last_transaction;
-+ unsigned short *s_mb_offsets, *s_mb_maxs;
-+
-+ /* tunables */
-+ unsigned long s_mb_factor;
-+ unsigned long s_stripe;
-+ unsigned long s_mb_small_req;
-+ unsigned long s_mb_large_req;
-+ unsigned long s_mb_max_to_scan;
-+ unsigned long s_mb_min_to_scan;
-+ unsigned long s_mb_max_groups_to_scan;
-+ unsigned long s_mb_stats;
-+ unsigned long s_mb_order2_reqs;
-+ unsigned long *s_mb_prealloc_table;
-+ unsigned long s_mb_prealloc_table_size;
-+ unsigned long s_mb_group_prealloc;
-+ /* where last allocation was done - for stream allocation */
-+ unsigned long s_mb_last_group;
-+ unsigned long s_mb_last_start;
-+
-+ /* history to debug policy */
-+ struct ext3_mb_history *s_mb_history;
-+ int s_mb_history_cur;
-+ int s_mb_history_max;
-+ int s_mb_history_num;
-+ struct proc_dir_entry *s_dev_proc;
-+ spinlock_t s_mb_history_lock;
-+ int s_mb_history_filter;
-+
-+ /* stats for buddy allocator */
-+ spinlock_t s_mb_pa_lock;
-+ atomic_t s_bal_reqs; /* number of reqs with len > 1 */
-+ atomic_t s_bal_success; /* we found long enough chunks */
-+ atomic_t s_bal_allocated; /* in blocks */
-+ atomic_t s_bal_ex_scanned; /* total extents scanned */
-+ atomic_t s_bal_goals; /* goal hits */
-+ atomic_t s_bal_breaks; /* too long searches */
-+ atomic_t s_bal_2orders; /* 2^order hits */
-+ spinlock_t s_bal_lock;
-+ unsigned long s_mb_buddies_generated;
-+ unsigned long long s_mb_generation_time;
-+ atomic_t s_mb_lost_chunks;
-+ atomic_t s_mb_preallocated;
-+ atomic_t s_mb_discarded;
-+
-+ /* locality groups */
-+ struct ext3_locality_group *s_locality_groups;
-+
- };
-
-+#define EXT3_GROUP_INFO(sb, group) \
-+ EXT3_SB(sb)->s_group_info[(group) >> EXT3_DESC_PER_BLOCK_BITS(sb)] \
-+ [(group) & (EXT3_DESC_PER_BLOCK(sb) - 1)]
-+
- #endif /* _LINUX_EXT3_FS_SB */
-Index: linux-stage/fs/ext3/super.c
-===================================================================
---- linux-stage.orig/fs/ext3/super.c
-+++ linux-stage/fs/ext3/super.c
-@@ -392,6 +392,7 @@ static void ext3_put_super (struct super
- struct ext3_super_block *es = sbi->s_es;
- int i, err;
-
-+ ext3_mb_release(sb);
- ext3_ext_release(sb);
- ext3_xattr_put_super(sb);
- err = journal_destroy(sbi->s_journal);
-@@ -438,6 +439,10 @@ static void ext3_put_super (struct super
- invalidate_bdev(sbi->journal_bdev, 0);
- ext3_blkdev_remove(sbi);
- }
-+ if (sbi->s_dev_proc) {
-+ remove_proc_entry(sbi->s_dev_proc->name, proc_root_ext3);
-+ sbi->s_dev_proc = NULL;
-+ }
- sb->s_fs_info = NULL;
- kfree(sbi);
- return;
-@@ -463,6 +468,8 @@ static struct inode *ext3_alloc_inode(st
- ei->vfs_inode.i_version = 1;
-
- memset(&ei->i_cached_extent, 0, sizeof(ei->i_cached_extent));
-+ INIT_LIST_HEAD(&ei->i_prealloc_list);
-+ spin_lock_init(&ei->i_prealloc_lock);
- return &ei->vfs_inode;
- }
-
-@@ -1465,6 +1472,7 @@ static int ext3_fill_super (struct super
- unsigned long journal_devnum = 0;
- unsigned long def_mount_opts;
- struct inode *root;
-+ char *devname;
- int blocksize;
- int hblock;
- int db_count;
-@@ -1480,6 +1488,22 @@ static int ext3_fill_super (struct super
- sbi->s_mount_opt = 0;
- sbi->s_resuid = EXT3_DEF_RESUID;
- sbi->s_resgid = EXT3_DEF_RESGID;
-+ devname = kstrdup(sb->s_id, GFP_KERNEL);
-+ if (devname) {
-+ char *p = devname;
-+ while ((p = strchr(p, '/')))
-+ *p = '!';
-+ sbi->s_dev_proc = proc_mkdir(devname, proc_root_ext3);
-+ if (sbi->s_dev_proc == NULL)
-+ printk(KERN_WARNING "EXT3-fs warning: unable to create "
-+ "procfs entry for %s(%s)\n",
-+ sb->s_id, devname);
-+ kfree(devname);
-+ } else {
-+ printk(KERN_WARNING "EXT3-fs warning: cannot allocate memory "
-+ "to create procfs entry for %s\n",
-+ sb->s_id);
-+ }
-
- unlock_kernel();
-
-@@ -1885,6 +1909,10 @@ failed_mount:
- ext3_blkdev_remove(sbi);
- brelse(bh);
- out_fail:
-+ if (sbi->s_dev_proc) {
-+ remove_proc_entry(sbi->s_dev_proc->name, proc_root_ext3);
-+ sbi->s_dev_proc = NULL;
-+ }
- sb->s_fs_info = NULL;
- kfree(sbi);
- lock_kernel();
-@@ -2850,9 +2878,46 @@ static struct file_system_type ext3_fs_t
- |FS_HAS_TRYTOFREE,
- };
-
-+#define EXT3_ROOT "ext3"
-+struct proc_dir_entry *proc_root_ext3;
-+
-+int __init init_ext3_proc(void)
-+{
-+ int ret;
-+
-+ if ((ret = init_ext3_mb_proc()))
-+ goto out;
-+
-+ proc_root_ext3 = proc_mkdir(EXT3_ROOT, proc_root_fs);
-+ if (proc_root_ext3 == NULL) {
-+ printk(KERN_ERR "EXT3-fs: Unable to create %s\n", EXT3_ROOT);
-+ ret = -ENOMEM;
-+ goto out_mb_proc;
-+ }
-+
-+ return 0;
-+
-+out_mb_proc:
-+ exit_ext3_mb_proc();
-+out:
-+ return ret;
-+}
-+
-+void exit_ext3_proc(void)
-+{
-+ exit_ext3_mb_proc();
-+ remove_proc_entry(EXT3_ROOT, proc_root_fs);
-+}
-+
- static int __init init_ext3_fs(void)
- {
-- int err = init_ext3_xattr();
-+ int err;
-+
-+ err = init_ext3_proc();
-+ if (err)
-+ return err;
-+
-+ err = init_ext3_xattr();
- if (err)
- return err;
- err = init_inodecache();
-@@ -2877,6 +2942,7 @@ static void __exit exit_ext3_fs(void)
- unregister_filesystem(&ext3_fs_type);
- destroy_inodecache();
- exit_ext3_xattr();
-+ exit_ext3_proc();
- }
-
- int ext3_map_inode_page(struct inode *inode, struct page *page,
-Index: linux-stage/fs/ext3/mballoc.c
-===================================================================
---- /dev/null
-+++ linux-stage/fs/ext3/mballoc.c
-@@ -0,0 +1,4486 @@
-+/*
-+ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
-+ * Written by Alex Zhuravlev <alex.zhuravlev@sun.com>
-+ *
-+ * 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-
-+ */
-+
-+
-+/*
-+ * mballoc.c contains the multiblocks allocation routines
-+ */
-+
-+#include <linux/time.h>
-+#include <linux/fs.h>
-+#include <linux/namei.h>
-+#include <linux/ext3_jbd.h>
-+#include <linux/jbd.h>
-+#include <linux/ext3_fs.h>
-+#include <linux/quotaops.h>
-+#include <linux/buffer_head.h>
-+#include <linux/module.h>
-+#include <linux/swap.h>
-+#include <linux/proc_fs.h>
-+#include <linux/pagemap.h>
-+#include <linux/seq_file.h>
-+#include <linux/version.h>
-+
-+/*
-+ * MUSTDO:
-+ * - test ext3_ext_search_left() and ext3_ext_search_right()
-+ * - search for metadata in few groups
-+ *
-+ * TODO v4:
-+ * - normalization should take into account whether file is still open
-+ * - discard preallocations if no free space left (policy?)
-+ * - don't normalize tails
-+ * - quota
-+ * - reservation for superuser
-+ *
-+ * TODO v3:
-+ * - bitmap read-ahead (proposed by Oleg Drokin aka green)
-+ * - track min/max extents in each group for better group selection
-+ * - mb_mark_used() may allocate chunk right after splitting buddy
-+ * - tree of groups sorted by number of free blocks
-+ * - error handling
-+ */
-+
-+/*
-+ * mballoc operates on the following data:
-+ * - on-disk bitmap
-+ * - in-core buddy (actually includes buddy and bitmap)
-+ * - preallocation descriptors (PAs)
-+ *
-+ * there are two types of preallocations:
-+ * - inode
-+ * assiged to specific inode and can be used for this inode only.
-+ * it describes part of inode's space preallocated to specific
-+ * physical blocks. any block from that preallocated can be used
-+ * independent. the descriptor just tracks number of blocks left
-+ * unused. so, before taking some block from descriptor, one must
-+ * make sure corresponded logical block isn't allocated yet. this
-+ * also means that freeing any block within descriptor's range
-+ * must discard all preallocated blocks.
-+ * - locality group
-+ * assigned to specific locality group which does not translate to
-+ * permanent set of inodes: inode can join and leave group. space
-+ * from this type of preallocation can be used for any inode. thus
-+ * it's consumed from the beginning to the end.
-+ *
-+ * relation between them can be expressed as:
-+ * in-core buddy = on-disk bitmap + preallocation descriptors
-+ *
-+ * this mean blocks mballoc considers used are:
-+ * - allocated blocks (persistent)
-+ * - preallocated blocks (non-persistent)
-+ *
-+ * consistency in mballoc world means that at any time a block is either
-+ * free or used in ALL structures. notice: "any time" should not be read
-+ * literally -- time is discrete and delimited by locks.
-+ *
-+ * to keep it simple, we don't use block numbers, instead we count number of
-+ * blocks: how many blocks marked used/free in on-disk bitmap, buddy and PA.
-+ *
-+ * all operations can be expressed as:
-+ * - init buddy: buddy = on-disk + PAs
-+ * - new PA: buddy += N; PA = N
-+ * - use inode PA: on-disk += N; PA -= N
-+ * - discard inode PA buddy -= on-disk - PA; PA = 0
-+ * - use locality group PA on-disk += N; PA -= N
-+ * - discard locality group PA buddy -= PA; PA = 0
-+ * note: 'buddy -= on-disk - PA' is used to show that on-disk bitmap
-+ * is used in real operation because we can't know actual used
-+ * bits from PA, only from on-disk bitmap
-+ *
-+ * if we follow this strict logic, then all operations above should be atomic.
-+ * given some of them can block, we'd have to use something like semaphores
-+ * killing performance on high-end SMP hardware. let's try to relax it using
-+ * the following knowledge:
-+ * 1) if buddy is referenced, it's already initialized
-+ * 2) while block is used in buddy and the buddy is referenced,
-+ * nobody can re-allocate that block
-+ * 3) we work on bitmaps and '+' actually means 'set bits'. if on-disk has
-+ * bit set and PA claims same block, it's OK. IOW, one can set bit in
-+ * on-disk bitmap if buddy has same bit set or/and PA covers corresponded
-+ * block
-+ *
-+ * so, now we're building a concurrency table:
-+ * - init buddy vs.
-+ * - new PA
-+ * blocks for PA are allocated in the buddy, buddy must be referenced
-+ * until PA is linked to allocation group to avoid concurrent buddy init
-+ * - use inode PA
-+ * we need to make sure that either on-disk bitmap or PA has uptodate data
-+ * given (3) we care that PA-=N operation doesn't interfere with init
-+ * - discard inode PA
-+ * the simplest way would be to have buddy initialized by the discard
-+ * - use locality group PA
-+ * again PA-=N must be serialized with init
-+ * - discard locality group PA
-+ * the simplest way would be to have buddy initialized by the discard
-+ * - new PA vs.
-+ * - use inode PA
-+ * i_truncate_mutex serializes them
-+ * - discard inode PA
-+ * discard process must wait until PA isn't used by another process
-+ * - use locality group PA
-+ * some mutex should serialize them
-+ * - discard locality group PA
-+ * discard process must wait until PA isn't used by another process
-+ * - use inode PA
-+ * - use inode PA
-+ * i_truncate_mutex or another mutex should serializes them
-+ * - discard inode PA
-+ * discard process must wait until PA isn't used by another process
-+ * - use locality group PA
-+ * nothing wrong here -- they're different PAs covering different blocks
-+ * - discard locality group PA
-+ * discard process must wait until PA isn't used by another process
-+ *
-+ * now we're ready to make few consequences:
-+ * - PA is referenced and while it is no discard is possible
-+ * - PA is referenced until block isn't marked in on-disk bitmap
-+ * - PA changes only after on-disk bitmap
-+ * - discard must not compete with init. either init is done before
-+ * any discard or they're serialized somehow
-+ * - buddy init as sum of on-disk bitmap and PAs is done atomically
-+ *
-+ * a special case when we've used PA to emptiness. no need to modify buddy
-+ * in this case, but we should care about concurrent init
-+ *
-+ */
-+
-+ /*
-+ * Logic in few words:
-+ *
-+ * - allocation:
-+ * load group
-+ * find blocks
-+ * mark bits in on-disk bitmap
-+ * release group
-+ *
-+ * - use preallocation:
-+ * find proper PA (per-inode or group)
-+ * load group
-+ * mark bits in on-disk bitmap
-+ * release group
-+ * release PA
-+ *
-+ * - free:
-+ * load group
-+ * mark bits in on-disk bitmap
-+ * release group
-+ *
-+ * - discard preallocations in group:
-+ * mark PAs deleted
-+ * move them onto local list
-+ * load on-disk bitmap
-+ * load group
-+ * remove PA from object (inode or locality group)
-+ * mark free blocks in-core
-+ *
-+ * - discard inode's preallocations:
-+ */
-+
-+/*
-+ * Locking rules
-+ *
-+ * Locks:
-+ * - bitlock on a group (group)
-+ * - object (inode/locality) (object)
-+ * - per-pa lock (pa)
-+ *
-+ * Paths:
-+ * - new pa
-+ * object
-+ * group
-+ *
-+ * - find and use pa:
-+ * pa
-+ *
-+ * - release consumed pa:
-+ * pa
-+ * group
-+ * object
-+ *
-+ * - generate in-core bitmap:
-+ * group
-+ * pa
-+ *
-+ * - discard all for given object (inode, locality group):
-+ * object
-+ * pa
-+ * group
-+ *
-+ * - discard all for given group:
-+ * group
-+ * pa
-+ * group
-+ * object
-+ *
-+ */
-+
-+/*
-+ * with AGGRESSIVE_CHECK allocator runs consistency checks over
-+ * structures. these checks slow things down a lot
-+ */
-+#define AGGRESSIVE_CHECK__
-+
-+/*
-+ * with DOUBLE_CHECK defined mballoc creates persistent in-core
-+ * bitmaps, maintains and uses them to check for double allocations
-+ */
-+#define DOUBLE_CHECK__
-+
-+/*
-+ */
-+#define MB_DEBUG__
-+#ifdef MB_DEBUG
-+#define mb_debug(fmt,a...) printk(fmt, ##a)
-+#else
-+#define mb_debug(fmt,a...)
-+#endif
-+
-+/*
-+ * with EXT3_MB_HISTORY mballoc stores last N allocations in memory
-+ * and you can monitor it in /proc/fs/ext3/<dev>/mb_history
-+ */
-+#define EXT3_MB_HISTORY
-+#define EXT3_MB_HISTORY_ALLOC 1 /* allocation */
-+#define EXT3_MB_HISTORY_PREALLOC 2 /* preallocated blocks used */
-+#define EXT3_MB_HISTORY_DISCARD 4 /* preallocation discarded */
-+#define EXT3_MB_HISTORY_FREE 8 /* free */
-+
-+#define EXT3_MB_HISTORY_DEFAULT (EXT3_MB_HISTORY_ALLOC | \
-+ EXT3_MB_HISTORY_PREALLOC | \
-+ EXT3_MB_HISTORY_DISCARD | \
-+ EXT3_MB_HISTORY_FREE)
-+
-+/*
-+ * How long mballoc can look for a best extent (in found extents)
-+ */
-+#define MB_DEFAULT_MAX_TO_SCAN 200
-+
-+/*
-+ * How long mballoc must look for a best extent
-+ */
-+#define MB_DEFAULT_MIN_TO_SCAN 10
-+
-+/*
-+ * How many groups mballoc will scan looking for the best chunk
-+ */
-+#define MB_DEFAULT_MAX_GROUPS_TO_SCAN 5
-+
-+/*
-+ * with 'ext3_mb_stats' allocator will collect stats that will be
-+ * shown at umount. The collecting costs though!
-+ */
-+#define MB_DEFAULT_STATS 1
-+
-+/*
-+ * files smaller than MB_DEFAULT_STREAM_THRESHOLD are served
-+ * by the stream allocator, which purpose is to pack requests
-+ * as close each to other as possible to produce smooth I/O traffic
-+ */
-+#define MB_DEFAULT_STREAM_THRESHOLD 16 /* 64K */
-+
-+/*
-+ * for which requests use 2^N search using buddies
-+ */
-+#define MB_DEFAULT_ORDER2_REQS 8
-+
-+/*
-+ * default stripe size = 1MB
-+ */
-+#define MB_DEFAULT_STRIPE 256
-+
-+static kmem_cache_t *ext3_pspace_cachep = NULL;
-+
-+#ifdef EXT3_BB_MAX_BLOCKS
-+#undef EXT3_BB_MAX_BLOCKS
-+#endif
-+#define EXT3_BB_MAX_BLOCKS 30
-+
-+struct ext3_free_metadata {
-+ unsigned group;
-+ unsigned short num;
-+ unsigned short blocks[EXT3_BB_MAX_BLOCKS];
-+ struct list_head list;
-+};
-+
-+struct ext3_group_info {
-+ unsigned long bb_state;
-+ unsigned long bb_tid;
-+ struct ext3_free_metadata *bb_md_cur;
-+ unsigned short bb_first_free;
-+ unsigned short bb_free;
-+ unsigned short bb_fragments;
-+ struct list_head bb_prealloc_list;
-+ unsigned long bb_prealloc_nr;
-+#ifdef DOUBLE_CHECK
-+ void *bb_bitmap;
-+#endif
-+ unsigned short bb_counters[];
-+};
-+
-+#define EXT3_GROUP_INFO_NEED_INIT_BIT 0
-+#define EXT3_GROUP_INFO_LOCKED_BIT 1
-+
-+#define EXT3_MB_GRP_NEED_INIT(grp) \
-+ (test_bit(EXT3_GROUP_INFO_NEED_INIT_BIT, &(grp)->bb_state))
-+
-+
-+struct ext3_prealloc_space {
-+ struct list_head pa_inode_list;
-+ struct list_head pa_group_list;
-+ union {
-+ struct list_head pa_tmp_list;
-+ struct rcu_head pa_rcu;
-+ } u;
-+ spinlock_t pa_lock;
-+ atomic_t pa_count;
-+ unsigned pa_deleted;
-+ unsigned long pa_pstart; /* phys. block */
-+ unsigned long pa_lstart; /* log. block */
-+ unsigned short pa_len; /* len of preallocated chunk */
-+ unsigned short pa_free; /* how many blocks are free */
-+ unsigned short pa_linear; /* consumed in one direction
-+ * strictly, for group prealloc */
-+ spinlock_t *pa_obj_lock;
-+ struct inode *pa_inode; /* hack, for history only */
-+};
-+
-+
-+struct ext3_free_extent {
-+ unsigned long fe_logical;
-+ unsigned long fe_start;
-+ unsigned long fe_group;
-+ unsigned long fe_len;
-+};
-+
-+/*
-+ * Locality group:
-+ * we try to group all related changes together
-+ * so that writeback can flush/allocate them together as well
-+ */
-+struct ext3_locality_group {
-+ /* for allocator */
-+ struct semaphore lg_sem; /* to serialize allocates */
-+ struct list_head lg_prealloc_list;/* list of preallocations */
-+ spinlock_t lg_prealloc_lock;
-+};
-+
-+struct ext3_allocation_context {
-+ struct inode *ac_inode;
-+ struct super_block *ac_sb;
-+
-+ /* original request */
-+ struct ext3_free_extent ac_o_ex;
-+
-+ /* goal request (after normalization) */
-+ struct ext3_free_extent ac_g_ex;
-+
-+ /* the best found extent */
-+ struct ext3_free_extent ac_b_ex;
-+
-+ /* copy of the bext found extent taken before preallocation efforts */
-+ struct ext3_free_extent ac_f_ex;
-+
-+ /* number of iterations done. we have to track to limit searching */
-+ unsigned long ac_ex_scanned;
-+ __u16 ac_groups_scanned;
-+ __u16 ac_found;
-+ __u16 ac_tail;
-+ __u16 ac_buddy;
-+ __u16 ac_flags; /* allocation hints */
-+ __u8 ac_status;
-+ __u8 ac_criteria;
-+ __u8 ac_repeats;
-+ __u8 ac_2order; /* if request is to allocate 2^N blocks and
-+ * N > 0, the field stores N, otherwise 0 */
-+ __u8 ac_op; /* operation, for history only */
-+ struct page *ac_bitmap_page;
-+ struct page *ac_buddy_page;
-+ struct ext3_prealloc_space *ac_pa;
-+ struct ext3_locality_group *ac_lg;
-+};
-+
-+#define AC_STATUS_CONTINUE 1
-+#define AC_STATUS_FOUND 2
-+#define AC_STATUS_BREAK 3
-+
-+struct ext3_mb_history {
-+ struct ext3_free_extent orig; /* orig allocation */
-+ struct ext3_free_extent goal; /* goal allocation */
-+ struct ext3_free_extent result; /* result allocation */
-+ unsigned pid;
-+ unsigned ino;
-+ __u16 found; /* how many extents have been found */
-+ __u16 groups; /* how many groups have been scanned */
-+ __u16 tail; /* what tail broke some buddy */
-+ __u16 buddy; /* buddy the tail ^^^ broke */
-+ __u16 flags;
-+ __u8 cr:8; /* which phase the result extent was found at */
-+ __u8 op:4;
-+ __u8 merged:1;
-+};
-+
-+struct ext3_buddy {
-+ struct page *bd_buddy_page;
-+ void *bd_buddy;
-+ struct page *bd_bitmap_page;
-+ void *bd_bitmap;
-+ struct ext3_group_info *bd_info;
-+ struct super_block *bd_sb;
-+ unsigned bd_group;
-+ unsigned bd_blkbits;
-+};
-+#define EXT3_MB_BITMAP(e3b) ((e3b)->bd_bitmap)
-+#define EXT3_MB_BUDDY(e3b) ((e3b)->bd_buddy)
-+
-+#ifndef EXT3_MB_HISTORY
-+#define ext3_mb_store_history(ac)
-+#else
-+static void ext3_mb_store_history(struct ext3_allocation_context *ac);
-+#endif
-+
-+#define in_range(b, first, len) ((b) >= (first) && (b) <= (first) + (len) - 1)
-+
-+int ext3_create (struct inode *, struct dentry *, int, struct nameidata *);
-+struct buffer_head * read_block_bitmap(struct super_block *, unsigned int);
-+unsigned long ext3_new_blocks_old(handle_t *handle, struct inode *inode,
-+ unsigned long goal, unsigned long *count, int *errp);
-+void ext3_mb_release_blocks(struct super_block *, int);
-+void ext3_mb_poll_new_transaction(struct super_block *, handle_t *);
-+void ext3_mb_free_committed_blocks(struct super_block *);
-+int ext3_mb_generate_from_pa(struct super_block *sb, void *bitmap, int group);
-+void ext3_mb_free_consumed_preallocations(struct ext3_allocation_context *ac);
-+void ext3_mb_return_to_preallocation(struct inode *inode, struct ext3_buddy *e3b,
-+ sector_t block, int count);
-+void ext3_mb_show_ac(struct ext3_allocation_context *ac);
-+void ext3_mb_check_with_pa(struct ext3_buddy *e3b, int first, int count);
-+void ext3_mb_put_pa(struct ext3_allocation_context *, struct super_block *, struct ext3_prealloc_space *pa);
-+int ext3_mb_init_per_dev_proc(struct super_block *sb);
-+int ext3_mb_destroy_per_dev_proc(struct super_block *sb);
-+
-+/*
-+ * Calculate the block group number and offset, given a block number
-+ */
-+static void ext3_get_group_no_and_offset(struct super_block *sb,
-+ unsigned long blocknr,
-+ unsigned long *blockgrpp,
-+ unsigned long *offsetp)
-+{
-+ struct ext3_super_block *es = EXT3_SB(sb)->s_es;
-+ unsigned long offset;
-+
-+ blocknr = blocknr - le32_to_cpu(es->s_first_data_block);
-+ offset = blocknr % EXT3_BLOCKS_PER_GROUP(sb);
-+ blocknr = blocknr / EXT3_BLOCKS_PER_GROUP(sb);
-+ if (offsetp)
-+ *offsetp = offset;
-+ if (blockgrpp)
-+ *blockgrpp = blocknr;
-+
-+}
-+
-+static inline void
-+ext3_lock_group(struct super_block *sb, int group)
-+{
-+ bit_spin_lock(EXT3_GROUP_INFO_LOCKED_BIT,
-+ &EXT3_GROUP_INFO(sb, group)->bb_state);
-+}
-+
-+static inline void
-+ext3_unlock_group(struct super_block *sb, int group)
-+{
-+ bit_spin_unlock(EXT3_GROUP_INFO_LOCKED_BIT,
-+ &EXT3_GROUP_INFO(sb, group)->bb_state);
-+}
-+
-+static inline int
-+ext3_is_group_locked(struct super_block *sb, int group)
-+{
-+ return bit_spin_is_locked(EXT3_GROUP_INFO_LOCKED_BIT,
-+ &EXT3_GROUP_INFO(sb, group)->bb_state);
-+}
-+
-+unsigned long ext3_grp_offs_to_block(struct super_block *sb,
-+ struct ext3_free_extent *fex)
-+{
-+ unsigned long block;
-+
-+ block = (unsigned long) fex->fe_group * EXT3_BLOCKS_PER_GROUP(sb)
-+ + fex->fe_start
-+ + le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block);
-+ return block;
-+}
-+
-+#if BITS_PER_LONG == 64
-+#define mb_correct_addr_and_bit(bit,addr) \
-+{ \
-+ bit += ((unsigned long) addr & 7UL) << 3; \
-+ addr = (void *) ((unsigned long) addr & ~7UL); \
-+}
-+#elif BITS_PER_LONG == 32
-+#define mb_correct_addr_and_bit(bit,addr) \
-+{ \
-+ bit += ((unsigned long) addr & 3UL) << 3; \
-+ addr = (void *) ((unsigned long) addr & ~3UL); \
-+}
-+#else
-+#error "how many bits you are?!"
-+#endif
-+
-+static inline int mb_test_bit(int bit, void *addr)
-+{
-+ mb_correct_addr_and_bit(bit,addr);
-+ return ext2_test_bit(bit, addr);
-+}
-+
-+static inline void mb_set_bit(int bit, void *addr)
-+{
-+ mb_correct_addr_and_bit(bit,addr);
-+ ext2_set_bit(bit, addr);
-+}
-+
-+static inline void mb_set_bit_atomic(spinlock_t *lock, int bit, void *addr)
-+{
-+ mb_correct_addr_and_bit(bit,addr);
-+ ext2_set_bit_atomic(lock, bit, addr);
-+}
-+
-+static inline void mb_clear_bit(int bit, void *addr)
-+{
-+ mb_correct_addr_and_bit(bit,addr);
-+ ext2_clear_bit(bit, addr);
-+}
-+
-+static inline void mb_clear_bit_atomic(spinlock_t *lock, int bit, void *addr)
-+{
-+ mb_correct_addr_and_bit(bit,addr);
-+ ext2_clear_bit_atomic(lock, bit, addr);
-+}
-+
-+static inline int mb_find_next_zero_bit(void *addr, int max, int start)
-+{
-+ int fix;
-+#if BITS_PER_LONG == 64
-+ fix = ((unsigned long) addr & 7UL) << 3;
-+ addr = (void *) ((unsigned long) addr & ~7UL);
-+#elif BITS_PER_LONG == 32
-+ fix = ((unsigned long) addr & 3UL) << 3;
-+ addr = (void *) ((unsigned long) addr & ~3UL);
-+#else
-+#error "how many bits you are?!"
-+#endif
-+ max += fix;
-+ start += fix;
-+ return ext2_find_next_zero_bit(addr, max, start) - fix;
-+}
-+
-+static inline int mb_find_next_bit(void *addr, int max, int start)
-+{
-+ int fix;
-+#if BITS_PER_LONG == 64
-+ fix = ((unsigned long) addr & 7UL) << 3;
-+ addr = (void *) ((unsigned long) addr & ~7UL);
-+#elif BITS_PER_LONG == 32
-+ fix = ((unsigned long) addr & 3UL) << 3;
-+ addr = (void *) ((unsigned long) addr & ~3UL);
-+#else
-+#error "how many bits you are?!"
-+#endif
-+ max += fix;
-+ start += fix;
-+
-+#ifdef __BIG_ENDIAN
-+#else
-+ return find_next_bit(addr, max, start) - fix;
-+#endif
-+}
-+
-+static inline void *mb_find_buddy(struct ext3_buddy *e3b, int order, int *max)
-+{
-+ char *bb;
-+
-+ BUG_ON(EXT3_MB_BITMAP(e3b) == EXT3_MB_BUDDY(e3b));
-+ BUG_ON(max == NULL);
-+
-+ if (order > e3b->bd_blkbits + 1) {
-+ *max = 0;
-+ return NULL;
-+ }
-+
-+ /* at order 0 we see each particular block */
-+ *max = 1 << (e3b->bd_blkbits + 3);
-+ if (order == 0)
-+ return EXT3_MB_BITMAP(e3b);
-+
-+ bb = EXT3_MB_BUDDY(e3b) + EXT3_SB(e3b->bd_sb)->s_mb_offsets[order];
-+ *max = EXT3_SB(e3b->bd_sb)->s_mb_maxs[order];
-+
-+ return bb;
-+}
-+
-+#ifdef DOUBLE_CHECK
-+void mb_free_blocks_double(struct inode *inode, struct ext3_buddy *e3b,
-+ int first, int count)
-+{
-+ int i;
-+ struct super_block *sb = e3b->bd_sb;
-+
-+ if (unlikely(e3b->bd_info->bb_bitmap == NULL))
-+ return;
-+ BUG_ON(!ext3_is_group_locked(sb, e3b->bd_group));
-+ for (i = 0; i < count; i++) {
-+ if (!mb_test_bit(first + i, e3b->bd_info->bb_bitmap)) {
-+ unsigned long blocknr;
-+ blocknr = e3b->bd_group * EXT3_BLOCKS_PER_GROUP(sb);
-+ blocknr += first + i;
-+ blocknr +=
-+ le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block);
-+
-+ ext3_error(sb, __FUNCTION__, "double-free of inode"
-+ " %lu's block %lu(bit %u in group %u)\n",
-+ inode ? inode->i_ino : 0, blocknr,
-+ first + i, e3b->bd_group);
-+ }
-+ mb_clear_bit(first + i, e3b->bd_info->bb_bitmap);
-+ }
-+}
-+
-+void mb_mark_used_double(struct ext3_buddy *e3b, int first, int count)
-+{
-+ int i;
-+ if (unlikely(e3b->bd_info->bb_bitmap == NULL))
-+ return;
-+ BUG_ON(!ext3_is_group_locked(e3b->bd_sb, e3b->bd_group));
-+ for (i = 0; i < count; i++) {
-+ BUG_ON(mb_test_bit(first + i, e3b->bd_info->bb_bitmap));
-+ mb_set_bit(first + i, e3b->bd_info->bb_bitmap);
-+ }
-+}
-+
-+void mb_cmp_bitmaps(struct ext3_buddy *e3b, void *bitmap)
-+{
-+ if (memcmp(e3b->bd_info->bb_bitmap, bitmap, e3b->bd_sb->s_blocksize)) {
-+ unsigned char *b1, *b2;
-+ int i;
-+ b1 = (unsigned char *) e3b->bd_info->bb_bitmap;
-+ b2 = (unsigned char *) bitmap;
-+ for (i = 0; i < e3b->bd_sb->s_blocksize; i++) {
-+ if (b1[i] != b2[i]) {
-+ printk("corruption in group %u at byte %u(%u): "
-+ "%x in copy != %x on disk/prealloc\n",
-+ e3b->bd_group, i, i * 8, b1[i], b2[i]);
-+ BUG();
-+ }
-+ }
-+ }
-+}
-+
-+#else
-+#define mb_free_blocks_double(a,b,c,d)
-+#define mb_mark_used_double(a,b,c)
-+#define mb_cmp_bitmaps(a,b)
-+#endif
-+
-+#ifdef AGGRESSIVE_CHECK
-+
-+#define MB_CHECK_ASSERT(assert) \
-+do { \
-+ if (!(assert)) { \
-+ printk (KERN_EMERG \
-+ "Assertion failure in %s() at %s:%d: \"%s\"\n", \
-+ function, file, line, # assert); \
-+ BUG(); \
-+ } \
-+} while (0)
-+
-+static int __mb_check_buddy(struct ext3_buddy *e3b, char *file,
-+ const char *function, int line)
-+{
-+ struct super_block *sb = e3b->bd_sb;
-+ int order = e3b->bd_blkbits + 1;
-+ int max, max2, i, j, k, count;
-+ struct ext3_group_info *grp;
-+ int fragments = 0, fstart;
-+ struct list_head *cur;
-+ void *buddy, *buddy2;
-+
-+ if (!test_opt(sb, MBALLOC))
-+ return 0;
-+
-+ {
-+ static int mb_check_counter = 0;
-+ if (mb_check_counter++ % 100 != 0)
-+ return 0;
-+ }
-+
-+ while (order > 1) {
-+ buddy = mb_find_buddy(e3b, order, &max);
-+ MB_CHECK_ASSERT(buddy);
-+ buddy2 = mb_find_buddy(e3b, order - 1, &max2);
-+ MB_CHECK_ASSERT(buddy2);
-+ MB_CHECK_ASSERT(buddy != buddy2);
-+ MB_CHECK_ASSERT(max * 2 == max2);
-+
-+ count = 0;
-+ for (i = 0; i < max; i++) {
-+
-+ if (mb_test_bit(i, buddy)) {
-+ /* only single bit in buddy2 may be 1 */
-+ if (!mb_test_bit(i << 1, buddy2))
-+ MB_CHECK_ASSERT(mb_test_bit((i<<1)+1, buddy2));
-+ else if (!mb_test_bit((i << 1) + 1, buddy2))
-+ MB_CHECK_ASSERT(mb_test_bit(i << 1, buddy2));
-+ continue;
-+ }
-+
-+ /* both bits in buddy2 must be 0 */
-+ MB_CHECK_ASSERT(mb_test_bit(i << 1, buddy2));
-+ MB_CHECK_ASSERT(mb_test_bit((i << 1) + 1, buddy2));
-+
-+ for (j = 0; j < (1 << order); j++) {
-+ k = (i * (1 << order)) + j;
-+ MB_CHECK_ASSERT(!mb_test_bit(k, EXT3_MB_BITMAP(e3b)));
-+ }
-+ count++;
-+ }
-+ MB_CHECK_ASSERT(e3b->bd_info->bb_counters[order] == count);
-+ order--;
-+ }
-+
-+ fstart = -1;
-+ buddy = mb_find_buddy(e3b, 0, &max);
-+ for (i = 0; i < max; i++) {
-+ if (!mb_test_bit(i, buddy)) {
-+ MB_CHECK_ASSERT(i >= e3b->bd_info->bb_first_free);
-+ if (fstart == -1) {
-+ fragments++;
-+ fstart = i;
-+ }
-+ continue;
-+ }
-+ fstart = -1;
-+ /* check used bits only */
-+ for (j = 0; j < e3b->bd_blkbits + 1; j++) {
-+ buddy2 = mb_find_buddy(e3b, j, &max2);
-+ k = i >> j;
-+ MB_CHECK_ASSERT(k < max2);
-+ MB_CHECK_ASSERT(mb_test_bit(k, buddy2));
-+ }
-+ }
-+ MB_CHECK_ASSERT(!EXT3_MB_GRP_NEED_INIT(e3b->bd_info));
-+ MB_CHECK_ASSERT(e3b->bd_info->bb_fragments == fragments);
-+
-+ grp = EXT3_GROUP_INFO(sb, e3b->bd_group);
-+ buddy = mb_find_buddy(e3b, 0, &max);
-+ list_for_each(cur, &grp->bb_prealloc_list) {
-+ unsigned long groupnr;
-+ struct ext3_prealloc_space *pa;
-+ pa = list_entry(cur, struct ext3_prealloc_space, group_list);
-+ ext3_get_group_no_and_offset(sb, pa->pstart, &groupnr, &k);
-+ MB_CHECK_ASSERT(groupnr == e3b->bd_group);
-+ for (i = 0; i < pa->len; i++)
-+ MB_CHECK_ASSERT(mb_test_bit(k + i, buddy));
-+ }
-+ return 0;
-+}
-+#undef MB_CHECK_ASSERT
-+#define mb_check_buddy(e3b) __mb_check_buddy(e3b,__FILE__,__FUNCTION__,__LINE__)
-+#else
-+#define mb_check_buddy(e3b)
-+#endif
-+
-+/* find most significant bit */
-+static int inline fmsb(unsigned short word)
-+{
-+ int order;
-+
-+ if (word > 255) {
-+ order = 7;
-+ word >>= 8;
-+ } else {
-+ order = -1;
-+ }
-+
-+ do {
-+ order++;
-+ word >>= 1;
-+ } while (word != 0);
-+
-+ return order;
-+}
-+
-+static void inline
-+ext3_mb_mark_free_simple(struct super_block *sb, void *buddy, unsigned first,
-+ int len, struct ext3_group_info *grp)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ unsigned short min, max, chunk, border;
-+
-+ BUG_ON(len >= EXT3_BLOCKS_PER_GROUP(sb));
-+
-+ border = 2 << sb->s_blocksize_bits;
-+
-+ while (len > 0) {
-+ /* find how many blocks can be covered since this position */
-+ max = ffs(first | border) - 1;
-+
-+ /* find how many blocks of power 2 we need to mark */
-+ min = fmsb(len);
-+
-+ if (max < min)
-+ min = max;
-+ chunk = 1 << min;
-+
-+ /* mark multiblock chunks only */
-+ grp->bb_counters[min]++;
-+ if (min > 0)
-+ mb_clear_bit(first >> min, buddy + sbi->s_mb_offsets[min]);
-+
-+ len -= chunk;
-+ first += chunk;
-+ }
-+}
-+
-+static int
-+ext3_mb_generate_buddy(struct super_block *sb, void *buddy, void *bitmap,
-+ int group)
-+{
-+ struct ext3_group_info *grp = EXT3_GROUP_INFO(sb, group);
-+ unsigned short max = EXT3_BLOCKS_PER_GROUP(sb);
-+ unsigned short i = 0, first, len;
-+ unsigned free = 0, fragments = 0;
-+ unsigned long long period = get_cycles();
-+
-+ /* initialize buddy from bitmap which is aggregation
-+ * of on-disk bitmap and preallocations */
-+ i = mb_find_next_zero_bit(bitmap, max, 0);
-+ grp->bb_first_free = i;
-+ while (i < max) {
-+ fragments++;
-+ first = i;
-+ i = ext2_find_next_le_bit(bitmap, max, i);
-+ if (i > max)
-+ i = max;
-+ len = i - first;
-+ free += len;
-+ if (len > 1)
-+ ext3_mb_mark_free_simple(sb, buddy, first, len, grp);
-+ else
-+ grp->bb_counters[0]++;
-+ if (i < max)
-+ i = mb_find_next_zero_bit(bitmap, max, i);
-+ }
-+ grp->bb_fragments = fragments;
-+
-+ if (free != grp->bb_free) {
-+ struct ext3_group_desc *gdp;
-+ gdp = ext3_get_group_desc (sb, group, NULL);
-+ ext3_error(sb, __FUNCTION__,
-+ "group %u: %u blocks in bitmap, %u in bb, "
-+ "%u in gd, %lu pa's\n", group, free, grp->bb_free,
-+ le16_to_cpu(gdp->bg_free_blocks_count),
-+ grp->bb_prealloc_nr);
-+ return -EIO;
-+ }
-+
-+ clear_bit(EXT3_GROUP_INFO_NEED_INIT_BIT, &grp->bb_state);
-+
-+ period = get_cycles() - period;
-+ spin_lock(&EXT3_SB(sb)->s_bal_lock);
-+ EXT3_SB(sb)->s_mb_buddies_generated++;
-+ EXT3_SB(sb)->s_mb_generation_time += period;
-+ spin_unlock(&EXT3_SB(sb)->s_bal_lock);
-+
-+ return 0;
-+}
-+
-+static int ext3_mb_init_cache(struct page *page, char *incore)
-+{
-+ int blocksize, blocks_per_page, groups_per_page;
-+ int err = 0, i, first_group, first_block;
-+ struct super_block *sb;
-+ struct buffer_head *bhs;
-+ struct buffer_head **bh;
-+ struct inode *inode;
-+ char *data, *bitmap;
-+
-+ mb_debug("init page %lu\n", page->index);
-+
-+ inode = page->mapping->host;
-+ sb = inode->i_sb;
-+ blocksize = 1 << inode->i_blkbits;
-+ blocks_per_page = PAGE_CACHE_SIZE / blocksize;
-+
-+ groups_per_page = blocks_per_page >> 1;
-+ if (groups_per_page == 0)
-+ groups_per_page = 1;
-+
-+ /* allocate buffer_heads to read bitmaps */
-+ if (groups_per_page > 1) {
-+ err = -ENOMEM;
-+ i = sizeof(struct buffer_head *) * groups_per_page;
-+ bh = kmalloc(i, GFP_NOFS);
-+ if (bh == NULL)
-+ goto out;
-+ memset(bh, 0, i);
-+ } else
-+ bh = &bhs;
-+
-+ first_group = page->index * blocks_per_page / 2;
-+
-+ /* read all groups the page covers into the cache */
-+ for (i = 0; i < groups_per_page; i++) {
-+ struct ext3_group_desc * desc;
-+
-+ if (first_group + i >= EXT3_SB(sb)->s_groups_count)
-+ break;
-+
-+ err = -EIO;
-+ desc = ext3_get_group_desc(sb, first_group + i, NULL);
-+ if (desc == NULL)
-+ goto out;
-+
-+ err = -ENOMEM;
-+ bh[i] = sb_getblk(sb, le32_to_cpu(desc->bg_block_bitmap));
-+ if (bh[i] == NULL)
-+ goto out;
-+
-+ if (buffer_uptodate(bh[i]))
-+ continue;
-+
-+ lock_buffer(bh[i]);
-+ if (buffer_uptodate(bh[i])) {
-+ unlock_buffer(bh[i]);
-+ continue;
-+ }
-+
-+ get_bh(bh[i]);
-+ bh[i]->b_end_io = end_buffer_read_sync;
-+ submit_bh(READ, bh[i]);
-+ mb_debug("read bitmap for group %u\n", first_group + i);
-+ }
-+
-+ /* wait for I/O completion */
-+ for (i = 0; i < groups_per_page && bh[i]; i++)
-+ wait_on_buffer(bh[i]);
-+
-+ err = -EIO;
-+ for (i = 0; i < groups_per_page && bh[i]; i++)
-+ if (!buffer_uptodate(bh[i]))
-+ goto out;
-+
-+ err = 0;
-+ first_block = page->index * blocks_per_page;
-+ for (i = 0; i < blocks_per_page && err == 0; i++) {
-+ int group;
-+
-+ group = (first_block + i) >> 1;
-+ if (group >= EXT3_SB(sb)->s_groups_count)
-+ break;
-+
-+ data = page_address(page) + (i * blocksize);
-+ bitmap = bh[group - first_group]->b_data;
-+
-+ if ((first_block + i) & 1) {
-+ /* this is block of buddy */
-+ BUG_ON(incore == NULL);
-+ mb_debug("put buddy for group %u in page %lu/%x\n",
-+ group, page->index, i * blocksize);
-+ memset(data, 0xff, blocksize);
-+ EXT3_GROUP_INFO(sb, group)->bb_fragments = 0;
-+ memset(EXT3_GROUP_INFO(sb, group)->bb_counters, 0,
-+ sizeof(unsigned short)*(sb->s_blocksize_bits+2));
-+ err = ext3_mb_generate_buddy(sb, data, incore, group);
-+ incore = NULL;
-+ } else {
-+ /* this is block of bitmap */
-+ BUG_ON(incore != NULL);
-+ mb_debug("put bitmap for group %u in page %lu/%x\n",
-+ group, page->index, i * blocksize);
-+
-+ /* see comments in ext3_mb_put_pa() */
-+ ext3_lock_group(sb, group);
-+ memcpy(data, bitmap, blocksize);
-+
-+ /* mark all preallocated blocks used in in-core bitmap */
-+ err = ext3_mb_generate_from_pa(sb, data, group);
-+ ext3_unlock_group(sb, group);
-+
-+ incore = data;
-+ }
-+ }
-+ if (likely(err == 0))
-+ SetPageUptodate(page);
-+
-+out:
-+ if (bh) {
-+ for (i = 0; i < groups_per_page && bh[i]; i++)
-+ brelse(bh[i]);
-+ if (bh != &bhs)
-+ kfree(bh);
-+ }
-+ return err;
-+}
-+
-+static int ext3_mb_load_buddy(struct super_block *sb, int group,
-+ struct ext3_buddy *e3b)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ struct inode *inode = sbi->s_buddy_cache;
-+ int blocks_per_page, block, pnum, poff;
-+ struct page *page;
-+
-+ mb_debug("load group %u\n", group);
-+
-+ blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize;
-+
-+ e3b->bd_blkbits = sb->s_blocksize_bits;
-+ e3b->bd_info = EXT3_GROUP_INFO(sb, group);
-+ e3b->bd_sb = sb;
-+ e3b->bd_group = group;
-+ e3b->bd_buddy_page = NULL;
-+ e3b->bd_bitmap_page = NULL;
-+
-+ block = group * 2;
-+ pnum = block / blocks_per_page;
-+ poff = block % blocks_per_page;
-+
-+ /* we could use find_or_create_page(), but it locks page
-+ * what we'd like to avoid in fast path ... */
-+ page = find_get_page(inode->i_mapping, pnum);
-+ if (page == NULL || !PageUptodate(page)) {
-+ if (page)
-+ page_cache_release(page);
-+ page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS);
-+ if (page) {
-+ BUG_ON(page->mapping != inode->i_mapping);
-+ if (!PageUptodate(page)) {
-+ ext3_mb_init_cache(page, NULL);
-+ mb_cmp_bitmaps(e3b, page_address(page) +
-+ (poff * sb->s_blocksize));
-+ }
-+ unlock_page(page);
-+ }
-+ }
-+ e3b->bd_bitmap_page = page;
-+ if (page == NULL || !PageUptodate(page))
-+ goto err;
-+ e3b->bd_bitmap = page_address(page) + (poff * sb->s_blocksize);
-+ mark_page_accessed(page);
-+
-+ block++;
-+ pnum = block / blocks_per_page;
-+ poff = block % blocks_per_page;
-+
-+ page = find_get_page(inode->i_mapping, pnum);
-+ if (page == NULL || !PageUptodate(page)) {
-+ if (page)
-+ page_cache_release(page);
-+ page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS);
-+ if (page) {
-+ BUG_ON(page->mapping != inode->i_mapping);
-+ if (!PageUptodate(page))
-+ ext3_mb_init_cache(page, e3b->bd_bitmap);
-+
-+ unlock_page(page);
-+ }
-+ }
-+ e3b->bd_buddy_page = page;
-+ if (page == NULL || !PageUptodate(page))
-+ goto err;
-+ e3b->bd_buddy = page_address(page) + (poff * sb->s_blocksize);
-+ mark_page_accessed(page);
-+
-+ BUG_ON(e3b->bd_bitmap_page == NULL);
-+ BUG_ON(e3b->bd_buddy_page == NULL);
-+
-+ return 0;
-+
-+err:
-+ if (e3b->bd_bitmap_page)
-+ page_cache_release(e3b->bd_bitmap_page);
-+ if (e3b->bd_buddy_page)
-+ page_cache_release(e3b->bd_buddy_page);
-+ e3b->bd_buddy = NULL;
-+ e3b->bd_bitmap = NULL;
-+ return -EIO;
-+}
-+
-+static void ext3_mb_release_desc(struct ext3_buddy *e3b)
-+{
-+ if (e3b->bd_bitmap_page)
-+ page_cache_release(e3b->bd_bitmap_page);
-+ if (e3b->bd_buddy_page)
-+ page_cache_release(e3b->bd_buddy_page);
-+}
-+
-+
-+static int mb_find_order_for_block(struct ext3_buddy *e3b, int block)
-+{
-+ int order = 1;
-+ void *bb;
-+
-+ BUG_ON(EXT3_MB_BITMAP(e3b) == EXT3_MB_BUDDY(e3b));
-+ BUG_ON(block >= (1 << (e3b->bd_blkbits + 3)));
-+
-+ bb = EXT3_MB_BUDDY(e3b);
-+ while (order <= e3b->bd_blkbits + 1) {
-+ block = block >> 1;
-+ if (!mb_test_bit(block, bb)) {
-+ /* this block is part of buddy of order 'order' */
-+ return order;
-+ }
-+ bb += 1 << (e3b->bd_blkbits - order);
-+ order++;
-+ }
-+ return 0;
-+}
-+
-+static inline void mb_clear_bits(spinlock_t *lock, void *bm, int cur, int len)
-+{
-+ __u32 *addr;
-+
-+ len = cur + len;
-+ while (cur < len) {
-+ if ((cur & 31) == 0 && (len - cur) >= 32) {
-+ /* fast path: clear whole word at once */
-+ addr = bm + (cur >> 3);
-+ *addr = 0;
-+ cur += 32;
-+ continue;
-+ }
-+ if (lock)
-+ mb_clear_bit_atomic(lock, cur, bm);
-+ else
-+ mb_clear_bit(cur, bm);
-+ cur++;
-+ }
-+}
-+
-+static inline void mb_set_bits(spinlock_t *lock, void *bm, int cur, int len)
-+{
-+ __u32 *addr;
-+
-+ len = cur + len;
-+ while (cur < len) {
-+ if ((cur & 31) == 0 && (len - cur) >= 32) {
-+ /* fast path: clear whole word at once */
-+ addr = bm + (cur >> 3);
-+ *addr = 0xffffffff;
-+ cur += 32;
-+ continue;
-+ }
-+ if (lock)
-+ mb_set_bit_atomic(lock, cur, bm);
-+ else
-+ mb_set_bit(cur, bm);
-+ cur++;
-+ }
-+}
-+
-+static int mb_free_blocks(struct inode *inode, struct ext3_buddy *e3b,
-+ int first, int count)
-+{
-+ int block = 0, max = 0, order;
-+ void *buddy, *buddy2;
-+ struct super_block *sb = e3b->bd_sb;
-+
-+ BUG_ON(first + count > (sb->s_blocksize << 3));
-+ BUG_ON(!ext3_is_group_locked(sb, e3b->bd_group));
-+ mb_check_buddy(e3b);
-+ mb_free_blocks_double(inode, e3b, first, count);
-+
-+ e3b->bd_info->bb_free += count;
-+ if (first < e3b->bd_info->bb_first_free)
-+ e3b->bd_info->bb_first_free = first;
-+
-+ /* let's maintain fragments counter */
-+ if (first != 0)
-+ block = !mb_test_bit(first - 1, EXT3_MB_BITMAP(e3b));
-+ if (first + count < EXT3_SB(sb)->s_mb_maxs[0])
-+ max = !mb_test_bit(first + count, EXT3_MB_BITMAP(e3b));
-+ if (block && max)
-+ e3b->bd_info->bb_fragments--;
-+ else if (!block && !max)
-+ e3b->bd_info->bb_fragments++;
-+
-+ /* let's maintain buddy itself */
-+ while (count-- > 0) {
-+ block = first++;
-+ order = 0;
-+
-+ if (!mb_test_bit(block, EXT3_MB_BITMAP(e3b))) {
-+ unsigned long blocknr;
-+ blocknr = e3b->bd_group * EXT3_BLOCKS_PER_GROUP(sb);
-+ blocknr += block;
-+ blocknr +=
-+ le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block);
-+
-+ ext3_error(sb, __FUNCTION__, "double-free of inode"
-+ " %lu's block %lu(bit %u in group %u)\n",
-+ inode ? inode->i_ino : 0, blocknr, block,
-+ e3b->bd_group);
-+ }
-+ mb_clear_bit(block, EXT3_MB_BITMAP(e3b));
-+ e3b->bd_info->bb_counters[order]++;
-+
-+ /* start of the buddy */
-+ buddy = mb_find_buddy(e3b, order, &max);
-+
-+ do {
-+ block &= ~1UL;
-+ if (mb_test_bit(block, buddy) ||
-+ mb_test_bit(block + 1, buddy))
-+ break;
-+
-+ /* both the buddies are free, try to coalesce them */
-+ buddy2 = mb_find_buddy(e3b, order + 1, &max);
-+
-+ if (!buddy2)
-+ break;
-+
-+ if (order > 0) {
-+ /* for special purposes, we don't set
-+ * free bits in bitmap */
-+ mb_set_bit(block, buddy);
-+ mb_set_bit(block + 1, buddy);
-+ }
-+ e3b->bd_info->bb_counters[order]--;
-+ e3b->bd_info->bb_counters[order]--;
-+
-+ block = block >> 1;
-+ order++;
-+ e3b->bd_info->bb_counters[order]++;
-+
-+ mb_clear_bit(block, buddy2);
-+ buddy = buddy2;
-+ } while (1);
-+ }
-+ mb_check_buddy(e3b);
-+
-+ return 0;
-+}
-+
-+static int mb_find_extent(struct ext3_buddy *e3b, int order, int block,
-+ int needed, struct ext3_free_extent *ex)
-+{
-+ int next = block, max, ord;
-+ void *buddy;
-+
-+ BUG_ON(!ext3_is_group_locked(e3b->bd_sb, e3b->bd_group));
-+ BUG_ON(ex == NULL);
-+
-+ buddy = mb_find_buddy(e3b, order, &max);
-+ BUG_ON(buddy == NULL);
-+ BUG_ON(block >= max);
-+ if (mb_test_bit(block, buddy)) {
-+ ex->fe_len = 0;
-+ ex->fe_start = 0;
-+ ex->fe_group = 0;
-+ return 0;
-+ }
-+
-+ if (likely(order == 0)) {
-+ /* find actual order */
-+ order = mb_find_order_for_block(e3b, block);
-+ block = block >> order;
-+ }
-+
-+ ex->fe_len = 1 << order;
-+ ex->fe_start = block << order;
-+ ex->fe_group = e3b->bd_group;
-+
-+ /* calc difference from given start */
-+ next = next - ex->fe_start;
-+ ex->fe_len -= next;
-+ ex->fe_start += next;
-+
-+ while (needed > ex->fe_len && (buddy = mb_find_buddy(e3b, order, &max))) {
-+
-+ if (block + 1 >= max)
-+ break;
-+
-+ next = (block + 1) * (1 << order);
-+ if (mb_test_bit(next, EXT3_MB_BITMAP(e3b)))
-+ break;
-+
-+ ord = mb_find_order_for_block(e3b, next);
-+
-+ order = ord;
-+ block = next >> order;
-+ ex->fe_len += 1 << order;
-+ }
-+
-+ BUG_ON(ex->fe_start + ex->fe_len > (1 << (e3b->bd_blkbits + 3)));
-+ return ex->fe_len;
-+}
-+
-+static int mb_mark_used(struct ext3_buddy *e3b, struct ext3_free_extent *ex)
-+{
-+ int ord, mlen = 0, max = 0, cur;
-+ int start = ex->fe_start;
-+ int len = ex->fe_len;
-+ unsigned ret = 0;
-+ int len0 = len;
-+ void *buddy;
-+
-+ BUG_ON(start + len > (e3b->bd_sb->s_blocksize << 3));
-+ BUG_ON(e3b->bd_group != ex->fe_group);
-+ BUG_ON(!ext3_is_group_locked(e3b->bd_sb, e3b->bd_group));
-+ spin_lock(sb_bgl_lock(EXT3_SB(e3b->bd_sb), ex->fe_group));
-+ mb_check_buddy(e3b);
-+ mb_mark_used_double(e3b, start, len);
-+
-+ e3b->bd_info->bb_free -= len;
-+ if (e3b->bd_info->bb_first_free == start)
-+ e3b->bd_info->bb_first_free += len;
-+
-+ /* let's maintain fragments counter */
-+ if (start != 0)
-+ mlen = !mb_test_bit(start - 1, EXT3_MB_BITMAP(e3b));
-+ if (start + len < EXT3_SB(e3b->bd_sb)->s_mb_maxs[0])
-+ max = !mb_test_bit(start + len, EXT3_MB_BITMAP(e3b));
-+ if (mlen && max)
-+ e3b->bd_info->bb_fragments++;
-+ else if (!mlen && !max)
-+ e3b->bd_info->bb_fragments--;
-+
-+ /* let's maintain buddy itself */
-+ while (len) {
-+ ord = mb_find_order_for_block(e3b, start);
-+
-+ if (((start >> ord) << ord) == start && len >= (1 << ord)) {
-+ /* the whole chunk may be allocated at once! */
-+ mlen = 1 << ord;
-+ buddy = mb_find_buddy(e3b, ord, &max);
-+ BUG_ON((start >> ord) >= max);
-+ mb_set_bit(start >> ord, buddy);
-+ e3b->bd_info->bb_counters[ord]--;
-+ start += mlen;
-+ len -= mlen;
-+ BUG_ON(len < 0);
-+ continue;
-+ }
-+
-+ /* store for history */
-+ if (ret == 0)
-+ ret = len | (ord << 16);
-+
-+ /* we have to split large buddy */
-+ BUG_ON(ord <= 0);
-+ buddy = mb_find_buddy(e3b, ord, &max);
-+ mb_set_bit(start >> ord, buddy);
-+ e3b->bd_info->bb_counters[ord]--;
-+
-+ ord--;
-+ cur = (start >> ord) & ~1U;
-+ buddy = mb_find_buddy(e3b, ord, &max);
-+ mb_clear_bit(cur, buddy);
-+ mb_clear_bit(cur + 1, buddy);
-+ e3b->bd_info->bb_counters[ord]++;
-+ e3b->bd_info->bb_counters[ord]++;
-+ }
-+
-+ mb_set_bits(NULL, EXT3_MB_BITMAP(e3b), ex->fe_start, len0);
-+ mb_check_buddy(e3b);
-+ spin_unlock(sb_bgl_lock(EXT3_SB(e3b->bd_sb), ex->fe_group));
-+
-+ return ret;
-+}
-+
-+/*
-+ * Must be called under group lock!
-+ */
-+static void ext3_mb_use_best_found(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ unsigned long ret;
-+
-+ BUG_ON(ac->ac_b_ex.fe_group != e3b->bd_group);
-+ BUG_ON(ac->ac_status == AC_STATUS_FOUND);
-+
-+ ac->ac_b_ex.fe_len = min(ac->ac_b_ex.fe_len, ac->ac_g_ex.fe_len);
-+ ac->ac_b_ex.fe_logical = ac->ac_g_ex.fe_logical;
-+ ret = mb_mark_used(e3b, &ac->ac_b_ex);
-+
-+ /* preallocation can change ac_b_ex, thus we store actually
-+ * allocated blocks for history */
-+ ac->ac_f_ex = ac->ac_b_ex;
-+
-+ ac->ac_status = AC_STATUS_FOUND;
-+ ac->ac_tail = ret & 0xffff;
-+ ac->ac_buddy = ret >> 16;
-+
-+ /* XXXXXXX: SUCH A HORRIBLE **CK */
-+ ac->ac_bitmap_page = e3b->bd_bitmap_page;
-+ get_page(ac->ac_bitmap_page);
-+ ac->ac_buddy_page = e3b->bd_buddy_page;
-+ get_page(ac->ac_buddy_page);
-+
-+ /* store last allocated for subsequent stream allocation */
-+ if ((ac->ac_flags & EXT3_MB_HINT_DATA)) {
-+ spin_lock(&sbi->s_md_lock);
-+ sbi->s_mb_last_group = ac->ac_f_ex.fe_group;
-+ sbi->s_mb_last_start = ac->ac_f_ex.fe_start;
-+ spin_unlock(&sbi->s_md_lock);
-+ }
-+}
-+
-+/*
-+ * regular allocator, for general purposes allocation
-+ */
-+
-+void ext3_mb_check_limits(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b,
-+ int finish_group)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ struct ext3_free_extent *bex = &ac->ac_b_ex;
-+ struct ext3_free_extent *gex = &ac->ac_g_ex;
-+ struct ext3_free_extent ex;
-+ int max;
-+
-+ /*
-+ * We don't want to scan for a whole year
-+ */
-+ if (ac->ac_found > sbi->s_mb_max_to_scan &&
-+ !(ac->ac_flags & EXT3_MB_HINT_FIRST)) {
-+ ac->ac_status = AC_STATUS_BREAK;
-+ return;
-+ }
-+
-+ /*
-+ * Haven't found good chunk so far, let's continue
-+ */
-+ if (bex->fe_len < gex->fe_len)
-+ return;
-+
-+ if ((finish_group || ac->ac_found > sbi->s_mb_min_to_scan)
-+ && bex->fe_group == e3b->bd_group) {
-+ /* recheck chunk's availability - we don't know
-+ * when it was found (within this lock-unlock
-+ * period or not) */
-+ max = mb_find_extent(e3b, 0, bex->fe_start, gex->fe_len, &ex);
-+ if (max >= gex->fe_len) {
-+ ext3_mb_use_best_found(ac, e3b);
-+ return;
-+ }
-+ }
-+}
-+
-+/*
-+ * The routine checks whether found extent is good enough. If it is,
-+ * then the extent gets marked used and flag is set to the context
-+ * to stop scanning. Otherwise, the extent is compared with the
-+ * previous found extent and if new one is better, then it's stored
-+ * in the context. Later, the best found extent will be used, if
-+ * mballoc can't find good enough extent.
-+ *
-+ * FIXME: real allocation policy is to be designed yet!
-+ */
-+static void ext3_mb_measure_extent(struct ext3_allocation_context *ac,
-+ struct ext3_free_extent *ex,
-+ struct ext3_buddy *e3b)
-+{
-+ struct ext3_free_extent *bex = &ac->ac_b_ex;
-+ struct ext3_free_extent *gex = &ac->ac_g_ex;
-+
-+ BUG_ON(ex->fe_len <= 0);
-+ BUG_ON(ex->fe_len >= EXT3_BLOCKS_PER_GROUP(ac->ac_sb));
-+ BUG_ON(ex->fe_start >= EXT3_BLOCKS_PER_GROUP(ac->ac_sb));
-+ BUG_ON(ac->ac_status != AC_STATUS_CONTINUE);
-+
-+ ac->ac_found++;
-+
-+ /*
-+ * The special case - take what you catch first
-+ */
-+ if (unlikely(ac->ac_flags & EXT3_MB_HINT_FIRST)) {
-+ *bex = *ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ return;
-+ }
-+
-+ /*
-+ * Let's check whether the chuck is good enough
-+ */
-+ if (ex->fe_len == gex->fe_len) {
-+ *bex = *ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ return;
-+ }
-+
-+ /*
-+ * If this is first found extent, just store it in the context
-+ */
-+ if (bex->fe_len == 0) {
-+ *bex = *ex;
-+ return;
-+ }
-+
-+ /*
-+ * If new found extent is better, store it in the context
-+ */
-+ if (bex->fe_len < gex->fe_len) {
-+ /* if the request isn't satisfied, any found extent
-+ * larger than previous best one is better */
-+ if (ex->fe_len > bex->fe_len)
-+ *bex = *ex;
-+ } else if (ex->fe_len > gex->fe_len) {
-+ /* if the request is satisfied, then we try to find
-+ * an extent that still satisfy the request, but is
-+ * smaller than previous one */
-+ if (ex->fe_len < bex->fe_len)
-+ *bex = *ex;
-+ }
-+
-+ ext3_mb_check_limits(ac, e3b, 0);
-+}
-+
-+static int ext3_mb_try_best_found(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ struct ext3_free_extent ex = ac->ac_b_ex;
-+ int group = ex.fe_group, max, err;
-+
-+ BUG_ON(ex.fe_len <= 0);
-+ err = ext3_mb_load_buddy(ac->ac_sb, group, e3b);
-+ if (err)
-+ return err;
-+
-+ ext3_lock_group(ac->ac_sb, group);
-+ max = mb_find_extent(e3b, 0, ex.fe_start, ex.fe_len, &ex);
-+
-+ if (max > 0) {
-+ ac->ac_b_ex = ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ }
-+
-+ ext3_unlock_group(ac->ac_sb, group);
-+ ext3_mb_release_desc(e3b);
-+
-+ return 0;
-+}
-+
-+static int ext3_mb_find_by_goal(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ int group = ac->ac_g_ex.fe_group, max, err;
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ struct ext3_super_block *es = sbi->s_es;
-+ struct ext3_free_extent ex;
-+
-+ if (!(ac->ac_flags & EXT3_MB_HINT_TRY_GOAL))
-+ return 0;
-+
-+ err = ext3_mb_load_buddy(ac->ac_sb, group, e3b);
-+ if (err)
-+ return err;
-+
-+ ext3_lock_group(ac->ac_sb, group);
-+ max = mb_find_extent(e3b, 0, ac->ac_g_ex.fe_start,
-+ ac->ac_g_ex.fe_len, &ex);
-+
-+ if (max >= ac->ac_g_ex.fe_len && ac->ac_g_ex.fe_len == sbi->s_stripe) {
-+ unsigned long start;
-+ start = (e3b->bd_group * EXT3_BLOCKS_PER_GROUP(ac->ac_sb) +
-+ ex.fe_start + le32_to_cpu(es->s_first_data_block));
-+ if (start % sbi->s_stripe == 0) {
-+ ac->ac_found++;
-+ ac->ac_b_ex = ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ }
-+ } else if (max >= ac->ac_g_ex.fe_len) {
-+ BUG_ON(ex.fe_len <= 0);
-+ BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group);
-+ BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start);
-+ ac->ac_found++;
-+ ac->ac_b_ex = ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ } else if (max > 0 && (ac->ac_flags & EXT3_MB_HINT_MERGE)) {
-+ /* Sometimes, caller may want to merge even small
-+ * number of blocks to an existing extent */
-+ BUG_ON(ex.fe_len <= 0);
-+ BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group);
-+ BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start);
-+ ac->ac_found++;
-+ ac->ac_b_ex = ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ }
-+ ext3_unlock_group(ac->ac_sb, group);
-+ ext3_mb_release_desc(e3b);
-+
-+ return 0;
-+}
-+
-+/*
-+ * The routine scans buddy structures (not bitmap!) from given order
-+ * to max order and tries to find big enough chunk to satisfy the req
-+ */
-+static void ext3_mb_simple_scan_group(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ struct ext3_group_info *grp = e3b->bd_info;
-+ void *buddy;
-+ int i, k, max;
-+
-+ BUG_ON(ac->ac_2order <= 0);
-+ for (i = ac->ac_2order; i <= sb->s_blocksize_bits + 1; i++) {
-+ if (grp->bb_counters[i] == 0)
-+ continue;
-+
-+ buddy = mb_find_buddy(e3b, i, &max);
-+ BUG_ON(buddy == NULL);
-+
-+ k = mb_find_next_zero_bit(buddy, max, 0);
-+ BUG_ON(k >= max);
-+
-+ ac->ac_found++;
-+
-+ ac->ac_b_ex.fe_len = 1 << i;
-+ ac->ac_b_ex.fe_start = k << i;
-+ ac->ac_b_ex.fe_group = e3b->bd_group;
-+
-+ ext3_mb_use_best_found(ac, e3b);
-+
-+ BUG_ON(ac->ac_b_ex.fe_len != ac->ac_g_ex.fe_len);
-+
-+ if (EXT3_SB(sb)->s_mb_stats)
-+ atomic_inc(&EXT3_SB(sb)->s_bal_2orders);
-+
-+ break;
-+ }
-+}
-+
-+/*
-+ * The routine scans the group and measures all found extents.
-+ * In order to optimize scanning, caller must pass number of
-+ * free blocks in the group, so the routine can know upper limit.
-+ */
-+static void ext3_mb_complex_scan_group(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ void *bitmap = EXT3_MB_BITMAP(e3b);
-+ struct ext3_free_extent ex;
-+ int i, free;
-+
-+ free = e3b->bd_info->bb_free;
-+ BUG_ON(free <= 0);
-+
-+ i = e3b->bd_info->bb_first_free;
-+
-+ while (free && ac->ac_status == AC_STATUS_CONTINUE) {
-+ i = mb_find_next_zero_bit(bitmap, EXT3_BLOCKS_PER_GROUP(sb), i);
-+ if (i >= EXT3_BLOCKS_PER_GROUP(sb)) {
-+ BUG_ON(free != 0);
-+ break;
-+ }
-+
-+ mb_find_extent(e3b, 0, i, ac->ac_g_ex.fe_len, &ex);
-+ BUG_ON(ex.fe_len <= 0);
-+ BUG_ON(free < ex.fe_len);
-+
-+ ext3_mb_measure_extent(ac, &ex, e3b);
-+
-+ i += ex.fe_len;
-+ free -= ex.fe_len;
-+ }
-+
-+ ext3_mb_check_limits(ac, e3b, 1);
-+}
-+
-+/*
-+ * This is a special case for storages like raid5
-+ * we try to find stripe-aligned chunks for stripe-size requests
-+ */
-+static void ext3_mb_scan_aligned(struct ext3_allocation_context *ac,
-+ struct ext3_buddy *e3b)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ void *bitmap = EXT3_MB_BITMAP(e3b);
-+ struct ext3_free_extent ex;
-+ unsigned long i, max;
-+
-+ BUG_ON(sbi->s_stripe == 0);
-+
-+ /* find first stripe-aligned block */
-+ i = e3b->bd_group * EXT3_BLOCKS_PER_GROUP(sb)
-+ + le32_to_cpu(sbi->s_es->s_first_data_block);
-+ i = ((i + sbi->s_stripe - 1) / sbi->s_stripe) * sbi->s_stripe;
-+ i = (i - le32_to_cpu(sbi->s_es->s_first_data_block))
-+ % EXT3_BLOCKS_PER_GROUP(sb);
-+
-+ while (i < EXT3_BLOCKS_PER_GROUP(sb)) {
-+ if (!mb_test_bit(i, bitmap)) {
-+ max = mb_find_extent(e3b, 0, i, sbi->s_stripe, &ex);
-+ if (max >= sbi->s_stripe) {
-+ ac->ac_found++;
-+ ac->ac_b_ex = ex;
-+ ext3_mb_use_best_found(ac, e3b);
-+ break;
-+ }
-+ }
-+ i += sbi->s_stripe;
-+ }
-+}
-+
-+static int ext3_mb_good_group(struct ext3_allocation_context *ac,
-+ int group, int cr)
-+{
-+ struct ext3_group_info *grp = EXT3_GROUP_INFO(ac->ac_sb, group);
-+ unsigned free, fragments, i, bits;
-+
-+ BUG_ON(cr < 0 || cr >= 4);
-+ BUG_ON(EXT3_MB_GRP_NEED_INIT(grp));
-+
-+ free = grp->bb_free;
-+ fragments = grp->bb_fragments;
-+ if (free == 0)
-+ return 0;
-+ if (fragments == 0)
-+ return 0;
-+
-+ switch (cr) {
-+ case 0:
-+ BUG_ON(ac->ac_2order == 0);
-+ bits = ac->ac_sb->s_blocksize_bits + 1;
-+ for (i = ac->ac_2order; i <= bits; i++)
-+ if (grp->bb_counters[i] > 0)
-+ return 1;
-+ break;
-+ case 1:
-+ if ((free / fragments) >= ac->ac_g_ex.fe_len)
-+ return 1;
-+ break;
-+ case 2:
-+ if (free >= ac->ac_g_ex.fe_len)
-+ return 1;
-+ break;
-+ case 3:
-+ return 1;
-+ default:
-+ BUG();
-+ }
-+
-+ return 0;
-+}
-+
-+int ext3_mb_regular_allocator(struct ext3_allocation_context *ac)
-+{
-+ int group, i, cr, err = 0;
-+ struct ext3_sb_info *sbi;
-+ struct super_block *sb;
-+ struct ext3_buddy e3b;
-+
-+ sb = ac->ac_sb;
-+ sbi = EXT3_SB(sb);
-+ BUG_ON(ac->ac_status == AC_STATUS_FOUND);
-+
-+ /* first, try the goal */
-+ err = ext3_mb_find_by_goal(ac, &e3b);
-+ if (err || ac->ac_status == AC_STATUS_FOUND)
-+ goto out;
-+
-+ if (unlikely(ac->ac_flags & EXT3_MB_HINT_GOAL_ONLY))
-+ goto out;
-+
-+ i = ffs(ac->ac_g_ex.fe_len);
-+ ac->ac_2order = 0;
-+ if (i >= sbi->s_mb_order2_reqs) {
-+ i--;
-+ if ((ac->ac_g_ex.fe_len & (~(1 << i))) == 0)
-+ ac->ac_2order = i;
-+ }
-+
-+ /* if stream allocation is enabled, use global goal */
-+ if ((ac->ac_g_ex.fe_len < sbi->s_mb_large_req) &&
-+ (ac->ac_flags & EXT3_MB_HINT_DATA)) {
-+ /* TBD: may be hot point */
-+ spin_lock(&sbi->s_md_lock);
-+ ac->ac_g_ex.fe_group = sbi->s_mb_last_group;
-+ ac->ac_g_ex.fe_start = sbi->s_mb_last_start;
-+ spin_unlock(&sbi->s_md_lock);
-+ }
-+
-+ group = ac->ac_g_ex.fe_group;
-+
-+ /* Let's just scan groups to find more-less suitable blocks */
-+ cr = ac->ac_2order ? 0 : 1;
-+repeat:
-+ for (; cr < 4 && ac->ac_status == AC_STATUS_CONTINUE; cr++) {
-+ ac->ac_criteria = cr;
-+ for (i = 0; i < EXT3_SB(sb)->s_groups_count; group++, i++) {
-+ struct ext3_group_info *grp;
-+
-+ if (group == EXT3_SB(sb)->s_groups_count)
-+ group = 0;
-+
-+ /* If there's no chance that this group has a better
-+ * extent, just skip it instead of seeking to read
-+ * block bitmap from disk. Initially ac_b_ex.fe_len = 0,
-+ * so this always skips groups with no free space. */
-+ grp = EXT3_GROUP_INFO(ac->ac_sb, group);
-+ if (grp->bb_free <= ac->ac_b_ex.fe_len)
-+ continue;
-+
-+ if (EXT3_MB_GRP_NEED_INIT(EXT3_GROUP_INFO(sb, group))) {
-+ /* we need full data about the group
-+ * to make a good selection */
-+ err = ext3_mb_load_buddy(sb, group, &e3b);
-+ if (err)
-+ goto out;
-+ ext3_mb_release_desc(&e3b);
-+ }
-+
-+ /* check is group good for our criteries */
-+ if (!ext3_mb_good_group(ac, group, cr))
-+ continue;
-+
-+ err = ext3_mb_load_buddy(sb, group, &e3b);
-+ if (err)
-+ goto out;
-+
-+ ext3_lock_group(sb, group);
-+ if (!ext3_mb_good_group(ac, group, cr)) {
-+ /* someone did allocation from this group */
-+ ext3_unlock_group(sb, group);
-+ ext3_mb_release_desc(&e3b);
-+ continue;
-+ }
-+
-+ ac->ac_groups_scanned++;
-+ if (cr == 0)
-+ ext3_mb_simple_scan_group(ac, &e3b);
-+ else if (cr == 1 && ac->ac_g_ex.fe_len == sbi->s_stripe)
-+ ext3_mb_scan_aligned(ac, &e3b);
-+ else
-+ ext3_mb_complex_scan_group(ac, &e3b);
-+
-+ ext3_unlock_group(sb, group);
-+ ext3_mb_release_desc(&e3b);
-+
-+ if (ac->ac_status != AC_STATUS_CONTINUE)
-+ break;
-+ }
-+ }
-+
-+ if (ac->ac_b_ex.fe_len > 0 && ac->ac_status != AC_STATUS_FOUND &&
-+ !(ac->ac_flags & EXT3_MB_HINT_FIRST)) {
-+ /*
-+ * We've been searching too long. Let's try to allocate
-+ * the best chunk we've found so far
-+ */
-+
-+ ext3_mb_try_best_found(ac, &e3b);
-+ if (ac->ac_status != AC_STATUS_FOUND) {
-+ /*
-+ * Someone more lucky has already allocated it.
-+ * The only thing we can do is just take first
-+ * found block(s)
-+ printk(KERN_DEBUG "EXT3-fs: someone won our chunk\n");
-+ */
-+ ac->ac_b_ex.fe_group = 0;
-+ ac->ac_b_ex.fe_start = 0;
-+ ac->ac_b_ex.fe_len = 0;
-+ ac->ac_status = AC_STATUS_CONTINUE;
-+ ac->ac_flags |= EXT3_MB_HINT_FIRST;
-+ cr = 3;
-+ atomic_inc(&sbi->s_mb_lost_chunks);
-+ goto repeat;
-+ }
-+ }
-+out:
-+ return err;
-+}
-+
-+#ifdef EXT3_MB_HISTORY
-+struct ext3_mb_proc_session {
-+ struct ext3_mb_history *history;
-+ struct super_block *sb;
-+ int start;
-+ int max;
-+};
-+
-+static void *ext3_mb_history_skip_empty(struct ext3_mb_proc_session *s,
-+ struct ext3_mb_history *hs,
-+ int first)
-+{
-+ if (hs == s->history + s->max)
-+ hs = s->history;
-+ if (!first && hs == s->history + s->start)
-+ return NULL;
-+ while (hs->orig.fe_len == 0) {
-+ hs++;
-+ if (hs == s->history + s->max)
-+ hs = s->history;
-+ if (hs == s->history + s->start)
-+ return NULL;
-+ }
-+ return hs;
-+}
-+
-+static void *ext3_mb_seq_history_start(struct seq_file *seq, loff_t *pos)
-+{
-+ struct ext3_mb_proc_session *s = seq->private;
-+ struct ext3_mb_history *hs;
-+ int l = *pos;
-+
-+ if (l == 0)
-+ return SEQ_START_TOKEN;
-+ hs = ext3_mb_history_skip_empty(s, s->history + s->start, 1);
-+ if (!hs)
-+ return NULL;
-+ while (--l && (hs = ext3_mb_history_skip_empty(s, ++hs, 0)) != NULL);
-+ return hs;
-+}
-+
-+static void *ext3_mb_seq_history_next(struct seq_file *seq, void *v, loff_t *pos)
-+{
-+ struct ext3_mb_proc_session *s = seq->private;
-+ struct ext3_mb_history *hs = v;
-+
-+ ++*pos;
-+ if (v == SEQ_START_TOKEN)
-+ return ext3_mb_history_skip_empty(s, s->history + s->start, 1);
-+ else
-+ return ext3_mb_history_skip_empty(s, ++hs, 0);
-+}
-+
-+static int ext3_mb_seq_history_show(struct seq_file *seq, void *v)
-+{
-+ char buf[25], buf2[25], buf3[25], *fmt;
-+ struct ext3_mb_history *hs = v;
-+
-+ if (v == SEQ_START_TOKEN) {
-+ seq_printf(seq, "%-5s %-8s %-23s %-23s %-23s %-5s "
-+ "%-5s %-2s %-5s %-5s %-5s %-6s\n",
-+ "pid", "inode", "original", "goal", "result","found",
-+ "grps", "cr", "flags", "merge", "tail", "broken");
-+ return 0;
-+ }
-+
-+ if (hs->op == EXT3_MB_HISTORY_ALLOC) {
-+ fmt = "%-5u %-8u %-23s %-23s %-23s %-5u %-5u %-2u "
-+ "%-5u %-5s %-5u %-6u\n";
-+ sprintf(buf2, "%lu/%lu/%lu@%lu", hs->result.fe_group,
-+ hs->result.fe_start, hs->result.fe_len,
-+ hs->result.fe_logical);
-+ sprintf(buf, "%lu/%lu/%lu@%lu", hs->orig.fe_group,
-+ hs->orig.fe_start, hs->orig.fe_len,
-+ hs->orig.fe_logical);
-+ sprintf(buf3, "%lu/%lu/%lu@%lu", hs->goal.fe_group,
-+ hs->goal.fe_start, hs->goal.fe_len,
-+ hs->goal.fe_logical);
-+ seq_printf(seq, fmt, hs->pid, hs->ino, buf, buf3, buf2,
-+ hs->found, hs->groups, hs->cr, hs->flags,
-+ hs->merged ? "M" : "", hs->tail,
-+ hs->buddy ? 1 << hs->buddy : 0);
-+ } else if (hs->op == EXT3_MB_HISTORY_PREALLOC) {
-+ fmt = "%-5u %-8u %-23s %-23s %-23s\n";
-+ sprintf(buf2, "%lu/%lu/%lu@%lu", hs->result.fe_group,
-+ hs->result.fe_start, hs->result.fe_len,
-+ hs->result.fe_logical);
-+ sprintf(buf, "%lu/%lu/%lu@%lu", hs->orig.fe_group,
-+ hs->orig.fe_start, hs->orig.fe_len,
-+ hs->orig.fe_logical);
-+ seq_printf(seq, fmt, hs->pid, hs->ino, buf, "", buf2);
-+ } else if (hs->op == EXT3_MB_HISTORY_DISCARD) {
-+ sprintf(buf2, "%lu/%lu/%lu", hs->result.fe_group,
-+ hs->result.fe_start, hs->result.fe_len);
-+ seq_printf(seq, "%-5u %-8u %-23s discard\n",
-+ hs->pid, hs->ino, buf2);
-+ } else if (hs->op == EXT3_MB_HISTORY_FREE) {
-+ sprintf(buf2, "%lu/%lu/%lu", hs->result.fe_group,
-+ hs->result.fe_start, hs->result.fe_len);
-+ seq_printf(seq, "%-5u %-8u %-23s free\n",
-+ hs->pid, hs->ino, buf2);
-+ } else {
-+ seq_printf(seq, "unknown op %d\n", hs->op);
-+ }
-+ return 0;
-+}
-+
-+static void ext3_mb_seq_history_stop(struct seq_file *seq, void *v)
-+{
-+}
-+
-+static struct seq_operations ext3_mb_seq_history_ops = {
-+ .start = ext3_mb_seq_history_start,
-+ .next = ext3_mb_seq_history_next,
-+ .stop = ext3_mb_seq_history_stop,
-+ .show = ext3_mb_seq_history_show,
-+};
-+
-+static int ext3_mb_seq_history_open(struct inode *inode, struct file *file)
-+{
-+ struct super_block *sb = PDE(inode)->data;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ struct ext3_mb_proc_session *s;
-+ int rc, size;
-+
-+ s = kmalloc(sizeof(*s), GFP_KERNEL);
-+ if (s == NULL)
-+ return -ENOMEM;
-+ s->sb = sb;
-+ size = sizeof(struct ext3_mb_history) * sbi->s_mb_history_max;
-+ s->history = kmalloc(size, GFP_KERNEL);
-+ if (s->history == NULL) {
-+ kfree(s);
-+ return -ENOMEM;
-+ }
-+
-+ spin_lock(&sbi->s_mb_history_lock);
-+ memcpy(s->history, sbi->s_mb_history, size);
-+ s->max = sbi->s_mb_history_max;
-+ s->start = sbi->s_mb_history_cur % s->max;
-+ spin_unlock(&sbi->s_mb_history_lock);
-+
-+ rc = seq_open(file, &ext3_mb_seq_history_ops);
-+ if (rc == 0) {
-+ struct seq_file *m = (struct seq_file *)file->private_data;
-+ m->private = s;
-+ } else {
-+ kfree(s->history);
-+ kfree(s);
-+ }
-+ return rc;
-+
-+}
-+
-+static int ext3_mb_seq_history_release(struct inode *inode, struct file *file)
-+{
-+ struct seq_file *seq = (struct seq_file *)file->private_data;
-+ struct ext3_mb_proc_session *s = seq->private;
-+ kfree(s->history);
-+ kfree(s);
-+ return seq_release(inode, file);
-+}
-+
-+static ssize_t ext3_mb_seq_history_write(struct file *file,
-+ const char __user *buffer,
-+ size_t count, loff_t *ppos)
-+{
-+ struct seq_file *seq = (struct seq_file *)file->private_data;
-+ struct ext3_mb_proc_session *s = seq->private;
-+ struct super_block *sb = s->sb;
-+ char str[32];
-+ int value;
-+
-+ if (count >= sizeof(str)) {
-+ printk(KERN_ERR "EXT3-fs: %s string too long, max %u bytes\n",
-+ "mb_history", (int)sizeof(str));
-+ return -EOVERFLOW;
-+ }
-+
-+ if (copy_from_user(str, buffer, count))
-+ return -EFAULT;
-+
-+ value = simple_strtol(str, NULL, 0);
-+ if (value < 0)
-+ return -ERANGE;
-+ EXT3_SB(sb)->s_mb_history_filter = value;
-+
-+ return count;
-+}
-+
-+static struct file_operations ext3_mb_seq_history_fops = {
-+ .owner = THIS_MODULE,
-+ .open = ext3_mb_seq_history_open,
-+ .read = seq_read,
-+ .write = ext3_mb_seq_history_write,
-+ .llseek = seq_lseek,
-+ .release = ext3_mb_seq_history_release,
-+};
-+
-+static void *ext3_mb_seq_groups_start(struct seq_file *seq, loff_t *pos)
-+{
-+ struct super_block *sb = seq->private;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ long group;
-+
-+ if (*pos < 0 || *pos >= sbi->s_groups_count)
-+ return NULL;
-+
-+ group = *pos + 1;
-+ return (void *) group;
-+}
-+
-+static void *ext3_mb_seq_groups_next(struct seq_file *seq, void *v, loff_t *pos)
-+{
-+ struct super_block *sb = seq->private;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ long group;
-+
-+ ++*pos;
-+ if (*pos < 0 || *pos >= sbi->s_groups_count)
-+ return NULL;
-+ group = *pos + 1;
-+ return (void *) group;;
-+}
-+
-+static int ext3_mb_seq_groups_show(struct seq_file *seq, void *v)
-+{
-+ struct super_block *sb = seq->private;
-+ struct ext3_group_desc *gdp;
-+ long group = (long) v;
-+ int i, err, free = 0;
-+ struct ext3_buddy e3b;
-+ struct sg {
-+ struct ext3_group_info info;
-+ unsigned short counters[16];
-+ } sg;
-+
-+ group--;
-+ if (group == 0)
-+ seq_printf(seq, "#%-5s: %-5s %-5s %-5s %-5s %-5s "
-+ "[ %-5s %-5s %-5s %-5s %-5s %-5s %-5s "
-+ "%-5s %-5s %-5s %-5s %-5s %-5s %-5s ]\n",
-+ "group", "free", "ingd", "frags", "first", "pa",
-+ "2^0", "2^1", "2^2", "2^3", "2^4", "2^5","2^6",
-+ "2^7", "2^8", "2^9", "2^10", "2^11", "2^12", "2^13");
-+
-+ i = (sb->s_blocksize_bits + 2) * sizeof(sg.info.bb_counters[0]) +
-+ sizeof(struct ext3_group_info);
-+ err = ext3_mb_load_buddy(sb, group, &e3b);
-+ if (err) {
-+ seq_printf(seq, "#%-5lu: I/O error\n", group);
-+ return 0;
-+ }
-+
-+ gdp = ext3_get_group_desc(sb, group, NULL);
-+ if (gdp != NULL)
-+ free = le16_to_cpu(gdp->bg_free_blocks_count);
-+
-+ ext3_lock_group(sb, group);
-+ memcpy(&sg, EXT3_GROUP_INFO(sb, group), i);
-+ ext3_unlock_group(sb, group);
-+ ext3_mb_release_desc(&e3b);
-+
-+ seq_printf(seq, "#%-5lu: %-5u %-5u %-5u %-5u %-5lu [", group,
-+ sg.info.bb_free, free,
-+ sg.info.bb_fragments, sg.info.bb_first_free,
-+ sg.info.bb_prealloc_nr);
-+ for (i = 0; i <= 13; i++)
-+ seq_printf(seq, " %-5u", i <= sb->s_blocksize_bits + 1 ?
-+ sg.info.bb_counters[i] : 0);
-+ seq_printf(seq, " ]\n");
-+
-+ return 0;
-+}
-+
-+static void ext3_mb_seq_groups_stop(struct seq_file *seq, void *v)
-+{
-+}
-+
-+static struct seq_operations ext3_mb_seq_groups_ops = {
-+ .start = ext3_mb_seq_groups_start,
-+ .next = ext3_mb_seq_groups_next,
-+ .stop = ext3_mb_seq_groups_stop,
-+ .show = ext3_mb_seq_groups_show,
-+};
-+
-+static int ext3_mb_seq_groups_open(struct inode *inode, struct file *file)
-+{
-+ struct super_block *sb = PDE(inode)->data;
-+ int rc;
-+
-+ rc = seq_open(file, &ext3_mb_seq_groups_ops);
-+ if (rc == 0) {
-+ struct seq_file *m = (struct seq_file *)file->private_data;
-+ m->private = sb;
-+ }
-+ return rc;
-+
-+}
-+
-+static struct file_operations ext3_mb_seq_groups_fops = {
-+ .owner = THIS_MODULE,
-+ .open = ext3_mb_seq_groups_open,
-+ .read = seq_read,
-+ .llseek = seq_lseek,
-+ .release = seq_release,
-+};
-+
-+static void ext3_mb_history_release(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+
-+ remove_proc_entry("mb_groups", sbi->s_dev_proc);
-+ remove_proc_entry("mb_history", sbi->s_dev_proc);
-+
-+ if (sbi->s_mb_history)
-+ kfree(sbi->s_mb_history);
-+}
-+
-+static void ext3_mb_history_init(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ int i;
-+
-+ if (sbi->s_dev_proc != NULL) {
-+ struct proc_dir_entry *p;
-+ p = create_proc_entry("mb_history", S_IRUGO, sbi->s_dev_proc);
-+ if (p) {
-+ p->proc_fops = &ext3_mb_seq_history_fops;
-+ p->data = sb;
-+ }
-+ p = create_proc_entry("mb_groups", S_IRUGO, sbi->s_dev_proc);
-+ if (p) {
-+ p->proc_fops = &ext3_mb_seq_groups_fops;
-+ p->data = sb;
-+ }
-+ }
-+
-+ sbi->s_mb_history_max = 1000;
-+ sbi->s_mb_history_cur = 0;
-+ spin_lock_init(&sbi->s_mb_history_lock);
-+ i = sbi->s_mb_history_max * sizeof(struct ext3_mb_history);
-+ sbi->s_mb_history = kmalloc(i, GFP_KERNEL);
-+ if (likely(sbi->s_mb_history != NULL))
-+ memset(sbi->s_mb_history, 0, i);
-+ /* if we can't allocate history, then we simple won't use it */
-+}
-+
-+static void
-+ext3_mb_store_history(struct ext3_allocation_context *ac)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ struct ext3_mb_history h;
-+
-+ if (unlikely(sbi->s_mb_history == NULL))
-+ return;
-+
-+ if (!(ac->ac_op & sbi->s_mb_history_filter))
-+ return;
-+
-+ h.op = ac->ac_op;
-+ h.pid = current->pid;
-+ h.ino = ac->ac_inode ? ac->ac_inode->i_ino : 0;
-+ h.orig = ac->ac_o_ex;
-+ h.result = ac->ac_b_ex;
-+ h.flags = ac->ac_flags;
-+ h.found = ac->ac_found;
-+ h.groups = ac->ac_groups_scanned;
-+ h.cr = ac->ac_criteria;
-+ h.tail = ac->ac_tail;
-+ h.buddy = ac->ac_buddy;
-+ h.merged = 0;
-+ h.cr = ac->ac_criteria;
-+ if (ac->ac_op == EXT3_MB_HISTORY_ALLOC) {
-+ if (ac->ac_g_ex.fe_start == ac->ac_b_ex.fe_start &&
-+ ac->ac_g_ex.fe_group == ac->ac_b_ex.fe_group)
-+ h.merged = 1;
-+ h.goal = ac->ac_g_ex;
-+ h.result = ac->ac_f_ex;
-+ }
-+
-+ spin_lock(&sbi->s_mb_history_lock);
-+ memcpy(sbi->s_mb_history + sbi->s_mb_history_cur, &h, sizeof(h));
-+ if (++sbi->s_mb_history_cur >= sbi->s_mb_history_max)
-+ sbi->s_mb_history_cur = 0;
-+ spin_unlock(&sbi->s_mb_history_lock);
-+}
-+
-+#else
-+#define ext3_mb_history_release(sb)
-+#define ext3_mb_history_init(sb)
-+#endif
-+
-+int ext3_mb_init_backend(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ int i, j, len, metalen;
-+ int num_meta_group_infos =
-+ (sbi->s_groups_count + EXT3_DESC_PER_BLOCK(sb) - 1) >>
-+ EXT3_DESC_PER_BLOCK_BITS(sb);
-+ struct ext3_group_info **meta_group_info;
-+
-+ /* An 8TB filesystem with 64-bit pointers requires a 4096 byte
-+ * kmalloc. A 128kb malloc should suffice for a 256TB filesystem.
-+ * So a two level scheme suffices for now. */
-+ sbi->s_group_info = kmalloc(sizeof(*sbi->s_group_info) *
-+ num_meta_group_infos, GFP_KERNEL);
-+ if (sbi->s_group_info == NULL) {
-+ printk(KERN_ERR "EXT3-fs: can't allocate buddy meta group\n");
-+ return -ENOMEM;
-+ }
-+ sbi->s_buddy_cache = new_inode(sb);
-+ if (sbi->s_buddy_cache == NULL) {
-+ printk(KERN_ERR "EXT3-fs: can't get new inode\n");
-+ goto err_freesgi;
-+ }
-+ EXT3_I(sbi->s_buddy_cache)->i_disksize = 0;
-+
-+ metalen = sizeof(*meta_group_info) << EXT3_DESC_PER_BLOCK_BITS(sb);
-+ for (i = 0; i < num_meta_group_infos; i++) {
-+ if ((i + 1) == num_meta_group_infos)
-+ metalen = sizeof(*meta_group_info) *
-+ (sbi->s_groups_count -
-+ (i << EXT3_DESC_PER_BLOCK_BITS(sb)));
-+ meta_group_info = kmalloc(metalen, GFP_KERNEL);
-+ if (meta_group_info == NULL) {
-+ printk(KERN_ERR "EXT3-fs: can't allocate mem for a "
-+ "buddy group\n");
-+ goto err_freemeta;
-+ }
-+ sbi->s_group_info[i] = meta_group_info;
-+ }
-+
-+ /*
-+ * calculate needed size. if change bb_counters size,
-+ * don't forget about ext3_mb_generate_buddy()
-+ */
-+ len = sizeof(struct ext3_group_info);
-+ len += sizeof(unsigned short) * (sb->s_blocksize_bits + 2);
-+ for (i = 0; i < sbi->s_groups_count; i++) {
-+ struct ext3_group_desc * desc;
-+
-+ meta_group_info =
-+ sbi->s_group_info[i >> EXT3_DESC_PER_BLOCK_BITS(sb)];
-+ j = i & (EXT3_DESC_PER_BLOCK(sb) - 1);
-+
-+ meta_group_info[j] = kmalloc(len, GFP_KERNEL);
-+ if (meta_group_info[j] == NULL) {
-+ printk(KERN_ERR "EXT3-fs: can't allocate buddy mem\n");
-+ i--;
-+ goto err_freebuddy;
-+ }
-+ desc = ext3_get_group_desc(sb, i, NULL);
-+ if (desc == NULL) {
-+ printk(KERN_ERR"EXT3-fs: can't read descriptor %u\n",i);
-+ goto err_freebuddy;
-+ }
-+ memset(meta_group_info[j], 0, len);
-+ set_bit(EXT3_GROUP_INFO_NEED_INIT_BIT,
-+ &meta_group_info[j]->bb_state);
-+
-+ /* initialize bb_free to be able to skip
-+ * empty groups without initialization */
-+ meta_group_info[j]->bb_free =
-+ le16_to_cpu(desc->bg_free_blocks_count);
-+
-+ INIT_LIST_HEAD(&meta_group_info[j]->bb_prealloc_list);
-+
-+#ifdef DOUBLE_CHECK
-+ {
-+ struct buffer_head *bh;
-+ meta_group_info[j]->bb_bitmap =
-+ kmalloc(sb->s_blocksize, GFP_KERNEL);
-+ BUG_ON(meta_group_info[j]->bb_bitmap == NULL);
-+ bh = read_block_bitmap(sb, i);
-+ BUG_ON(bh == NULL);
-+ memcpy(meta_group_info[j]->bb_bitmap, bh->b_data,
-+ sb->s_blocksize);
-+ brelse(bh);
-+ }
-+#endif
-+
-+ }
-+
-+ return 0;
-+
-+err_freebuddy:
-+ while (i >= 0) {
-+ kfree(EXT3_GROUP_INFO(sb, i));
-+ i--;
-+ }
-+ i = num_meta_group_infos;
-+err_freemeta:
-+ while (--i >= 0)
-+ kfree(sbi->s_group_info[i]);
-+ iput(sbi->s_buddy_cache);
-+err_freesgi:
-+ kfree(sbi->s_group_info);
-+ return -ENOMEM;
-+}
-+
-+static void ext3_mb_prealloc_table_add(struct ext3_sb_info *sbi, int value)
-+{
-+ int i;
-+
-+ if (value > (sbi->s_blocks_per_group - 1 - 1 - sbi->s_itb_per_group))
-+ return;
-+
-+ for (i = 0; i < sbi->s_mb_prealloc_table_size; i++) {
-+ if (sbi->s_mb_prealloc_table[i] == 0) {
-+ sbi->s_mb_prealloc_table[i] = value;
-+ return;
-+ }
-+
-+ /* they should add values in order */
-+ if (value <= sbi->s_mb_prealloc_table[i])
-+ return;
-+ }
-+}
-+
-+int ext3_mb_init(struct super_block *sb, int needs_recovery)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ unsigned i, offset, max;
-+
-+ if (!test_opt(sb, MBALLOC))
-+ return 0;
-+
-+ i = (sb->s_blocksize_bits + 2) * sizeof(unsigned short);
-+
-+ sbi->s_mb_offsets = kmalloc(i, GFP_KERNEL);
-+ if (sbi->s_mb_offsets == NULL) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ return -ENOMEM;
-+ }
-+ sbi->s_mb_maxs = kmalloc(i, GFP_KERNEL);
-+ if (sbi->s_mb_maxs == NULL) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ kfree(sbi->s_mb_maxs);
-+ return -ENOMEM;
-+ }
-+
-+ /* order 0 is regular bitmap */
-+ sbi->s_mb_maxs[0] = sb->s_blocksize << 3;
-+ sbi->s_mb_offsets[0] = 0;
-+
-+ i = 1;
-+ offset = 0;
-+ max = sb->s_blocksize << 2;
-+ do {
-+ sbi->s_mb_offsets[i] = offset;
-+ sbi->s_mb_maxs[i] = max;
-+ offset += 1 << (sb->s_blocksize_bits - i);
-+ max = max >> 1;
-+ i++;
-+ } while (i <= sb->s_blocksize_bits + 1);
-+
-+ /* init file for buddy data */
-+ if ((i = ext3_mb_init_backend(sb))) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ kfree(sbi->s_mb_offsets);
-+ kfree(sbi->s_mb_maxs);
-+ return i;
-+ }
-+
-+ spin_lock_init(&sbi->s_md_lock);
-+ INIT_LIST_HEAD(&sbi->s_active_transaction);
-+ INIT_LIST_HEAD(&sbi->s_closed_transaction);
-+ INIT_LIST_HEAD(&sbi->s_committed_transaction);
-+ spin_lock_init(&sbi->s_bal_lock);
-+
-+ sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
-+ sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
-+ sbi->s_mb_max_groups_to_scan = MB_DEFAULT_MAX_GROUPS_TO_SCAN;
-+ sbi->s_mb_stats = MB_DEFAULT_STATS;
-+ sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS;
-+ sbi->s_mb_history_filter = EXT3_MB_HISTORY_DEFAULT;
-+
-+ if (sbi->s_stripe == 0) {
-+ sbi->s_mb_prealloc_table_size = 8;
-+ i = sbi->s_mb_prealloc_table_size * sizeof(unsigned long);
-+ sbi->s_mb_prealloc_table = kmalloc(i, GFP_NOFS);
-+ if (sbi->s_mb_prealloc_table == NULL) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ kfree(sbi->s_mb_offsets);
-+ kfree(sbi->s_mb_maxs);
-+ return -ENOMEM;
-+ }
-+ memset(sbi->s_mb_prealloc_table, 0, i);
-+
-+ ext3_mb_prealloc_table_add(sbi, 4);
-+ ext3_mb_prealloc_table_add(sbi, 8);
-+ ext3_mb_prealloc_table_add(sbi, 16);
-+ ext3_mb_prealloc_table_add(sbi, 32);
-+ ext3_mb_prealloc_table_add(sbi, 64);
-+ ext3_mb_prealloc_table_add(sbi, 128);
-+ ext3_mb_prealloc_table_add(sbi, 256);
-+ ext3_mb_prealloc_table_add(sbi, 512);
-+
-+ sbi->s_mb_small_req = 256;
-+ sbi->s_mb_large_req = 1024;
-+ sbi->s_mb_group_prealloc = 256;
-+ } else {
-+ sbi->s_mb_prealloc_table_size = 3;
-+ i = sbi->s_mb_prealloc_table_size * sizeof(unsigned long);
-+ sbi->s_mb_prealloc_table = kmalloc(i, GFP_NOFS);
-+ if (sbi->s_mb_prealloc_table == NULL) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ kfree(sbi->s_mb_offsets);
-+ kfree(sbi->s_mb_maxs);
-+ return -ENOMEM;
-+ }
-+ memset(sbi->s_mb_prealloc_table, 0, i);
-+
-+ ext3_mb_prealloc_table_add(sbi, sbi->s_stripe);
-+ ext3_mb_prealloc_table_add(sbi, sbi->s_stripe * 2);
-+ ext3_mb_prealloc_table_add(sbi, sbi->s_stripe * 4);
-+
-+ sbi->s_mb_small_req = sbi->s_stripe;
-+ sbi->s_mb_large_req = sbi->s_stripe * 8;
-+ sbi->s_mb_group_prealloc = sbi->s_stripe;
-+ }
-+
-+ i = sizeof(struct ext3_locality_group) * num_possible_cpus();
-+ sbi->s_locality_groups = kmalloc(i, GFP_NOFS);
-+ if (sbi->s_locality_groups == NULL) {
-+ clear_opt(sbi->s_mount_opt, MBALLOC);
-+ kfree(sbi->s_mb_prealloc_table);
-+ kfree(sbi->s_mb_offsets);
-+ kfree(sbi->s_mb_maxs);
-+ return -ENOMEM;
-+ }
-+ for (i = 0; i < num_possible_cpus(); i++) {
-+ struct ext3_locality_group *lg;
-+ lg = &sbi->s_locality_groups[i];
-+ sema_init(&lg->lg_sem, 1);
-+ INIT_LIST_HEAD(&lg->lg_prealloc_list);
-+ spin_lock_init(&lg->lg_prealloc_lock);
-+ }
-+
-+ ext3_mb_init_per_dev_proc(sb);
-+ ext3_mb_history_init(sb);
-+
-+ printk("EXT3-fs: mballoc enabled\n");
-+ return 0;
-+}
-+
-+void ext3_mb_cleanup_pa(struct ext3_group_info *grp)
-+{
-+ struct ext3_prealloc_space *pa;
-+ struct list_head *cur, *tmp;
-+ int count = 0;
-+
-+ list_for_each_safe(cur, tmp, &grp->bb_prealloc_list) {
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_group_list);
-+ list_del_rcu(&pa->pa_group_list);
-+ count++;
-+ kfree(pa);
-+ }
-+ if (count)
-+ mb_debug("mballoc: %u PAs left\n", count);
-+
-+}
-+
-+int ext3_mb_release(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ int i, num_meta_group_infos;
-+
-+ if (!test_opt(sb, MBALLOC))
-+ return 0;
-+
-+ /* release freed, non-committed blocks */
-+ spin_lock(&sbi->s_md_lock);
-+ list_splice_init(&sbi->s_closed_transaction,
-+ &sbi->s_committed_transaction);
-+ list_splice_init(&sbi->s_active_transaction,
-+ &sbi->s_committed_transaction);
-+ spin_unlock(&sbi->s_md_lock);
-+ ext3_mb_free_committed_blocks(sb);
-+
-+ if (sbi->s_group_info) {
-+ for (i = 0; i < sbi->s_groups_count; i++) {
-+#ifdef DOUBLE_CHECK
-+ if (EXT3_GROUP_INFO(sb, i)->bb_bitmap)
-+ kfree(EXT3_GROUP_INFO(sb, i)->bb_bitmap);
-+#endif
-+ ext3_mb_cleanup_pa(EXT3_GROUP_INFO(sb, i));
-+ kfree(EXT3_GROUP_INFO(sb, i));
-+ }
-+ num_meta_group_infos = (sbi->s_groups_count +
-+ EXT3_DESC_PER_BLOCK(sb) - 1) >>
-+ EXT3_DESC_PER_BLOCK_BITS(sb);
-+ for (i = 0; i < num_meta_group_infos; i++)
-+ kfree(sbi->s_group_info[i]);
-+ kfree(sbi->s_group_info);
-+ }
-+ if (sbi->s_mb_offsets)
-+ kfree(sbi->s_mb_offsets);
-+ if (sbi->s_mb_maxs)
-+ kfree(sbi->s_mb_maxs);
-+ if (sbi->s_buddy_cache)
-+ iput(sbi->s_buddy_cache);
-+ if (sbi->s_mb_stats) {
-+ printk("EXT3-fs: mballoc: %u blocks %u reqs (%u success)\n",
-+ atomic_read(&sbi->s_bal_allocated),
-+ atomic_read(&sbi->s_bal_reqs),
-+ atomic_read(&sbi->s_bal_success));
-+ printk("EXT3-fs: mballoc: %u extents scanned, %u goal hits, "
-+ "%u 2^N hits, %u breaks, %u lost\n",
-+ atomic_read(&sbi->s_bal_ex_scanned),
-+ atomic_read(&sbi->s_bal_goals),
-+ atomic_read(&sbi->s_bal_2orders),
-+ atomic_read(&sbi->s_bal_breaks),
-+ atomic_read(&sbi->s_mb_lost_chunks));
-+ printk("EXT3-fs: mballoc: %lu generated and it took %Lu\n",
-+ sbi->s_mb_buddies_generated++,
-+ sbi->s_mb_generation_time);
-+ printk("EXT3-fs: mballoc: %u preallocated, %u discarded\n",
-+ atomic_read(&sbi->s_mb_preallocated),
-+ atomic_read(&sbi->s_mb_discarded));
-+ }
-+
-+ if (sbi->s_locality_groups)
-+ kfree(sbi->s_locality_groups);
-+
-+ ext3_mb_history_release(sb);
-+ ext3_mb_destroy_per_dev_proc(sb);
-+
-+ return 0;
-+}
-+
-+void ext3_mb_free_committed_blocks(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ int err, i, count = 0, count2 = 0;
-+ struct ext3_free_metadata *md;
-+ struct ext3_buddy e3b;
-+
-+ if (list_empty(&sbi->s_committed_transaction))
-+ return;
-+
-+ /* there is committed blocks to be freed yet */
-+ do {
-+ /* get next array of blocks */
-+ md = NULL;
-+ spin_lock(&sbi->s_md_lock);
-+ if (!list_empty(&sbi->s_committed_transaction)) {
-+ md = list_entry(sbi->s_committed_transaction.next,
-+ struct ext3_free_metadata, list);
-+ list_del(&md->list);
-+ }
-+ spin_unlock(&sbi->s_md_lock);
-+
-+ if (md == NULL)
-+ break;
-+
-+ mb_debug("gonna free %u blocks in group %u (0x%p):",
-+ md->num, md->group, md);
-+
-+ err = ext3_mb_load_buddy(sb, md->group, &e3b);
-+ /* we expect to find existing buddy because it's pinned */
-+ BUG_ON(err != 0);
-+
-+ /* there are blocks to put in buddy to make them really free */
-+ count += md->num;
-+ count2++;
-+ ext3_lock_group(sb, md->group);
-+ for (i = 0; i < md->num; i++) {
-+ mb_debug(" %u", md->blocks[i]);
-+ err = mb_free_blocks(NULL, &e3b, md->blocks[i], 1);
-+ BUG_ON(err != 0);
-+ }
-+ mb_debug("\n");
-+ ext3_unlock_group(sb, md->group);
-+
-+ /* balance refcounts from ext3_mb_free_metadata() */
-+ page_cache_release(e3b.bd_buddy_page);
-+ page_cache_release(e3b.bd_bitmap_page);
-+
-+ kfree(md);
-+ ext3_mb_release_desc(&e3b);
-+
-+ } while (md);
-+
-+ mb_debug("freed %u blocks in %u structures\n", count, count2);
-+}
-+
-+#define EXT3_MB_STATS_NAME "stats"
-+#define EXT3_MB_MAX_TO_SCAN_NAME "max_to_scan"
-+#define EXT3_MB_MIN_TO_SCAN_NAME "min_to_scan"
-+#define EXT3_MB_ORDER2_REQ "order2_req"
-+#define EXT3_MB_SMALL_REQ "small_req"
-+#define EXT3_MB_LARGE_REQ "large_req"
-+#define EXT3_MB_PREALLOC_TABLE "prealloc_table"
-+#define EXT3_MB_GROUP_PREALLOC "group_prealloc"
-+
-+static int ext3_mb_read_prealloc_table(char *page, char **start, off_t off,
-+ int count, int *eof, void *data)
-+{
-+ struct ext3_sb_info *sbi = data;
-+ int len = 0;
-+ int i;
-+
-+ *eof = 1;
-+ if (off != 0)
-+ return 0;
-+
-+ for (i = 0; i < sbi->s_mb_prealloc_table_size; i++)
-+ len += sprintf(page + len, "%ld ",
-+ sbi->s_mb_prealloc_table[i]);
-+ len += sprintf(page + len, "\n");
-+
-+ *start = page;
-+ return len;
-+}
-+
-+static int ext3_mb_write_prealloc_table(struct file *file,
-+ const char __user *buf,
-+ unsigned long cnt, void *data)
-+{
-+ struct ext3_sb_info *sbi = data;
-+ unsigned long value;
-+ unsigned long prev = 0;
-+ char str[128];
-+ char *cur;
-+ char *end;
-+ unsigned long *new_table;
-+ int num = 0;
-+ int i = 0;
-+
-+ if (cnt >= sizeof(str))
-+ return -EINVAL;
-+ if (copy_from_user(str, buf, cnt))
-+ return -EFAULT;
-+
-+ num = 0;
-+ cur = str;
-+ end = str + cnt;
-+ while (cur < end) {
-+ while ((cur < end) && (*cur == ' ')) cur++;
-+ value = simple_strtol(cur, &cur, 0);
-+ if (value == 0)
-+ break;
-+ if (value <= prev)
-+ return -EINVAL;
-+ prev = value;
-+ num++;
-+ }
-+
-+ new_table = kmalloc(num * sizeof(*new_table), GFP_KERNEL);
-+ if (new_table == NULL)
-+ return -ENOMEM;
-+ kfree(sbi->s_mb_prealloc_table);
-+ memset(new_table, 0, num * sizeof(*new_table));
-+ sbi->s_mb_prealloc_table = new_table;
-+ sbi->s_mb_prealloc_table_size = num;
-+ cur = str;
-+ end = str + cnt;
-+ while (cur < end && i < num) {
-+ while ((cur < end) && (*cur == ' ')) cur++;
-+ value = simple_strtol(cur, &cur, 0);
-+ ext3_mb_prealloc_table_add(sbi, value);
-+ i++;
-+ }
-+
-+ return cnt;
-+}
-+
-+#define MB_PROC_VALUE_READ(name) \
-+static int ext3_mb_read_##name(char *page, char **start, \
-+ off_t off, int count, int *eof, void *data) \
-+{ \
-+ struct ext3_sb_info *sbi = data; \
-+ int len; \
-+ *eof = 1; \
-+ if (off != 0) \
-+ return 0; \
-+ len = sprintf(page, "%ld\n", sbi->s_mb_##name); \
-+ *start = page; \
-+ return len; \
-+}
-+
-+#define MB_PROC_VALUE_WRITE(name) \
-+static int ext3_mb_write_##name(struct file *file, \
-+ const char __user *buf, unsigned long cnt, void *data) \
-+{ \
-+ struct ext3_sb_info *sbi = data; \
-+ char str[32]; \
-+ long value; \
-+ if (cnt >= sizeof(str)) \
-+ return -EINVAL; \
-+ if (copy_from_user(str, buf, cnt)) \
-+ return -EFAULT; \
-+ value = simple_strtol(str, NULL, 0); \
-+ if (value <= 0) \
-+ return -ERANGE; \
-+ sbi->s_mb_##name = value; \
-+ return cnt; \
-+}
-+
-+MB_PROC_VALUE_READ(stats);
-+MB_PROC_VALUE_WRITE(stats);
-+MB_PROC_VALUE_READ(max_to_scan);
-+MB_PROC_VALUE_WRITE(max_to_scan);
-+MB_PROC_VALUE_READ(min_to_scan);
-+MB_PROC_VALUE_WRITE(min_to_scan);
-+MB_PROC_VALUE_READ(order2_reqs);
-+MB_PROC_VALUE_WRITE(order2_reqs);
-+MB_PROC_VALUE_READ(small_req);
-+MB_PROC_VALUE_WRITE(small_req);
-+MB_PROC_VALUE_READ(large_req);
-+MB_PROC_VALUE_WRITE(large_req);
-+MB_PROC_VALUE_READ(group_prealloc);
-+MB_PROC_VALUE_WRITE(group_prealloc);
-+
-+#define MB_PROC_HANDLER(name, var) \
-+do { \
-+ proc = create_proc_entry(name, mode, sbi->s_dev_proc); \
-+ if (proc == NULL) { \
-+ printk(KERN_ERR "EXT3-fs: can't to create %s\n", name); \
-+ goto err_out; \
-+ } \
-+ proc->data = sbi; \
-+ proc->read_proc = ext3_mb_read_##var ; \
-+ proc->write_proc = ext3_mb_write_##var; \
-+} while (0)
-+
-+int ext3_mb_init_per_dev_proc(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ mode_t mode = S_IFREG | S_IRUGO | S_IWUSR;
-+ struct proc_dir_entry *proc;
-+
-+ MB_PROC_HANDLER(EXT3_MB_STATS_NAME, stats);
-+ MB_PROC_HANDLER(EXT3_MB_MAX_TO_SCAN_NAME, max_to_scan);
-+ MB_PROC_HANDLER(EXT3_MB_MIN_TO_SCAN_NAME, min_to_scan);
-+ MB_PROC_HANDLER(EXT3_MB_ORDER2_REQ, order2_reqs);
-+ MB_PROC_HANDLER(EXT3_MB_SMALL_REQ, small_req);
-+ MB_PROC_HANDLER(EXT3_MB_LARGE_REQ, large_req);
-+ MB_PROC_HANDLER(EXT3_MB_PREALLOC_TABLE, prealloc_table);
-+ MB_PROC_HANDLER(EXT3_MB_GROUP_PREALLOC, group_prealloc);
-+
-+ return 0;
-+
-+err_out:
-+ remove_proc_entry(EXT3_MB_GROUP_PREALLOC, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_PREALLOC_TABLE, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_LARGE_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_SMALL_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_ORDER2_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_MIN_TO_SCAN_NAME, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_MAX_TO_SCAN_NAME, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_STATS_NAME, sbi->s_dev_proc);
-+
-+ return -ENOMEM;
-+}
-+
-+int ext3_mb_destroy_per_dev_proc(struct super_block *sb)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+
-+ if (sbi->s_dev_proc == NULL)
-+ return -EINVAL;
-+
-+ remove_proc_entry(EXT3_MB_GROUP_PREALLOC, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_PREALLOC_TABLE, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_SMALL_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_LARGE_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_ORDER2_REQ, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_MIN_TO_SCAN_NAME, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_MAX_TO_SCAN_NAME, sbi->s_dev_proc);
-+ remove_proc_entry(EXT3_MB_STATS_NAME, sbi->s_dev_proc);
-+
-+ return 0;
-+}
-+
-+int __init init_ext3_mb_proc(void)
-+{
-+ ext3_pspace_cachep =
-+ kmem_cache_create("ext3_prealloc_space",
-+ sizeof(struct ext3_prealloc_space),
-+ 0, SLAB_RECLAIM_ACCOUNT, NULL, NULL);
-+ if (ext3_pspace_cachep == NULL)
-+ return -ENOMEM;
-+
-+ return 0;
-+}
-+
-+void exit_ext3_mb_proc(void)
-+{
-+ /* XXX: synchronize_rcu(); */
-+ kmem_cache_destroy(ext3_pspace_cachep);
-+}
-+
-+
-+/*
-+ * Check quota and mark choosed space (ac->ac_b_ex) non-free in bitmaps
-+ * Returns 0 if success or error code
-+ */
-+int ext3_mb_mark_diskspace_used(struct ext3_allocation_context *ac, handle_t *handle)
-+{
-+ struct buffer_head *bitmap_bh = NULL;
-+ struct ext3_super_block *es;
-+ struct ext3_group_desc *gdp;
-+ struct buffer_head *gdp_bh;
-+ struct ext3_sb_info *sbi;
-+ struct super_block *sb;
-+ sector_t block;
-+ int err;
-+
-+ BUG_ON(ac->ac_status != AC_STATUS_FOUND);
-+ BUG_ON(ac->ac_b_ex.fe_len <= 0);
-+
-+ sb = ac->ac_sb;
-+ sbi = EXT3_SB(sb);
-+ es = sbi->s_es;
-+
-+ ext3_debug("using block group %d(%d)\n", ac->ac_b_group.group,
-+ gdp->bg_free_blocks_count);
-+
-+ err = -EIO;
-+ bitmap_bh = read_block_bitmap(sb, ac->ac_b_ex.fe_group);
-+ if (!bitmap_bh)
-+ goto out_err;
-+
-+ err = ext3_journal_get_write_access(handle, bitmap_bh);
-+ if (err)
-+ goto out_err;
-+
-+ err = -EIO;
-+ gdp = ext3_get_group_desc(sb, ac->ac_b_ex.fe_group, &gdp_bh);
-+ if (!gdp)
-+ goto out_err;
-+
-+ err = ext3_journal_get_write_access(handle, gdp_bh);
-+ if (err)
-+ goto out_err;
-+
-+ block = ac->ac_b_ex.fe_group * EXT3_BLOCKS_PER_GROUP(sb)
-+ + ac->ac_b_ex.fe_start
-+ + le32_to_cpu(es->s_first_data_block);
-+
-+ if (block == le32_to_cpu(gdp->bg_block_bitmap) ||
-+ block == le32_to_cpu(gdp->bg_inode_bitmap) ||
-+ in_range(block, le32_to_cpu(gdp->bg_inode_table),
-+ EXT3_SB(sb)->s_itb_per_group))
-+ ext3_error(sb, __FUNCTION__,
-+ "Allocating block in system zone - block = %lu",
-+ (unsigned long) block);
-+ ext3_lock_group(sb, ac->ac_b_ex.fe_group);
-+ spin_lock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
-+#ifdef AGGRESSIVE_CHECK
-+ {
-+ int i;
-+ for (i = 0; i < ac->ac_b_ex.fe_len; i++) {
-+ BUG_ON(mb_test_bit(ac->ac_b_ex.fe_start + i,
-+ bitmap_bh->b_data));
-+ }
-+ }
-+#endif
-+ mb_set_bits(NULL, bitmap_bh->b_data,
-+ ac->ac_b_ex.fe_start, ac->ac_b_ex.fe_len);
-+
-+ gdp->bg_free_blocks_count =
-+ cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count)
-+ - ac->ac_b_ex.fe_len);
-+ spin_unlock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
-+ percpu_counter_mod(&sbi->s_freeblocks_counter, - ac->ac_b_ex.fe_len);
-+ ext3_unlock_group(sb, ac->ac_b_ex.fe_group);
-+
-+ err = ext3_journal_dirty_metadata(handle, bitmap_bh);
-+ if (err)
-+ goto out_err;
-+ err = ext3_journal_dirty_metadata(handle, gdp_bh);
-+
-+out_err:
-+ sb->s_dirt = 1;
-+ brelse(bitmap_bh);
-+ return err;
-+}
-+
-+/*
-+ * here we normalize request for locality group
-+ * XXX: should we try to preallocate more than the group has now?
-+ */
-+void ext3_mb_normalize_group_request(struct ext3_allocation_context *ac)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ struct ext3_locality_group *lg = ac->ac_lg;
-+
-+ BUG_ON(lg == NULL);
-+ ac->ac_g_ex.fe_len = EXT3_SB(sb)->s_mb_group_prealloc;
-+
-+ mb_debug("#%u: goal %u blocks for locality group\n",
-+ current->pid, ac->ac_g_ex.fe_len);
-+}
-+
-+/*
-+ * Normalization means making request better in terms of
-+ * size and alignment
-+ */
-+void ext3_mb_normalize_request(struct ext3_allocation_context *ac,
-+ struct ext3_allocation_request *ar)
-+{
-+ struct ext3_inode_info *ei = EXT3_I(ac->ac_inode);
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ loff_t start, end, size, orig_size, orig_start;
-+ struct list_head *cur;
-+ int bsbits, i, wind;
-+
-+ /* do normalize only data requests, metadata requests
-+ do not need preallocation */
-+ if (!(ac->ac_flags & EXT3_MB_HINT_DATA))
-+ return;
-+
-+ /* sometime caller may want exact blocks */
-+ if (unlikely(ac->ac_flags & EXT3_MB_HINT_GOAL_ONLY))
-+ return;
-+
-+ /* caller may indicate that preallocation isn't
-+ * required (it's a tail, for example) */
-+ if (ac->ac_flags & EXT3_MB_HINT_NOPREALLOC)
-+ return;
-+
-+ if (ac->ac_flags & EXT3_MB_HINT_GROUP_ALLOC)
-+ return ext3_mb_normalize_group_request(ac);
-+
-+ bsbits = ac->ac_sb->s_blocksize_bits;
-+
-+ /* first, let's learn actual file size
-+ * given current request is allocated */
-+ size = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len;
-+ size = size << bsbits;
-+ if (size < i_size_read(ac->ac_inode))
-+ size = i_size_read(ac->ac_inode);
-+ size = (size + ac->ac_sb->s_blocksize - 1) >> bsbits;
-+
-+ start = 0;
-+ wind = 0;
-+
-+ /* let's choose preallocation window depending on file size */
-+ for (i = 0; i < sbi->s_mb_prealloc_table_size; i++) {
-+ if (size <= sbi->s_mb_prealloc_table[i]) {
-+ wind = sbi->s_mb_prealloc_table[i];
-+ break;
-+ }
-+ }
-+ size = wind;
-+
-+ if (wind == 0) {
-+ __u64 tstart, tend;
-+ /* file is quite large, we now preallocate with
-+ * the biggest configured window with regart to
-+ * logical offset */
-+ wind = sbi->s_mb_prealloc_table[i - 1];
-+ tstart = ac->ac_o_ex.fe_logical;
-+ do_div(tstart, wind);
-+ start = tstart * wind;
-+ tend = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len - 1;
-+ do_div(tend, wind);
-+ tend = tend * wind + wind;
-+ size = tend - start;
-+ }
-+ orig_size = size;
-+ orig_start = start;
-+
-+ /* don't cover already allocated blocks in selected range */
-+ if (ar->pleft && start <= ar->lleft) {
-+ size -= ar->lleft + 1 - start;
-+ start = ar->lleft + 1;
-+ }
-+ if (ar->pright && start + size - 1 >= ar->lright)
-+ size -= start + size - ar->lright;
-+
-+ end = start + size;
-+
-+ /* check we don't cross already preallocated blocks */
-+ rcu_read_lock();
-+ list_for_each_rcu(cur, &ei->i_prealloc_list) {
-+ struct ext3_prealloc_space *pa;
-+ unsigned long pa_end;
-+
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_inode_list);
-+
-+ if (pa->pa_deleted)
-+ continue;
-+ spin_lock(&pa->pa_lock);
-+ if (pa->pa_deleted) {
-+ spin_unlock(&pa->pa_lock);
-+ continue;
-+ }
-+
-+ pa_end = pa->pa_lstart + pa->pa_len;
-+
-+ /* PA must not overlap original request */
-+ BUG_ON(!(ac->ac_o_ex.fe_logical >= pa_end ||
-+ ac->ac_o_ex.fe_logical < pa->pa_lstart));
-+
-+ /* skip PA normalized request doesn't overlap with */
-+ if (pa->pa_lstart >= end) {
-+ spin_unlock(&pa->pa_lock);
-+ continue;
-+ }
-+ if (pa_end <= start) {
-+ spin_unlock(&pa->pa_lock);
-+ continue;
-+ }
-+ BUG_ON(pa->pa_lstart <= start && pa_end >= end);
-+
-+ if (pa_end <= ac->ac_o_ex.fe_logical) {
-+ BUG_ON(pa_end < start);
-+ start = pa_end;
-+ }
-+
-+ if (pa->pa_lstart > ac->ac_o_ex.fe_logical) {
-+ BUG_ON(pa->pa_lstart > end);
-+ end = pa->pa_lstart;
-+ }
-+ spin_unlock(&pa->pa_lock);
-+ }
-+ rcu_read_unlock();
-+ size = end - start;
-+
-+ /* XXX: extra loop to check we really don't overlap preallocations */
-+ rcu_read_lock();
-+ list_for_each_rcu(cur, &ei->i_prealloc_list) {
-+ struct ext3_prealloc_space *pa;
-+ unsigned long pa_end;
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_inode_list);
-+ spin_lock(&pa->pa_lock);
-+ if (pa->pa_deleted == 0) {
-+ pa_end = pa->pa_lstart + pa->pa_len;
-+ BUG_ON(!(start >= pa_end || end <= pa->pa_lstart));
-+ }
-+ spin_unlock(&pa->pa_lock);
-+ }
-+ rcu_read_unlock();
-+
-+ if (start + size <= ac->ac_o_ex.fe_logical &&
-+ start > ac->ac_o_ex.fe_logical) {
-+ printk("start %lu, size %lu, fe_logical %lu\n",
-+ (unsigned long) start, (unsigned long) size,
-+ (unsigned long) ac->ac_o_ex.fe_logical);
-+ }
-+ BUG_ON(start + size <= ac->ac_o_ex.fe_logical &&
-+ start > ac->ac_o_ex.fe_logical);
-+
-+ /* now prepare goal request */
-+
-+ /* XXX: is it better to align blocks WRT to logical
-+ * placement or satisfy big request as is */
-+ ac->ac_g_ex.fe_logical = start;
-+ ac->ac_g_ex.fe_len = size;
-+
-+ /* define goal start in order to merge */
-+ if (ar->pright && (ar->lright == (start + size))) {
-+ /* merge to the right */
-+ ext3_get_group_no_and_offset(ac->ac_sb, ar->pright - size,
-+ &ac->ac_f_ex.fe_group,
-+ &ac->ac_f_ex.fe_start);
-+ ac->ac_flags |= EXT3_MB_HINT_TRY_GOAL;
-+ }
-+ if (ar->pleft && (ar->lleft + 1 == start)) {
-+ /* merge to the left */
-+ ext3_get_group_no_and_offset(ac->ac_sb, ar->pleft + 1,
-+ &ac->ac_f_ex.fe_group,
-+ &ac->ac_f_ex.fe_start);
-+ ac->ac_flags |= EXT3_MB_HINT_TRY_GOAL;
-+ }
-+
-+ mb_debug("goal: %u(was %u) blocks at %u\n", (unsigned) size,
-+ (unsigned) orig_size, (unsigned) start);
-+}
-+
-+void ext3_mb_collect_stats(struct ext3_allocation_context *ac)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+
-+ if (sbi->s_mb_stats && ac->ac_g_ex.fe_len > 1) {
-+ atomic_inc(&sbi->s_bal_reqs);
-+ atomic_add(ac->ac_b_ex.fe_len, &sbi->s_bal_allocated);
-+ if (ac->ac_o_ex.fe_len >= ac->ac_g_ex.fe_len)
-+ atomic_inc(&sbi->s_bal_success);
-+ atomic_add(ac->ac_found, &sbi->s_bal_ex_scanned);
-+ if (ac->ac_g_ex.fe_start == ac->ac_b_ex.fe_start &&
-+ ac->ac_g_ex.fe_group == ac->ac_b_ex.fe_group)
-+ atomic_inc(&sbi->s_bal_goals);
-+ if (ac->ac_found > sbi->s_mb_max_to_scan)
-+ atomic_inc(&sbi->s_bal_breaks);
-+ }
-+
-+ ext3_mb_store_history(ac);
-+}
-+
-+/*
-+ * use blocks preallocated to inode
-+ */
-+void ext3_mb_use_inode_pa(struct ext3_allocation_context *ac,
-+ struct ext3_prealloc_space *pa)
-+{
-+ unsigned long start, len;
-+
-+ /* found preallocated blocks, use them */
-+ start = pa->pa_pstart + (ac->ac_o_ex.fe_logical - pa->pa_lstart);
-+ len = min(pa->pa_pstart + pa->pa_len, start + ac->ac_o_ex.fe_len);
-+ len = len - start;
-+ ext3_get_group_no_and_offset(ac->ac_sb, start, &ac->ac_b_ex.fe_group,
-+ &ac->ac_b_ex.fe_start);
-+ ac->ac_b_ex.fe_len = len;
-+ ac->ac_status = AC_STATUS_FOUND;
-+ ac->ac_pa = pa;
-+
-+ BUG_ON(start < pa->pa_pstart);
-+ BUG_ON(start + len > pa->pa_pstart + pa->pa_len);
-+ BUG_ON(pa->pa_free < len);
-+ pa->pa_free -= len;
-+
-+ mb_debug("use %lu/%lu from inode pa %p\n", start, len, pa);
-+}
-+
-+/*
-+ * use blocks preallocated to locality group
-+ */
-+void ext3_mb_use_group_pa(struct ext3_allocation_context *ac,
-+ struct ext3_prealloc_space *pa)
-+{
-+ unsigned len = ac->ac_o_ex.fe_len;
-+
-+ ext3_get_group_no_and_offset(ac->ac_sb, pa->pa_pstart,
-+ &ac->ac_b_ex.fe_group,
-+ &ac->ac_b_ex.fe_start);
-+ ac->ac_b_ex.fe_len = len;
-+ ac->ac_status = AC_STATUS_FOUND;
-+ ac->ac_pa = pa;
-+
-+ /* we don't correct pa_pstart or pa_plen here to avoid
-+ * possible race when tte group is being loaded concurrently
-+ * instead we correct pa later, after blocks are marked
-+ * in on-disk bitmap -- see ext3_mb_release_context() */
-+ mb_debug("use %lu/%lu from group pa %p\n", pa->pa_lstart-len, len, pa);
-+}
-+
-+/*
-+ * search goal blocks in preallocated space
-+ */
-+int ext3_mb_use_preallocated(struct ext3_allocation_context *ac)
-+{
-+ struct ext3_inode_info *ei = EXT3_I(ac->ac_inode);
-+ struct ext3_locality_group *lg;
-+ struct ext3_prealloc_space *pa;
-+ struct list_head *cur;
-+
-+ /* only data can be preallocated */
-+ if (!(ac->ac_flags & EXT3_MB_HINT_DATA))
-+ return 0;
-+
-+ /* first, try per-file preallocation */
-+ rcu_read_lock();
-+ list_for_each_rcu(cur, &ei->i_prealloc_list) {
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_inode_list);
-+
-+ /* all fields in this condition don't change,
-+ * so we can skip locking for them */
-+ if (ac->ac_o_ex.fe_logical < pa->pa_lstart ||
-+ ac->ac_o_ex.fe_logical >= pa->pa_lstart + pa->pa_len)
-+ continue;
-+
-+ /* found preallocated blocks, use them */
-+ spin_lock(&pa->pa_lock);
-+ if (pa->pa_deleted == 0 && pa->pa_free) {
-+ atomic_inc(&pa->pa_count);
-+ ext3_mb_use_inode_pa(ac, pa);
-+ spin_unlock(&pa->pa_lock);
-+ ac->ac_criteria = 10;
-+ rcu_read_unlock();
-+ return 1;
-+ }
-+ spin_unlock(&pa->pa_lock);
-+ }
-+ rcu_read_unlock();
-+
-+ /* can we use group allocation? */
-+ if (!(ac->ac_flags & EXT3_MB_HINT_GROUP_ALLOC))
-+ return 0;
-+
-+ /* inode may have no locality group for some reason */
-+ lg = ac->ac_lg;
-+ if (lg == NULL)
-+ return 0;
-+
-+ rcu_read_lock();
-+ list_for_each_rcu(cur, &lg->lg_prealloc_list) {
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_inode_list);
-+ spin_lock(&pa->pa_lock);
-+ if (pa->pa_deleted == 0 && pa->pa_free >= ac->ac_o_ex.fe_len) {
-+ atomic_inc(&pa->pa_count);
-+ ext3_mb_use_group_pa(ac, pa);
-+ spin_unlock(&pa->pa_lock);
-+ ac->ac_criteria = 20;
-+ rcu_read_unlock();
-+ return 1;
-+ }
-+ spin_unlock(&pa->pa_lock);
-+ }
-+ rcu_read_unlock();
-+
-+ return 0;
-+}
-+
-+/*
-+ * check free blocks in bitmap match free block in group descriptor
-+ * do this before taking preallocated blocks into account to be able
-+ * to detect on-disk corruptions
-+ */
-+int ext3_mb_check_ondisk_bitmap(struct super_block *sb, void *bitmap,
-+ struct ext3_group_desc *gdp, int group)
-+{
-+ unsigned short max = EXT3_BLOCKS_PER_GROUP(sb);
-+ unsigned short i, first, free = 0;
-+
-+ spin_lock(sb_bgl_lock(EXT3_SB(sb), group));
-+ i = mb_find_next_zero_bit(bitmap, max, 0);
-+
-+ while (i < max) {
-+ first = i;
-+ i = ext2_find_next_le_bit(bitmap, max, i);
-+ if (i > max)
-+ i = max;
-+ free += i - first;
-+ if (i < max)
-+ i = mb_find_next_zero_bit(bitmap, max, i);
-+ }
-+
-+ if (free != le16_to_cpu(gdp->bg_free_blocks_count)) {
-+ spin_unlock(sb_bgl_lock(EXT3_SB(sb), group));
-+ ext3_error(sb, __FUNCTION__, "on-disk bitmap for group %d"
-+ "corrupted: %u blocks free in bitmap, %u - in gd\n",
-+ group, free, le16_to_cpu(gdp->bg_free_blocks_count));
-+ return -EIO;
-+ }
-+ spin_unlock(sb_bgl_lock(EXT3_SB(sb), group));
-+ return 0;
-+}
-+
-+/*
-+ * the function goes through all preallocation in this group and marks them
-+ * used in in-core bitmap. buddy must be generated from this bitmap
-+ */
-+int ext3_mb_generate_from_pa(struct super_block *sb, void *bitmap, int group)
-+{
-+ struct ext3_group_info *grp = EXT3_GROUP_INFO(sb, group);
-+ struct ext3_prealloc_space *pa;
-+ struct ext3_group_desc *gdp;
-+ struct list_head *cur;
-+ unsigned long groupnr;
-+ unsigned long start;
-+ int preallocated = 0, count = 0, len, skip = 0, err;
-+
-+ gdp = ext3_get_group_desc (sb, group, NULL);
-+ if (gdp == NULL)
-+ return -EIO;
-+
-+ /* before applying preallocations, check bitmap consistency */
-+ err = ext3_mb_check_ondisk_bitmap(sb, bitmap, gdp, group);
-+ if (err)
-+ return err;
-+
-+ /* all form of preallocation discards first load group,
-+ * so the only competing code is preallocation use.
-+ * we don't need any locking here
-+ * notice we do NOT ignore preallocations with pa_deleted
-+ * otherwise we could leave used blocks available for
-+ * allocation in buddy when concurrent ext3_mb_put_pa()
-+ * is dropping preallocation
-+ */
-+ list_for_each_rcu(cur, &grp->bb_prealloc_list) {
-+ pa = list_entry(cur, struct ext3_prealloc_space, pa_group_list);
-+ spin_lock(&pa->pa_lock);
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart, &groupnr, &start);
-+ len = pa->pa_len;
-+ spin_unlock(&pa->pa_lock);
-+ if (unlikely(len == 0)) {
-+ skip++;
-+ continue;
-+ }
-+ BUG_ON(groupnr != group && len != 0);
-+ mb_set_bits(sb_bgl_lock(EXT3_SB(sb), group), bitmap, start,len);
-+ preallocated += len;
-+ count++;
-+ }
-+ if (count + skip != grp->bb_prealloc_nr) {
-+ ext3_error(sb, __FUNCTION__, "lost preallocations: "
-+ "count %d, bb_prealloc_nr %lu, skip %d\n",
-+ count, grp->bb_prealloc_nr, skip);
-+ return -EIO;
-+ }
-+ mb_debug("prellocated %u for group %u\n", preallocated, group);
-+ return 0;
-+}
-+
-+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,5)
-+static void ext3_mb_pa_callback(struct rcu_head *head)
-+{
-+ struct ext3_prealloc_space *pa;
-+ pa = container_of(head, struct ext3_prealloc_space, u.pa_rcu);
-+ kmem_cache_free(ext3_pspace_cachep, pa);
-+}
-+#define mb_call_rcu(__pa) call_rcu(&(__pa)->u.pa_rcu, ext3_mb_pa_callback)
-+#else
-+static void ext3_mb_pa_callback(void *pa)
-+{
-+ kmem_cache_free(ext3_pspace_cachep, pa);
-+}
-+#define mb_call_rcu(__pa) call_rcu(&(__pa)->u.pa_rcu, ext3_mb_pa_callback, pa)
-+#endif
-+
-+/*
-+ * drops a reference to preallocated space descriptor
-+ * if this was the last reference and the space is consumed
-+ */
-+void ext3_mb_put_pa(struct ext3_allocation_context *ac,
-+ struct super_block *sb, struct ext3_prealloc_space *pa)
-+{
-+ unsigned long grp;
-+
-+ if (!atomic_dec_and_test(&pa->pa_count) || pa->pa_free != 0)
-+ return;
-+
-+ /* in this short window concurrent discard can set pa_deleted */
-+ spin_lock(&pa->pa_lock);
-+ if (pa->pa_deleted == 1) {
-+ spin_unlock(&pa->pa_lock);
-+ return;
-+ }
-+
-+ pa->pa_deleted = 1;
-+ spin_unlock(&pa->pa_lock);
-+
-+ /* -1 is to protect from crossing allocation group */
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart - 1, &grp, NULL);
-+
-+ /*
-+ * possible race:
-+ *
-+ * P1 (buddy init) P2 (regular allocation)
-+ * find block B in PA
-+ * copy on-disk bitmap to buddy
-+ * mark B in on-disk bitmap
-+ * drop PA from group
-+ * mark all PAs in buddy
-+ *
-+ * thus, P1 initializes buddy with B available. to prevent this
-+ * we make "copy" and "mark all PAs" atomic and serialize "drop PA"
-+ * against that pair
-+ */
-+ ext3_lock_group(sb, grp);
-+ list_del_rcu(&pa->pa_group_list);
-+ EXT3_GROUP_INFO(sb, grp)->bb_prealloc_nr--;
-+ ext3_unlock_group(sb, grp);
-+
-+ spin_lock(pa->pa_obj_lock);
-+ list_del_rcu(&pa->pa_inode_list);
-+ spin_unlock(pa->pa_obj_lock);
-+
-+ mb_call_rcu(pa);
-+}
-+
-+/*
-+ * creates new preallocated space for given inode
-+ */
-+int ext3_mb_new_inode_pa(struct ext3_allocation_context *ac)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ struct ext3_prealloc_space *pa;
-+ struct ext3_group_info *grp;
-+ struct ext3_inode_info *ei;
-+
-+ /* preallocate only when found space is larger then requested */
-+ BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len);
-+ BUG_ON(ac->ac_status != AC_STATUS_FOUND);
-+ BUG_ON(!S_ISREG(ac->ac_inode->i_mode));
-+
-+ pa = kmem_cache_alloc(ext3_pspace_cachep, GFP_NOFS);
-+ if (pa == NULL)
-+ return -ENOMEM;
-+
-+ if (ac->ac_b_ex.fe_len < ac->ac_g_ex.fe_len) {
-+ int winl, wins, win, offs;
-+
-+ /* we can't allocate as much as normalizer wants.
-+ * so, found space must get proper lstart
-+ * to cover original request */
-+ BUG_ON(ac->ac_g_ex.fe_logical > ac->ac_o_ex.fe_logical);
-+ BUG_ON(ac->ac_g_ex.fe_len < ac->ac_o_ex.fe_len);
-+
-+ /* we're limited by original request in that
-+ * logical block must be covered any way
-+ * winl is window we can move our chunk within */
-+ winl = ac->ac_o_ex.fe_logical - ac->ac_g_ex.fe_logical;
-+
-+ /* also, we should cover whole original request */
-+ wins = ac->ac_b_ex.fe_len - ac->ac_o_ex.fe_len;
-+
-+ /* the smallest one defines real window */
-+ win = min(winl, wins);
-+
-+ offs = ac->ac_o_ex.fe_logical % ac->ac_b_ex.fe_len;
-+ if (offs && offs < win)
-+ win = offs;
-+
-+ ac->ac_b_ex.fe_logical = ac->ac_o_ex.fe_logical - win;
-+ BUG_ON(ac->ac_o_ex.fe_logical < ac->ac_b_ex.fe_logical);
-+ BUG_ON(ac->ac_o_ex.fe_len > ac->ac_b_ex.fe_len);
-+ }
-+
-+ /* preallocation can change ac_b_ex, thus we store actually
-+ * allocated blocks for history */
-+ ac->ac_f_ex = ac->ac_b_ex;
-+
-+ pa->pa_lstart = ac->ac_b_ex.fe_logical;
-+ pa->pa_pstart = ext3_grp_offs_to_block(sb, &ac->ac_b_ex);
-+ pa->pa_len = ac->ac_b_ex.fe_len;
-+ pa->pa_free = pa->pa_len;
-+ atomic_set(&pa->pa_count, 1);
-+ spin_lock_init(&pa->pa_lock);
-+ pa->pa_deleted = 0;
-+ pa->pa_linear = 0;
-+
-+ mb_debug("new inode pa %p: %lu/%lu for %lu\n", pa,
-+ pa->pa_pstart, pa->pa_len, pa->pa_lstart);
-+
-+ ext3_mb_use_inode_pa(ac, pa);
-+ atomic_add(pa->pa_free, &EXT3_SB(sb)->s_mb_preallocated);
-+
-+ ei = EXT3_I(ac->ac_inode);
-+ grp = EXT3_GROUP_INFO(sb, ac->ac_b_ex.fe_group);
-+
-+ pa->pa_obj_lock = &ei->i_prealloc_lock;
-+ pa->pa_inode = ac->ac_inode;
-+
-+ ext3_lock_group(sb, ac->ac_b_ex.fe_group);
-+ list_add_rcu(&pa->pa_group_list, &grp->bb_prealloc_list);
-+ grp->bb_prealloc_nr++;
-+ ext3_unlock_group(sb, ac->ac_b_ex.fe_group);
-+
-+ spin_lock(pa->pa_obj_lock);
-+ list_add_rcu(&pa->pa_inode_list, &ei->i_prealloc_list);
-+ spin_unlock(pa->pa_obj_lock);
-+
-+ return 0;
-+}
-+
-+/*
-+ * creates new preallocated space for locality group inodes belongs to
-+ */
-+int ext3_mb_new_group_pa(struct ext3_allocation_context *ac)
-+{
-+ struct super_block *sb = ac->ac_sb;
-+ struct ext3_locality_group *lg;
-+ struct ext3_prealloc_space *pa;
-+ struct ext3_group_info *grp;
-+
-+ /* preallocate only when found space is larger then requested */
-+ BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len);
-+ BUG_ON(ac->ac_status != AC_STATUS_FOUND);
-+ BUG_ON(!S_ISREG(ac->ac_inode->i_mode));
-+
-+ BUG_ON(ext3_pspace_cachep == NULL);
-+ pa = kmem_cache_alloc(ext3_pspace_cachep, GFP_NOFS);
-+ if (pa == NULL)
-+ return -ENOMEM;
-+
-+ /* preallocation can change ac_b_ex, thus we store actually
-+ * allocated blocks for history */
-+ ac->ac_f_ex = ac->ac_b_ex;
-+
-+ pa->pa_pstart = ext3_grp_offs_to_block(sb, &ac->ac_b_ex);
-+ pa->pa_lstart = pa->pa_pstart;
-+ pa->pa_len = ac->ac_b_ex.fe_len;
-+ pa->pa_free = pa->pa_len;
-+ atomic_set(&pa->pa_count, 1);
-+ spin_lock_init(&pa->pa_lock);
-+ pa->pa_deleted = 0;
-+ pa->pa_linear = 1;
-+
-+ mb_debug("new group pa %p: %lu/%lu for %lu\n", pa,
-+ pa->pa_pstart, pa->pa_len, pa->pa_lstart);
-+
-+ ext3_mb_use_group_pa(ac, pa);
-+ atomic_add(pa->pa_free, &EXT3_SB(sb)->s_mb_preallocated);
-+
-+ grp = EXT3_GROUP_INFO(sb, ac->ac_b_ex.fe_group);
-+ lg = ac->ac_lg;
-+ BUG_ON(lg == NULL);
-+
-+ pa->pa_obj_lock = &lg->lg_prealloc_lock;
-+ pa->pa_inode = NULL;
-+
-+ ext3_lock_group(sb, ac->ac_b_ex.fe_group);
-+ list_add_rcu(&pa->pa_group_list, &grp->bb_prealloc_list);
-+ grp->bb_prealloc_nr++;
-+ ext3_unlock_group(sb, ac->ac_b_ex.fe_group);
-+
-+ spin_lock(pa->pa_obj_lock);
-+ list_add_tail_rcu(&pa->pa_inode_list, &lg->lg_prealloc_list);
-+ spin_unlock(pa->pa_obj_lock);
-+
-+ return 0;
-+}
-+
-+int ext3_mb_new_preallocation(struct ext3_allocation_context *ac)
-+{
-+ int err;
-+
-+ if (ac->ac_flags & EXT3_MB_HINT_GROUP_ALLOC)
-+ err = ext3_mb_new_group_pa(ac);
-+ else
-+ err = ext3_mb_new_inode_pa(ac);
-+ return err;
-+}
-+
-+/*
-+ * finds all unused blocks in on-disk bitmap, frees them in
-+ * in-core bitmap and buddy.
-+ * @pa must be unlinked from inode and group lists, so that
-+ * nobody else can find/use it.
-+ * the caller MUST hold group/inode locks.
-+ * TODO: optimize the case when there are no in-core structures yet
-+ */
-+int ext3_mb_release_inode_pa(struct ext3_buddy *e3b,
-+ struct buffer_head *bitmap_bh,
-+ struct ext3_prealloc_space *pa)
-+{
-+ struct ext3_allocation_context ac;
-+ struct super_block *sb = e3b->bd_sb;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ unsigned long bit, end, next, group;
-+ sector_t start;
-+ int err = 0, free = 0;
-+
-+ BUG_ON(pa->pa_deleted == 0);
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit);
-+ BUG_ON(group != e3b->bd_group && pa->pa_len != 0);
-+ end = bit + pa->pa_len;
-+
-+ ac.ac_sb = sb;
-+ ac.ac_inode = pa->pa_inode;
-+ ac.ac_op = EXT3_MB_HISTORY_DISCARD;
-+ ac.ac_o_ex.fe_len = 1;
-+
-+ while (bit < end) {
-+ bit = mb_find_next_zero_bit(bitmap_bh->b_data, end, bit);
-+ if (bit >= end)
-+ break;
-+ next = mb_find_next_bit(bitmap_bh->b_data, end, bit);
-+ if (next > end)
-+ next = end;
-+ start = group * EXT3_BLOCKS_PER_GROUP(sb) + bit +
-+ le32_to_cpu(sbi->s_es->s_first_data_block);
-+ mb_debug(" free preallocated %u/%u in group %u\n",
-+ (unsigned) start, (unsigned) next - bit,
-+ (unsigned) group);
-+ free += next - bit;
-+
-+ ac.ac_b_ex.fe_group = group;
-+ ac.ac_b_ex.fe_start = bit;
-+ ac.ac_b_ex.fe_len = next - bit;
-+ ac.ac_b_ex.fe_logical = 0;
-+ ext3_mb_store_history(&ac);
-+
-+ mb_free_blocks(pa->pa_inode, e3b, bit, next - bit);
-+ bit = next + 1;
-+ }
-+ if (free != pa->pa_free) {
-+ printk("pa %p: logic %lu, phys. %lu, len %lu\n",
-+ pa, (unsigned long) pa->pa_lstart,
-+ (unsigned long) pa->pa_pstart,
-+ (unsigned long) pa->pa_len);
-+ printk("free %u, pa_free %u\n", free, pa->pa_free);
-+ }
-+ BUG_ON(free != pa->pa_free);
-+ atomic_add(free, &sbi->s_mb_discarded);
-+
-+ return err;
-+}
-+
-+int ext3_mb_release_group_pa(struct ext3_buddy *e3b,
-+ struct ext3_prealloc_space *pa)
-+{
-+ struct ext3_allocation_context ac;
-+ struct super_block *sb = e3b->bd_sb;
-+ unsigned long bit, group;
-+
-+ ac.ac_op = EXT3_MB_HISTORY_DISCARD;
-+
-+ BUG_ON(pa->pa_deleted == 0);
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit);
-+ BUG_ON(group != e3b->bd_group && pa->pa_len != 0);
-+ mb_free_blocks(pa->pa_inode, e3b, bit, pa->pa_len);
-+ atomic_add(pa->pa_len, &EXT3_SB(sb)->s_mb_discarded);
-+
-+ ac.ac_sb = sb;
-+ ac.ac_inode = NULL;
-+ ac.ac_b_ex.fe_group = group;
-+ ac.ac_b_ex.fe_start = bit;
-+ ac.ac_b_ex.fe_len = pa->pa_len;
-+ ac.ac_b_ex.fe_logical = 0;
-+ ext3_mb_store_history(&ac);
-+
-+ return 0;
-+}
-+
-+/*
-+ * releases all preallocations in given group
-+ *
-+ * first, we need to decide discard policy:
-+ * - when do we discard
-+ * 1) ENOSPC
-+ * - how many do we discard
-+ * 1) how many requested
-+ */
-+int ext3_mb_discard_group_preallocations(struct super_block *sb,
-+ int group, int needed)
-+{
-+ struct ext3_group_info *grp = EXT3_GROUP_INFO(sb, group);
-+ struct buffer_head *bitmap_bh = NULL;
-+ struct ext3_prealloc_space *pa, *tmp;
-+ struct list_head list;
-+ struct ext3_buddy e3b;
-+ int err, busy, free = 0;
-+
-+ mb_debug("discard preallocation for group %lu\n", group);
-+
-+ if (list_empty(&grp->bb_prealloc_list))
-+ return 0;
-+
-+ bitmap_bh = read_block_bitmap(sb, group);
-+ if (bitmap_bh == NULL) {
-+ /* error handling here */
-+ ext3_mb_release_desc(&e3b);
-+ BUG_ON(bitmap_bh == NULL);
-+ }
-+
-+ err = ext3_mb_load_buddy(sb, group, &e3b);
-+ if (err) {
-+ brelse(bitmap_bh);
-+ return err;
-+ }
-+
-+ if (needed == 0)
-+ needed = EXT3_BLOCKS_PER_GROUP(sb) + 1;
-+
-+ grp = EXT3_GROUP_INFO(sb, group);
-+ INIT_LIST_HEAD(&list);
-+
-+repeat:
-+ busy = 0;
-+ ext3_lock_group(sb, group);
-+ list_for_each_entry_safe (pa, tmp, &grp->bb_prealloc_list, pa_group_list) {
-+ spin_lock(&pa->pa_lock);
-+ if (atomic_read(&pa->pa_count)) {
-+ spin_unlock(&pa->pa_lock);
-+ busy = 1;
-+ continue;
-+ }
-+ if (pa->pa_deleted) {
-+ spin_unlock(&pa->pa_lock);
-+ continue;
-+ }
-+
-+ /* seems this one can be freed ... */
-+ pa->pa_deleted = 1;
-+
-+ /* we can trust pa_free ... */
-+ free += pa->pa_free;
-+
-+ spin_unlock(&pa->pa_lock);
-+
-+ BUG_ON(grp->bb_prealloc_nr == 0);
-+ grp->bb_prealloc_nr--;
-+ list_del_rcu(&pa->pa_group_list);
-+ list_add(&pa->u.pa_tmp_list, &list);
-+ }
-+
-+ /* if we still need more blocks and some PAs were used, try again */
-+ if (free < needed && busy) {
-+ ext3_unlock_group(sb, group);
-+ goto repeat;
-+ }
-+
-+ /* found anything to free? */
-+ if (list_empty(&list)) {
-+ BUG_ON(free != 0);
-+ goto out;
-+ }
-+
-+ /* now free all selected PAs */
-+ list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) {
-+
-+ /* remove from object (inode or locality group) */
-+ spin_lock(pa->pa_obj_lock);
-+ list_del_rcu(&pa->pa_inode_list);
-+ spin_unlock(pa->pa_obj_lock);
-+
-+ if (pa->pa_linear)
-+ ext3_mb_release_group_pa(&e3b, pa);
-+ else
-+ ext3_mb_release_inode_pa(&e3b, bitmap_bh, pa);
-+
-+ list_del(&pa->u.pa_tmp_list);
-+ mb_call_rcu(pa);
-+ }
-+
-+out:
-+ ext3_unlock_group(sb, group);
-+ ext3_mb_release_desc(&e3b);
-+ brelse(bitmap_bh);
-+ return free;
-+}
-+
-+/*
-+ * releases all non-used preallocated blocks for given inode
-+ */
-+void ext3_mb_discard_inode_preallocations(struct inode *inode)
-+{
-+ struct ext3_inode_info *ei = EXT3_I(inode);
-+ struct super_block *sb = inode->i_sb;
-+ struct buffer_head *bitmap_bh = NULL;
-+ struct ext3_prealloc_space *pa, *tmp;
-+ unsigned long group = 0;
-+ struct list_head list;
-+ struct ext3_buddy e3b;
-+ int err;
-+
-+ if (!test_opt(sb, MBALLOC) || !S_ISREG(inode->i_mode)) {
-+ /*BUG_ON(!list_empty(&ei->i_prealloc_list));*/
-+ return;
-+ }
-+
-+ mb_debug("discard preallocation for inode %lu\n", inode->i_ino);
-+
-+ INIT_LIST_HEAD(&list);
-+
-+repeat:
-+ /* first, collect all pa's in the inode */
-+ spin_lock(&ei->i_prealloc_lock);
-+ while (!list_empty(&ei->i_prealloc_list)) {
-+ pa = list_entry(ei->i_prealloc_list.next,
-+ struct ext3_prealloc_space, pa_inode_list);
-+ BUG_ON(pa->pa_obj_lock != &ei->i_prealloc_lock);
-+ spin_lock(&pa->pa_lock);
-+ if (atomic_read(&pa->pa_count)) {
-+ /* this shouldn't happen often - nobody should
-+ * use preallocation while we're discarding it */
-+ spin_unlock(&pa->pa_lock);
-+ spin_unlock(&ei->i_prealloc_lock);
-+ current->state = TASK_UNINTERRUPTIBLE;
-+ schedule_timeout(HZ);
-+ goto repeat;
-+
-+ }
-+ if (pa->pa_deleted == 0) {
-+ pa->pa_deleted = 1;
-+ spin_unlock(&pa->pa_lock);
-+ list_del_rcu(&pa->pa_inode_list);
-+ list_add(&pa->u.pa_tmp_list, &list);
-+ continue;
-+ }
-+
-+ /* someone is deleting pa right now */
-+ spin_unlock(&pa->pa_lock);
-+ spin_unlock(&ei->i_prealloc_lock);
-+
-+ /* we have to wait here because pa_deleted
-+ * doesn't mean pa is already unlinked from
-+ * the list. as we might be called from
-+ * ->clear_inode() the inode will get freed
-+ * and concurrent thread which is unlinking
-+ * pa from inode's list may access already
-+ * freed memory, bad-bad-bad */
-+
-+ /* XXX: if this happens too often, we can
-+ * add a flag to force wait only in case
-+ * of ->clear_inode(), but not in case of
-+ * regular truncate */
-+ current->state = TASK_UNINTERRUPTIBLE;
-+ schedule_timeout(HZ);
-+ goto repeat;
-+ }
-+ spin_unlock(&ei->i_prealloc_lock);
-+
-+ list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) {
-+ BUG_ON(pa->pa_linear != 0);
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart, &group, NULL);
-+
-+ err = ext3_mb_load_buddy(sb, group, &e3b);
-+ if (err)
-+ return;
-+
-+ bitmap_bh = read_block_bitmap(sb, group);
-+
-+ ext3_lock_group(sb, group);
-+ BUG_ON(e3b.bd_info->bb_prealloc_nr == 0);
-+ e3b.bd_info->bb_prealloc_nr--;
-+ list_del_rcu(&pa->pa_group_list);
-+
-+ /* can be NULL due to IO error, at worst
-+ * we leave some free blocks unavailable
-+ * do not go RO - no need for */
-+ if (bitmap_bh != NULL)
-+ ext3_mb_release_inode_pa(&e3b, bitmap_bh, pa);
-+ ext3_unlock_group(sb, group);
-+
-+ ext3_mb_release_desc(&e3b);
-+ brelse(bitmap_bh);
-+
-+ list_del(&pa->u.pa_tmp_list);
-+ mb_call_rcu(pa);
-+ }
-+}
-+
-+/*
-+ * finds all preallocated spaces and return blocks being freed to them
-+ * if preallocated space becomes full (no block is used from the space)
-+ * then the function frees space in buddy
-+ * XXX: at the moment, truncate (which is the only way to free blocks)
-+ * discards all preallocations
-+ */
-+void ext3_mb_return_to_preallocation(struct inode *inode, struct ext3_buddy *e3b,
-+ sector_t block, int count)
-+{
-+ BUG_ON(!list_empty(&EXT3_I(inode)->i_prealloc_list));
-+}
-+
-+void ext3_mb_show_ac(struct ext3_allocation_context *ac)
-+{
-+#if 0
-+ struct super_block *sb = ac->ac_sb;
-+ int i;
-+
-+ printk(KERN_ERR "EXT3-fs: can't allocate: status %d flags %d\n",
-+ ac->ac_status, ac->ac_flags);
-+ printk(KERN_ERR "EXT3-fs: orig %lu/%lu/%lu@%lu, goal %lu/%lu/%lu@%lu, "
-+ "best %lu/%lu/%lu@%lu cr %d\n",
-+ ac->ac_o_ex.fe_group, ac->ac_o_ex.fe_start,
-+ ac->ac_o_ex.fe_len, ac->ac_o_ex.fe_logical,
-+ ac->ac_g_ex.fe_group, ac->ac_g_ex.fe_start,
-+ ac->ac_g_ex.fe_len, ac->ac_g_ex.fe_logical,
-+ ac->ac_b_ex.fe_group, ac->ac_b_ex.fe_start,
-+ ac->ac_b_ex.fe_len, ac->ac_b_ex.fe_logical,
-+ ac->ac_criteria);
-+ printk(KERN_ERR "EXT3-fs: %lu scanned, %d found\n", ac->ac_ex_scanned,
-+ ac->ac_found);
-+ printk("EXT3-fs: groups: ");
-+ for (i = 0; i < EXT3_SB(sb)->s_groups_count; i++) {
-+ struct ext3_group_info *grp = EXT3_GROUP_INFO(sb, i);
-+ struct ext3_prealloc_space *pa;
-+ unsigned long start;
-+ struct list_head *cur;
-+ list_for_each_rcu(cur, &grp->bb_prealloc_list) {
-+ pa = list_entry(cur, struct ext3_prealloc_space,
-+ pa_group_list);
-+ spin_lock(&pa->pa_lock);
-+ ext3_get_group_no_and_offset(sb, pa->pa_pstart, NULL, &start);
-+ spin_unlock(&pa->pa_lock);
-+ printk("PA:%u:%lu:%u ", i, start, pa->pa_len);
-+ }
-+
-+ if (grp->bb_free == 0)
-+ continue;
-+ printk("%d: %d/%d ", i, grp->bb_free, grp->bb_fragments);
-+ }
-+ printk("\n");
-+ //dump_stack();
-+#endif
-+}
-+
-+void ext3_mb_group_or_file(struct ext3_allocation_context *ac)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(ac->ac_sb);
-+ loff_t size;
-+ int bsbits;
-+
-+ if (!(ac->ac_flags & EXT3_MB_HINT_DATA))
-+ return;
-+
-+ if (ac->ac_o_ex.fe_len >= sbi->s_mb_small_req)
-+ return;
-+
-+ if (unlikely(ac->ac_flags & EXT3_MB_HINT_GOAL_ONLY))
-+ return;
-+
-+ /* request is so large that we don't care about
-+ * streaming - it overweights any possible seek */
-+ if (ac->ac_o_ex.fe_len >= sbi->s_mb_large_req)
-+ return;
-+
-+ bsbits = ac->ac_sb->s_blocksize_bits;
-+
-+ size = ac->ac_o_ex.fe_logical + ac->ac_o_ex.fe_len;
-+ size = size << bsbits;
-+ if (size < i_size_read(ac->ac_inode))
-+ size = i_size_read(ac->ac_inode);
-+ size = (size + ac->ac_sb->s_blocksize - 1) >> bsbits;
-+
-+ /* don't use group allocation for large files */
-+ if (size >= sbi->s_mb_large_req)
-+ return;
-+
-+ BUG_ON(ac->ac_lg != NULL);
-+ ac->ac_lg = &sbi->s_locality_groups[smp_processor_id()];
-+
-+ /* we're going to use group allocation */
-+ ac->ac_flags |= EXT3_MB_HINT_GROUP_ALLOC;
-+
-+ /* serialize all allocations in the group */
-+ down(&ac->ac_lg->lg_sem);
-+}
-+
-+int ext3_mb_initialize_context(struct ext3_allocation_context *ac,
-+ struct ext3_allocation_request *ar)
-+{
-+ struct super_block *sb = ar->inode->i_sb;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ struct ext3_super_block *es = sbi->s_es;
-+ unsigned long group, len, goal;
-+ unsigned long block;
-+
-+ /* we can't allocate > group size */
-+ len = ar->len;
-+ if (len >= EXT3_BLOCKS_PER_GROUP(sb) - 10)
-+ len = EXT3_BLOCKS_PER_GROUP(sb) - 10;
-+
-+ /* start searching from the goal */
-+ goal = ar->goal;
-+ if (goal < le32_to_cpu(es->s_first_data_block) ||
-+ goal >= le32_to_cpu(es->s_blocks_count))
-+ goal = le32_to_cpu(es->s_first_data_block);
-+ ext3_get_group_no_and_offset(sb, goal, &group, &block);
-+
-+ /* set up allocation goals */
-+ ac->ac_b_ex.fe_logical = ar->logical;
-+ ac->ac_b_ex.fe_group = 0;
-+ ac->ac_b_ex.fe_start = 0;
-+ ac->ac_b_ex.fe_len = 0;
-+ ac->ac_status = AC_STATUS_CONTINUE;
-+ ac->ac_groups_scanned = 0;
-+ ac->ac_ex_scanned = 0;
-+ ac->ac_found = 0;
-+ ac->ac_sb = sb;
-+ ac->ac_inode = ar->inode;
-+ ac->ac_o_ex.fe_logical = ar->logical;
-+ ac->ac_o_ex.fe_group = group;
-+ ac->ac_o_ex.fe_start = block;
-+ ac->ac_o_ex.fe_len = len;
-+ ac->ac_g_ex.fe_logical = ar->logical;
-+ ac->ac_g_ex.fe_group = group;
-+ ac->ac_g_ex.fe_start = block;
-+ ac->ac_g_ex.fe_len = len;
-+ ac->ac_f_ex.fe_len = 0;
-+ ac->ac_flags = ar->flags;
-+ ac->ac_2order = 0;
-+ ac->ac_criteria = 0;
-+ ac->ac_pa = NULL;
-+ ac->ac_bitmap_page = NULL;
-+ ac->ac_buddy_page = NULL;
-+ ac->ac_lg = NULL;
-+
-+ /* we have to define context: we'll we work with a file or
-+ * locality group. this is a policy, actually */
-+ ext3_mb_group_or_file(ac);
-+
-+ mb_debug("init ac: %u blocks @ %llu, goal %llu, flags %x, 2^%d, "
-+ "left: %llu/%llu, right %llu/%llu to %swritable\n",
-+ (unsigned) ar->len, (unsigned) ar->logical,
-+ (unsigned) ar->goal, ac->ac_flags, ac->ac_2order,
-+ (unsigned) ar->lleft, (unsigned) ar->pleft,
-+ (unsigned) ar->lright, (unsigned) ar->pright,
-+ atomic_read(&ar->inode->i_writecount) ? "" : "non-");
-+ return 0;
-+
-+}
-+
-+/*
-+ * release all resource we used in allocation
-+ */
-+int ext3_mb_release_context(struct ext3_allocation_context *ac)
-+{
-+ if (ac->ac_pa) {
-+ if (ac->ac_pa->pa_linear) {
-+ /* see comment in ext3_mb_use_group_pa() */
-+ spin_lock(&ac->ac_pa->pa_lock);
-+ ac->ac_pa->pa_pstart += ac->ac_b_ex.fe_len;
-+ ac->ac_pa->pa_lstart += ac->ac_b_ex.fe_len;
-+ ac->ac_pa->pa_free -= ac->ac_b_ex.fe_len;
-+ ac->ac_pa->pa_len -= ac->ac_b_ex.fe_len;
-+ spin_unlock(&ac->ac_pa->pa_lock);
-+ }
-+ ext3_mb_put_pa(ac, ac->ac_sb, ac->ac_pa);
-+ }
-+ if (ac->ac_bitmap_page)
-+ page_cache_release(ac->ac_bitmap_page);
-+ if (ac->ac_buddy_page)
-+ page_cache_release(ac->ac_buddy_page);
-+ if (ac->ac_flags & EXT3_MB_HINT_GROUP_ALLOC)
-+ up(&ac->ac_lg->lg_sem);
-+ ext3_mb_collect_stats(ac);
-+ return 0;
-+}
-+
-+int ext3_mb_discard_preallocations(struct super_block *sb, int needed)
-+{
-+ int i, ret, freed = 0;
-+
-+ for (i = 0; i < EXT3_SB(sb)->s_groups_count && needed > 0; i++) {
-+ ret = ext3_mb_discard_group_preallocations(sb, i, needed);
-+ freed += ret;
-+ needed -= ret;
-+ }
-+
-+ return freed;
-+}
-+
-+/*
-+ * Main entry point into mballoc to allocate blocks
-+ * it tries to use preallocation first, then falls back
-+ * to usual allocation
-+ */
-+unsigned long ext3_mb_new_blocks(handle_t *handle,
-+ struct ext3_allocation_request *ar, int *errp)
-+{
-+ struct ext3_allocation_context ac;
-+ struct ext3_sb_info *sbi;
-+ struct super_block *sb;
-+ unsigned long block = 0;
-+ int freed, inquota;
-+
-+ sb = ar->inode->i_sb;
-+ sbi = EXT3_SB(sb);
-+
-+ if (!test_opt(sb, MBALLOC)) {
-+ static int ext3_mballoc_warning = 0;
-+ if (ext3_mballoc_warning++ == 0)
-+ printk(KERN_ERR "EXT3-fs: multiblock request with "
-+ "mballoc disabled!\n");
-+ ar->len = 1;
-+ block = ext3_new_block_old(handle, ar->inode, ar->goal, errp);
-+ return block;
-+ }
-+
-+ while (ar->len && DQUOT_ALLOC_BLOCK(ar->inode, ar->len)) {
-+ ar->flags |= EXT3_MB_HINT_NOPREALLOC;
-+ ar->len--;
-+ }
-+ if (ar->len == 0) {
-+ *errp = -EDQUOT;
-+ return 0;
-+ }
-+ inquota = ar->len;
-+
-+ ext3_mb_poll_new_transaction(sb, handle);
-+
-+ if ((*errp = ext3_mb_initialize_context(&ac, ar))) {
-+ ar->len = 0;
-+ goto out;
-+ }
-+
-+ ac.ac_op = EXT3_MB_HISTORY_PREALLOC;
-+ if (!ext3_mb_use_preallocated(&ac)) {
-+
-+ ac.ac_op = EXT3_MB_HISTORY_ALLOC;
-+ ext3_mb_normalize_request(&ac, ar);
-+
-+repeat:
-+ /* allocate space in core */
-+ ext3_mb_regular_allocator(&ac);
-+
-+ /* as we've just preallocated more space than
-+ * user requested orinally, we store allocated
-+ * space in a special descriptor */
-+ if (ac.ac_status == AC_STATUS_FOUND &&
-+ ac.ac_o_ex.fe_len < ac.ac_b_ex.fe_len)
-+ ext3_mb_new_preallocation(&ac);
-+ }
-+
-+ if (likely(ac.ac_status == AC_STATUS_FOUND)) {
-+ ext3_mb_mark_diskspace_used(&ac, handle);
-+ *errp = 0;
-+ block = ext3_grp_offs_to_block(sb, &ac.ac_b_ex);
-+ ar->len = ac.ac_b_ex.fe_len;
-+ } else {
-+ freed = ext3_mb_discard_preallocations(sb, ac.ac_o_ex.fe_len);
-+ if (freed)
-+ goto repeat;
-+ *errp = -ENOSPC;
-+ ac.ac_b_ex.fe_len = 0;
-+ ar->len = 0;
-+ ext3_mb_show_ac(&ac);
-+ }
-+
-+ ext3_mb_release_context(&ac);
-+
-+out:
-+ if (ar->len < inquota)
-+ DQUOT_FREE_BLOCK(ar->inode, inquota - ar->len);
-+
-+ return block;
-+}
-+EXPORT_SYMBOL(ext3_mb_new_blocks);
-+
-+int ext3_new_block(handle_t *handle, struct inode *inode,
-+ unsigned long goal, int *errp)
-+{
-+ struct ext3_allocation_request ar;
-+ unsigned long ret;
-+
-+ if (!test_opt(inode->i_sb, MBALLOC)) {
-+ ret = ext3_new_block_old(handle, inode, goal, errp);
-+ return ret;
-+ }
-+
-+ ar.inode = inode;
-+ ar.goal = goal;
-+ ar.len = 1;
-+ ar.logical = 0;
-+ ar.lleft = 0;
-+ ar.pleft = 0;
-+ ar.lright = 0;
-+ ar.pright = 0;
-+ ar.flags = 0;
-+ ret = ext3_mb_new_blocks(handle, &ar, errp);
-+ return ret;
-+}
-+
-+void ext3_mb_poll_new_transaction(struct super_block *sb, handle_t *handle)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+
-+ if (sbi->s_last_transaction == handle->h_transaction->t_tid)
-+ return;
-+
-+ /* new transaction! time to close last one and free blocks for
-+ * committed transaction. we know that only transaction can be
-+ * active, so previos transaction can be being logged and we
-+ * know that transaction before previous is known to be already
-+ * logged. this means that now we may free blocks freed in all
-+ * transactions before previous one. hope I'm clear enough ... */
-+
-+ spin_lock(&sbi->s_md_lock);
-+ if (sbi->s_last_transaction != handle->h_transaction->t_tid) {
-+ mb_debug("new transaction %lu, old %lu\n",
-+ (unsigned long) handle->h_transaction->t_tid,
-+ (unsigned long) sbi->s_last_transaction);
-+ list_splice_init(&sbi->s_closed_transaction,
-+ &sbi->s_committed_transaction);
-+ list_splice_init(&sbi->s_active_transaction,
-+ &sbi->s_closed_transaction);
-+ sbi->s_last_transaction = handle->h_transaction->t_tid;
-+ }
-+ spin_unlock(&sbi->s_md_lock);
-+
-+ ext3_mb_free_committed_blocks(sb);
-+}
-+
-+int ext3_mb_free_metadata(handle_t *handle, struct ext3_buddy *e3b,
-+ int group, int block, int count)
-+{
-+ struct ext3_group_info *db = e3b->bd_info;
-+ struct super_block *sb = e3b->bd_sb;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+ struct ext3_free_metadata *md;
-+ int i;
-+
-+ BUG_ON(e3b->bd_bitmap_page == NULL);
-+ BUG_ON(e3b->bd_buddy_page == NULL);
-+
-+ for (i = 0; i < count; i++) {
-+ md = db->bb_md_cur;
-+ if (md && db->bb_tid != handle->h_transaction->t_tid) {
-+ db->bb_md_cur = NULL;
-+ md = NULL;
-+ }
-+
-+ if (md == NULL) {
-+ ext3_unlock_group(sb, group);
-+ md = kmalloc(sizeof(*md), GFP_KERNEL);
-+ if (md == NULL)
-+ return -ENOMEM;
-+ md->num = 0;
-+ md->group = group;
-+
-+ ext3_lock_group(sb, group);
-+ if (db->bb_md_cur == NULL) {
-+ spin_lock(&sbi->s_md_lock);
-+ list_add(&md->list, &sbi->s_active_transaction);
-+ spin_unlock(&sbi->s_md_lock);
-+ /* protect buddy cache from being freed,
-+ * otherwise we'll refresh it from
-+ * on-disk bitmap and lose not-yet-available
-+ * blocks */
-+ page_cache_get(e3b->bd_buddy_page);
-+ page_cache_get(e3b->bd_bitmap_page);
-+ db->bb_md_cur = md;
-+ db->bb_tid = handle->h_transaction->t_tid;
-+ mb_debug("new md 0x%p for group %u\n",
-+ md, md->group);
-+ } else {
-+ kfree(md);
-+ md = db->bb_md_cur;
-+ }
-+ }
-+
-+ BUG_ON(md->num >= EXT3_BB_MAX_BLOCKS);
-+ md->blocks[md->num] = block + i;
-+ md->num++;
-+ if (md->num == EXT3_BB_MAX_BLOCKS) {
-+ /* no more space, put full container on a sb's list */
-+ db->bb_md_cur = NULL;
-+ }
-+ }
-+ return 0;
-+}
-+
-+/*
-+ * Main entry point into mballoc to free blocks
-+ */
-+void ext3_mb_free_blocks(handle_t *handle, struct inode *inode,
-+ unsigned long block, unsigned long count,
-+ int metadata, int *freed)
-+{
-+ struct buffer_head *bitmap_bh = NULL;
-+ struct super_block *sb = inode->i_sb;
-+ struct ext3_allocation_context ac;
-+ struct ext3_group_desc *gdp;
-+ struct ext3_super_block *es;
-+ unsigned long bit, overflow;
-+ struct buffer_head *gd_bh;
-+ unsigned long block_group;
-+ struct ext3_sb_info *sbi;
-+ struct ext3_buddy e3b;
-+ int err = 0, ret;
-+
-+ *freed = 0;
-+
-+ ext3_mb_poll_new_transaction(sb, handle);
-+
-+ sbi = EXT3_SB(sb);
-+ es = EXT3_SB(sb)->s_es;
-+ if (block < le32_to_cpu(es->s_first_data_block) ||
-+ block + count < block ||
-+ block + count > le32_to_cpu(es->s_blocks_count)) {
-+ ext3_error (sb, __FUNCTION__,
-+ "Freeing blocks not in datazone - "
-+ "block = %lu, count = %lu", block, count);
-+ goto error_return;
-+ }
-+
-+ ext3_debug("freeing block %lu\n", block);
-+
-+ ac.ac_op = EXT3_MB_HISTORY_FREE;
-+ ac.ac_inode = inode;
-+ ac.ac_sb = sb;
-+
-+do_more:
-+ overflow = 0;
-+ ext3_get_group_no_and_offset(sb, block, &block_group, &bit);
-+
-+ /*
-+ * Check to see if we are freeing blocks across a group
-+ * boundary.
-+ */
-+ if (bit + count > EXT3_BLOCKS_PER_GROUP(sb)) {
-+ overflow = bit + count - EXT3_BLOCKS_PER_GROUP(sb);
-+ count -= overflow;
-+ }
-+ brelse(bitmap_bh);
-+ bitmap_bh = read_block_bitmap(sb, block_group);
-+ if (!bitmap_bh)
-+ goto error_return;
-+ gdp = ext3_get_group_desc (sb, block_group, &gd_bh);
-+ if (!gdp)
-+ goto error_return;
-+
-+ if (in_range (le32_to_cpu(gdp->bg_block_bitmap), block, count) ||
-+ in_range (le32_to_cpu(gdp->bg_inode_bitmap), block, count) ||
-+ in_range (block, le32_to_cpu(gdp->bg_inode_table),
-+ EXT3_SB(sb)->s_itb_per_group) ||
-+ in_range (block + count - 1, le32_to_cpu(gdp->bg_inode_table),
-+ EXT3_SB(sb)->s_itb_per_group))
-+ ext3_error(sb, __FUNCTION__,
-+ "Freeing blocks in system zone - "
-+ "Block = %lu, count = %lu", block, count);
-+
-+ BUFFER_TRACE(bitmap_bh, "getting write access");
-+ err = ext3_journal_get_write_access(handle, bitmap_bh);
-+ if (err)
-+ goto error_return;
-+
-+ /*
-+ * We are about to modify some metadata. Call the journal APIs
-+ * to unshare ->b_data if a currently-committing transaction is
-+ * using it
-+ */
-+ BUFFER_TRACE(gd_bh, "get_write_access");
-+ err = ext3_journal_get_write_access(handle, gd_bh);
-+ if (err)
-+ goto error_return;
-+
-+ err = ext3_mb_load_buddy(sb, block_group, &e3b);
-+ if (err)
-+ goto error_return;
-+
-+ ext3_lock_group(sb, block_group);
-+ spin_lock(sb_bgl_lock(sbi, block_group));
-+#ifdef AGGRESSIVE_CHECK
-+ {
-+ int i;
-+ for (i = 0; i < count; i++)
-+ BUG_ON(!mb_test_bit(bit + i, bitmap_bh->b_data));
-+ }
-+#endif
-+ mb_clear_bits(NULL, bitmap_bh->b_data, bit, count);
-+ gdp->bg_free_blocks_count =
-+ cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) + count);
-+ spin_unlock(sb_bgl_lock(sbi, block_group));
-+ percpu_counter_mod(&sbi->s_freeblocks_counter, count);
-+
-+ if (metadata) {
-+ /* blocks being freed are metadata. these blocks shouldn't
-+ * be used until this transaction is committed */
-+ ext3_mb_free_metadata(handle, &e3b, block_group, bit, count);
-+ } else {
-+ err = mb_free_blocks(inode, &e3b, bit, count);
-+ ext3_mb_return_to_preallocation(inode, &e3b, block, count);
-+ BUG_ON(err != 0);
-+ }
-+ ext3_unlock_group(sb, block_group);
-+
-+ ac.ac_b_ex.fe_group = block_group;
-+ ac.ac_b_ex.fe_start = bit;
-+ ac.ac_b_ex.fe_len = count;
-+ ext3_mb_store_history(&ac);
-+
-+ /* We dirtied the bitmap block */
-+ BUFFER_TRACE(bitmap_bh, "dirtied bitmap block");
-+ err = ext3_journal_dirty_metadata(handle, bitmap_bh);
-+
-+ ext3_mb_release_desc(&e3b);
-+
-+ *freed += count;
-+
-+ /* And the group descriptor block */
-+ BUFFER_TRACE(gd_bh, "dirtied group descriptor block");
-+ ret = ext3_journal_dirty_metadata(handle, gd_bh);
-+ if (!err) err = ret;
-+
-+ if (overflow && !err) {
-+ block += count;
-+ count = overflow;
-+ goto do_more;
-+ }
-+ sb->s_dirt = 1;
-+error_return:
-+ brelse(bitmap_bh);
-+ ext3_std_error(sb, err);
-+ return;
-+}
-+
-+EXPORT_SYMBOL(ext3_free_blocks);
-+EXPORT_SYMBOL(ext3_mb_discard_inode_preallocations);
+++ /dev/null
-Index: linux-stage/fs/ext3/super.c
-===================================================================
---- linux-stage.orig/fs/ext3/super.c
-+++ linux-stage/fs/ext3/super.c
-@@ -35,6 +35,8 @@
- #include <linux/namei.h>
- #include <linux/quotaops.h>
- #include <linux/seq_file.h>
-+#include <linux/kthread.h>
-+#include <linux/utsname.h>
-
- #include <asm/uaccess.h>
-
-@@ -83,7 +85,7 @@ handle_t *ext3_journal_start_sb(struct s
- * take the FS itself readonly cleanly. */
- journal = EXT3_SB(sb)->s_journal;
- if (is_journal_aborted(journal)) {
-- ext3_abort(sb, __FUNCTION__,
-+ ext3_abort(sb, __func__,
- "Detected aborted journal");
- return ERR_PTR(-EROFS);
- }
-@@ -305,7 +307,7 @@ void ext3_update_dynamic_rev(struct supe
- if (le32_to_cpu(es->s_rev_level) > EXT3_GOOD_OLD_REV)
- return;
-
-- ext3_warning(sb, __FUNCTION__,
-+ ext3_warning(sb, __func__,
- "updating to rev %d because of new feature flag, "
- "running e2fsck is recommended",
- EXT3_DYNAMIC_REV);
-@@ -440,6 +442,8 @@ static void ext3_put_super (struct super
- invalidate_bdev(sbi->journal_bdev, 0);
- ext3_blkdev_remove(sbi);
- }
-+ if (sbi->s_mmp_tsk)
-+ kthread_stop(sbi->s_mmp_tsk);
- if (sbi->s_dev_proc) {
- remove_proc_entry(sbi->s_dev_proc->name, proc_root_ext3);
- sbi->s_dev_proc = NULL;
-@@ -1395,7 +1399,7 @@ static int ext3_check_descriptors (struc
- return 0;
- }
- if (!ext3_group_desc_csum_verify(sbi, i, gdp)) {
-- ext3_error(sb, __FUNCTION__,
-+ ext3_error(sb, __func__,
- "Checksum for group %d failed (%u!=%u)\n", i,
- le16_to_cpu(ext3_group_desc_csum(sbi,i,gdp)),
- le16_to_cpu(gdp->bg_checksum));
-@@ -1490,7 +1494,7 @@ static void ext3_orphan_cleanup (struct
- if (inode->i_nlink) {
- printk(KERN_DEBUG
- "%s: truncating inode %lu to %Ld bytes\n",
-- __FUNCTION__, inode->i_ino, inode->i_size);
-+ __func__, inode->i_ino, inode->i_size);
- jbd_debug(2, "truncating inode %lu to %Ld bytes\n",
- inode->i_ino, inode->i_size);
- ext3_truncate(inode);
-@@ -1498,7 +1502,7 @@ static void ext3_orphan_cleanup (struct
- } else {
- printk(KERN_DEBUG
- "%s: deleting unreferenced inode %lu\n",
-- __FUNCTION__, inode->i_ino);
-+ __func__, inode->i_ino);
- jbd_debug(2, "deleting unreferenced inode %lu\n",
- inode->i_ino);
- nr_orphans++;
-@@ -1568,6 +1572,355 @@ static ext3_fsblk_t descriptor_loc(struc
- return (has_super + ext3_group_first_block_no(sb, bg));
- }
-
-+/*
-+ * Write the MMP block using WRITE_SYNC to try to get the block on-disk
-+ * faster.
-+ */
-+static int write_mmp_block(struct buffer_head *bh)
-+{
-+ mark_buffer_dirty(bh);
-+ lock_buffer(bh);
-+ bh->b_end_io = end_buffer_write_sync;
-+ get_bh(bh);
-+ submit_bh(WRITE_SYNC, bh);
-+ wait_on_buffer(bh);
-+ if (unlikely(!buffer_uptodate(bh)))
-+ return 1;
-+
-+ return 0;
-+}
-+
-+/*
-+ * Read the MMP block. It _must_ be read from disk and hence we clear the
-+ * uptodate flag on the buffer.
-+ */
-+static int read_mmp_block(struct super_block *sb, struct buffer_head **bh,
-+ unsigned long mmp_block)
-+{
-+ struct mmp_struct *mmp;
-+
-+ if (*bh)
-+ clear_buffer_uptodate(*bh);
-+
-+#if 0
-+ brelse(*bh);
-+
-+ *bh = sb_bread(sb, mmp_block);
-+#else
-+ if (!*bh)
-+ *bh = sb_getblk(sb, mmp_block);
-+ if (*bh) {
-+ get_bh(*bh);
-+ lock_buffer(*bh);
-+ (*bh)->b_end_io = end_buffer_read_sync;
-+ submit_bh(READ_SYNC, *bh);
-+ wait_on_buffer(*bh);
-+ if (!buffer_uptodate(*bh)) {
-+ brelse(*bh);
-+ *bh = NULL;
-+ }
-+ }
-+#endif
-+ if (!*bh) {
-+ ext3_warning(sb, __func__,
-+ "Error while reading MMP block %lu", mmp_block);
-+ return -EIO;
-+ }
-+
-+ mmp = (struct mmp_struct *)((*bh)->b_data);
-+ if (le32_to_cpu(mmp->mmp_magic) != EXT3_MMP_MAGIC)
-+ return -EINVAL;
-+
-+ return 0;
-+}
-+
-+/*
-+ * Dump as much information as possible to help the admin.
-+ */
-+static void dump_mmp_msg(struct super_block *sb, struct mmp_struct *mmp,
-+ const char *function, const char *msg)
-+{
-+ ext3_warning(sb, function, "%s", msg);
-+ ext3_warning(sb, function, "MMP failure info: last update time: %llu, "
-+ "last update node: %s, last update device: %s\n",
-+ (long long unsigned int)le64_to_cpu(mmp->mmp_time),
-+ mmp->mmp_nodename, mmp->mmp_bdevname);
-+}
-+
-+/*
-+ * kmmpd will update the MMP sequence every s_mmp_update_interval seconds
-+ */
-+static int kmmpd(void *data)
-+{
-+ struct super_block *sb = ((struct mmpd_data *) data)->sb;
-+ struct buffer_head *bh = ((struct mmpd_data *) data)->bh;
-+ struct ext3_super_block *es = EXT3_SB(sb)->s_es;
-+ struct mmp_struct *mmp;
-+ unsigned long mmp_block;
-+ u32 seq = 0;
-+ unsigned long failed_writes = 0;
-+ int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval);
-+ unsigned mmp_check_interval;
-+ unsigned long last_update_time;
-+ unsigned long diff;
-+ int retval;
-+
-+ mmp_block = le64_to_cpu(es->s_mmp_block);
-+ mmp = (struct mmp_struct *)(bh->b_data);
-+ mmp->mmp_time = cpu_to_le64(get_seconds());
-+ /*
-+ * Start with the higher mmp_check_interval and reduce it if
-+ * the MMP block is being updated on time.
-+ */
-+ mmp_check_interval = max(5UL * mmp_update_interval,
-+ EXT3_MMP_MIN_CHECK_INTERVAL);
-+ mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
-+ bdevname(bh->b_bdev, mmp->mmp_bdevname);
-+
-+ down_read(&uts_sem);
-+ memcpy(mmp->mmp_nodename, system_utsname.nodename,
-+ sizeof(mmp->mmp_nodename));
-+ up_read(&uts_sem);
-+
-+ while (!kthread_should_stop()) {
-+ if (++seq > EXT3_MMP_SEQ_MAX)
-+ seq = 1;
-+
-+ mmp->mmp_seq = cpu_to_le32(seq);
-+ mmp->mmp_time = cpu_to_le64(get_seconds());
-+ last_update_time = jiffies;
-+
-+ retval = write_mmp_block(bh);
-+ /*
-+ * Don't spew too many error messages. Print one every
-+ * (s_mmp_update_interval * 60) seconds.
-+ */
-+ if (retval && (failed_writes % 60) == 0) {
-+ ext3_error(sb, __func__,
-+ "Error writing to MMP block");
-+ failed_writes++;
-+ }
-+
-+ if (!(le32_to_cpu(es->s_feature_incompat) &
-+ EXT3_FEATURE_INCOMPAT_MMP)) {
-+ ext3_warning(sb, __func__, "kmmpd being stopped "
-+ "since MMP feature has been disabled.");
-+ EXT3_SB(sb)->s_mmp_tsk = NULL;
-+ goto failed;
-+ }
-+
-+ if (sb->s_flags & MS_RDONLY) {
-+ ext3_warning(sb, __func__, "kmmpd being stopped "
-+ "since filesystem has been remounted as "
-+ "readonly.");
-+ EXT3_SB(sb)->s_mmp_tsk = NULL;
-+ goto failed;
-+ }
-+
-+ diff = jiffies - last_update_time;
-+ if (diff < mmp_update_interval * HZ)
-+ schedule_timeout_interruptible(mmp_update_interval*
-+ HZ - diff);
-+
-+ /*
-+ * We need to make sure that more than mmp_check_interval
-+ * seconds have not passed since writing. If that has happened
-+ * we need to check if the MMP block is as we left it.
-+ */
-+ diff = jiffies - last_update_time;
-+ if (diff > mmp_check_interval * HZ) {
-+ struct buffer_head *bh_check = NULL;
-+ struct mmp_struct *mmp_check;
-+
-+ retval = read_mmp_block(sb, &bh_check, mmp_block);
-+ if (retval) {
-+ ext3_error(sb, __func__, "error reading MMP"
-+ "data: %d", retval);
-+ EXT3_SB(sb)->s_mmp_tsk = NULL;
-+ goto failed;
-+ }
-+
-+ mmp_check = (struct mmp_struct *)(bh_check->b_data);
-+ if (mmp->mmp_seq != mmp_check->mmp_seq ||
-+ memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename,
-+ sizeof(mmp->mmp_nodename))) {
-+ dump_mmp_msg(sb, mmp_check, __func__,
-+ "Error while updating MMP info. "
-+ "The filesystem seems to have "
-+ "been multiply mounted.");
-+ ext3_error(sb, __func__, "abort");
-+ goto failed;
-+ }
-+ put_bh(bh_check);
-+ }
-+
-+ /*
-+ * Adjust the mmp_check_interval depending on how much time
-+ * it took for the MMP block to be written.
-+ */
-+ mmp_check_interval = max(min(5 * diff / HZ,
-+ EXT3_MMP_MAX_CHECK_INTERVAL),
-+ EXT3_MMP_MIN_CHECK_INTERVAL);
-+ mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
-+ }
-+
-+ /*
-+ * Unmount seems to be clean.
-+ */
-+ mmp->mmp_seq = cpu_to_le32(EXT3_MMP_SEQ_CLEAN);
-+ mmp->mmp_time = cpu_to_le64(get_seconds());
-+
-+ retval = write_mmp_block(bh);
-+
-+failed:
-+ brelse(bh);
-+ return retval;
-+}
-+
-+/*
-+ * Get a random new sequence number but make sure it is not greater than
-+ * EXT3_MMP_SEQ_MAX.
-+ */
-+static unsigned int mmp_new_seq(void)
-+{
-+ u32 new_seq;
-+
-+ do {
-+ get_random_bytes(&new_seq, sizeof(u32));
-+ } while (new_seq > EXT3_MMP_SEQ_MAX);
-+
-+ return new_seq;
-+}
-+
-+/*
-+ * Protect the filesystem from being mounted more than once.
-+ */
-+static int ext3_multi_mount_protect(struct super_block *sb,
-+ unsigned long mmp_block)
-+{
-+ struct ext3_super_block *es = EXT3_SB(sb)->s_es;
-+ struct buffer_head *bh = NULL;
-+ struct mmp_struct *mmp = NULL;
-+ struct mmpd_data *mmpd_data;
-+ u32 seq;
-+ unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval);
-+ unsigned int wait_time = 0;
-+ int retval;
-+
-+ if (mmp_block < le32_to_cpu(es->s_first_data_block) ||
-+ mmp_block >= le32_to_cpu(es->s_blocks_count)) {
-+ ext3_warning(sb, __func__,
-+ "Invalid MMP block in superblock");
-+ goto failed;
-+ }
-+
-+ retval = read_mmp_block(sb, &bh, mmp_block);
-+ if (retval)
-+ goto failed;
-+
-+ mmp = (struct mmp_struct *)(bh->b_data);
-+
-+ if (mmp_check_interval < EXT3_MMP_MIN_CHECK_INTERVAL)
-+ mmp_check_interval = EXT3_MMP_MIN_CHECK_INTERVAL;
-+
-+ /*
-+ * If check_interval in MMP block is larger, use that instead of
-+ * update_interval from the superblock.
-+ */
-+ if (mmp->mmp_check_interval > mmp_check_interval)
-+ mmp_check_interval = mmp->mmp_check_interval;
-+
-+ seq = le32_to_cpu(mmp->mmp_seq);
-+ if (seq == EXT3_MMP_SEQ_CLEAN)
-+ goto skip;
-+
-+ if (seq == EXT3_MMP_SEQ_FSCK) {
-+ dump_mmp_msg(sb, mmp, __func__,
-+ "fsck is running on the filesystem");
-+ goto failed;
-+ }
-+
-+ wait_time = min(mmp_check_interval * 2 + 1,
-+ mmp_check_interval + 60);
-+
-+ /* Print MMP interval if more than 20 secs. */
-+ if (wait_time > EXT3_MMP_MIN_CHECK_INTERVAL * 4)
-+ ext3_warning(sb, __func__, "MMP interval %u higher than "
-+ "expected, please wait.\n", wait_time * 2);
-+
-+ if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
-+ ext3_warning(sb, __func__, "MMP startup interrupted, failing "
-+ "mount\n");
-+ goto failed;
-+ }
-+
-+ retval = read_mmp_block(sb, &bh, mmp_block);
-+ if (retval)
-+ goto failed;
-+ mmp = (struct mmp_struct *)(bh->b_data);
-+ if (seq != le32_to_cpu(mmp->mmp_seq)) {
-+ dump_mmp_msg(sb, mmp, __func__,
-+ "Device is already active on another node.");
-+ goto failed;
-+ }
-+
-+skip:
-+ /*
-+ * write a new random sequence number.
-+ */
-+ mmp->mmp_seq = seq = cpu_to_le32(mmp_new_seq());
-+
-+ retval = write_mmp_block(bh);
-+ if (retval)
-+ goto failed;
-+
-+ /*
-+ * wait for MMP interval and check mmp_seq.
-+ */
-+ if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
-+ ext3_warning(sb, __func__, "MMP startup interrupted, failing "
-+ "mount\n");
-+ goto failed;
-+ }
-+
-+ retval = read_mmp_block(sb, &bh, mmp_block);
-+ if (retval)
-+ goto failed;
-+ mmp = (struct mmp_struct *)(bh->b_data);
-+ if (seq != le32_to_cpu(mmp->mmp_seq)) {
-+ dump_mmp_msg(sb, mmp, __func__,
-+ "Device is already active on another node.");
-+ goto failed;
-+ }
-+
-+ mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
-+ if (!mmpd_data) {
-+ ext3_warning(sb, KERN_ERR, "not enough memory for mmpd_data");
-+ goto failed;
-+ }
-+ mmpd_data->sb = sb;
-+ mmpd_data->bh = bh;
-+
-+ /*
-+ * Start a kernel thread to update the MMP block periodically.
-+ */
-+ EXT3_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, mmpd_data, "kmmpd-%s",
-+ bdevname(bh->b_bdev,
-+ mmp->mmp_bdevname));
-+ if (IS_ERR(EXT3_SB(sb)->s_mmp_tsk)) {
-+ EXT3_SB(sb)->s_mmp_tsk = NULL;
-+ ext3_warning(sb, __func__, "Unable to create kmmpd thread "
-+ "for %s.", sb->s_id);
-+ goto failed;
-+ }
-+
-+ return 0;
-+
-+failed:
-+ brelse(bh);
-+ return 1;
-+}
-+
-
- static int ext3_fill_super (struct super_block *sb, void *data, int silent)
- {
-@@ -1901,6 +2254,11 @@ static int ext3_fill_super (struct super
- EXT3_HAS_INCOMPAT_FEATURE(sb,
- EXT3_FEATURE_INCOMPAT_RECOVER));
-
-+ if (EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_MMP) &&
-+ !(sb->s_flags & MS_RDONLY))
-+ if (ext3_multi_mount_protect(sb, le64_to_cpu(es->s_mmp_block)))
-+ goto failed_mount3;
-+
- /*
- * The first inode we look at is the journal inode. Don't try
- * root first: it may be modified in the journal!
-@@ -2033,6 +2391,8 @@ cantfind_ext3:
- failed_mount4:
- journal_destroy(sbi->s_journal);
- failed_mount3:
-+ if (sbi->s_mmp_tsk)
-+ kthread_stop(sbi->s_mmp_tsk);
- percpu_counter_destroy(&sbi->s_freeblocks_counter);
- percpu_counter_destroy(&sbi->s_freeinodes_counter);
- percpu_counter_destroy(&sbi->s_dirs_counter);
-@@ -2409,9 +2769,9 @@ static void ext3_clear_journal_err(struc
- char nbuf[16];
-
- errstr = ext3_decode_error(sb, j_errno, nbuf);
-- ext3_warning(sb, __FUNCTION__, "Filesystem error recorded "
-+ ext3_warning(sb, __func__, "Filesystem error recorded "
- "from previous mount: %s", errstr);
-- ext3_warning(sb, __FUNCTION__, "Marking fs in need of "
-+ ext3_warning(sb, __func__, "Marking fs in need of "
- "filesystem check.");
-
- EXT3_SB(sb)->s_mount_state |= EXT3_ERROR_FS;
-@@ -2531,7 +2891,7 @@ static int ext3_remount (struct super_bl
- ext3_fsblk_t n_blocks_count = 0;
- unsigned long old_sb_flags;
- struct ext3_mount_options old_opts;
-- int err;
-+ int err = 0;
- #ifdef CONFIG_QUOTA
- int i;
- #endif
-@@ -2557,7 +2917,7 @@ static int ext3_remount (struct super_bl
- }
-
- if (sbi->s_mount_opt & EXT3_MOUNT_ABORT)
-- ext3_abort(sb, __FUNCTION__, "Abort forced by user");
-+ ext3_abort(sb, __func__, "Abort forced by user");
-
- sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
- ((sbi->s_mount_opt & EXT3_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
-@@ -2630,6 +2990,13 @@ static int ext3_remount (struct super_bl
- }
- if (!ext3_setup_super (sb, es, 0))
- sb->s_flags &= ~MS_RDONLY;
-+ if (EXT3_HAS_INCOMPAT_FEATURE(sb,
-+ EXT3_FEATURE_INCOMPAT_MMP))
-+ if (ext3_multi_mount_protect(sb,
-+ le64_to_cpu(es->s_mmp_block))) {
-+ err = -EROFS;
-+ goto restore_opts;
-+ }
- }
- }
- #ifdef CONFIG_QUOTA
-Index: linux-stage/include/linux/ext3_fs.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs.h
-+++ linux-stage/include/linux/ext3_fs.h
-@@ -594,13 +594,17 @@ struct ext3_super_block {
- __le32 s_first_meta_bg; /* First metablock block group */
- __le32 s_mkfs_time; /* When the filesystem was created */
- __le32 s_jnl_blocks[17]; /* Backup of the journal inode */
-- __le32 s_blocks_count_hi; /* Blocks count high 32 bits */
-+/*150*/ __le32 s_blocks_count_hi; /* Blocks count high 32 bits */
- __le32 s_r_blocks_count_hi; /* Reserved blocks count high 32 bits*/
- __le32 s_free_blocks_count_hi; /* Free blocks count high 32 bits */
- __le16 s_min_extra_isize; /* All inodes have at least # bytes */
- __le16 s_want_extra_isize; /* New inodes should reserve # bytes */
-- __le32 s_flags; /* Miscellaneous flags */
-- __u32 s_reserved[167]; /* Padding to the end of the block */
-+/*160*/ __le32 s_flags; /* Miscellaneous flags */
-+ __le16 s_raid_stride; /* RAID stride */
-+ __le16 s_mmp_update_interval; /* # seconds to wait in MMP checking */
-+ __le64 s_mmp_block; /* Block for multi-mount protection */
-+/*170*/ __le32 s_raid_stripe_width; /* blocks on all data disks (N*stride)*/
-+ __le32 s_reserved[163]; /* Padding to the end of the block */
- };
-
- #ifdef __KERNEL__
-@@ -703,12 +707,14 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_INCOMPAT_JOURNAL_DEV 0x0008 /* Journal device */
- #define EXT3_FEATURE_INCOMPAT_META_BG 0x0010
- #define EXT3_FEATURE_INCOMPAT_EXTENTS 0x0040 /* extents support */
-+#define EXT3_FEATURE_INCOMPAT_MMP 0x0100
-
- #define EXT3_FEATURE_COMPAT_SUPP EXT2_FEATURE_COMPAT_EXT_ATTR
- #define EXT3_FEATURE_INCOMPAT_SUPP (EXT3_FEATURE_INCOMPAT_FILETYPE| \
- EXT3_FEATURE_INCOMPAT_RECOVER| \
- EXT3_FEATURE_INCOMPAT_META_BG| \
-- EXT3_FEATURE_INCOMPAT_EXTENTS)
-+ EXT3_FEATURE_INCOMPAT_EXTENTS| \
-+ EXT3_FEATURE_INCOMPAT_MMP)
- #define EXT3_FEATURE_RO_COMPAT_SUPP (EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER| \
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
- EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
-@@ -871,6 +877,50 @@ ext3_group_first_block_no(struct super_b
- #define ERR_BAD_DX_DIR -75000
-
- /*
-+ * This structure will be used for multiple mount protection. It will be
-+ * written into the block number saved in the s_mmp_block field in the
-+ * superblock. Programs that check MMP should assume that if
-+ * SEQ_FSCK (or any unknown code above SEQ_MAX) is present then it is NOT safe
-+ * to use the filesystem, regardless of how old the timestamp is.
-+ */
-+#define EXT3_MMP_MAGIC 0x004D4D50U /* ASCII for MMP */
-+#define EXT3_MMP_SEQ_CLEAN 0xFF4D4D50U /* mmp_seq value for clean unmount */
-+#define EXT3_MMP_SEQ_FSCK 0xE24D4D50U /* mmp_seq value when being fscked */
-+#define EXT3_MMP_SEQ_MAX 0xE24D4D4FU /* maximum valid mmp_seq value */
-+
-+struct mmp_struct {
-+ __le32 mmp_magic;
-+ __le32 mmp_seq;
-+ __le64 mmp_time;
-+ char mmp_nodename[64];
-+ char mmp_bdevname[32];
-+ __le16 mmp_check_interval;
-+ __le16 mmp_pad1;
-+ __le32 mmp_pad2[227];
-+};
-+
-+/* arguments passed to the mmp thread */
-+struct mmpd_data {
-+ struct buffer_head *bh; /* bh from initial read_mmp_block() */
-+ struct super_block *sb; /* super block of the fs */
-+};
-+
-+/*
-+ * Default interval in seconds to update the MMP sequence number.
-+ */
-+#define EXT3_MMP_UPDATE_INTERVAL 1
-+
-+/*
-+ * Minimum interval for MMP checking in seconds.
-+ */
-+#define EXT3_MMP_MIN_CHECK_INTERVAL 5UL
-+
-+/*
-+ * Maximum interval for MMP checking in seconds.
-+ */
-+#define EXT3_MMP_MAX_CHECK_INTERVAL 300UL
-+
-+/*
- * Function prototypes
- */
-
-Index: linux-stage/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-stage.orig/include/linux/ext3_fs_sb.h
-+++ linux-stage/include/linux/ext3_fs_sb.h
-@@ -157,6 +157,8 @@ struct ext3_sb_info {
- /* locality groups */
- struct ext3_locality_group *s_locality_groups;
-
-+ /* Kernel thread for multiple mount protection */
-+ struct task_struct *s_mmp_tsk;
- };
-
- #define EXT3_GROUP_INFO(sb, group) \
+++ /dev/null
-Index: linux-2.6.18.8/fs/ext3/ialloc.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/ialloc.c 2007-06-20 18:54:59.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/ialloc.c 2007-06-20 18:54:59.000000000 +0200
-@@ -729,7 +729,8 @@ got:
- /* This is the optimal IO size (for stat), not the fs block size */
- inode->i_blksize = PAGE_SIZE;
- inode->i_blocks = 0;
-- inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_mtime = inode->i_atime = inode->i_ctime = ei->i_crtime =
-+ ext3_current_time(inode);
-
- memset(ei->i_data, 0, sizeof(ei->i_data));
- ei->i_dir_start_lookup = 0;
-@@ -761,9 +762,8 @@ got:
- spin_unlock(&sbi->s_next_gen_lock);
-
- ei->i_state = EXT3_STATE_NEW;
-- ei->i_extra_isize =
-- (EXT3_INODE_SIZE(inode->i_sb) > EXT3_GOOD_OLD_INODE_SIZE) ?
-- sizeof(struct ext3_inode) - EXT3_GOOD_OLD_INODE_SIZE : 0;
-+
-+ ei->i_extra_isize = EXT3_SB(sb)->s_want_extra_isize;
-
- ret = inode;
- if(DQUOT_ALLOC_INODE(inode)) {
-Index: linux-2.6.18.8/fs/ext3/inode.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/inode.c 2007-06-20 18:54:52.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/inode.c 2007-06-20 18:54:59.000000000 +0200
-@@ -727,7 +727,7 @@ static int ext3_splice_branch(handle_t *
-
- /* We are done with atomic stuff, now do the rest of housekeeping */
-
-- inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_ctime = ext3_current_time(inode);
- ext3_mark_inode_dirty(handle, inode);
-
- /* had we spliced it onto indirect block? */
-@@ -2375,7 +2375,7 @@ do_indirects:
- ext3_discard_reservation(inode);
-
- mutex_unlock(&ei->truncate_mutex);
-- inode->i_mtime = inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_mtime = inode->i_ctime = ext3_current_time(inode);
- ext3_mark_inode_dirty(handle, inode);
-
- /*
-@@ -2611,10 +2611,6 @@ void ext3_read_inode(struct inode * inod
- }
- inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
- inode->i_size = le32_to_cpu(raw_inode->i_size);
-- inode->i_atime.tv_sec = le32_to_cpu(raw_inode->i_atime);
-- inode->i_ctime.tv_sec = le32_to_cpu(raw_inode->i_ctime);
-- inode->i_mtime.tv_sec = le32_to_cpu(raw_inode->i_mtime);
-- inode->i_atime.tv_nsec = inode->i_ctime.tv_nsec = inode->i_mtime.tv_nsec = 0;
-
- ei->i_state = 0;
- ei->i_dir_start_lookup = 0;
-@@ -2689,6 +2685,11 @@ void ext3_read_inode(struct inode * inod
- } else
- ei->i_extra_isize = 0;
-
-+ EXT3_INODE_GET_XTIME(i_ctime, inode, raw_inode);
-+ EXT3_INODE_GET_XTIME(i_mtime, inode, raw_inode);
-+ EXT3_INODE_GET_XTIME(i_atime, inode, raw_inode);
-+ EXT3_EINODE_GET_XTIME(i_crtime, ei, raw_inode);
-+
- if (S_ISREG(inode->i_mode)) {
- inode->i_op = &ext3_file_inode_operations;
- inode->i_fop = &ext3_file_operations;
-@@ -2769,9 +2770,12 @@ static int ext3_do_update_inode(handle_t
- }
- raw_inode->i_links_count = cpu_to_le16(inode->i_nlink);
- raw_inode->i_size = cpu_to_le32(ei->i_disksize);
-- raw_inode->i_atime = cpu_to_le32(inode->i_atime.tv_sec);
-- raw_inode->i_ctime = cpu_to_le32(inode->i_ctime.tv_sec);
-- raw_inode->i_mtime = cpu_to_le32(inode->i_mtime.tv_sec);
-+
-+ EXT3_INODE_SET_XTIME(i_ctime, inode, raw_inode);
-+ EXT3_INODE_SET_XTIME(i_mtime, inode, raw_inode);
-+ EXT3_INODE_SET_XTIME(i_atime, inode, raw_inode);
-+ EXT3_EINODE_SET_XTIME(i_crtime, ei, raw_inode);
-+
- raw_inode->i_blocks = cpu_to_le32(inode->i_blocks);
- raw_inode->i_dtime = cpu_to_le32(ei->i_dtime);
- raw_inode->i_flags = cpu_to_le32(ei->i_flags);
-Index: linux-2.6.18.8/fs/ext3/ioctl.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/ioctl.c 2007-06-20 18:42:05.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/ioctl.c 2007-06-20 18:54:59.000000000 +0200
-@@ -120,7 +120,7 @@ int ext3_ioctl (struct inode * inode, st
- ei->i_flags = flags;
-
- ext3_set_inode_flags(inode);
-- inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_ctime = ext3_current_time(inode);
-
- err = ext3_mark_iloc_dirty(handle, inode, &iloc);
- flags_err:
-@@ -157,7 +157,7 @@ flags_err:
- return PTR_ERR(handle);
- err = ext3_reserve_inode_write(handle, inode, &iloc);
- if (err == 0) {
-- inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_ctime = ext3_current_time(inode);
- inode->i_generation = generation;
- err = ext3_mark_iloc_dirty(handle, inode, &iloc);
- }
-Index: linux-2.6.18.8/fs/ext3/namei.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/namei.c 2007-06-20 18:54:53.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/namei.c 2007-06-20 18:54:59.000000000 +0200
-@@ -1287,7 +1287,7 @@ static int add_dirent_to_buf(handle_t *h
- * happen is that the times are slightly out of date
- * and/or different from the directory change time.
- */
-- dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC;
-+ dir->i_mtime = dir->i_ctime = ext3_current_time(dir);
- ext3_update_dx_flag(dir);
- dir->i_version++;
- ext3_mark_inode_dirty(handle, dir);
-@@ -2079,7 +2079,7 @@ static int ext3_rmdir (struct inode * di
- inode->i_version++;
- inode->i_nlink = 0;
- ext3_orphan_add(handle, inode);
-- inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME_SEC;
-+ inode->i_ctime = dir->i_ctime = dir->i_mtime = ext3_current_time(inode);
- ext3_mark_inode_dirty(handle, inode);
- ext3_dec_count(handle, dir);
- ext3_update_dx_flag(dir);
-@@ -2129,13 +2129,13 @@ static int ext3_unlink(struct inode * di
- retval = ext3_delete_entry(handle, dir, de, bh);
- if (retval)
- goto end_unlink;
-- dir->i_ctime = dir->i_mtime = CURRENT_TIME_SEC;
-+ dir->i_ctime = dir->i_mtime = ext3_current_time(dir);
- ext3_update_dx_flag(dir);
- ext3_mark_inode_dirty(handle, dir);
- ext3_dec_count(handle, inode);
- if (!inode->i_nlink)
- ext3_orphan_add(handle, inode);
-- inode->i_ctime = dir->i_ctime;
-+ inode->i_ctime = ext3_current_time(inode);
- ext3_mark_inode_dirty(handle, inode);
- retval = 0;
-
-@@ -2237,7 +2237,7 @@ retry:
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_ctime = ext3_current_time(inode);
- ext3_inc_count(handle, inode);
- atomic_inc(&inode->i_count);
-
-@@ -2340,7 +2340,7 @@ static int ext3_rename (struct inode * o
- * Like most other Unix systems, set the ctime for inodes on a
- * rename.
- */
-- old_inode->i_ctime = CURRENT_TIME_SEC;
-+ old_inode->i_ctime = ext3_current_time(old_inode);
- ext3_mark_inode_dirty(handle, old_inode);
-
- /*
-@@ -2373,9 +2373,9 @@ static int ext3_rename (struct inode * o
-
- if (new_inode) {
- ext3_dec_count(handle, new_inode);
-- new_inode->i_ctime = CURRENT_TIME_SEC;
-+ new_inode->i_ctime = ext3_current_time(new_inode);
- }
-- old_dir->i_ctime = old_dir->i_mtime = CURRENT_TIME_SEC;
-+ old_dir->i_ctime = old_dir->i_mtime = ext3_current_time(old_dir);
- ext3_update_dx_flag(old_dir);
- if (dir_bh) {
- BUFFER_TRACE(dir_bh, "get_write_access");
-Index: linux-2.6.18.8/fs/ext3/super.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/super.c 2007-06-20 18:54:59.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/super.c 2007-06-20 18:54:59.000000000 +0200
-@@ -1713,6 +1713,8 @@ static int ext3_fill_super (struct super
- sbi->s_inode_size);
- goto failed_mount;
- }
-+ if (sbi->s_inode_size > EXT3_GOOD_OLD_INODE_SIZE)
-+ sb->s_time_gran = 1 << (EXT3_EPOCH_BITS - 2);
- }
- sbi->s_frag_size = EXT3_MIN_FRAG_SIZE <<
- le32_to_cpu(es->s_log_frag_size);
-@@ -1917,6 +1919,32 @@ static int ext3_fill_super (struct super
- }
-
- ext3_setup_super (sb, es, sb->s_flags & MS_RDONLY);
-+
-+ /* determine the minimum size of new large inodes, if present */
-+ if (sbi->s_inode_size > EXT3_GOOD_OLD_INODE_SIZE) {
-+ sbi->s_want_extra_isize = sizeof(struct ext3_inode) -
-+ EXT3_GOOD_OLD_INODE_SIZE;
-+ if (EXT3_HAS_RO_COMPAT_FEATURE(sb,
-+ EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE)) {
-+ if (sbi->s_want_extra_isize <
-+ le16_to_cpu(es->s_want_extra_isize))
-+ sbi->s_want_extra_isize =
-+ le16_to_cpu(es->s_want_extra_isize);
-+ if (sbi->s_want_extra_isize <
-+ le16_to_cpu(es->s_min_extra_isize))
-+ sbi->s_want_extra_isize =
-+ le16_to_cpu(es->s_min_extra_isize);
-+ }
-+ }
-+ /* Check if enough inode space is available */
-+ if (EXT3_GOOD_OLD_INODE_SIZE + sbi->s_want_extra_isize >
-+ sbi->s_inode_size) {
-+ sbi->s_want_extra_isize = sizeof(struct ext3_inode) -
-+ EXT3_GOOD_OLD_INODE_SIZE;
-+ printk(KERN_INFO "EXT3-fs: required extra inode space not"
-+ "available.\n");
-+ }
-+
- /*
- * akpm: core read_super() calls in here with the superblock locked.
- * That deadlocks, because orphan cleanup needs to lock the superblock
-Index: linux-2.6.18.8/fs/ext3/xattr.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/xattr.c 2007-06-20 18:54:52.000000000 +0200
-+++ linux-2.6.18.8/fs/ext3/xattr.c 2007-06-20 18:54:59.000000000 +0200
-@@ -1007,7 +1007,7 @@ ext3_xattr_set_handle(handle_t *handle,
- }
- if (!error) {
- ext3_xattr_update_super_block(handle, inode->i_sb);
-- inode->i_ctime = CURRENT_TIME_SEC;
-+ inode->i_ctime = ext3_current_time(inode);
- error = ext3_mark_iloc_dirty(handle, inode, &is.iloc);
- /*
- * The bh is consumed by ext3_mark_iloc_dirty, even with
-Index: linux-2.6.18.8/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs.h 2007-06-20 18:54:59.000000000 +0200
-+++ linux-2.6.18.8/include/linux/ext3_fs.h 2007-06-20 18:54:59.000000000 +0200
-@@ -288,7 +288,7 @@ struct ext3_inode {
- __le16 i_uid; /* Low 16 bits of Owner Uid */
- __le32 i_size; /* Size in bytes */
- __le32 i_atime; /* Access time */
-- __le32 i_ctime; /* Creation time */
-+ __le32 i_ctime; /* Inode Change time */
- __le32 i_mtime; /* Modification time */
- __le32 i_dtime; /* Deletion Time */
- __le16 i_gid; /* Low 16 bits of Group Id */
-@@ -337,10 +337,73 @@ struct ext3_inode {
- } osd2; /* OS dependent 2 */
- __le16 i_extra_isize;
- __le16 i_pad1;
-+ __le32 i_ctime_extra; /* extra Change time (nsec << 2 | epoch) */
-+ __le32 i_mtime_extra; /* extra Modification time(nsec << 2 | epoch) */
-+ __le32 i_atime_extra; /* extra Access time (nsec << 2 | epoch) */
-+ __le32 i_crtime; /* File Creation time */
-+ __le32 i_crtime_extra; /* extra File Creation time (nsec << 2 | epoch) */
- };
-
- #define i_size_high i_dir_acl
-
-+#define EXT3_EPOCH_BITS 2
-+#define EXT3_EPOCH_MASK ((1 << EXT3_EPOCH_BITS) - 1)
-+#define EXT3_NSEC_MASK (~0UL << EXT3_EPOCH_BITS)
-+
-+#define EXT3_FITS_IN_INODE(ext3_inode, einode, field) \
-+ ((offsetof(typeof(*ext3_inode), field) + \
-+ sizeof((ext3_inode)->field)) \
-+ <= (EXT3_GOOD_OLD_INODE_SIZE + \
-+ (einode)->i_extra_isize)) \
-+
-+static inline __le32 ext3_encode_extra_time(struct timespec *time)
-+{
-+ return cpu_to_le32((sizeof(time->tv_sec) > 4 ?
-+ time->tv_sec >> 32 : 0) |
-+ ((time->tv_nsec << 2) & EXT3_NSEC_MASK));
-+}
-+
-+static inline void ext3_decode_extra_time(struct timespec *time, __le32 extra) {
-+ if (sizeof(time->tv_sec) > 4)
-+ time->tv_sec |= (__u64)(le32_to_cpu(extra) & EXT3_EPOCH_MASK)
-+ << 32;
-+ time->tv_nsec = (le32_to_cpu(extra) & EXT3_NSEC_MASK) >> 2;
-+}
-+
-+#define EXT3_INODE_SET_XTIME(xtime, inode, raw_inode) \
-+do { \
-+ (raw_inode)->xtime = cpu_to_le32((inode)->xtime.tv_sec); \
-+ if (EXT3_FITS_IN_INODE(raw_inode, EXT3_I(inode), xtime ## _extra)) \
-+ (raw_inode)->xtime ## _extra = \
-+ ext3_encode_extra_time(&(inode)->xtime); \
-+} while (0)
-+
-+#define EXT3_EINODE_SET_XTIME(xtime, einode, raw_inode)\
-+do { \
-+ if (EXT3_FITS_IN_INODE(raw_inode, einode, xtime)) \
-+ (raw_inode)->xtime = cpu_to_le32((einode)->xtime.tv_sec); \
-+ if (EXT3_FITS_IN_INODE(raw_inode, einode, xtime ## _extra)) \
-+ (raw_inode)->xtime ## _extra = \
-+ ext3_encode_extra_time(&(einode)->xtime); \
-+} while (0)
-+
-+#define EXT3_INODE_GET_XTIME(xtime, inode, raw_inode) \
-+do { \
-+ (inode)->xtime.tv_sec = le32_to_cpu((raw_inode)->xtime); \
-+ if (EXT3_FITS_IN_INODE(raw_inode, EXT3_I(inode), xtime ## _extra)) \
-+ ext3_decode_extra_time(&(inode)->xtime, \
-+ raw_inode->xtime ## _extra); \
-+} while (0)
-+
-+#define EXT3_EINODE_GET_XTIME(xtime, einode, raw_inode) \
-+do { \
-+ if (EXT3_FITS_IN_INODE(raw_inode, einode, xtime)) \
-+ (einode)->xtime.tv_sec = le32_to_cpu((raw_inode)->xtime); \
-+ if (EXT3_FITS_IN_INODE(raw_inode, einode, xtime ## _extra)) \
-+ ext3_decode_extra_time(&(einode)->xtime, \
-+ raw_inode->xtime ## _extra); \
-+} while (0)
-+
- #if defined(__KERNEL__) || defined(__linux__)
- #define i_reserved1 osd1.linux1.l_i_reserved1
- #define i_frag osd2.linux2.l_i_frag
-@@ -520,11 +583,19 @@ struct ext3_super_block {
- __le32 s_last_orphan; /* start of list of inodes to delete */
- __le32 s_hash_seed[4]; /* HTREE hash seed */
- __u8 s_def_hash_version; /* Default hash version to use */
-- __u8 s_reserved_char_pad;
-- __u16 s_reserved_word_pad;
-+ __u8 s_jnl_backup_type; /* Default type of journal backup */
-+ __le16 s_desc_size; /* Group desc. size: INCOMPAT_64BIT */
- __le32 s_default_mount_opts;
-- __le32 s_first_meta_bg; /* First metablock block group */
-- __u32 s_reserved[190]; /* Padding to the end of the block */
-+ __le32 s_first_meta_bg; /* First metablock block group */
-+ __le32 s_mkfs_time; /* When the filesystem was created */
-+ __le32 s_jnl_blocks[17]; /* Backup of the journal inode */
-+ __le32 s_blocks_count_hi; /* Blocks count high 32 bits */
-+ __le32 s_r_blocks_count_hi; /* Reserved blocks count high 32 bits*/
-+ __le32 s_free_blocks_count_hi; /* Free blocks count high 32 bits */
-+ __le16 s_min_extra_isize; /* All inodes have at least # bytes */
-+ __le16 s_want_extra_isize; /* New inodes should reserve # bytes */
-+ __le32 s_flags; /* Miscellaneous flags */
-+ __u32 s_reserved[167]; /* Padding to the end of the block */
- };
-
- #ifdef __KERNEL__
-@@ -539,6 +610,13 @@ static inline struct ext3_inode_info *EX
- return container_of(inode, struct ext3_inode_info, vfs_inode);
- }
-
-+static inline struct timespec ext3_current_time(struct inode *inode)
-+{
-+ return (inode->i_sb->s_time_gran < NSEC_PER_SEC) ?
-+ current_fs_time(inode->i_sb) : CURRENT_TIME_SEC;
-+}
-+
-+
- static inline int ext3_valid_inum(struct super_block *sb, unsigned long ino)
- {
- return ino == EXT3_ROOT_INO ||
-@@ -611,6 +689,8 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_RO_COMPAT_BTREE_DIR 0x0004
- #define EXT4_FEATURE_RO_COMPAT_GDT_CSUM 0x0010
- #define EXT4_FEATURE_RO_COMPAT_DIR_NLINK 0x0020
-+#define EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE 0x0040
-+
-
- #define EXT3_FEATURE_INCOMPAT_COMPRESSION 0x0001
- #define EXT3_FEATURE_INCOMPAT_FILETYPE 0x0002
-@@ -628,6 +708,7 @@ static inline int ext3_valid_inum(struct
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
- EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
- EXT4_FEATURE_RO_COMPAT_DIR_NLINK| \
-+ EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE| \
- EXT3_FEATURE_RO_COMPAT_BTREE_DIR)
-
- /*
-Index: linux-2.6.18.8/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_sb.h 2007-06-20 18:54:54.000000000 +0200
-+++ linux-2.6.18.8/include/linux/ext3_fs_sb.h 2007-06-20 18:54:59.000000000 +0200
-@@ -68,6 +68,9 @@ struct ext3_sb_info {
- /* Last group used to allocate inode */
- int s_last_alloc_group;
-
-+ /* New inodes should reserve # bytes */
-+ unsigned int s_want_extra_isize;
-+
- /* root of the per fs reservation window tree */
- spinlock_t s_rsv_window_lock;
- struct rb_root s_rsv_window_root;
-Index: linux-2.6.18.8/include/linux/ext3_fs_i.h
-===================================================================
---- linux-2.6.18.8.orig/include/linux/ext3_fs_i.h 2007-06-20 18:54:57.000000000 +0200
-+++ linux-2.6.18.8/include/linux/ext3_fs_i.h 2007-06-20 18:54:59.000000000 +0200
-@@ -140,6 +140,8 @@ struct ext3_inode_info {
- /* on-disk additional length */
- __u16 i_extra_isize;
-
-+ struct timespec i_crtime;
-+
- /*
- * truncate_mutex is for serialising ext3_truncate() against
- * ext3_getblock(). In the 2.4 ext2 design, great chunks of inode's
+++ /dev/null
-Index: linux-2.6.18-53.1.14/fs/ext3/namei.c
-===================================================================
---- linux-2.6.18-53.1.14/fs/ext3/namei.c 2008-12-02 13:21:14.000000000 +0530
-+++ linux-2.6.18-53.1.14_new/fs/ext3/namei.c 2008-12-02 15:11:09.000000000 +0530
-@@ -1656,11 +1656,17 @@ static int ext3_delete_entry (handle_t *
- static inline void ext3_inc_count(handle_t *handle, struct inode *inode)
- {
- inode->i_nlink++;
-+ if (is_dx(inode) && inode->i_nlink > 1) {
-+ /* limit is 16-bit i_links_count */
-+ if (inode->i_nlink >= EXT3_LINK_MAX || inode->i_nlink == 2)
-+ inode->i_nlink = 1;
-+ }
- }
-
- static inline void ext3_dec_count(handle_t *handle, struct inode *inode)
- {
-- inode->i_nlink--;
-+ if (!S_ISDIR(inode->i_mode) || inode->i_nlink > 2)
-+ drop_nlink(inode);
- }
-
- static int ext3_add_nondir(handle_t *handle,
-@@ -1759,7 +1770,7 @@ static int ext3_mkdir(struct inode * dir
- struct ext3_dir_entry_2 * de;
- int err, retries = 0;
-
-- if (dir->i_nlink >= EXT3_LINK_MAX)
-+ if (EXT3_DIR_LINK_MAX(dir))
- return -EMLINK;
-
- retry:
-@@ -1782,7 +1793,7 @@ retry:
- inode->i_size = EXT3_I(inode)->i_disksize = inode->i_sb->s_blocksize;
- dir_block = ext3_bread (handle, inode, 0, 1, &err);
- if (!dir_block) {
-- inode->i_nlink--; /* is this nlink == 0? */
-+ drop_nlink(inode); /* is this nlink == 0? */
- ext3_mark_inode_dirty(handle, inode);
- iput (inode);
- goto out_stop;
-@@ -1814,7 +1825,7 @@ retry:
- iput (inode);
- goto out_stop;
- }
-- dir->i_nlink++;
-+ ext3_inc_count(handle, dir);
- ext3_update_dx_flag(dir);
- ext3_mark_inode_dirty(handle, dir);
- d_instantiate(dentry, inode);
-@@ -2079,10 +2090,10 @@ static int ext3_rmdir (struct inode * di
- retval = ext3_delete_entry(handle, dir, de, bh);
- if (retval)
- goto end_rmdir;
-- if (inode->i_nlink != 2)
-- ext3_warning (inode->i_sb, "ext3_rmdir",
-- "empty directory has nlink!=2 (%d)",
-- inode->i_nlink);
-+ if (!EXT3_DIR_LINK_EMPTY(inode))
-+ ext3_warning(inode->i_sb, "ext3_rmdir",
-+ "empty directory has too many links (%d)",
-+ inode->i_nlink);
- inode->i_version++;
- inode->i_nlink = 0;
- /* There's no need to set i_disksize: the fact that i_nlink is
-@@ -2092,7 +2103,7 @@ static int ext3_rmdir (struct inode * di
- ext3_orphan_add(handle, inode);
- inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME_SEC;
- ext3_mark_inode_dirty(handle, inode);
-- dir->i_nlink--;
-+ ext3_dec_count(handle, dir);
- ext3_update_dx_flag(dir);
- ext3_mark_inode_dirty(handle, dir);
-
-@@ -2143,7 +2154,7 @@ static int ext3_unlink(struct inode * di
- dir->i_ctime = dir->i_mtime = CURRENT_TIME_SEC;
- ext3_update_dx_flag(dir);
- ext3_mark_inode_dirty(handle, dir);
-- inode->i_nlink--;
-+ ext3_dec_count(handle, inode);
- if (!inode->i_nlink)
- ext3_orphan_add(handle, inode);
- inode->i_ctime = dir->i_ctime;
-@@ -2219,7 +2230,7 @@ static int ext3_link (struct dentry * ol
- struct inode *inode = old_dentry->d_inode;
- int err, retries = 0;
-
-- if (inode->i_nlink >= EXT3_LINK_MAX)
-+ if (EXT3_DIR_LINK_MAX(inode))
- return -EMLINK;
- /*
- * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing
-@@ -2313,8 +2324,8 @@ static int ext3_rename (struct inode * o
- if (le32_to_cpu(PARENT_INO(dir_bh->b_data)) != old_dir->i_ino)
- goto end_rename;
- retval = -EMLINK;
-- if (!new_inode && new_dir!=old_dir &&
-- new_dir->i_nlink >= EXT3_LINK_MAX)
-+ if (!new_inode && new_dir != old_dir &&
-+ EXT3_DIR_LINK_MAX(new_dir))
- goto end_rename;
- }
- if (!new_bh) {
-@@ -2371,7 +2382,7 @@ static int ext3_rename (struct inode * o
- }
-
- if (new_inode) {
-- new_inode->i_nlink--;
-+ ext3_dec_count(handle, new_inode);
- new_inode->i_ctime = CURRENT_TIME_SEC;
- }
- old_dir->i_ctime = old_dir->i_mtime = CURRENT_TIME_SEC;
-@@ -2382,11 +2393,13 @@ static int ext3_rename (struct inode * o
- PARENT_INO(dir_bh->b_data) = cpu_to_le32(new_dir->i_ino);
- BUFFER_TRACE(dir_bh, "call ext3_journal_dirty_metadata");
- ext3_journal_dirty_metadata(handle, dir_bh);
-- old_dir->i_nlink--;
-+ ext3_dec_count(handle, old_dir);
- if (new_inode) {
-- new_inode->i_nlink--;
-+ /* checked empty_dir above, can't have another parent,
-+ * ext3_dec_count() won't work for many-linked dirs */
-+ new_inode->i_nlink = 0;
- } else {
-- new_dir->i_nlink++;
-+ ext3_inc_count(handle, new_dir);
- ext3_update_dx_flag(new_dir);
- ext3_mark_inode_dirty(handle, new_dir);
- }
-Index: linux-2.6.18-53.1.14/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18-53.1.14/include/linux/ext3_fs.h 2008-12-02 13:21:23.000000000 +0530
-+++ linux-2.6.18-53.1.14_new/include/linux/ext3_fs.h 2008-12-02 13:22:04.000000000 +0530
-@@ -74,7 +74,7 @@
- /*
- * Maximal count of links to a file
- */
--#define EXT3_LINK_MAX 32000
-+#define EXT3_LINK_MAX 65000
-
- /*
- * Macro-instructions used to manage several block sizes
-@@ -563,6 +563,7 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER 0x0001
- #define EXT3_FEATURE_RO_COMPAT_LARGE_FILE 0x0002
- #define EXT3_FEATURE_RO_COMPAT_BTREE_DIR 0x0004
-+#define EXT4_FEATURE_RO_COMPAT_DIR_NLINK 0x0020
-
- #define EXT3_FEATURE_INCOMPAT_COMPRESSION 0x0001
- #define EXT3_FEATURE_INCOMPAT_FILETYPE 0x0002
-@@ -576,6 +577,7 @@ static inline int ext3_valid_inum(struct
- EXT3_FEATURE_INCOMPAT_META_BG)
- #define EXT3_FEATURE_RO_COMPAT_SUPP (EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER| \
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
-+ EXT4_FEATURE_RO_COMPAT_DIR_NLINK| \
- EXT3_FEATURE_RO_COMPAT_BTREE_DIR)
-
- /*
+++ /dev/null
-Index: linux-2.6.16.60-0.33_org/fs/ext3/hash.c
-===================================================================
---- linux-2.6.16.60-0.33_org/fs/ext3/hash.c 2009-07-01 18:16:50.000000000 +0530
-+++ linux-2.6.16.60-0.33_new/fs/ext3/hash.c 2009-07-01 18:26:58.000000000 +0530
-@@ -8,7 +8,7 @@
- * This file may be redistributed under the terms of the GNU Public
- * License.
- */
--
-+#include <linux/module.h>
- #include <linux/fs.h>
- #include <linux/jbd.h>
- #include <linux/sched.h>
-@@ -173,3 +173,4 @@ int ext3fs_dirhash(const char *name, int
- hinfo->minor_hash = minor_hash;
- return 0;
- }
-+EXPORT_SYMBOL(ext3fs_dirhash);
-Index: linux-2.6.16.60-0.33_org/fs/ext3/namei.c
-===================================================================
---- linux-2.6.16.60-0.33_org/fs/ext3/namei.c 2009-07-01 18:16:50.000000000 +0530
-+++ linux-2.6.16.60-0.33_new/fs/ext3/namei.c 2009-07-01 18:24:49.000000000 +0530
-@@ -75,6 +75,7 @@ struct buffer_head *ext3_append(handle_t
-
- return bh;
- }
-+EXPORT_SYMBOL(ext3_append);
-
- #ifndef assert
- #define assert(test) J_ASSERT(test)
-Index: linux-2.6.16.60-0.33_org/fs/ext3/super.c
-===================================================================
---- linux-2.6.16.60-0.33_org/fs/ext3/super.c 2009-07-01 18:16:50.000000000 +0530
-+++ linux-2.6.16.60-0.33_new/fs/ext3/super.c 2009-07-01 18:24:27.000000000 +0530
-@@ -260,6 +260,7 @@ void __ext3_std_error (struct super_bloc
-
- ext3_handle_error(sb);
- }
-+EXPORT_SYMBOL(__ext3_std_error);
-
- /*
- * ext3_abort is a much stronger failure handler than ext3_error. The
-Index: linux-2.6.16.60-0.33_org/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.16.60-0.33_org/include/linux/ext3_fs.h 2009-07-01 18:16:50.000000000 +0530
-+++ linux-2.6.16.60-0.33_new/include/linux/ext3_fs.h 2009-07-01 18:22:09.000000000 +0530
-@@ -1055,6 +1055,8 @@ extern void ext3_abort (struct super_blo
- extern void ext3_warning (struct super_block *, const char *, const char *, ...)
- __attribute__ ((format (printf, 3, 4)));
- extern void ext3_update_dynamic_rev (struct super_block *sb);
-+extern void __ext3_std_error (struct super_block * sb, const char * function,
-+ int errno);
-
- #define ext3_std_error(sb, errno) \
- do { \
-@@ -1076,6 +1078,8 @@ extern struct file_operations ext3_file_
- /* namei.c */
- extern struct inode_operations ext3_dir_inode_operations;
- extern struct inode_operations ext3_special_inode_operations;
-+extern struct buffer_head *ext3_append(handle_t *handle, struct inode *inode,
-+ u32 *block, int *err);
-
- /* symlink.c */
- extern struct inode_operations ext3_symlink_inode_operations;
-
+++ /dev/null
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/namei.c linux-2.6.18-128.1.6_2/fs/ext3/namei.c
---- linux-2.6.18-128.1.6_1/fs/ext3/namei.c 2009-08-13 19:07:07.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/namei.c 2009-08-13 19:10:43.000000000 +0530
-@@ -24,6 +24,7 @@
- * Theodore Ts'o, 2002
- */
-
-+#include <linux/module.h>
- #include <linux/fs.h>
- #include <linux/pagemap.h>
- #include <linux/jbd.h>
-@@ -841,7 +842,7 @@ static inline int search_dirblock(struct
- * The returned buffer_head has ->b_count elevated. The caller is expected
- * to brelse() it when appropriate.
- */
--static struct buffer_head * ext3_find_entry (struct dentry *dentry,
-+struct buffer_head * ext3_find_entry (struct dentry *dentry,
- struct ext3_dir_entry_2 ** res_dir)
- {
- struct super_block * sb;
-@@ -953,6 +954,7 @@ cleanup_and_exit:
- brelse (bh_use[ra_ptr]);
- return ret;
- }
-+EXPORT_SYMBOL(ext3_find_entry);
-
- #ifdef CONFIG_EXT3_INDEX
- static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry,
-@@ -1489,7 +1491,7 @@ static int make_indexed_dir(handle_t *ha
- * may not sleep between calling this and putting something into
- * the entry, as someone else might have used it while you slept.
- */
--static int ext3_add_entry (handle_t *handle, struct dentry *dentry,
-+int ext3_add_entry (handle_t *handle, struct dentry *dentry,
- struct inode *inode)
- {
- struct inode *dir = dentry->d_parent->d_inode;
-@@ -1542,6 +1544,7 @@ static int ext3_add_entry (handle_t *han
- de->rec_len = cpu_to_le16(blocksize);
- return add_dirent_to_buf(handle, dentry, inode, de, bh);
- }
-+EXPORT_SYMBOL(ext3_add_entry);
-
- #ifdef CONFIG_EXT3_INDEX
- /*
-@@ -1684,10 +1687,10 @@ cleanup:
- * ext3_delete_entry deletes a directory entry by merging it with the
- * previous entry
- */
--static int ext3_delete_entry (handle_t *handle,
-- struct inode * dir,
-- struct ext3_dir_entry_2 * de_del,
-- struct buffer_head * bh)
-+int ext3_delete_entry (handle_t *handle,
-+ struct inode * dir,
-+ struct ext3_dir_entry_2 * de_del,
-+ struct buffer_head * bh)
- {
- struct ext3_dir_entry_2 * de, * pde;
- int i;
-@@ -1719,6 +1722,7 @@ static int ext3_delete_entry (handle_t *
- }
- return -ENOENT;
- }
-+EXPORT_SYMBOL(ext3_delete_entry);
-
- /*
- * ext3_mark_inode_dirty is somewhat expensive, so unlike ext2 we
-@@ -1774,6 +1778,26 @@ static struct inode * ext3_new_inode_wan
- return ext3_new_inode(handle, dir, mode, inum);
- }
-
-+struct inode * ext3_create_inode(handle_t *handle, struct inode * dir, int mode)
-+{
-+ struct inode *inode;
-+
-+ inode = ext3_new_inode(handle, dir, mode, 0);
-+ if (!IS_ERR(inode)) {
-+ if (S_ISCHR(mode) || S_ISBLK(mode) || S_ISFIFO(mode)) {
-+#ifdef CONFIG_LDISKFS_FS_XATTR
-+ inode->i_op = &ext3_special_inode_operations;
-+#endif
-+ } else {
-+ inode->i_op = &ext3_file_inode_operations;
-+ inode->i_fop = &ext3_file_operations;
-+ ext3_set_aops(inode);
-+ }
-+ }
-+ return inode;
-+}
-+EXPORT_SYMBOL(ext3_create_inode);
-+
- /*
- * By the time this is called, we already have created
- * the directory cache entry for the new file, but it
-@@ -1848,42 +1872,28 @@ retry:
- return err;
- }
-
--static int ext3_mkdir(struct inode * dir, struct dentry * dentry, int mode)
-+/* Initialize @inode as a subdirectory of @dir, and add the
-+ * "." and ".." entries into the first directory block. */
-+int ext3_add_dot_dotdot(handle_t *handle, struct inode * dir,
-+ struct inode *inode)
- {
-- handle_t *handle;
-- struct inode * inode;
- struct buffer_head * dir_block;
- struct ext3_dir_entry_2 * de;
-- int err, retries = 0;
--
-- if (EXT3_DIR_LINK_MAX(dir))
-- return -EMLINK;
-+ int err = 0;
-
--retry:
-- handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
-- EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 +
-- 2*EXT3_QUOTA_INIT_BLOCKS(dir->i_sb));
- if (IS_ERR(handle))
- return PTR_ERR(handle);
-
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode = ext3_new_inode_wantedi (handle, dir, S_IFDIR | mode, dentry);
-- err = PTR_ERR(inode);
-- if (IS_ERR(inode))
-- goto out_stop;
--
- inode->i_op = &ext3_dir_inode_operations;
- inode->i_fop = &ext3_dir_operations;
- inode->i_size = EXT3_I(inode)->i_disksize = inode->i_sb->s_blocksize;
- dir_block = ext3_bread (handle, inode, 0, 1, &err);
-- if (!dir_block) {
-- drop_nlink(inode); /* is this nlink == 0? */
-- ext3_mark_inode_dirty(handle, inode);
-- iput (inode);
-- goto out_stop;
-- }
-+ if (!dir_block)
-+ goto get_out;
-+
- BUFFER_TRACE(dir_block, "get_write_access");
- ext3_journal_get_write_access(handle, dir_block);
- de = (struct ext3_dir_entry_2 *) dir_block->b_data;
-@@ -1904,6 +1914,45 @@ retry:
- ext3_journal_dirty_metadata(handle, dir_block);
- brelse (dir_block);
- ext3_mark_inode_dirty(handle, inode);
-+
-+get_out:
-+ return err;
-+
-+}
-+EXPORT_SYMBOL(ext3_add_dot_dotdot);
-+
-+static int ext3_mkdir(struct inode * dir, struct dentry * dentry, int mode)
-+{
-+ handle_t *handle;
-+ struct inode * inode;
-+ int err, retries = 0;
-+
-+ if (EXT3_DIR_LINK_MAX(dir))
-+ return -EMLINK;
-+
-+retry:
-+ handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +
-+ EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 +
-+ 2*EXT3_QUOTA_INIT_BLOCKS(dir->i_sb));
-+ if (IS_ERR(handle))
-+ return PTR_ERR(handle);
-+
-+ if (IS_DIRSYNC(dir))
-+ handle->h_sync = 1;
-+
-+ inode = ext3_new_inode_wantedi (handle, dir, S_IFDIR | mode, dentry);
-+ err = PTR_ERR(inode);
-+ if (IS_ERR(inode))
-+ goto out_stop;
-+
-+ err = ext3_add_dot_dotdot(handle, dir, inode);
-+ if (err) {
-+ inode->i_nlink = 0;
-+ ext3_mark_inode_dirty(handle, inode);
-+ iput (inode);
-+ goto out_stop;
-+ }
-+
- err = ext3_add_entry (handle, dentry, inode);
- if (err) {
- inode->i_nlink = 0;
-diff -rupN linux-2.6.18-128.1.6_1/include/linux/ext3_fs.h linux-2.6.18-128.1.6_2/include/linux/ext3_fs.h
---- linux-2.6.18-128.1.6_1/include/linux/ext3_fs.h 2009-08-13 19:07:07.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/include/linux/ext3_fs.h 2009-08-13 19:14:07.000000000 +0530
-@@ -1076,6 +1076,10 @@ extern int ext3_group_extend(struct supe
- ext3_fsblk_t n_blocks_count);
-
- /* super.c */
-+extern int ext3_xattr_set_handle(handle_t *handle, struct inode *inode,
-+ int name_index, const char *name,
-+ const void *value, size_t value_len,
-+ int flags);
- extern struct proc_dir_entry *proc_root_ext3;
- extern int __init init_ext3_proc(void);
- extern void exit_ext3_proc(void);
-@@ -1107,6 +1111,20 @@ extern struct inode_operations ext3_file
- extern const struct file_operations ext3_file_operations;
-
- /* namei.c */
-+extern struct inode *ext3_create_inode(handle_t *handle,
-+ struct inode * dir, int mode);
-+extern int ext3_add_entry(handle_t *handle, struct dentry *dentry,
-+ struct inode *inode);
-+extern int ext3_delete_entry(handle_t *handle,
-+ struct inode * dir,
-+ struct ext3_dir_entry_2 * de_del,
-+ struct buffer_head * bh);
-+extern struct buffer_head * ext3_find_entry(struct dentry *dentry,
-+ struct ext3_dir_entry_2
-+ ** res_dir);
-+#define ll_ext3_find_entry(inode, dentry, res_dir) ext3_find_entry(dentry, res_dir)
-+extern int ext3_add_dot_dotdot(handle_t *handle, struct inode *dir,
-+ struct inode *inode);
- extern struct inode_operations ext3_dir_inode_operations;
- extern struct inode_operations ext3_special_inode_operations;
-
-
+++ /dev/null
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/namei.c linux-2.6.18-128.1.6_2/fs/ext3/namei.c
---- linux-2.6.18-128.1.6_1/fs/ext3/namei.c 2009-08-13 19:27:12.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/namei.c 2009-08-13 19:33:34.000000000 +0530
-@@ -51,19 +51,25 @@
- #define NAMEI_RA_SIZE (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS)
- #define NAMEI_RA_INDEX(c,b) (((c) * NAMEI_RA_BLOCKS) + (b))
-
--static struct buffer_head *ext3_append(handle_t *handle,
-+struct buffer_head *ext3_append(handle_t *handle,
- struct inode *inode,
- u32 *block, int *err)
- {
- struct buffer_head *bh;
-+ struct ext3_inode_info *ei = EXT3_I(inode);
-+
-+ /* with parallel dir operations all appends
-+ * have to be serialized -bzzz */
-+ down(&ei->i_append_sem);
-
- *block = inode->i_size >> inode->i_sb->s_blocksize_bits;
-
-- if ((bh = ext3_bread(handle, inode, *block, 1, err))) {
-+ bh = ext3_bread(handle, inode, *block, 1, err);
-+ if (bh != NULL) {
- inode->i_size += inode->i_sb->s_blocksize;
-- EXT3_I(inode)->i_disksize = inode->i_size;
-- ext3_journal_get_write_access(handle,bh);
-+ ei->i_disksize = inode->i_size;
- }
-+ up(&ei->i_append_sem);
- return bh;
- }
-
-diff -rupN linux-2.6.18-128.1.6_1/fs/ext3/super.c linux-2.6.18-128.1.6_2/fs/ext3/super.c
---- linux-2.6.18-128.1.6_1/fs/ext3/super.c 2009-08-13 19:27:12.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/fs/ext3/super.c 2009-08-13 19:27:40.000000000 +0530
-@@ -481,6 +481,9 @@ static struct inode *ext3_alloc_inode(st
- ei->i_acl = EXT3_ACL_NOT_CACHED;
- ei->i_default_acl = EXT3_ACL_NOT_CACHED;
- #endif
-+ dynlock_init(&ei->i_htree_lock);
-+ sema_init(&ei->i_append_sem, 1);
-+
- ei->i_block_alloc_info = NULL;
- ei->vfs_inode.i_version = 1;
-
-diff -rupN linux-2.6.18-128.1.6_1/include/linux/ext3_fs_i.h linux-2.6.18-128.1.6_2/include/linux/ext3_fs_i.h
---- linux-2.6.18-128.1.6_1/include/linux/ext3_fs_i.h 2009-08-13 19:27:12.000000000 +0530
-+++ linux-2.6.18-128.1.6_2/include/linux/ext3_fs_i.h 2009-08-13 19:31:22.000000000 +0530
-@@ -16,6 +16,7 @@
- #ifndef _LINUX_EXT3_FS_I
- #define _LINUX_EXT3_FS_I
-
-+#include <linux/dynlocks.h>
- #include <linux/rwsem.h>
- #include <linux/rbtree.h>
- #include <linux/seqlock.h>
-@@ -104,6 +105,10 @@ struct ext3_inode_info {
- /* block reservation info */
- struct ext3_block_alloc_info *i_block_alloc_info;
-
-+ /* following fields for parallel directory operations -bzzz */
-+ struct dynlock i_htree_lock;
-+ struct semaphore i_append_sem;
-+
- __u32 i_dir_start_lookup;
- #ifdef CONFIG_EXT3_FS_XATTR
- /*
+++ /dev/null
-Index: linux-2.6.18.8/fs/ext3/namei.c
-===================================================================
---- linux-2.6.18.8.orig/fs/ext3/namei.c
-+++ linux-2.6.18.8/fs/ext3/namei.c
-@@ -347,8 +347,8 @@ dx_probe(struct dentry *dentry, struct i
- root->info.hash_version != DX_HASH_HALF_MD4 &&
- root->info.hash_version != DX_HASH_LEGACY) {
- ext3_warning(dir->i_sb, __FUNCTION__,
-- "Unrecognised inode hash code %d",
-- root->info.hash_version);
-+ "Unrecognised inode hash code %d for directory "
-+ "#%lu", root->info.hash_version, dir->i_ino);
- brelse(bh);
- *err = ERR_BAD_DX_DIR;
- goto fail;
-
+++ /dev/null
-Index: linux-stage/fs/ext3/ialloc.c
-===================================================================
---- linux-stage.orig/fs/ext3/ialloc.c 2005-06-26 10:59:43.048185981 +0200
-+++ linux-stage/fs/ext3/ialloc.c 2005-06-26 11:01:21.317716027 +0200
-@@ -775,7 +775,6 @@
- if (!gdp)
- continue;
- desc_count += le16_to_cpu(gdp->bg_free_inodes_count);
-- cond_resched();
- }
- return desc_count;
- #endif
-Index: linux-stage/fs/ext3/super.c
-===================================================================
---- linux-stage.orig/fs/ext3/super.c 2005-06-26 10:59:43.205412542 +0200
-+++ linux-stage/fs/ext3/super.c 2005-06-26 11:02:29.599941754 +0200
-@@ -2236,11 +2232,9 @@
- * block group descriptors. If the sparse superblocks
- * feature is turned on, then not all groups have this.
- */
-- for (i = 0; i < ngroups; i++) {
-+ for (i = 0; i < ngroups; i++)
- overhead += ext3_bg_has_super(sb, i) +
- ext3_bg_num_gdb(sb, i);
-- cond_resched();
-- }
-
- /*
- * Every block group has an inode bitmap, a block
+++ /dev/null
-Index: linux-2.6.18-8.1.8/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-8.1.8.orig/fs/ext3/super.c
-+++ linux-2.6.18-8.1.8/fs/ext3/super.c
-@@ -2923,18 +2923,18 @@ static int ext3_statfs (struct dentry *
- struct super_block *sb = dentry->d_sb;
- struct ext3_sb_info *sbi = EXT3_SB(sb);
- struct ext3_super_block *es = sbi->s_es;
-- ext3_fsblk_t overhead;
-- int i;
-
-- if (test_opt (sb, MINIX_DF))
-- overhead = 0;
-- else {
-- unsigned long ngroups;
-- ngroups = EXT3_SB(sb)->s_groups_count;
-+ if (test_opt(sb, MINIX_DF)) {
-+ sbi->s_overhead_last = 0;
-+ } else if (sbi->s_blocks_last != le32_to_cpu(es->s_blocks_count)) {
-+ unsigned long ngroups = sbi->s_groups_count, i;
-+ ext3_fsblk_t overhead = 0;
- smp_rmb();
-
- /*
-- * Compute the overhead (FS structures)
-+ * Compute the overhead (FS structures). This is constant
-+ * for a given filesystem unless the number of block groups
-+ * changes so we cache the previous value until it does.
- */
-
- /*
-@@ -2956,18 +2956,23 @@ static int ext3_statfs (struct dentry *
- * Every block group has an inode bitmap, a block
- * bitmap, and an inode table.
- */
-- overhead += (ngroups * (2 + EXT3_SB(sb)->s_itb_per_group));
-+ overhead += ngroups * (2 + sbi->s_itb_per_group);
-+ sbi->s_overhead_last = overhead;
-+ smp_wmb();
-+ sbi->s_blocks_last = le32_to_cpu(es->s_blocks_count);
- }
-
- buf->f_type = EXT3_SUPER_MAGIC;
- buf->f_bsize = sb->s_blocksize;
-- buf->f_blocks = le32_to_cpu(es->s_blocks_count) - overhead;
-+ buf->f_blocks = le32_to_cpu(es->s_blocks_count) - sbi->s_overhead_last;
- buf->f_bfree = percpu_counter_sum(&sbi->s_freeblocks_counter);
-+ es->s_free_blocks_count = cpu_to_le32(buf->f_bfree);
- buf->f_bavail = buf->f_bfree - le32_to_cpu(es->s_r_blocks_count);
- if (buf->f_bfree < le32_to_cpu(es->s_r_blocks_count))
- buf->f_bavail = 0;
- buf->f_files = le32_to_cpu(es->s_inodes_count);
- buf->f_ffree = percpu_counter_sum(&sbi->s_freeinodes_counter);
-+ es->s_free_inodes_count = cpu_to_le32(buf->f_ffree);
- buf->f_namelen = EXT3_NAME_LEN;
- return 0;
- }
-Index: linux-2.6.18-8.1.8/include/linux/ext3_fs_sb.h
-===================================================================
---- linux-2.6.18-8.1.8.orig/include/linux/ext3_fs_sb.h
-+++ linux-2.6.18-8.1.8/include/linux/ext3_fs_sb.h
-@@ -45,6 +45,8 @@ struct ext3_sb_info {
- unsigned long s_gdb_count; /* Number of group descriptor blocks */
- unsigned long s_desc_per_block; /* Number of group descriptors per block */
- unsigned long s_groups_count; /* Number of groups in the fs */
-+ unsigned long s_overhead_last; /* Last calculated overhead */
-+ unsigned long s_blocks_last; /* Last seen block count */
- struct buffer_head * s_sbh; /* Buffer containing the super block */
- struct ext3_super_block * s_es; /* Pointer to the super block in the buffer */
- struct buffer_head ** s_group_desc;
+++ /dev/null
-Add support for the uninit_groups feature to the kernel.
-
-Keep a high water mark of used inodes for each group to improve e2fsck time.
-Block and inode bitmaps can be uninitialized on disk via a flag in the
-group descriptor to avoid reading or scanning them at e2fsck time.
-A checksum of each group descriptor is used to ensure that corruption in
-the group descriptor's bit flags does not cause incorrect operation.
-
-Index: linux-2.6.18-53.1.14/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.18-53.1.14.orig/include/linux/ext3_fs.h
-+++ linux-2.6.18-53.1.14/include/linux/ext3_fs.h
-@@ -150,16 +150,22 @@ struct ext3_allocation_request {
- */
- struct ext3_group_desc
- {
-- __le32 bg_block_bitmap; /* Blocks bitmap block */
-- __le32 bg_inode_bitmap; /* Inodes bitmap block */
-+ __le32 bg_block_bitmap; /* Blocks bitmap block */
-+ __le32 bg_inode_bitmap; /* Inodes bitmap block */
- __le32 bg_inode_table; /* Inodes table block */
- __le16 bg_free_blocks_count; /* Free blocks count */
- __le16 bg_free_inodes_count; /* Free inodes count */
- __le16 bg_used_dirs_count; /* Directories count */
-- __u16 bg_pad;
-- __le32 bg_reserved[3];
-+ __le16 bg_flags; /* EXT3_BG_flags (UNINIT, etc) */
-+ __le32 bg_reserved[2]; /* Likely block/inode bitmap checksum */
-+ __le16 bg_itable_unused; /* Unused inodes count */
-+ __le16 bg_checksum; /* crc16(sb_uuid+group+desc) */
- };
-
-+#define EXT3_BG_INODE_UNINIT 0x0001 /* Inode table/bitmap not in use */
-+#define EXT3_BG_BLOCK_UNINIT 0x0002 /* Block bitmap not in use */
-+#define EXT3_BG_INODE_ZEROED 0x0004 /* On-disk itable initialized to zero */
-+
- /*
- * Macro-instructions used to manage group descriptors
- */
-@@ -603,6 +609,7 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER 0x0001
- #define EXT3_FEATURE_RO_COMPAT_LARGE_FILE 0x0002
- #define EXT3_FEATURE_RO_COMPAT_BTREE_DIR 0x0004
-+#define EXT4_FEATURE_RO_COMPAT_GDT_CSUM 0x0010
- #define EXT4_FEATURE_RO_COMPAT_DIR_NLINK 0x0020
-
- #define EXT3_FEATURE_INCOMPAT_COMPRESSION 0x0001
-@@ -619,6 +626,7 @@ static inline int ext3_valid_inum(struct
- EXT3_FEATURE_INCOMPAT_EXTENTS)
- #define EXT3_FEATURE_RO_COMPAT_SUPP (EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER| \
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
-+ EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
- EXT4_FEATURE_RO_COMPAT_DIR_NLINK| \
- EXT3_FEATURE_RO_COMPAT_BTREE_DIR)
-
-Index: linux-2.6.18-53.1.14/fs/ext3/resize.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/resize.c
-+++ linux-2.6.18-53.1.14/fs/ext3/resize.c
-@@ -18,6 +18,7 @@
- #include <linux/errno.h>
- #include <linux/slab.h>
-
-+#include "group.h"
-
- #define outside(b, first, last) ((b) < (first) || (b) >= (last))
- #define inside(b, first, last) ((b) >= (first) && (b) < (last))
-@@ -137,25 +138,6 @@ static struct buffer_head *bclean(handle
- }
-
- /*
-- * To avoid calling the atomic setbit hundreds or thousands of times, we only
-- * need to use it within a single byte (to ensure we get endianness right).
-- * We can use memset for the rest of the bitmap as there are no other users.
-- */
--static void mark_bitmap_end(int start_bit, int end_bit, char *bitmap)
--{
-- int i;
--
-- if (start_bit >= end_bit)
-- return;
--
-- ext3_debug("mark end bits +%d through +%d used\n", start_bit, end_bit);
-- for (i = start_bit; i < ((start_bit + 7) & ~7UL); i++)
-- ext3_set_bit(i, bitmap);
-- if (i < end_bit)
-- memset(bitmap + (i >> 3), 0xff, (end_bit - i) >> 3);
--}
--
--/*
- * If we have fewer than thresh credits, extend by EXT3_MAX_TRANS_DATA.
- * If that fails, restart the transaction & regain write access for the
- * buffer head which is used for block_bitmap modifications.
-@@ -834,6 +816,7 @@ int ext3_group_add(struct super_block *s
- gdp->bg_inode_table = cpu_to_le32(input->inode_table);
- gdp->bg_free_blocks_count = cpu_to_le16(input->free_blocks_count);
- gdp->bg_free_inodes_count = cpu_to_le16(EXT3_INODES_PER_GROUP(sb));
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi, input->group, gdp);
-
- /*
- * Make the new blocks and inodes valid next. We do this before
-Index: linux-2.6.18-53.1.14/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/super.c
-+++ linux-2.6.18-53.1.14/fs/ext3/super.c
-@@ -41,6 +41,7 @@
- #include "xattr.h"
- #include "acl.h"
- #include "namei.h"
-+#include "group.h"
-
- static int ext3_load_journal(struct super_block *, struct ext3_super_block *,
- unsigned long journal_devnum);
-@@ -1227,6 +1228,91 @@ static int ext3_setup_super(struct super
- return res;
- }
-
-+#if !defined(CONFIG_CRC16) && !defined(CONFIG_CRC16_MODULE)
-+/** CRC table for the CRC-16. The poly is 0x8005 (x^16 + x^15 + x^2 + 1) */
-+__u16 const crc16_table[256] = {
-+ 0x0000, 0xC0C1, 0xC181, 0x0140, 0xC301, 0x03C0, 0x0280, 0xC241,
-+ 0xC601, 0x06C0, 0x0780, 0xC741, 0x0500, 0xC5C1, 0xC481, 0x0440,
-+ 0xCC01, 0x0CC0, 0x0D80, 0xCD41, 0x0F00, 0xCFC1, 0xCE81, 0x0E40,
-+ 0x0A00, 0xCAC1, 0xCB81, 0x0B40, 0xC901, 0x09C0, 0x0880, 0xC841,
-+ 0xD801, 0x18C0, 0x1980, 0xD941, 0x1B00, 0xDBC1, 0xDA81, 0x1A40,
-+ 0x1E00, 0xDEC1, 0xDF81, 0x1F40, 0xDD01, 0x1DC0, 0x1C80, 0xDC41,
-+ 0x1400, 0xD4C1, 0xD581, 0x1540, 0xD701, 0x17C0, 0x1680, 0xD641,
-+ 0xD201, 0x12C0, 0x1380, 0xD341, 0x1100, 0xD1C1, 0xD081, 0x1040,
-+ 0xF001, 0x30C0, 0x3180, 0xF141, 0x3300, 0xF3C1, 0xF281, 0x3240,
-+ 0x3600, 0xF6C1, 0xF781, 0x3740, 0xF501, 0x35C0, 0x3480, 0xF441,
-+ 0x3C00, 0xFCC1, 0xFD81, 0x3D40, 0xFF01, 0x3FC0, 0x3E80, 0xFE41,
-+ 0xFA01, 0x3AC0, 0x3B80, 0xFB41, 0x3900, 0xF9C1, 0xF881, 0x3840,
-+ 0x2800, 0xE8C1, 0xE981, 0x2940, 0xEB01, 0x2BC0, 0x2A80, 0xEA41,
-+ 0xEE01, 0x2EC0, 0x2F80, 0xEF41, 0x2D00, 0xEDC1, 0xEC81, 0x2C40,
-+ 0xE401, 0x24C0, 0x2580, 0xE541, 0x2700, 0xE7C1, 0xE681, 0x2640,
-+ 0x2200, 0xE2C1, 0xE381, 0x2340, 0xE101, 0x21C0, 0x2080, 0xE041,
-+ 0xA001, 0x60C0, 0x6180, 0xA141, 0x6300, 0xA3C1, 0xA281, 0x6240,
-+ 0x6600, 0xA6C1, 0xA781, 0x6740, 0xA501, 0x65C0, 0x6480, 0xA441,
-+ 0x6C00, 0xACC1, 0xAD81, 0x6D40, 0xAF01, 0x6FC0, 0x6E80, 0xAE41,
-+ 0xAA01, 0x6AC0, 0x6B80, 0xAB41, 0x6900, 0xA9C1, 0xA881, 0x6840,
-+ 0x7800, 0xB8C1, 0xB981, 0x7940, 0xBB01, 0x7BC0, 0x7A80, 0xBA41,
-+ 0xBE01, 0x7EC0, 0x7F80, 0xBF41, 0x7D00, 0xBDC1, 0xBC81, 0x7C40,
-+ 0xB401, 0x74C0, 0x7580, 0xB541, 0x7700, 0xB7C1, 0xB681, 0x7640,
-+ 0x7200, 0xB2C1, 0xB381, 0x7340, 0xB101, 0x71C0, 0x7080, 0xB041,
-+ 0x5000, 0x90C1, 0x9181, 0x5140, 0x9301, 0x53C0, 0x5280, 0x9241,
-+ 0x9601, 0x56C0, 0x5780, 0x9741, 0x5500, 0x95C1, 0x9481, 0x5440,
-+ 0x9C01, 0x5CC0, 0x5D80, 0x9D41, 0x5F00, 0x9FC1, 0x9E81, 0x5E40,
-+ 0x5A00, 0x9AC1, 0x9B81, 0x5B40, 0x9901, 0x59C0, 0x5880, 0x9841,
-+ 0x8801, 0x48C0, 0x4980, 0x8941, 0x4B00, 0x8BC1, 0x8A81, 0x4A40,
-+ 0x4E00, 0x8EC1, 0x8F81, 0x4F40, 0x8D01, 0x4DC0, 0x4C80, 0x8C41,
-+ 0x4400, 0x84C1, 0x8581, 0x4540, 0x8701, 0x47C0, 0x4680, 0x8641,
-+ 0x8201, 0x42C0, 0x4380, 0x8341, 0x4100, 0x81C1, 0x8081, 0x4040
-+};
-+
-+static inline __u16 crc16_byte(__u16 crc, const __u8 data)
-+{
-+ return (crc >> 8) ^ crc16_table[(crc ^ data) & 0xff];
-+}
-+
-+__u16 crc16(__u16 crc, __u8 const *buffer, size_t len)
-+{
-+ while (len--)
-+ crc = crc16_byte(crc, *buffer++);
-+ return crc;
-+}
-+#endif
-+
-+__le16 ext3_group_desc_csum(struct ext3_sb_info *sbi, __u32 block_group,
-+ struct ext3_group_desc *gdp)
-+{
-+ __u16 crc = 0;
-+
-+ if (sbi->s_es->s_feature_ro_compat &
-+ cpu_to_le32(EXT4_FEATURE_RO_COMPAT_GDT_CSUM)) {
-+ int offset = offsetof(struct ext3_group_desc, bg_checksum);
-+ __le32 le_group = cpu_to_le32(block_group);
-+
-+ crc = crc16(~0, sbi->s_es->s_uuid, sizeof(sbi->s_es->s_uuid));
-+ crc = crc16(crc, (__u8 *)&le_group, sizeof(le_group));
-+ crc = crc16(crc, (__u8 *)gdp, offset);
-+ offset += sizeof(gdp->bg_checksum); /* skip checksum */
-+ BUG_ON(offset != sizeof(*gdp)); /* XXX handle s_desc_size */
-+ /* for checksum of struct ext4_group_desc do the rest...
-+ if ((sbi->s_es->s_feature_incompat &
-+ cpu_to_le32(EXT4_FEATURE_INCOMPAT_64BIT)) &&
-+ offset < le16_to_cpu(sbi->s_es->s_desc_size)) {
-+ crc = crc16(crc, (__u8 *)gdp + offset,
-+ le16_to_cpu(sbi->s_es->s_desc_size) -
-+ offset);
-+ */
-+ }
-+
-+ return cpu_to_le16(crc);
-+}
-+
-+int ext3_group_desc_csum_verify(struct ext3_sb_info *sbi, __u32 block_group,
-+ struct ext3_group_desc *gdp)
-+{
-+ return (gdp->bg_checksum ==
-+ ext3_group_desc_csum(sbi, block_group, gdp));
-+}
-+
- /* Called at mount-time, super-block is locked */
- static int ext3_check_descriptors (struct super_block * sb)
- {
-@@ -1281,6 +1367,13 @@ static int ext3_check_descriptors (struc
- le32_to_cpu(gdp->bg_inode_table));
- return 0;
- }
-+ if (!ext3_group_desc_csum_verify(sbi, i, gdp)) {
-+ ext3_error(sb, __FUNCTION__,
-+ "Checksum for group %d failed (%u!=%u)\n", i,
-+ le16_to_cpu(ext3_group_desc_csum(sbi,i,gdp)),
-+ le16_to_cpu(gdp->bg_checksum));
-+ return 0;
-+ }
- first_block += EXT3_BLOCKS_PER_GROUP(sb);
- gdp++;
- }
-Index: linux-2.6.18-53.1.14/fs/ext3/group.h
-===================================================================
---- /dev/null
-+++ linux-2.6.18-53.1.14/fs/ext3/group.h
-@@ -0,0 +1,30 @@
-+/*
-+ * linux/fs/ext3/group.h
-+ *
-+ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
-+ *
-+ * Author: Andreas Dilger <adilger@clusterfs.com>
-+ */
-+
-+#ifndef _LINUX_EXT3_GROUP_H
-+#define _LINUX_EXT3_GROUP_H
-+#if defined(CONFIG_CRC16) || defined(CONFIG_CRC16_MODULE)
-+#include <linux/crc16.h>
-+#endif
-+
-+extern __le16 ext3_group_desc_csum(struct ext3_sb_info *sbi, __u32 group,
-+ struct ext3_group_desc *gdp);
-+extern int ext3_group_desc_csum_verify(struct ext3_sb_info *sbi, __u32 group,
-+ struct ext3_group_desc *gdp);
-+struct buffer_head *read_block_bitmap(struct super_block *sb,
-+ unsigned int block_group);
-+extern unsigned ext3_init_block_bitmap(struct super_block *sb,
-+ struct buffer_head *bh, int group,
-+ struct ext3_group_desc *desc);
-+#define ext3_free_blocks_after_init(sb, group, desc) \
-+ ext3_init_block_bitmap(sb, NULL, group, desc)
-+extern unsigned ext3_init_inode_bitmap(struct super_block *sb,
-+ struct buffer_head *bh, int group,
-+ struct ext3_group_desc *desc);
-+extern void mark_bitmap_end(int start_bit, int end_bit, char *bitmap);
-+#endif /* _LINUX_EXT3_GROUP_H */
-Index: linux-2.6.18-53.1.14/fs/ext3/ialloc.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/ialloc.c
-+++ linux-2.6.18-53.1.14/fs/ext3/ialloc.c
-@@ -28,6 +28,7 @@
-
- #include "xattr.h"
- #include "acl.h"
-+#include "group.h"
-
- /*
- * ialloc.c contains the inodes allocation and deallocation routines
-@@ -43,6 +44,52 @@
- * the free blocks count in the block.
- */
-
-+/*
-+ * To avoid calling the atomic setbit hundreds or thousands of times, we only
-+ * need to use it within a single byte (to ensure we get endianness right).
-+ * We can use memset for the rest of the bitmap as there are no other users.
-+ */
-+void mark_bitmap_end(int start_bit, int end_bit, char *bitmap)
-+{
-+ int i;
-+
-+ if (start_bit >= end_bit)
-+ return;
-+
-+ ext3_debug("mark end bits +%d through +%d used\n", start_bit, end_bit);
-+ for (i = start_bit; i < ((start_bit + 7) & ~7UL); i++)
-+ ext3_set_bit(i, bitmap);
-+ if (i < end_bit)
-+ memset(bitmap + (i >> 3), 0xff, (end_bit - i) >> 3);
-+}
-+
-+/* Initializes an uninitialized inode bitmap */
-+unsigned ext3_init_inode_bitmap(struct super_block *sb,
-+ struct buffer_head *bh, int block_group,
-+ struct ext3_group_desc *gdp)
-+{
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+
-+ J_ASSERT_BH(bh, buffer_locked(bh));
-+
-+ /* If checksum is bad mark all blocks and inodes used to prevent
-+ * allocation, essentially implementing a per-group read-only flag. */
-+ if (!ext3_group_desc_csum_verify(sbi, block_group, gdp)) {
-+ ext3_error(sb, __FUNCTION__, "Checksum bad for group %u\n",
-+ block_group);
-+ gdp->bg_free_blocks_count = 0;
-+ gdp->bg_free_inodes_count = 0;
-+ gdp->bg_itable_unused = 0;
-+ memset(bh->b_data, 0xff, sb->s_blocksize);
-+ return 0;
-+ }
-+
-+ memset(bh->b_data, 0, (EXT3_INODES_PER_GROUP(sb) + 7) / 8);
-+ mark_bitmap_end(EXT3_INODES_PER_GROUP(sb), EXT3_BLOCKS_PER_GROUP(sb),
-+ bh->b_data);
-+
-+ return EXT3_INODES_PER_GROUP(sb);
-+}
-
- /*
- * Read the inode allocation bitmap for a given block_group, reading
-@@ -59,8 +106,19 @@ read_inode_bitmap(struct super_block * s
- desc = ext3_get_group_desc(sb, block_group, NULL);
- if (!desc)
- goto error_out;
--
-- bh = sb_bread(sb, le32_to_cpu(desc->bg_inode_bitmap));
-+ if (desc->bg_flags & cpu_to_le16(EXT3_BG_INODE_UNINIT)) {
-+ bh = sb_getblk(sb, le32_to_cpu(desc->bg_inode_bitmap));
-+ if (!buffer_uptodate(bh)) {
-+ lock_buffer(bh);
-+ if (!buffer_uptodate(bh)) {
-+ ext3_init_inode_bitmap(sb, bh,block_group,desc);
-+ set_buffer_uptodate(bh);
-+ }
-+ unlock_buffer(bh);
-+ }
-+ } else {
-+ bh = sb_bread(sb, le32_to_cpu(desc->bg_inode_bitmap));
-+ }
- if (!bh)
- ext3_error(sb, "read_inode_bitmap",
- "Cannot read inode bitmap - "
-@@ -169,6 +227,8 @@ void ext3_free_inode (handle_t *handle,
- if (is_directory)
- gdp->bg_used_dirs_count = cpu_to_le16(
- le16_to_cpu(gdp->bg_used_dirs_count) - 1);
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi,block_group,
-+ gdp);
- spin_unlock(sb_bgl_lock(sbi, block_group));
- percpu_counter_inc(&sbi->s_freeinodes_counter);
- if (is_directory)
-@@ -454,7 +514,7 @@ struct inode *ext3_new_inode(handle_t *h
- struct ext3_sb_info *sbi;
- int err = 0;
- struct inode *ret;
-- int i;
-+ int i, free = 0;
-
- /* Cannot create files in a deleted directory */
- if (!dir || !dir->i_nlink)
-@@ -571,11 +631,13 @@ repeat_in_this_group:
- goto out;
-
- got:
-- ino += group * EXT3_INODES_PER_GROUP(sb) + 1;
-- if (ino < EXT3_FIRST_INO(sb) || ino > le32_to_cpu(es->s_inodes_count)) {
-- ext3_error (sb, "ext3_new_inode",
-- "reserved inode or inode > inodes count - "
-- "block_group = %d, inode=%lu", group, ino);
-+ ino++;
-+ if ((group == 0 && ino < EXT3_FIRST_INO(sb)) ||
-+ ino > EXT3_INODES_PER_GROUP(sb)) {
-+ ext3_error(sb, __FUNCTION__,
-+ "reserved inode or inode > inodes count - "
-+ "block_group = %d, inode=%lu", group,
-+ ino + group * EXT3_INODES_PER_GROUP(sb));
- err = -EIO;
- goto fail;
- }
-@@ -583,13 +645,64 @@ got:
- BUFFER_TRACE(bh2, "get_write_access");
- err = ext3_journal_get_write_access(handle, bh2);
- if (err) goto fail;
-+
-+ /* We may have to initialize the block bitmap if it isn't already */
-+ if (EXT3_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_GDT_CSUM) &&
-+ gdp->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-+ struct buffer_head *block_bh = read_block_bitmap(sb, group);
-+
-+ BUFFER_TRACE(block_bh, "get block bitmap access");
-+ err = ext3_journal_get_write_access(handle, block_bh);
-+ if (err) {
-+ brelse(block_bh);
-+ goto fail;
-+ }
-+
-+ free = 0;
-+ spin_lock(sb_bgl_lock(sbi, group));
-+ /* recheck and clear flag under lock if we still need to */
-+ if (gdp->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-+ gdp->bg_flags &= cpu_to_le16(~EXT3_BG_BLOCK_UNINIT);
-+ free = ext3_free_blocks_after_init(sb, group, gdp);
-+ gdp->bg_free_blocks_count = cpu_to_le16(free);
-+ }
-+ spin_unlock(sb_bgl_lock(sbi, group));
-+
-+ /* Don't need to dirty bitmap block if we didn't change it */
-+ if (free) {
-+ BUFFER_TRACE(block_bh, "dirty block bitmap");
-+ err = ext3_journal_dirty_metadata(handle, block_bh);
-+ }
-+
-+ brelse(block_bh);
-+ if (err)
-+ goto fail;
-+ }
-+
- spin_lock(sb_bgl_lock(sbi, group));
-+ /* If we didn't allocate from within the initialized part of the inode
-+ * table then we need to initialize up to this inode. */
-+ if (EXT3_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_GDT_CSUM)) {
-+ if (gdp->bg_flags & cpu_to_le16(EXT3_BG_INODE_UNINIT)) {
-+ gdp->bg_flags &= cpu_to_le16(~EXT3_BG_INODE_UNINIT);
-+ free = 0;
-+ } else {
-+ free = EXT3_INODES_PER_GROUP(sb) -
-+ le16_to_cpu(gdp->bg_itable_unused);
-+ }
-+
-+ if (ino > free)
-+ gdp->bg_itable_unused =
-+ cpu_to_le16(EXT3_INODES_PER_GROUP(sb) - ino);
-+ }
-+
- gdp->bg_free_inodes_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_free_inodes_count) - 1);
- if (S_ISDIR(mode)) {
- gdp->bg_used_dirs_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_used_dirs_count) + 1);
- }
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi, group, gdp);
- spin_unlock(sb_bgl_lock(sbi, group));
- BUFFER_TRACE(bh2, "call ext3_journal_dirty_metadata");
- err = ext3_journal_dirty_metadata(handle, bh2);
-@@ -611,7 +724,7 @@ got:
- inode->i_gid = current->fsgid;
- inode->i_mode = mode;
-
-- inode->i_ino = ino;
-+ inode->i_ino = ino + group * EXT3_INODES_PER_GROUP(sb);
- /* This is the optimal IO size (for stat), not the fs block size */
- inode->i_blocks = 0;
- inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC;
-Index: linux-2.6.18-53.1.14/fs/ext3/mballoc.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/mballoc.c
-+++ linux-2.6.18-53.1.14/fs/ext3/mballoc.c
-@@ -36,6 +36,8 @@
- #include <linux/seq_file.h>
- #include <linux/version.h>
-
-+#include "group.h"
-+
- /*
- * MUSTDO:
- * - test ext3_ext_search_left() and ext3_ext_search_right()
-@@ -323,6 +325,7 @@ struct ext3_group_info {
- unsigned long bb_state;
- unsigned long bb_tid;
- struct ext3_free_metadata *bb_md_cur;
-+ struct ext3_group_desc *bb_gdp;
- unsigned short bb_first_free;
- unsigned short bb_free;
- unsigned short bb_fragments;
-@@ -941,10 +944,7 @@ static int ext3_mb_init_cache(struct pag
- if (first_group + i >= EXT3_SB(sb)->s_groups_count)
- break;
-
-- err = -EIO;
-- desc = ext3_get_group_desc(sb, first_group + i, NULL);
-- if (desc == NULL)
-- goto out;
-+ desc = EXT3_GROUP_INFO(sb, first_group + i)->bb_gdp;
-
- err = -ENOMEM;
- bh[i] = sb_getblk(sb, le32_to_cpu(desc->bg_block_bitmap));
-@@ -959,7 +959,12 @@ static int ext3_mb_init_cache(struct pag
- unlock_buffer(bh[i]);
- continue;
- }
--
-+ if (desc->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-+ ext3_init_block_bitmap(sb, bh[i], first_group + i,desc);
-+ set_buffer_uptodate(bh[i]);
-+ unlock_buffer(bh[i]);
-+ continue;
-+ }
- get_bh(bh[i]);
- bh[i]->b_end_io = end_buffer_read_sync;
- submit_bh(READ, bh[i]);
-@@ -1731,6 +1736,10 @@ static int ext3_mb_good_group(struct ext
- switch (cr) {
- case 0:
- BUG_ON(ac->ac_2order == 0);
-+ /* If this group is uninitialized, skip it initially */
-+ if (grp->bb_gdp->bg_flags &
-+ cpu_to_le16(EXT3_BG_BLOCK_UNINIT))
-+ return 0;
- bits = ac->ac_sb->s_blocksize_bits + 1;
- for (i = ac->ac_2order; i <= bits; i++)
- if (grp->bb_counters[i] > 0)
-@@ -1824,7 +1833,9 @@ repeat:
- }
-
- ac->ac_groups_scanned++;
-- if (cr == 0)
-+ if (cr == 0 || (e3b.bd_info->bb_gdp->bg_flags &
-+ cpu_to_le16(EXT3_BG_BLOCK_UNINIT) &&
-+ ac->ac_2order != 0))
- ext3_mb_simple_scan_group(ac, &e3b);
- else if (cr == 1 && ac->ac_g_ex.fe_len == sbi->s_stripe)
- ext3_mb_scan_aligned(ac, &e3b);
-@@ -2304,12 +2315,13 @@ int ext3_mb_init_backend(struct super_bl
- i--;
- goto err_freebuddy;
- }
-+ memset(meta_group_info[j], 0, len);
- desc = ext3_get_group_desc(sb, i, NULL);
-+ meta_group_info[j]->bb_gdp = desc;
- if (desc == NULL) {
- printk(KERN_ERR"EXT3-fs: can't read descriptor %u\n",i);
- goto err_freebuddy;
- }
-- memset(meta_group_info[j], 0, len);
- set_bit(EXT3_GROUP_INFO_NEED_INIT_BIT,
- &meta_group_info[j]->bb_state);
-
-@@ -2943,9 +2955,17 @@ int ext3_mb_mark_diskspace_used(struct e
- mb_set_bits(NULL, bitmap_bh->b_data,
- ac->ac_b_ex.fe_start, ac->ac_b_ex.fe_len);
-
-+ if (gdp->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-+ gdp->bg_flags &= cpu_to_le16(~EXT3_BG_BLOCK_UNINIT);
-+ gdp->bg_free_blocks_count =
-+ cpu_to_le16(ext3_free_blocks_after_init(sb,
-+ ac->ac_b_ex.fe_group,
-+ gdp));
-+ }
- gdp->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count)
- - ac->ac_b_ex.fe_len);
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi, ac->ac_b_ex.fe_group, gdp);
- spin_unlock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
- percpu_counter_mod(&sbi->s_freeblocks_counter, - ac->ac_b_ex.fe_len);
-
-@@ -4355,6 +4375,7 @@ do_more:
- spin_lock(sb_bgl_lock(sbi, block_group));
- gdp->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) + count);
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi, block_group, gdp);
- spin_unlock(sb_bgl_lock(sbi, block_group));
- percpu_counter_mod(&sbi->s_freeblocks_counter, count);
-
-Index: linux-2.6.18-53.1.14/fs/ext3/balloc.c
-===================================================================
---- linux-2.6.18-53.1.14.orig/fs/ext3/balloc.c
-+++ linux-2.6.18-53.1.14/fs/ext3/balloc.c
-@@ -20,6 +20,7 @@
- #include <linux/quotaops.h>
- #include <linux/buffer_head.h>
-
-+#include "group.h"
- /*
- * balloc.c contains the blocks allocation and deallocation routines
- */
-@@ -73,6 +74,83 @@ struct ext3_group_desc * ext3_get_group_
- return desc + offset;
- }
-
-+/* Initializes an uninitialized block bitmap if given, and returns the
-+ * number of blocks free in the group. */
-+unsigned ext3_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
-+ int block_group, struct ext3_group_desc *gdp)
-+{
-+ unsigned long start;
-+ int bit, bit_max;
-+ unsigned free_blocks;
-+ struct ext3_sb_info *sbi = EXT3_SB(sb);
-+
-+ if (bh) {
-+ J_ASSERT_BH(bh, buffer_locked(bh));
-+
-+ /* If checksum is bad mark all blocks use to prevent allocation,
-+ * essentially implementing a per-group read-only flag. */
-+ if (!ext3_group_desc_csum_verify(sbi, block_group, gdp)) {
-+ ext3_error(sb, __FUNCTION__,
-+ "Checksum bad for group %u\n", block_group);
-+ gdp->bg_free_blocks_count = 0;
-+ gdp->bg_free_inodes_count = 0;
-+ gdp->bg_itable_unused = 0;
-+ memset(bh->b_data, 0xff, sb->s_blocksize);
-+ return 0;
-+ }
-+ memset(bh->b_data, 0, sb->s_blocksize);
-+ }
-+
-+ /* Check for superblock and gdt backups in this group */
-+ bit_max = ext3_bg_has_super(sb, block_group);
-+
-+ if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_META_BG) ||
-+ block_group < le32_to_cpu(sbi->s_es->s_first_meta_bg) *
-+ sbi->s_desc_per_block) {
-+ if (bit_max) {
-+ bit_max += ext3_bg_num_gdb(sb, block_group);
-+ bit_max +=le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks);
-+ }
-+ } else { /* For META_BG_BLOCK_GROUPS */
-+ int group_rel = (block_group -
-+ le32_to_cpu(sbi->s_es->s_first_meta_bg)) %
-+ EXT3_DESC_PER_BLOCK(sb);
-+ if (group_rel == 0 || group_rel == 1 ||
-+ (group_rel == EXT3_DESC_PER_BLOCK(sb) - 1))
-+ bit_max += 1;
-+ }
-+
-+ /* Last and first groups are always initialized */
-+ free_blocks = EXT3_BLOCKS_PER_GROUP(sb) - bit_max;
-+
-+ if (bh) {
-+ for (bit = 0; bit < bit_max; bit++)
-+ ext3_set_bit(bit, bh->b_data);
-+
-+ start = block_group * EXT3_BLOCKS_PER_GROUP(sb) +
-+ le32_to_cpu(sbi->s_es->s_first_data_block);
-+
-+ /* Set bits for block and inode bitmaps, and inode table */
-+ ext3_set_bit(le32_to_cpu(gdp->bg_block_bitmap) - start,
-+ bh->b_data);
-+ ext3_set_bit(le32_to_cpu(gdp->bg_inode_bitmap) - start,
-+ bh->b_data);
-+ for (bit = le32_to_cpu(gdp->bg_inode_table) - start,
-+ bit_max = bit + sbi->s_itb_per_group; bit < bit_max; bit++)
-+ ext3_set_bit(bit, bh->b_data);
-+
-+ /*
-+ * Also if the number of blocks within the group is
-+ * less than the blocksize * 8 ( which is the size
-+ * of bitmap ), set rest of the block bitmap to 1
-+ */
-+ mark_bitmap_end(EXT3_BLOCKS_PER_GROUP(sb), sb->s_blocksize * 8,
-+ bh->b_data);
-+ }
-+
-+ return free_blocks - sbi->s_itb_per_group - 2;
-+}
-+
- /*
- * Read the bitmap for a given block_group, reading into the specified
- * slot in the superblock's bitmap cache.
-@@ -88,7 +166,19 @@ read_block_bitmap(struct super_block *sb
- desc = ext3_get_group_desc (sb, block_group, NULL);
- if (!desc)
- goto error_out;
-- bh = sb_bread(sb, le32_to_cpu(desc->bg_block_bitmap));
-+ if (desc->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT)) {
-+ bh = sb_getblk(sb, le32_to_cpu(desc->bg_block_bitmap));
-+ if (!buffer_uptodate(bh)) {
-+ lock_buffer(bh);
-+ if (!buffer_uptodate(bh)) {
-+ ext3_init_block_bitmap(sb, bh,block_group,desc);
-+ set_buffer_uptodate(bh);
-+ }
-+ unlock_buffer(bh);
-+ }
-+ } else {
-+ bh = sb_bread(sb, le32_to_cpu(desc->bg_block_bitmap));
-+ }
- if (!bh)
- ext3_error (sb, "read_block_bitmap",
- "Cannot read block bitmap - "
-@@ -467,6 +557,7 @@ do_more:
- desc->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(desc->bg_free_blocks_count) +
- group_freed);
-+ desc->bg_checksum = ext3_group_desc_csum(sbi, block_group, desc);
- spin_unlock(sb_bgl_lock(sbi, block_group));
- percpu_counter_mod(&sbi->s_freeblocks_counter, count);
-
-@@ -1434,8 +1525,11 @@ allocated:
- ret_block, goal_hits, goal_attempts);
-
- spin_lock(sb_bgl_lock(sbi, group_no));
-+ if (gdp->bg_flags & cpu_to_le16(EXT3_BG_BLOCK_UNINIT))
-+ gdp->bg_flags &= cpu_to_le16(~EXT3_BG_BLOCK_UNINIT);
- gdp->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) - num);
-+ gdp->bg_checksum = ext3_group_desc_csum(sbi, group_no, gdp);
- spin_unlock(sb_bgl_lock(sbi, group_no));
- percpu_counter_mod(&sbi->s_freeblocks_counter, -num);
-
+++ /dev/null
-Index: linux-2.6.18-128.1.6/fs/ext3/super.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext3/super.c 2009-04-14 23:05:45.000000000 -0400
-+++ linux-2.6.18-128.1.6/fs/ext3/super.c 2009-07-24 01:30:55.000000000 -0400
-@@ -2785,6 +2785,9 @@ static int __init init_ext3_fs(void)
- err = register_filesystem(&ext3_fs_type);
- if (err)
- goto out;
-+
-+ printk(KERN_INFO "ldiskfs created from ""ext""3-2.6-rhel5\n");
-+
- return 0;
- out:
- destroy_inodecache();
---- /dev/null 2009-09-21 17:11:24.467285554 +0800
-+++ linux-2.6.27.21-0.1/fs/ext3/fiemap.h
-@@ -0,0 +1,2 @@
-+
-+#include_next <fiemap.h>
+++ /dev/null
-Index: linux-2.6.9/fs/ext3/ialloc.c
-===================================================================
---- linux-2.6.9.orig/fs/ext3/ialloc.c 2007-03-13 00:47:22.000000000 +0100
-+++ linux-2.6.9/fs/ext3/ialloc.c 2007-07-26 09:23:37.000000000 +0200
-@@ -419,7 +419,8 @@ static int find_group_other(struct super
- * For other inodes, search forward from the parent directory's block
- * group to find a free inode.
- */
--struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, int mode)
-+struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, int mode,
-+ unsigned long goal)
- {
- struct super_block *sb;
- struct buffer_head *bitmap_bh = NULL;
-@@ -447,6 +448,41 @@ struct inode *ext3_new_inode(handle_t *h
-
- sbi = EXT3_SB(sb);
- es = sbi->s_es;
-+ if (goal) {
-+ group = (goal - 1) / EXT3_INODES_PER_GROUP(sb);
-+ ino = (goal - 1) % EXT3_INODES_PER_GROUP(sb);
-+ err = -EIO;
-+
-+ gdp = ext3_get_group_desc(sb, group, &bh2);
-+ if (!gdp)
-+ goto fail;
-+
-+ bitmap_bh = read_inode_bitmap (sb, group);
-+ if (!bitmap_bh)
-+ goto fail;
-+
-+ BUFFER_TRACE(bh, "get_write_access");
-+ err = ext3_journal_get_write_access(handle, bitmap_bh);
-+ if (err) goto fail;
-+
-+ if (ext3_set_bit_atomic(sb_bgl_lock(sbi, group),
-+ ino, bitmap_bh->b_data)) {
-+ printk(KERN_ERR "goal inode %lu unavailable\n", goal);
-+ /* Oh well, we tried. */
-+ goto continue_allocation;
-+ }
-+
-+ BUFFER_TRACE(bh, "call ext3_journal_dirty_metadata");
-+ err = ext3_journal_dirty_metadata(handle, bitmap_bh);
-+ if (err) goto fail;
-+
-+ /* We've shortcircuited the allocation system successfully,
-+ * now finish filling in the inode.
-+ */
-+ goto got;
-+ }
-+
-+continue_allocation:
- if (S_ISDIR(mode)) {
- if (test_opt (sb, OLDALLOC))
- group = find_group_dir(sb, dir);
-Index: linux-2.6.9/fs/ext3/ioctl.c
-===================================================================
---- linux-2.6.9.orig/fs/ext3/ioctl.c 2007-03-13 00:47:22.000000000 +0100
-+++ linux-2.6.9/fs/ext3/ioctl.c 2007-07-26 09:25:24.000000000 +0200
-@@ -13,6 +13,7 @@
- #include <linux/ext3_jbd.h>
- #include <linux/time.h>
- #include <asm/uaccess.h>
-+#include <linux/namei.h>
-
-
- int ext3_ioctl (struct inode * inode, struct file * filp, unsigned int cmd,
-@@ -25,6 +26,31 @@ int ext3_ioctl (struct inode * inode, st
- ext3_debug ("cmd = %u, arg = %lu\n", cmd, arg);
-
- switch (cmd) {
-+ case EXT3_IOC_CREATE_INUM: {
-+ char name[32];
-+ struct dentry *dchild, *dparent;
-+ int rc = 0;
-+
-+ dparent = list_entry(inode->i_dentry.next, struct dentry,
-+ d_alias);
-+ snprintf(name, sizeof name, "%lu", arg);
-+ dchild = lookup_one_len(name, dparent, strlen(name));
-+ if (dchild->d_inode) {
-+ printk(KERN_ERR "%*s/%lu already exists (ino %lu)\n",
-+ dparent->d_name.len, dparent->d_name.name, arg,
-+ dchild->d_inode->i_ino);
-+ rc = -EEXIST;
-+ } else {
-+ dchild->d_fsdata = (void *)arg;
-+ rc = vfs_create(inode, dchild, 0644, NULL);
-+ if (rc)
-+ printk(KERN_ERR "vfs_create: %d\n", rc);
-+ else if (dchild->d_inode->i_ino != arg)
-+ rc = -EEXIST;
-+ }
-+ dput(dchild);
-+ return rc;
-+ }
- case EXT3_IOC_GETFLAGS:
- flags = ei->i_flags & EXT3_FL_USER_VISIBLE;
- return put_user(flags, (int __user *) arg);
-Index: linux-2.6.9/fs/ext3/namei.c
-===================================================================
---- linux-2.6.9.orig/fs/ext3/namei.c 2007-03-13 00:47:27.000000000 +0100
-+++ linux-2.6.9/fs/ext3/namei.c 2007-07-26 09:23:37.000000000 +0200
-@@ -97,6 +97,7 @@ struct dx_entry
- __le32 block;
- };
-
-+
- /*
- * dx_root_info is laid out so that if it should somehow get overlaid by a
- * dirent the two low bits of the hash version will be zero. Therefore, the
-@@ -141,6 +142,14 @@ struct dx_map_entry
- u32 offs;
- };
-
-+#define LVFS_DENTRY_PARAM_MAGIC 20070216UL
-+struct lvfs_dentry_params
-+{
-+ unsigned long p_inum;
-+ void *p_ptr;
-+ u32 magic;
-+};
-+
- #ifdef CONFIG_EXT3_INDEX
- static inline unsigned dx_get_block (struct dx_entry *entry);
- static void dx_set_block (struct dx_entry *entry, unsigned value);
-@@ -1633,6 +1642,20 @@ static int ext3_add_nondir(handle_t *han
- return err;
- }
-
-+static struct inode * ext3_new_inode_wantedi(handle_t *handle, struct inode *dir,
-+ int mode, struct dentry *dentry)
-+{
-+ unsigned long inum = 0;
-+
-+ if (dentry->d_fsdata != NULL) {
-+ struct lvfs_dentry_params *param = dentry->d_fsdata;
-+
-+ if (param->magic == LVFS_DENTRY_PARAM_MAGIC)
-+ inum = param->p_inum;
-+ }
-+ return ext3_new_inode(handle, dir, mode, inum);
-+}
-+
- /*
- * By the time this is called, we already have created
- * the directory cache entry for the new file, but it
-@@ -1658,7 +1681,7 @@ retry:
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode = ext3_new_inode (handle, dir, mode);
-+ inode = ext3_new_inode_wantedi (handle, dir, mode, dentry);
- err = PTR_ERR(inode);
- if (!IS_ERR(inode)) {
- inode->i_op = &ext3_file_inode_operations;
-@@ -1692,7 +1715,7 @@ retry:
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode = ext3_new_inode (handle, dir, mode);
-+ inode = ext3_new_inode_wantedi (handle, dir, mode, dentry);
- err = PTR_ERR(inode);
- if (!IS_ERR(inode)) {
- init_special_inode(inode, inode->i_mode, rdev);
-@@ -1728,7 +1751,7 @@ retry:
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode = ext3_new_inode (handle, dir, S_IFDIR | mode);
-+ inode = ext3_new_inode_wantedi (handle, dir, S_IFDIR | mode, dentry);
- err = PTR_ERR(inode);
- if (IS_ERR(inode))
- goto out_stop;
-@@ -2133,7 +2156,7 @@ retry:
- if (IS_DIRSYNC(dir))
- handle->h_sync = 1;
-
-- inode = ext3_new_inode (handle, dir, S_IFLNK|S_IRWXUGO);
-+ inode = ext3_new_inode_wantedi (handle, dir, S_IFLNK|S_IRWXUGO, dentry);
- err = PTR_ERR(inode);
- if (IS_ERR(inode))
- goto out_stop;
-Index: linux-2.6.9/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.9.orig/include/linux/ext3_fs.h 2007-03-13 00:47:22.000000000 +0100
-+++ linux-2.6.9/include/linux/ext3_fs.h 2007-07-26 09:23:37.000000000 +0200
-@@ -741,7 +741,8 @@ extern int ext3fs_dirhash(const char *na
- dx_hash_info *hinfo);
-
- /* ialloc.c */
--extern struct inode * ext3_new_inode (handle_t *, struct inode *, int);
-+extern struct inode * ext3_new_inode (handle_t *, struct inode *, int,
-+ unsigned long);
- extern void ext3_free_inode (handle_t *, struct inode *);
- extern struct inode * ext3_orphan_get (struct super_block *, unsigned long);
- extern unsigned long ext3_count_free_inodes (struct super_block *);
-@@ -833,4 +834,6 @@ extern struct inode_operations ext3_fast
-
- #endif /* __KERNEL__ */
-
-+/* EXT3_IOC_CREATE_INUM at bottom of file (visible to kernel and user). */
-+#define EXT3_IOC_CREATE_INUM _IOW('f', 5, long)
- #endif /* _LINUX_EXT3_FS_H */
+++ /dev/null
-Index: linux-2.6.22.14/include/linux/ext3_fs.h
-===================================================================
---- linux-2.6.22.14.orig/include/linux/ext3_fs.h
-+++ linux-2.6.22.14/include/linux/ext3_fs.h
-@@ -923,6 +923,13 @@ struct mmp_struct {
- #define EXT3_MMP_MIN_CHECK_INTERVAL 5
-
- /*
-+ * Indicates that ctime should not be updated in ext3_xattr_set_handle()
-+ */
-+#ifndef XATTR_NO_CTIME
-+#define XATTR_NO_CTIME 0x80
-+#endif
-+
-+/*
- * Function prototypes
- */
-
-Index: linux-2.6.22.14/fs/ext3/xattr.c
-===================================================================
---- linux-2.6.22.14.orig/fs/ext3/xattr.c
-+++ linux-2.6.22.14/fs/ext3/xattr.c
-@@ -1039,7 +1039,8 @@ ext3_xattr_set_handle(handle_t *handle,
- }
- if (!error) {
- ext3_xattr_update_super_block(handle, inode->i_sb);
-- inode->i_ctime = ext3_current_time(inode);
-+ if (!(flags & XATTR_NO_CTIME))
-+ inode->i_ctime = ext3_current_time(inode);
- error = ext3_mark_iloc_dirty(handle, inode, &is.iloc);
- if (!value)
- EXT3_I(inode)->i_state &= ~EXT3_STATE_NO_EXPAND;
+++ /dev/null
-this patch implements feature which allows ext3 fs users (e.g. Lustre)
-to store data in ext3 dirent.
-data is stored in ext3 dirent after file-name, this space is accounted
-in de->rec_len. flag EXT3_DIRENT_LUFID added to d_type if extra data
-is present.
-
-make use of dentry->d_fsdata to pass fid to ext3. so no
-changes in ext3_add_entry() interface required.
-
-Index: b/fs/ext3/namei.c
-===================================================================
---- a/fs/ext3/namei.c
-+++ b/fs/ext3/namei.c
-@@ -174,7 +174,8 @@ static unsigned dx_get_count (struct dx_
- static unsigned dx_get_limit (struct dx_entry *entries);
- static void dx_set_count (struct dx_entry *entries, unsigned value);
- static void dx_set_limit (struct dx_entry *entries, unsigned value);
--static unsigned dx_root_limit (struct inode *dir, unsigned infosize);
-+static unsigned dx_root_limit (__u32 blocksize, struct ext3_dir_entry_2*,
-+ unsigned infosize);
- static unsigned dx_node_limit (struct inode *dir);
- static struct dx_frame *dx_probe(struct dentry *dentry,
- struct inode *dir,
-@@ -204,11 +205,13 @@ static int ext3_dx_add_entry(handle_t *h
- */
- struct dx_root_info * dx_get_dx_info(struct ext3_dir_entry_2 *de)
- {
-+ BUG_ON(de->name_len != 1);
-+
- /* get dotdot first */
- de = (struct ext3_dir_entry_2 *)((char *)de + le16_to_cpu(de->rec_len));
-
- /* dx root info is after dotdot entry */
-- de = (struct ext3_dir_entry_2 *)((char *)de + EXT3_DIR_REC_LEN(2));
-+ de = (struct ext3_dir_entry_2 *)((char *)de + EXT3_DIR_REC_LEN(de));
-
- return (struct dx_root_info *) de;
- }
-@@ -253,16 +256,24 @@ static inline void dx_set_limit (struct
- ((struct dx_countlimit *) entries)->limit = cpu_to_le16(value);
- }
-
--static inline unsigned dx_root_limit (struct inode *dir, unsigned infosize)
-+static inline unsigned dx_root_limit (__u32 blocksize,
-+ struct ext3_dir_entry_2 * dot_de, unsigned infosize)
- {
-- unsigned entry_space = dir->i_sb->s_blocksize - EXT3_DIR_REC_LEN(1) -
-- EXT3_DIR_REC_LEN(2) - infosize;
-- return 0? 20: entry_space / sizeof(struct dx_entry);
-+ struct ext3_dir_entry_2 * dotdot_de;
-+ unsigned entry_space;
-+
-+ BUG_ON(dot_de->name_len != 1);
-+ dotdot_de = (struct ext3_dir_entry_2 *)((char *)dot_de
-+ + le16_to_cpu(dot_de->rec_len));
-+ entry_space = blocksize - EXT3_DIR_REC_LEN(dot_de) -
-+ EXT3_DIR_REC_LEN(dotdot_de) - infosize;
-+
-+ return entry_space / sizeof(struct dx_entry);
- }
-
- static inline unsigned dx_node_limit (struct inode *dir)
- {
-- unsigned entry_space = dir->i_sb->s_blocksize - EXT3_DIR_REC_LEN(0);
-+ unsigned entry_space = dir->i_sb->s_blocksize - __EXT3_DIR_REC_LEN(0);
- return 0? 22: entry_space / sizeof(struct dx_entry);
- }
-
-@@ -309,7 +320,7 @@ static struct stats dx_show_leaf(struct
- printk(":%x.%u ", h.hash,
- ((char *) de - base));
- }
-- space += EXT3_DIR_REC_LEN(de->name_len);
-+ space += EXT3_DIR_REC_LEN(de);
- names++;
- }
- de = (struct ext3_dir_entry_2 *) ((char *) de + le16_to_cpu(de->rec_len));
-@@ -411,7 +422,9 @@ dx_probe(struct dentry *dentry, struct i
-
- entries = (struct dx_entry *) (((char *)info) + info->info_length);
-
-- if (dx_get_limit(entries) != dx_root_limit(dir, info->info_length)) {
-+ if (dx_get_limit(entries) != dx_root_limit(dir->i_sb->s_blocksize,
-+ (struct ext3_dir_entry_2*)bh->b_data,
-+ info->info_length)) {
- ext3_warning(dir->i_sb, __FUNCTION__,
- "dx entry: limit != root limit");
- brelse(bh);
-@@ -468,14 +481,17 @@ dx_probe(struct dentry *dentry, struct i
- if (!indirect--) return frame;
- if (!(bh = ext3_bread (NULL,dir, dx_get_block(at), 0, err)))
- goto fail2;
-- at = entries = ((struct dx_node *) bh->b_data)->entries;
-+ entries = ((struct dx_node *) bh->b_data)->entries;
- if (dx_get_limit(entries) != dx_node_limit (dir)) {
- ext3_warning(dir->i_sb, __FUNCTION__,
-- "dx entry: limit != node limit");
-+ "block %u(%lu): limit %u != node limit %u",
-+ dx_get_block(at), (long)bh->b_blocknr,
-+ dx_get_limit(entries), dx_node_limit(dir));
- brelse(bh);
- *err = ERR_BAD_DX_DIR;
- goto fail2;
- }
-+ at = entries;
- frame++;
- frame->bh = NULL;
- }
-@@ -608,7 +624,7 @@ static int htree_dirblock_to_tree(struct
- de = (struct ext3_dir_entry_2 *) bh->b_data;
- top = (struct ext3_dir_entry_2 *) ((char *) de +
- dir->i_sb->s_blocksize -
-- EXT3_DIR_REC_LEN(0));
-+ __EXT3_DIR_REC_LEN(0));
- for (; de < top; de = ext3_next_entry(de)) {
- if (!ext3_check_dir_entry("htree_dirblock_to_tree", dir, de, bh,
- (block<<EXT3_BLOCK_SIZE_BITS(dir->i_sb))
-@@ -1001,7 +1015,7 @@ static struct buffer_head * ext3_dx_find
- goto errout;
- de = (struct ext3_dir_entry_2 *) bh->b_data;
- top = (struct ext3_dir_entry_2 *) ((char *) de + sb->s_blocksize -
-- EXT3_DIR_REC_LEN(0));
-+ __EXT3_DIR_REC_LEN(0));
- for (; de < top; de = ext3_next_entry(de))
- if (ext3_match (namelen, name, de)) {
- if (!ext3_check_dir_entry("ext3_find_entry",
-@@ -1183,7 +1197,7 @@ dx_move_dirents(char *from, char *to, st
-
- while (count--) {
- struct ext3_dir_entry_2 *de = (struct ext3_dir_entry_2 *) (from + map->offs);
-- rec_len = EXT3_DIR_REC_LEN(de->name_len);
-+ rec_len = EXT3_DIR_REC_LEN(de);
- memcpy (to, de, rec_len);
- ((struct ext3_dir_entry_2 *) to)->rec_len =
- cpu_to_le16(rec_len);
-@@ -1208,7 +1222,7 @@ static struct ext3_dir_entry_2* dx_pack_
- next = (struct ext3_dir_entry_2 *) ((char *) de +
- le16_to_cpu(de->rec_len));
- if (de->inode && de->name_len) {
-- rec_len = EXT3_DIR_REC_LEN(de->name_len);
-+ rec_len = EXT3_DIR_REC_LEN(de);
- if (de > to)
- memmove(to, de, rec_len);
- to->rec_len = cpu_to_le16(rec_len);
-@@ -1334,12 +1348,18 @@ static int add_dirent_to_buf(handle_t *h
- struct inode *dir = dentry->d_parent->d_inode;
- const char *name = dentry->d_name.name;
- int namelen = dentry->d_name.len;
-+ unsigned char *data;
- unsigned long offset = 0;
- unsigned short reclen;
-- int nlen, rlen, err;
-+ int nlen, rlen, err, dlen = 0;
- char *top;
-
-- reclen = EXT3_DIR_REC_LEN(namelen);
-+ data = ext3_dentry_get_data(inode->i_sb, (struct ext3_dentry_param *)
-+ dentry->d_fsdata);
-+
-+ if (data)
-+ dlen = *data + 1;
-+ reclen = __EXT3_DIR_REC_LEN(namelen + dlen);
- if (!de) {
- de = (struct ext3_dir_entry_2 *)bh->b_data;
- top = bh->b_data + dir->i_sb->s_blocksize - reclen;
-@@ -1353,7 +1373,7 @@ static int add_dirent_to_buf(handle_t *h
- brelse (bh);
- return -EEXIST;
- }
-- nlen = EXT3_DIR_REC_LEN(de->name_len);
-+ nlen = EXT3_DIR_REC_LEN(de);
- rlen = le16_to_cpu(de->rec_len);
- if ((de->inode? rlen - nlen: rlen) >= reclen)
- break;
-@@ -1372,7 +1392,7 @@ static int add_dirent_to_buf(handle_t *h
- }
-
- /* By now the buffer is marked for journaling */
-- nlen = EXT3_DIR_REC_LEN(de->name_len);
-+ nlen = EXT3_DIR_REC_LEN(de);
- rlen = le16_to_cpu(de->rec_len);
- if (de->inode) {
- struct ext3_dir_entry_2 *de1 = (struct ext3_dir_entry_2 *)((char *)de + nlen);
-@@ -1388,6 +1408,11 @@ static int add_dirent_to_buf(handle_t *h
- de->inode = 0;
- de->name_len = namelen;
- memcpy (de->name, name, namelen);
-+ if (data) {
-+ de->name[namelen] = 0;
-+ memcpy(&de->name[namelen + 1], data, *(char*) data);
-+ de->file_type |= EXT3_DIRENT_LUFID;
-+ }
- /*
- * XXX shouldn't update any times until successful
- * completion of syscall, but too many callers depend
-@@ -1474,7 +1499,8 @@ static int make_indexed_dir(handle_t *ha
- entries = (void *)dx_info + sizeof(*dx_info);
- dx_set_block (entries, 1);
- dx_set_count (entries, 1);
-- dx_set_limit (entries, dx_root_limit(dir, sizeof(*dx_info)));
-+ dx_set_limit (entries, dx_root_limit(dir->i_sb->s_blocksize,
-+ dot_de, sizeof(*dx_info)));
-
- /* Initialize as for dx_probe */
- hinfo.hash_version = dx_info->hash_version;
-@@ -1502,6 +1528,8 @@ static int ext3_update_dotdot(handle_t *
- struct buffer_head * dir_block;
- struct ext3_dir_entry_2 * de;
- int len, journal = 0, err = 0;
-+ int dlen = 0;
-+ char *data;
-
- if (IS_ERR(handle))
- return PTR_ERR(handle);
-@@ -1517,19 +1545,23 @@ static int ext3_update_dotdot(handle_t *
- /* the first item must be "." */
- assert(de->name_len == 1 && de->name[0] == '.');
- len = le16_to_cpu(de->rec_len);
-- assert(len >= EXT3_DIR_REC_LEN(1));
-- if (len > EXT3_DIR_REC_LEN(1)) {
-+ assert(len >= __EXT3_DIR_REC_LEN(1));
-+ if (len > __EXT3_DIR_REC_LEN(1)) {
- BUFFER_TRACE(dir_block, "get_write_access");
- err = ext3_journal_get_write_access(handle, dir_block);
- if (err)
- goto out_journal;
-
- journal = 1;
-- de->rec_len = cpu_to_le16(EXT3_DIR_REC_LEN(1));
-+ de->rec_len = cpu_to_le16(EXT3_DIR_REC_LEN(de));
- }
-
-- len -= EXT3_DIR_REC_LEN(1);
-- assert(len == 0 || len >= EXT3_DIR_REC_LEN(2));
-+ len -= EXT3_DIR_REC_LEN(de);
-+ data = ext3_dentry_get_data(dir->i_sb,
-+ (struct ext3_dentry_param *) dentry->d_fsdata);
-+ if (data)
-+ dlen = *data + 1;
-+ assert(len == 0 || len >= __EXT3_DIR_REC_LEN(2 + dlen));
- de = (struct ext3_dir_entry_2 *)
- ((char *) de + le16_to_cpu(de->rec_len));
- if (!journal) {
-@@ -1543,11 +1575,15 @@ static int ext3_update_dotdot(handle_t *
- if (len > 0)
- de->rec_len = cpu_to_le16(len);
- else
-- assert(le16_to_cpu(de->rec_len) >= EXT3_DIR_REC_LEN(2));
-+ assert(le16_to_cpu(de->rec_len) >= __EXT3_DIR_REC_LEN(2));
- de->name_len = 2;
- strcpy (de->name, "..");
- ext3_set_de_type(dir->i_sb, de, S_IFDIR);
--
-+ if (data) {
-+ de->name[2] = 0;
-+ memcpy(&de->name[2 + 1], data, dlen);
-+ de->file_type |= EXT3_DIRENT_LUFID;
-+ }
- out_journal:
- if (journal) {
- BUFFER_TRACE(dir_block, "call ext3_journal_dirty_metadata");
-@@ -1958,12 +1994,12 @@ retry:
-
- /* Initialize @inode as a subdirectory of @dir, and add the
- * "." and ".." entries into the first directory block. */
--int ext3_add_dot_dotdot(handle_t *handle, struct inode * dir,
-- struct inode *inode)
-+int ext3_add_dot_dotdot(handle_t *handle, struct inode * dir, struct inode *inode,
-+ const void *data1, const void *data2)
- {
- struct buffer_head * dir_block;
- struct ext3_dir_entry_2 * de;
-- int err = 0;
-+ int err = 0, dot_reclen;
-
- if (IS_ERR(handle))
- return PTR_ERR(handle);
-@@ -1983,16 +2019,36 @@ int ext3_add_dot_dotdot(handle_t *handle
- de = (struct ext3_dir_entry_2 *) dir_block->b_data;
- de->inode = cpu_to_le32(inode->i_ino);
- de->name_len = 1;
-- de->rec_len = cpu_to_le16(EXT3_DIR_REC_LEN(de->name_len));
-- strcpy (de->name, ".");
- ext3_set_de_type(dir->i_sb, de, S_IFDIR);
-+ strcpy (de->name, ".");
-+
-+ /* get packed fid data*/
-+ data1 = ext3_dentry_get_data(dir->i_sb,
-+ (struct ext3_dentry_param *) data1);
-+ if (data1) {
-+ de->name[1] = 0;
-+ memcpy(&de->name[2], data1, *(char*) data1);
-+ de->file_type |= EXT3_DIRENT_LUFID;
-+ }
-+ de->rec_len = cpu_to_le16(EXT3_DIR_REC_LEN(de));
-+ dot_reclen = cpu_to_le16(de->rec_len);
-+
- de = (struct ext3_dir_entry_2 *)
- ((char *) de + le16_to_cpu(de->rec_len));
- de->inode = cpu_to_le32(dir->i_ino);
-- de->rec_len = cpu_to_le16(inode->i_sb->s_blocksize-EXT3_DIR_REC_LEN(1));
-+ de->rec_len = cpu_to_le16(inode->i_sb->s_blocksize-dot_reclen);
- de->name_len = 2;
- strcpy (de->name, "..");
- ext3_set_de_type(dir->i_sb, de, S_IFDIR);
-+
-+ /* get packed fid data*/
-+ data2 = ext3_dentry_get_data(dir->i_sb,
-+ (struct ext3_dentry_param *) data2);
-+ if (data2) {
-+ de->name[2] = 0;
-+ memcpy(&de->name[3], data2, *(char*) data2);
-+ de->file_type |= EXT3_DIRENT_LUFID;
-+ }
- inode->i_nlink = 2;
- BUFFER_TRACE(dir_block, "call ext3_journal_dirty_metadata");
- ext3_journal_dirty_metadata(handle, dir_block);
-@@ -2029,7 +2085,7 @@ retry:
- if (IS_ERR(inode))
- goto out_stop;
-
-- err = ext3_add_dot_dotdot(handle, dir, inode);
-+ err = ext3_add_dot_dotdot(handle, dir, inode, NULL, NULL);
- if (err) {
- inode->i_nlink = 0;
- ext3_mark_inode_dirty(handle, inode);
-@@ -2067,7 +2123,7 @@ static int empty_dir (struct inode * ino
- int err = 0;
-
- sb = inode->i_sb;
-- if (inode->i_size < EXT3_DIR_REC_LEN(1) + EXT3_DIR_REC_LEN(2) ||
-+ if (inode->i_size < __EXT3_DIR_REC_LEN(1) + __EXT3_DIR_REC_LEN(2) ||
- !(bh = ext3_bread (NULL, inode, 0, 0, &err))) {
- if (err)
- ext3_error(inode->i_sb, __FUNCTION__,
-Index: b/include/linux/ext3_fs.h
-===================================================================
---- a/include/linux/ext3_fs.h
-+++ b/include/linux/ext3_fs.h
-@@ -717,13 +717,16 @@ static inline int ext3_valid_inum(struct
- #define EXT3_FEATURE_INCOMPAT_META_BG 0x0010
- #define EXT3_FEATURE_INCOMPAT_EXTENTS 0x0040 /* extents support */
- #define EXT3_FEATURE_INCOMPAT_MMP 0x0100
-+#define EXT3_FEATURE_INCOMPAT_DIRDATA 0x1000
-
- #define EXT3_FEATURE_COMPAT_SUPP EXT2_FEATURE_COMPAT_EXT_ATTR
- #define EXT3_FEATURE_INCOMPAT_SUPP (EXT3_FEATURE_INCOMPAT_FILETYPE| \
- EXT3_FEATURE_INCOMPAT_RECOVER| \
- EXT3_FEATURE_INCOMPAT_META_BG| \
- EXT3_FEATURE_INCOMPAT_EXTENTS| \
-- EXT3_FEATURE_INCOMPAT_MMP)
-+ EXT3_FEATURE_INCOMPAT_MMP| \
-+ EXT3_FEATURE_INCOMPAT_DIRDATA)
-+
- #define EXT3_FEATURE_RO_COMPAT_SUPP (EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER| \
- EXT3_FEATURE_RO_COMPAT_LARGE_FILE| \
- EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
-@@ -790,7 +793,44 @@ struct ext3_dir_entry_2 {
- #define EXT3_FT_SYMLINK 7
-
- #define EXT3_FT_MAX 8
-+#define EXT3_FT_MASK 0xf
-+
-+#if EXT3_FT_MAX > EXT3_FT_MASK
-+#error "conflicting EXT3_FT_MAX and EXT3_FT_MASK"
-+#endif
-+
-+#define EXT3_LUFID_MAGIC 0xAD200907UL
-+struct ext3_dentry_param {
-+ __u32 edp_magic; /* EXT3_LUFID_MAGIC */
-+ char edp_len; /* size of edp_data in bytes */
-+ char edp_data[0]; /* packed array of data */
-+} __attribute__((packed));
-+
-+static inline unsigned char *ext3_dentry_get_data(struct super_block *sb,
-+ struct ext3_dentry_param* p)
-+
-+{
-+ if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_DIRDATA))
-+ return NULL;
-+ if (p && p->edp_magic == EXT3_LUFID_MAGIC)
-+ return &p->edp_len;
-+ else
-+ return NULL;
-+}
-+
-+/*
-+ * d_type has 4 unused bits, so it can hold four types data. these different
-+ * type of data (e.g. lustre data, high 32 bits of 64-bit inode number) can be
-+ * stored, in flag order, after file-name in ext3 dirent.
-+*/
-+/*
-+ * this flag is added to d_type if ext3 dirent has extra data after
-+ * filename. this data length is variable and length is stored in first byte
-+ * of data. data start after filename NUL byte.
-+ * This is used by Lustre FS.
-+ */
-
-+#define EXT3_DIRENT_LUFID 0x10
- /*
- * EXT3_DIR_PAD defines the directory entries boundaries
- *
-@@ -798,8 +838,12 @@ struct ext3_dir_entry_2 {
- */
- #define EXT3_DIR_PAD 4
- #define EXT3_DIR_ROUND (EXT3_DIR_PAD - 1)
--#define EXT3_DIR_REC_LEN(name_len) (((name_len) + 8 + EXT3_DIR_ROUND) & \
-+#define __EXT3_DIR_REC_LEN(len) (((len) + 8 + EXT3_DIR_ROUND) & \
- ~EXT3_DIR_ROUND)
-+
-+#define EXT3_DIR_REC_LEN(de) (__EXT3_DIR_REC_LEN(de->name_len +\
-+ ext3_get_dirent_data_len(de)))
-+
- /*
- * Hash Tree Directory indexing
- * (c) Daniel Phillips, 2001
-@@ -1125,7 +1169,8 @@ extern struct buffer_head * ext3_find_en
- struct ext3_dir_entry_2
- ** res_dir);
- extern int ext3_add_dot_dotdot(handle_t *handle, struct inode *dir,
-- struct inode *inode);
-+ struct inode *inode, const void *data1,
-+ const void *data2);
- extern struct inode_operations ext3_dir_inode_operations;
- extern struct inode_operations ext3_special_inode_operations;
- extern struct buffer_head *ext3_append(handle_t *handle, struct inode *inode,
-@@ -1159,6 +1204,29 @@ ext3_get_blocks_wrap(handle_t *handle, s
- return ret;
- }
-
-+/*
-+ * Compute the total directory entry data length.
-+ * This includes the filename and an implicit NUL terminator (always present),
-+ * and optional extensions. Each extension has a bit set in the high 4 bits of
-+ * de->file_type, and the extension length is the first byte in each entry.
-+ */
-+
-+static inline int ext3_get_dirent_data_len(struct ext3_dir_entry_2 *de)
-+{
-+ char *len = de->name + de->name_len + 1 /* NUL terminator */;
-+ int dlen = 0;
-+ __u8 extra_data_flags = (de->file_type & ~EXT3_FT_MASK) >> 4;
-+
-+ while (extra_data_flags) {
-+ if (extra_data_flags & 1) {
-+ dlen += *len + (dlen == 0);
-+ len += *len;
-+ }
-+ extra_data_flags >>= 1;
-+ }
-+ return dlen;
-+}
-+
- #endif /* __KERNEL__ */
-
- /* EXT3_IOC_CREATE_INUM at bottom of file (visible to kernel and user). */
-Index: b/fs/ext3/dir.c
-===================================================================
---- a/fs/ext3/dir.c
-+++ b/fs/ext3/dir.c
-@@ -53,11 +53,17 @@ const struct file_operations ext3_dir_op
-
- static unsigned char get_dtype(struct super_block *sb, int filetype)
- {
-+ int fl_index = filetype & EXT3_FT_MASK;
-+
- if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_FILETYPE) ||
-- (filetype >= EXT3_FT_MAX))
-+ (fl_index >= EXT3_FT_MAX))
- return DT_UNKNOWN;
-
-- return (ext3_filetype_table[filetype]);
-+ if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_DIRDATA))
-+ return (ext3_filetype_table[fl_index]);
-+
-+ return (ext3_filetype_table[fl_index]) |
-+ (filetype & EXT3_DIRENT_LUFID);
- }
-
-
-@@ -69,11 +75,11 @@ int ext3_check_dir_entry (const char * f
- const char * error_msg = NULL;
- const int rlen = le16_to_cpu(de->rec_len);
-
-- if (rlen < EXT3_DIR_REC_LEN(1))
-+ if (rlen < __EXT3_DIR_REC_LEN(1))
- error_msg = "rec_len is smaller than minimal";
- else if (rlen % 4 != 0)
- error_msg = "rec_len % 4 != 0";
-- else if (rlen < EXT3_DIR_REC_LEN(de->name_len))
-+ else if (rlen < EXT3_DIR_REC_LEN(de))
- error_msg = "rec_len is too small for name_len";
- else if (((char *) de - bh->b_data) + rlen > dir->i_sb->s_blocksize)
- error_msg = "directory entry across blocks";
-@@ -177,7 +183,7 @@ revalidate:
- * failure will be detected in the
- * dirent test below. */
- if (le16_to_cpu(de->rec_len) <
-- EXT3_DIR_REC_LEN(1))
-+ __EXT3_DIR_REC_LEN(1))
- break;
- i += le16_to_cpu(de->rec_len);
- }
-@@ -210,7 +216,6 @@ revalidate:
- * during the copy operation.
- */
- unsigned long version = filp->f_version;
--
- error = filldir(dirent, de->name,
- de->name_len,
- filp->f_pos,
-@@ -342,13 +347,17 @@ int ext3_htree_store_dirent(struct file
- struct rb_node **p, *parent = NULL;
- struct fname * fname, *new_fn;
- struct dir_private_info *info;
-+ int extra_data = 1;
- int len;
-
- info = (struct dir_private_info *) dir_file->private_data;
- p = &info->root.rb_node;
-
- /* Create and allocate the fname structure */
-- len = sizeof(struct fname) + dirent->name_len + 1;
-+ if (dirent->file_type & EXT3_DIRENT_LUFID)
-+ extra_data = ext3_get_dirent_data_len(dirent);
-+
-+ len = sizeof(struct fname) + dirent->name_len + extra_data;
- new_fn = kmalloc(len, GFP_KERNEL);
- if (!new_fn)
- return -ENOMEM;
-@@ -358,7 +367,7 @@ int ext3_htree_store_dirent(struct file
- new_fn->inode = le32_to_cpu(dirent->inode);
- new_fn->name_len = dirent->name_len;
- new_fn->file_type = dirent->file_type;
-- memcpy(new_fn->name, dirent->name, dirent->name_len);
-+ memcpy(new_fn->name, dirent->name, dirent->name_len + extra_data);
- new_fn->name[dirent->name_len] = 0;
-
- while (*p) {
+++ /dev/null
-Index: b/fs/ext3/inode.c
-===================================================================
---- a/fs/ext3/inode.c
-+++ b/fs/ext3/inode.c
-@@ -2676,6 +2676,25 @@ void ext3_set_inode_flags(struct inode *
- inode->i_flags |= S_DIRSYNC;
- }
-
-+/* Propagate flags from i_flags to EXT3_I(inode)->i_flags */
-+void ext3_get_inode_flags(struct ldiskfs_inode_info *ei)
-+{
-+ unsigned int flags = ei->vfs_inode.i_flags;
-+
-+ ei->i_flags &= ~(EXT3_SYNC_FL|EXT3_APPEND_FL|
-+ EXT3_IMMUTABLE_FL|EXT3_NOATIME_FL|EXT3_DIRSYNC_FL);
-+ if (flags & S_SYNC)
-+ ei->i_flags |= EXT3_SYNC_FL;
-+ if (flags & S_APPEND)
-+ ei->i_flags |= EXT3_APPEND_FL;
-+ if (flags & S_IMMUTABLE)
-+ ei->i_flags |= EXT3_IMMUTABLE_FL;
-+ if (flags & S_NOATIME)
-+ ei->i_flags |= EXT3_NOATIME_FL;
-+ if (flags & S_DIRSYNC)
-+ ei->i_flags |= EXT3_DIRSYNC_FL;
-+}
-+
- void ext3_read_inode(struct inode * inode)
- {
- struct ext3_iloc iloc;
-@@ -2841,6 +2860,7 @@ static int ext3_do_update_inode(handle_t
- if (ei->i_state & EXT3_STATE_NEW)
- memset(raw_inode, 0, EXT3_SB(inode->i_sb)->s_inode_size);
-
-+ ext3_get_inode_flags(ei);
- raw_inode->i_mode = cpu_to_le16(inode->i_mode);
- if(!(test_opt(inode->i_sb, NO_UID32))) {
- raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid));
-Index: b/fs/ext3/ioctl.c
-===================================================================
---- a/fs/ext3/ioctl.c
-+++ b/fs/ext3/ioctl.c
-@@ -117,6 +117,7 @@ int ext3_ioctl (struct inode * inode, st
- return rc;
- }
- case EXT3_IOC_GETFLAGS:
-+ ext3_get_inode_flags(ei);
- flags = ei->i_flags & EXT3_FL_USER_VISIBLE;
- return put_user(flags, (int __user *) arg);
- case EXT3_IOC_SETFLAGS: {
-Index: b/include/linux/ext3_fs.h
-===================================================================
---- a/include/linux/ext3_fs.h
-+++ b/include/linux/ext3_fs.h
-@@ -1092,6 +1092,7 @@ extern int ext3_get_inode_loc(struct ino
- extern int ext3_can_truncate(struct inode *inode);
- extern void ext3_truncate (struct inode *);
- extern void ext3_set_inode_flags(struct inode *);
-+extern void ext3_get_inode_flags(struct ldiskfs_inode_info *ei);
- extern void ext3_set_aops(struct inode *inode);
- extern int ext3_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
- u64 start, u64 len);
+++ /dev/null
-Index: linux-2.6.18-128.1.6/fs/ext4/balloc.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext4/balloc.c
-+++ linux-2.6.18-128.1.6/fs/ext4/balloc.c
-@@ -321,16 +321,16 @@ ext4_read_block_bitmap(struct super_bloc
- unlock_buffer(bh);
- return bh;
- }
-- spin_lock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_lock_group(sb, block_group);
- if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
- ext4_init_block_bitmap(sb, bh, block_group, desc);
- set_bitmap_uptodate(bh);
- set_buffer_uptodate(bh);
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_unlock_group(sb, block_group);
- unlock_buffer(bh);
- return bh;
- }
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_unlock_group(sb, block_group);
- if (buffer_uptodate(bh)) {
- /*
- * if not uninit if bh is uptodate,
-@@ -787,7 +788,7 @@ do_more:
- * the allocator uses.
- */
- BUFFER_TRACE(bitmap_bh, "clear bit");
-- if (!ext4_clear_bit_atomic(sb_bgl_lock(sbi, block_group),
-+ if (!ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group),
- bit + i, bitmap_bh->b_data)) {
- jbd_unlock_bh_state(bitmap_bh);
- ext4_error(sb, __func__,
-@@ -801,18 +802,18 @@ do_more:
- blocks_freed++;
- }
- }
-- spin_lock(sb_bgl_lock(sbi, block_group));
-+ ext4_lock_group(sb, block_group);
- blk_free_count = blocks_freed + ext4_free_blks_count(sb, desc);
- ext4_free_blks_set(sb, desc, blk_free_count);
- desc->bg_checksum = ext4_group_desc_csum(sbi, block_group, desc);
-- spin_unlock(sb_bgl_lock(sbi, block_group));
-+ ext4_unlock_group(sb, block_group);
- percpu_counter_add(&sbi->s_freeblocks_counter, blocks_freed);
-
- if (sbi->s_log_groups_per_flex) {
- ext4_group_t flex_group = ext4_flex_group(sbi, block_group);
-- spin_lock(sb_bgl_lock(sbi, flex_group));
-+ ext4_lock_group(sb, block_group);
- sbi->s_flex_groups[flex_group].free_blocks += blocks_freed;
-- spin_unlock(sb_bgl_lock(sbi, flex_group));
-+ ext4_unlock_group(sb, block_group);
- }
-
- /* We dirtied the bitmap block */
-Index: linux-2.6.18-128.1.6/fs/ext4/ialloc.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext4/ialloc.c
-+++ linux-2.6.18-128.1.6/fs/ext4/ialloc.c
-@@ -118,16 +118,16 @@ ext4_read_inode_bitmap(struct super_bloc
- unlock_buffer(bh);
- return bh;
- }
-- spin_lock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_lock_group(sb, block_group);
- if (desc->bg_flags & cpu_to_le16(EXT4_BG_INODE_UNINIT)) {
- ext4_init_inode_bitmap(sb, bh, block_group, desc);
- set_bitmap_uptodate(bh);
- set_buffer_uptodate(bh);
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_unlock_group(sb, block_group);
- unlock_buffer(bh);
- return bh;
- }
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
-+ ext4_unlock_group(sb, block_group);
- if (buffer_uptodate(bh)) {
- /*
- * if not uninit if bh is uptodate,
-@@ -221,9 +221,9 @@ void ext4_free_inode (handle_t *handle,
- goto error_return;
-
- /* Ok, now we can actually update the inode bitmaps.. */
-- spin_lock(sb_bgl_lock(sbi, block_group));
-+ ext4_lock_group(sb, block_group);
- cleared = ext4_clear_bit(bit, bitmap_bh->b_data);
-- spin_unlock(sb_bgl_lock(sbi, block_group));
-+ ext4_unlock_group(sb, block_group);
- if (!cleared)
- ext4_error(sb, "ext4_free_inode",
- "bit already cleared for inode %lu", ino);
-@@ -233,7 +233,7 @@ void ext4_free_inode (handle_t *handle,
- if (fatal) goto error_return;
-
- if (gdp) {
-- spin_lock(sb_bgl_lock(sbi, block_group));
-+ ext4_lock_group(sb, block_group);
- count = ext4_free_inodes_count(sb, gdp) + 1;
- ext4_free_inodes_set(sb, gdp, count);
- if (is_directory) {
-@@ -233,16 +233,16 @@ void ext4_free_inode (handle_t *handle,
- }
- gdp->bg_checksum = ext4_group_desc_csum(sbi,
- block_group, gdp);
-- spin_unlock(sb_bgl_lock(sbi, block_group));
-+ ext4_unlock_group(sb, block_group);
- percpu_counter_inc(&sbi->s_freeinodes_counter);
- if (is_directory)
- percpu_counter_dec(&sbi->s_dirs_counter);
-
- if (sbi->s_log_groups_per_flex) {
- flex_group = ext4_flex_group(sbi, block_group);
-- spin_lock(sb_bgl_lock(sbi, flex_group));
-+ ext4_lock_group(sb, flex_group);
- sbi->s_flex_groups[flex_group].free_inodes++;
-- spin_unlock(sb_bgl_lock(sbi, flex_group));
-+ ext4_unlock_group(sb, flex_group);
- }
- }
- BUFFER_TRACE(bh2, "call ext4_journal_dirty_metadata");
-@@ -630,7 +630,7 @@
- struct ext4_sb_info *sbi = EXT4_SB(sb);
- struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, NULL);
-
-- spin_lock(sb_bgl_lock(sbi, group));
-+ ext4_lock_group(sb, group);
- if (ext4_set_bit(ino, inode_bitmap_bh->b_data)) {
- /* not a free inode */
- retval = 1;
-@@ -691,7 +691,7 @@
- ino++;
- if ((group == 0 && ino < EXT4_FIRST_INO(sb)) ||
- ino > EXT4_INODES_PER_GROUP(sb)) {
-- spin_unlock(sb_bgl_lock(sbi, group));
-+ ext4_unlock_group(sb, group);
- ext4_error(sb, __func__,
- "reserved inode or inode > inodes count - "
- "block_group = %u, inode=%lu", group,
-@@ -692,7 +692,7 @@
- }
- gdp->bg_checksum = ext4_group_desc_csum(sbi, group, gdp);
- err_ret:
-- spin_unlock(sb_bgl_lock(sbi, group));
-+ ext4_unlock_group(sb, group);
- return retval;
- }
-
-@@ -751,16 +751,16 @@ got:
- }
-
- free = 0;
-- spin_lock(sb_bgl_lock(sbi, group));
-+ ext4_lock_group(sb, group);
- /* recheck and clear flag under lock if we still need to */
- if (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
- free = ext4_free_blocks_after_init(sb, group, gdp);
- gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
- ext4_free_blks_set(sb, gdp, free);
- gdp->bg_checksum = ext4_group_desc_csum(sbi, group,
- gdp);
- }
-- spin_unlock(sb_bgl_lock(sbi, group));
-+ ext4_unlock_group(sb, group);
-
- /* Don't need to dirty bitmap block if we didn't change it */
- if (free) {
-@@ -819,9 +819,9 @@ got:
-
- if (sbi->s_log_groups_per_flex) {
- flex_group = ext4_flex_group(sbi, group);
-- spin_lock(sb_bgl_lock(sbi, flex_group));
-+ ext4_lock_group(sb, flex_group);
- sbi->s_flex_groups[flex_group].free_inodes--;
-- spin_unlock(sb_bgl_lock(sbi, flex_group));
-+ ext4_unlock_group(sb, flex_group);
- }
-
- inode->i_uid = current->fsuid;
-Index: linux-2.6.18-128.1.6/fs/ext4/mballoc.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext4/mballoc.c
-+++ linux-2.6.18-128.1.6/fs/ext4/mballoc.c
-@@ -361,24 +361,12 @@ static inline void mb_set_bit(int bit, v
- ext4_set_bit(bit, addr);
- }
-
--static inline void mb_set_bit_atomic(spinlock_t *lock, int bit, void *addr)
--{
-- addr = mb_correct_addr_and_bit(&bit, addr);
-- ext4_set_bit_atomic(lock, bit, addr);
--}
--
- static inline void mb_clear_bit(int bit, void *addr)
- {
- addr = mb_correct_addr_and_bit(&bit, addr);
- ext4_clear_bit(bit, addr);
- }
-
--static inline void mb_clear_bit_atomic(spinlock_t *lock, int bit, void *addr)
--{
-- addr = mb_correct_addr_and_bit(&bit, addr);
-- ext4_clear_bit_atomic(lock, bit, addr);
--}
--
- static inline int mb_find_next_zero_bit(void *addr, int max, int start)
- {
- int fix = 0, ret, tmpmax;
-@@ -789,17 +777,17 @@ static int ext4_mb_init_cache(struct pag
- unlock_buffer(bh[i]);
- continue;
- }
-- spin_lock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
-+ ext4_lock_group(sb, first_group + i);
- if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
- ext4_init_block_bitmap(sb, bh[i],
- first_group + i, desc);
- set_bitmap_uptodate(bh[i]);
- set_buffer_uptodate(bh[i]);
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
-+ ext4_unlock_group(sb, first_group + i);
- unlock_buffer(bh[i]);
- continue;
- }
-- spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
-+ ext4_unlock_group(sb, first_group + i);
- if (buffer_uptodate(bh[i])) {
- /*
- * if not uninit if bh is uptodate,
-@@ -1021,7 +1009,7 @@ static int mb_find_order_for_block(struc
- return 0;
- }
-
--static void mb_clear_bits(spinlock_t *lock, void *bm, int cur, int len)
-+static void mb_clear_bits(void *bm, int cur, int len)
- {
- __u32 *addr;
-
-@@ -1034,15 +1022,12 @@ static void mb_clear_bits(spinlock_t *lo
- cur += 32;
- continue;
- }
-- if (lock)
-- mb_clear_bit_atomic(lock, cur, bm);
-- else
-- mb_clear_bit(cur, bm);
-+ mb_clear_bit(cur, bm);
- cur++;
- }
- }
-
--static void mb_set_bits(spinlock_t *lock, void *bm, int cur, int len)
-+static void mb_set_bits(void *bm, int cur, int len)
- {
- __u32 *addr;
-
-@@ -1052,10 +1040,7 @@ static void mb_set_bits(spinlock_t *lock
- cur += 32;
- continue;
- }
-- if (lock)
-- mb_set_bit_atomic(lock, cur, bm);
-- else
-- mb_set_bit(cur, bm);
-+ mb_set_bit(cur, bm);
- cur++;
- }
- }
-@@ -1268,8 +1256,7 @@ static int mb_mark_used(struct ext4_budd
- e4b->bd_info->bb_counters[ord]++;
- }
-
-- mb_set_bits(sb_bgl_lock(EXT4_SB(e4b->bd_sb), ex->fe_group),
-- EXT4_MB_BITMAP(e4b), ex->fe_start, len0);
-+ mb_set_bits(EXT4_MB_BITMAP(e4b), ex->fe_start, len0);
- mb_check_buddy(e4b);
-
- return ret;
-@@ -2651,7 +2638,7 @@ int ext4_mb_init(struct super_block *sb,
- return 0;
- }
-
--/* need to called with ext4 group lock (ext4_lock_group) */
-+/* need to called with the ext4 group lock held */
- static void ext4_mb_cleanup_pa(struct ext4_group_info *grp)
- {
- struct ext4_prealloc_space *pa;
-@@ -3130,14 +3117,17 @@ ext4_mb_mark_diskspace_used(struct ext4_
- * Fix the bitmap and repeat the block allocation
- * We leak some of the blocks here.
- */
-- mb_set_bits(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group),
-- bitmap_bh->b_data, ac->ac_b_ex.fe_start,
-- ac->ac_b_ex.fe_len);
-+ ext4_lock_group(sb, ac->ac_b_ex.fe_group);
-+ mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start,
-+ ac->ac_b_ex.fe_len);
-+ ext4_unlock_group(sb, ac->ac_b_ex.fe_group);
- err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh);
- if (!err)
- err = -EAGAIN;
- goto out_err;
- }
-+
-+ ext4_lock_group(sb, ac->ac_b_ex.fe_group);
- #ifdef AGGRESSIVE_CHECK
- {
- int i;
-@@ -3147,9 +3137,7 @@ ext4_mb_mark_diskspace_used(struct ext4_
- }
- }
- #endif
-- spin_lock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
-- mb_set_bits(NULL, bitmap_bh->b_data,
-- ac->ac_b_ex.fe_start, ac->ac_b_ex.fe_len);
-+ mb_set_bits(bitmap_bh->b_data, ac->ac_b_ex.fe_start,ac->ac_b_ex.fe_len);
- if (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
- gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
- gdp->bg_free_blocks_count =
-@@ -3160,7 +3148,8 @@ ext4_mb_mark_diskspace_used(struct ext4_
- len = ext4_free_blks_count(sb, gdp) - ac->ac_b_ex.fe_len;
- ext4_free_blks_set(sb, gdp, len);
- gdp->bg_checksum = ext4_group_desc_csum(sbi, ac->ac_b_ex.fe_group, gdp);
-- spin_unlock(sb_bgl_lock(sbi, ac->ac_b_ex.fe_group));
-+
-+ ext4_unlock_group(sb, ac->ac_b_ex.fe_group);
- percpu_counter_sub(&sbi->s_freeblocks_counter, ac->ac_b_ex.fe_len);
- /*
- * Now reduce the dirty block count also. Should not go negative
-@@ -3161,9 +3148,9 @@ ext4_mb_mark_diskspace_used(struct ext4_
- if (sbi->s_log_groups_per_flex) {
- ext4_group_t flex_group = ext4_flex_group(sbi,
- ac->ac_b_ex.fe_group);
-- spin_lock(sb_bgl_lock(sbi, flex_group));
-+ ext4_lock_group(sb, flex_group);
- sbi->s_flex_groups[flex_group].free_blocks -= ac->ac_b_ex.fe_len;
-- spin_unlock(sb_bgl_lock(sbi, flex_group));
-+ ext4_unlock_group(sb, flex_group);
- }
-
- err = ext4_journal_dirty_metadata(handle, bitmap_bh);
-@@ -3500,9 +3589,7 @@ int ext4_mb_check_ondisk_bitmap(struct s
-
- while (n) {
- entry = rb_entry(n, struct ext4_free_data, node);
-- mb_set_bits(sb_bgl_lock(EXT4_SB(sb), group),
-- bitmap, entry->start_blk,
-- entry->count);
-+ mb_set_bits(bitmap, entry->start_blk, entry->count);
- n = rb_next(n);
- }
- return;
-@@ -3600,7 +3589,7 @@ int ext4_mb_check_ondisk_bitmap(struct s
- /*
- * the function goes through all preallocation in this group and marks them
- * used in in-core bitmap. buddy must be generated from this bitmap
-- * Need to be called with ext4 group lock (ext4_lock_group)
-+ * Need to be called with ext4 group lock held.
- */
- static int ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap,
- ext4_group_t group)
-@@ -3646,8 +3635,7 @@ static int ext4_mb_generate_from_pa(stru
- continue;
- }
- BUG_ON(groupnr != group);
-- mb_set_bits(sb_bgl_lock(EXT4_SB(sb), group),
-- bitmap, start, len);
-+ mb_set_bits(bitmap, start, len);
- preallocated += len;
- count++;
- }
-@@ -4742,6 +4730,7 @@ static void ext4_mb_poll_new_transaction
- ext4_mb_free_committed_blocks(sb);
- }
-
-+/* need to be called with ldiskfs group lock held */
- static noinline_for_stack int
- ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
- ext4_group_t group, ext4_grpblk_t block, int count)
-@@ -4912,35 +4908,30 @@ do_more:
- new_entry->count = count;
- new_entry->t_tid = handle->h_transaction->t_tid;
- ext4_lock_group(sb, block_group);
-- mb_clear_bits(sb_bgl_lock(sbi, block_group), bitmap_bh->b_data,
-- bit, count);
-+ mb_clear_bits(bitmap_bh->b_data, bit, count);
- ext4_mb_free_metadata(handle, &e4b, new_entry);
-- ext4_unlock_group(sb, block_group);
- } else {
- ext4_lock_group(sb, block_group);
- /* need to update group_info->bb_free and bitmap
- * with group lock held. generate_buddy look at
- * them with group lock_held
- */
-- mb_clear_bits(sb_bgl_lock(sbi, block_group), bitmap_bh->b_data,
-- bit, count);
-+ mb_clear_bits(bitmap_bh->b_data, bit, count);
- mb_free_blocks(inode, &e4b, bit, count);
- ext4_mb_return_to_preallocation(inode, &e4b, block, count);
-- ext4_unlock_group(sb, block_group);
- }
-
-- spin_lock(sb_bgl_lock(sbi, block_group));
- ret = ext4_free_blks_count(sb, gdp) + count;
- ext4_free_blks_set(sb, gdp, ret);
- gdp->bg_checksum = ext4_group_desc_csum(sbi, block_group, gdp);
-- spin_unlock(sb_bgl_lock(sbi, block_group));
-+ ext4_unlock_group(sb, block_group);
- percpu_counter_add(&sbi->s_freeblocks_counter, count);
-
- if (sbi->s_log_groups_per_flex) {
- ext4_group_t flex_group = ext4_flex_group(sbi, block_group);
-- spin_lock(sb_bgl_lock(sbi, flex_group));
-+ ext4_lock_group(sb, flex_group);
- sbi->s_flex_groups[flex_group].free_blocks += count;
-- spin_unlock(sb_bgl_lock(sbi, flex_group));
-+ ext4_unlock_group(sb, flex_group);
- }
-
- ext4_mb_release_desc(&e4b);
-Index: linux-2.6.18-128.1.6/fs/ext4/super.c
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext4/super.c
-+++ linux-2.6.18-128.1.6/fs/ext4/super.c
-@@ -1934,18 +1934,18 @@ static int ext4_check_descriptors(struct
- "(block %llu)!\n", i, inode_table);
- return 0;
- }
-- spin_lock(sb_bgl_lock(sbi, i));
-+ ext4_lock_group(sb, i);
- if (!ext4_group_desc_csum_verify(sbi, i, gdp)) {
- printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: "
- "Checksum for group %u failed (%u!=%u)\n",
- i, le16_to_cpu(ext4_group_desc_csum(sbi, i,
- gdp)), le16_to_cpu(gdp->bg_checksum));
- if (!(sb->s_flags & MS_RDONLY)) {
-- spin_unlock(sb_bgl_lock(sbi, i));
-+ ext4_unlock_group(sb, i);
- return 0;
- }
- }
-- spin_unlock(sb_bgl_lock(sbi, i));
-+ ext4_unlock_group(sb, i);
- if (!flexbg_flag)
- first_block += EXT4_BLOCKS_PER_GROUP(sb);
- }
-Index: linux-2.6.18-128.1.6/fs/ext4/ext4.h
-===================================================================
---- linux-2.6.18-128.1.6.orig/fs/ext4/ext4.h
-+++ linux-2.6.18-128.1.6/fs/ext4/ext4.h
-@@ -127,35 +127,9 @@ struct ext4_group_info {
- };
-
- #define EXT4_GROUP_INFO_NEED_INIT_BIT 0
--#define EXT4_GROUP_INFO_LOCKED_BIT 1
-
- #define EXT4_MB_GRP_NEED_INIT(grp) \
- (test_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &((grp)->bb_state)))
--
--static inline void ext4_lock_group(struct super_block *sb, ext4_group_t group)
--{
-- struct ext4_group_info *grinfo = ext4_get_group_info(sb, group);
--
-- bit_spin_lock(EXT4_GROUP_INFO_LOCKED_BIT, &(grinfo->bb_state));
--}
--
--static inline void ext4_unlock_group(struct super_block *sb,
-- ext4_group_t group)
--{
-- struct ext4_group_info *grinfo = ext4_get_group_info(sb, group);
--
-- bit_spin_unlock(EXT4_GROUP_INFO_LOCKED_BIT, &(grinfo->bb_state));
--}
--
--static inline int ext4_is_group_locked(struct super_block *sb,
-- ext4_group_t group)
--{
-- struct ext4_group_info *grinfo = ext4_get_group_info(sb, group);
--
-- return bit_spin_is_locked(EXT4_GROUP_INFO_LOCKED_BIT,
-- &(grinfo->bb_state));
--}
--
- /*
- * Inodes and files operations
- */
-@@ -1303,6 +1303,32 @@ extern int ext4_get_blocks_wrap(handle_t
- set_bit(BH_BITMAP_UPTODATE, &(bh)->b_state);
- }
-
-+static inline spinlock_t *ext4_group_lock_ptr(struct super_block *sb,
-+ ext4_group_t group)
-+{
-+ struct blockgroup_lock *bgl = &EXT4_SB(sb)->s_blockgroup_lock;
-+ return &bgl->locks[group & (NR_BG_LOCKS-1)].lock;
-+}
-+
-+static inline void ext4_lock_group(struct super_block *sb, ext4_group_t group)
-+{
-+ spin_lock(ext4_group_lock_ptr(sb, group));
-+}
-+
-+static inline void ext4_unlock_group(struct super_block *sb,
-+ ext4_group_t group)
-+{
-+ spin_unlock(ext4_group_lock_ptr(sb, group));
-+}
-+
-+static inline int ext4_is_group_locked(struct super_block *sb,
-+ ext4_group_t group)
-+{
-+ return spin_is_locked(ext4_group_lock_ptr(sb, group));
-+}
-+
-+
-+
- #endif /* __KERNEL__ */
-
- #endif /* _EXT4_H */
+++ /dev/null
-ext3-version-2.6-rhel5.patch
-ext3-wantedi-2.6-rhel4.patch
-ext3-map_inode_page-2.6.18.patch
-export-ext3-2.6-rhel4.patch
-ext3-include-fixes-2.6-rhel4.patch
-ext3-extents-2.6.18-vanilla.patch
-ext3-mballoc3-core.patch
-ext3-mballoc3-2.6.18.patch
-ext3-nlinks-2.6-rhel5.patch
-ext3-remove-cond_resched-calls-2.6.12.patch
-ext3-uninit-2.6.18.patch
-ext3-nanosecond-2.6.18-vanilla.patch
-ext3-inode-version-2.6.18-vanilla.patch
-ext3-ea-expand-lose-block.patch
-ext3-mmp-2.6.18-vanilla.patch
-ext3-statfs-2.6-rhel5.patch
-ext3-lookup-dotdot-2.6.9.patch
-ext3-max-dir-size.patch
-ext3-print-inum-in-htree-warning.patch
-ext3-block-bitmap-validation-2.6-rhel5.patch
-ext3-xattr-no-update-ctime-2.6.22-vanilla.patch
-ext3-journal-chksum-2.6.18-vanilla.patch
-ext3-get-raid-stripe-from-sb.patch
-ext3-big-endian-check-2.6-rhel5.patch
-alloc-policy-2.6-rhlel5.diff
-ext3-force_over_8tb-rhel5.patch
-ext3-pdir-fix.patch
-ext3-osd-iop-common.patch
-ext3-osd-iam-exports.patch
-ext3-dynlocks-common.patch
-ext3-dynlocks-2.6-rhel5.patch
-ext3-hash-indexed-dir-dotdot-update.patch
-ext3-corrupted-orphans-2.6.patch
-ext3-kill-dx_root.patch
-ext3-fiemap-2.6-rhel5.patch
-ext3-mballoc-pa_free-mismatch.patch
-ext3_data_in_dirent.patch
-ext3_fix_i_flags.patch
-ext3-disable-mb-cache.patch
-ext3-export-64bit-name-hash.patch