2007-06-12 17:07:21 +04:00
/*
* Copyright ( C ) 2007 Oracle . All rights reserved .
*
* This program is free software ; you can redistribute it and / or
* modify it under the terms of the GNU General Public
* License v2 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
* License along with this program ; if not , write to the
* Free Software Foundation , Inc . , 59 Temple Place - Suite 330 ,
* Boston , MA 021110 - 1307 , USA .
*/
2007-04-02 18:50:19 +04:00
# ifndef __BTRFS_I__
# define __BTRFS_I__
2007-08-28 00:49:44 +04:00
# include "extent_map.h"
2008-01-25 00:13:08 +03:00
# include "extent_io.h"
2008-07-17 20:53:50 +04:00
# include "ordered-data.h"
2007-08-28 00:49:44 +04:00
2007-06-14 00:18:26 +04:00
/* in memory btrfs inode */
2007-04-02 18:50:19 +04:00
struct btrfs_inode {
2008-09-29 23:18:18 +04:00
/* which subvolume this inode belongs to */
2007-04-06 23:37:36 +04:00
struct btrfs_root * root ;
2008-09-29 23:18:18 +04:00
/* key used to find this inode on disk. This is used by the code
* to read in roots of subvolumes
*/
2007-04-06 23:37:36 +04:00
struct btrfs_key location ;
2008-09-29 23:18:18 +04:00
/* the extent_tree has caches of all the extent mappings to disk */
2007-08-28 00:49:44 +04:00
struct extent_map_tree extent_tree ;
2008-09-29 23:18:18 +04:00
/* the io_tree does range state (DIRTY, LOCKED etc) */
2008-01-25 00:13:08 +03:00
struct extent_io_tree io_tree ;
2008-09-29 23:18:18 +04:00
/* special utility tree used to record which mirrors have already been
* tried when checksums fail for a given block
*/
2008-04-10 00:28:12 +04:00
struct extent_io_tree io_failure_tree ;
2008-09-29 23:18:18 +04:00
/* held while logging the inode in tree-log.c */
2008-09-06 00:13:11 +04:00
struct mutex log_mutex ;
2008-09-29 23:18:18 +04:00
/* used to order data wrt metadata */
2008-07-17 20:53:50 +04:00
struct btrfs_ordered_inode_tree ordered_tree ;
2007-08-11 00:22:09 +04:00
2008-07-24 20:17:14 +04:00
/* for keeping track of orphaned inodes */
struct list_head i_orphan ;
2008-09-29 23:18:18 +04:00
/* list of all the delalloc inodes in the FS. There are times we need
* to write all the delalloc pages to disk , and this list is used
* to walk them all .
*/
2008-08-05 07:17:27 +04:00
struct list_head delalloc_inodes ;
2009-03-31 21:27:11 +04:00
/*
* list for tracking inodes that must be sent to disk before a
* rename or truncate commit
*/
struct list_head ordered_operations ;
Btrfs: Mixed back reference (FORWARD ROLLING FORMAT CHANGE)
This commit introduces a new kind of back reference for btrfs metadata.
Once a filesystem has been mounted with this commit, IT WILL NO LONGER
BE MOUNTABLE BY OLDER KERNELS.
When a tree block in subvolume tree is cow'd, the reference counts of all
extents it points to are increased by one. At transaction commit time,
the old root of the subvolume is recorded in a "dead root" data structure,
and the btree it points to is later walked, dropping reference counts
and freeing any blocks where the reference count goes to 0.
The increments done during cow and decrements done after commit cancel out,
and the walk is a very expensive way to go about freeing the blocks that
are no longer referenced by the new btree root. This commit reduces the
transaction overhead by avoiding the need for dead root records.
When a non-shared tree block is cow'd, we free the old block at once, and the
new block inherits old block's references. When a tree block with reference
count > 1 is cow'd, we increase the reference counts of all extents
the new block points to by one, and decrease the old block's reference count by
one.
This dead tree avoidance code removes the need to modify the reference
counts of lower level extents when a non-shared tree block is cow'd.
But we still need to update back ref for all pointers in the block.
This is because the location of the block is recorded in the back ref
item.
We can solve this by introducing a new type of back ref. The new
back ref provides information about pointer's key, level and in which
tree the pointer lives. This information allow us to find the pointer
by searching the tree. The shortcoming of the new back ref is that it
only works for pointers in tree blocks referenced by their owner trees.
This is mostly a problem for snapshots, where resolving one of these
fuzzy back references would be O(number_of_snapshots) and quite slow.
The solution used here is to use the fuzzy back references in the common
case where a given tree block is only referenced by one root,
and use the full back references when multiple roots have a reference
on a given block.
This commit adds per subvolume red-black tree to keep trace of cached
inodes. The red-black tree helps the balancing code to find cached
inodes whose inode numbers within a given range.
This commit improves the balancing code by introducing several data
structures to keep the state of balancing. The most important one
is the back ref cache. It caches how the upper level tree blocks are
referenced. This greatly reduce the overhead of checking back ref.
The improved balancing code scales significantly better with a large
number of snapshots.
This is a very large commit and was written in a number of
pieces. But, they depend heavily on the disk format change and were
squashed together to make sure git bisect didn't end up in a
bad state wrt space balancing or the format change.
Signed-off-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
2009-06-10 18:45:14 +04:00
/* node for the red-black tree that links inodes in subvolume root */
struct rb_node rb_node ;
2009-02-20 19:00:09 +03:00
/* the space_info for where this inode's data allocations are done */
struct btrfs_space_info * space_info ;
2008-09-29 23:18:18 +04:00
/* full 64 bit generation number, struct vfs_inode doesn't have a big
* enough field for this .
*/
2008-09-06 00:13:11 +04:00
u64 generation ;
2008-12-09 00:40:21 +03:00
/* sequence number for NFS changes */
u64 sequence ;
2007-08-11 00:22:09 +04:00
/*
* transid of the trans_handle that last modified this inode
*/
u64 last_trans ;
2009-10-13 21:21:08 +04:00
/*
* log transid when this inode was last modified
*/
u64 last_sub_trans ;
2008-09-06 00:13:11 +04:00
/*
* transid that last logged this inode
*/
u64 logged_trans ;
2008-09-11 23:53:12 +04:00
2008-09-29 23:18:18 +04:00
/* total number of bytes pending delalloc, used by stat to calc the
* real block usage of the file
*/
2008-02-08 21:49:28 +03:00
u64 delalloc_bytes ;
2008-09-29 23:18:18 +04:00
2009-02-20 19:00:09 +03:00
/* total number of bytes that may be used for this inode for
* delalloc
*/
u64 reserved_bytes ;
2008-09-29 23:18:18 +04:00
/*
* the size of the file stored in the metadata on disk . data = ordered
* means the in - memory i_size might be larger than the size on disk
* because not all the blocks are written yet .
*/
2008-07-17 20:54:05 +04:00
u64 disk_i_size ;
2008-09-29 23:18:18 +04:00
/* flags field from the on disk inode */
2008-01-08 23:54:37 +03:00
u32 flags ;
2008-07-24 20:12:38 +04:00
/*
* if this is a directory then index_cnt is the counter for the index
* number for new files that are created
*/
u64 index_cnt ;
2008-09-29 23:18:18 +04:00
2008-12-12 00:30:39 +03:00
/* the start of block group preferred for allocations. */
u64 block_group ;
2009-03-24 17:24:20 +03:00
/* the fsync log has some corner cases that mean we have to check
* directories to see if any unlinks have been done before
* the directory was logged . See tree - log . c for all the
* details
*/
u64 last_unlink_trans ;
2009-09-12 00:12:44 +04:00
/*
2009-10-08 21:34:05 +04:00
* Counters to keep track of the number of extent item ' s we may use due
* to delalloc and such . outstanding_extents is the number of extent
* items we think we ' ll end up using , and reserved_extents is the number
* of extent items we ' ve reserved metadata for .
2009-09-12 00:12:44 +04:00
*/
2010-05-16 18:48:47 +04:00
atomic_t outstanding_extents ;
2011-01-26 00:30:38 +03:00
atomic_t reserved_extents ;
2009-09-12 00:12:44 +04:00
2009-03-31 21:27:11 +04:00
/*
* ordered_data_close is set by truncate when a file that used
* to have good data has been truncated to zero . When it is set
* the btrfs file release call will add this inode to the
* ordered operations list so that we make sure to flush out any
* new data the application may have written before commit .
*
* yes , its silly to have a single bitflag , but we might grow more
* of these .
*/
unsigned ordered_data_close : 1 ;
2010-05-16 18:49:58 +04:00
unsigned orphan_meta_reserved : 1 ;
2009-09-21 23:56:00 +04:00
unsigned dummy_inode : 1 ;
2009-03-31 21:27:11 +04:00
2010-03-11 17:42:04 +03:00
/*
* always compress this one file
*/
2010-12-17 09:21:50 +03:00
unsigned force_compress : 4 ;
2010-03-11 17:42:04 +03:00
2008-09-29 23:18:18 +04:00
struct inode vfs_inode ;
2007-04-02 18:50:19 +04:00
} ;
2008-07-17 20:54:05 +04:00
2007-04-02 18:50:19 +04:00
static inline struct btrfs_inode * BTRFS_I ( struct inode * inode )
{
return container_of ( inode , struct btrfs_inode , vfs_inode ) ;
}
2008-07-17 20:54:05 +04:00
static inline void btrfs_i_size_write ( struct inode * inode , u64 size )
{
2009-11-12 12:34:21 +03:00
i_size_write ( inode , size ) ;
2008-07-17 20:54:05 +04:00
BTRFS_I ( inode ) - > disk_i_size = size ;
}
2007-04-02 18:50:19 +04:00
# endif