At least for the vm workload. Currently on fsync we will 1) Truncate all items in the log tree for the given inode if they exist and 2) Copy all items for a given inode into the log The problem with this is that for things like VMs you can have lots of extents from the fragmented writing behavior, and worst yet you may have only modified a few extents, not the entire thing. This patch fixes this problem by tracking which transid modified our extent, and then when we do the tree logging we find all of the extents we've modified in our current transaction, sort them and commit them. We also only truncate up to the xattrs of the inode and copy that stuff in normally, and then just drop any extents in the range we have that exist in the log already. Here are some numbers of a 50 meg fio job that does random writes and fsync()s after every write Original Patched SATA drive 82KB/s 140KB/s Fusion drive 431KB/s 2532KB/s So around 2-6 times faster depending on your hardware. There are a few corner cases, for example if you truncate at all we have to do it the old way since there is no way to be sure what is in the log is ok. This probably could be done smarter, but if you write-fsync-truncate-write-fsync you deserve what you get. All this work is in RAM of course so if your inode gets evicted from cache and you read it in and fsync it we'll do it the slow way if we are still in the same transaction that we last modified the inode in. The biggest cool part of this is that it requires no changes to the recovery code, so if you fsync with this patch and crash and load an old kernel, it will run the recovery and be a-ok. I have tested this pretty thoroughly with an fsync tester and everything comes back fine, as well as xfstests. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
218 lines
5.7 KiB
C
218 lines
5.7 KiB
C
/*
|
|
* 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.
|
|
*/
|
|
|
|
#ifndef __BTRFS_I__
|
|
#define __BTRFS_I__
|
|
|
|
#include "extent_map.h"
|
|
#include "extent_io.h"
|
|
#include "ordered-data.h"
|
|
#include "delayed-inode.h"
|
|
|
|
/*
|
|
* 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.
|
|
*/
|
|
#define BTRFS_INODE_ORDERED_DATA_CLOSE 0
|
|
#define BTRFS_INODE_ORPHAN_META_RESERVED 1
|
|
#define BTRFS_INODE_DUMMY 2
|
|
#define BTRFS_INODE_IN_DEFRAG 3
|
|
#define BTRFS_INODE_DELALLOC_META_RESERVED 4
|
|
#define BTRFS_INODE_HAS_ORPHAN_ITEM 5
|
|
#define BTRFS_INODE_HAS_ASYNC_EXTENT 6
|
|
#define BTRFS_INODE_NEEDS_FULL_SYNC 7
|
|
|
|
/* in memory btrfs inode */
|
|
struct btrfs_inode {
|
|
/* which subvolume this inode belongs to */
|
|
struct btrfs_root *root;
|
|
|
|
/* key used to find this inode on disk. This is used by the code
|
|
* to read in roots of subvolumes
|
|
*/
|
|
struct btrfs_key location;
|
|
|
|
/* Lock for counters */
|
|
spinlock_t lock;
|
|
|
|
/* the extent_tree has caches of all the extent mappings to disk */
|
|
struct extent_map_tree extent_tree;
|
|
|
|
/* the io_tree does range state (DIRTY, LOCKED etc) */
|
|
struct extent_io_tree io_tree;
|
|
|
|
/* special utility tree used to record which mirrors have already been
|
|
* tried when checksums fail for a given block
|
|
*/
|
|
struct extent_io_tree io_failure_tree;
|
|
|
|
/* held while logging the inode in tree-log.c */
|
|
struct mutex log_mutex;
|
|
|
|
/* held while doing delalloc reservations */
|
|
struct mutex delalloc_mutex;
|
|
|
|
/* used to order data wrt metadata */
|
|
struct btrfs_ordered_inode_tree ordered_tree;
|
|
|
|
/* 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.
|
|
*/
|
|
struct list_head delalloc_inodes;
|
|
|
|
/*
|
|
* list for tracking inodes that must be sent to disk before a
|
|
* rename or truncate commit
|
|
*/
|
|
struct list_head ordered_operations;
|
|
|
|
/* node for the red-black tree that links inodes in subvolume root */
|
|
struct rb_node rb_node;
|
|
|
|
unsigned long runtime_flags;
|
|
|
|
/* full 64 bit generation number, struct vfs_inode doesn't have a big
|
|
* enough field for this.
|
|
*/
|
|
u64 generation;
|
|
|
|
/*
|
|
* transid of the trans_handle that last modified this inode
|
|
*/
|
|
u64 last_trans;
|
|
|
|
/*
|
|
* log transid when this inode was last modified
|
|
*/
|
|
u64 last_sub_trans;
|
|
|
|
/*
|
|
* transid that last logged this inode
|
|
*/
|
|
u64 logged_trans;
|
|
|
|
/* total number of bytes pending delalloc, used by stat to calc the
|
|
* real block usage of the file
|
|
*/
|
|
u64 delalloc_bytes;
|
|
|
|
/*
|
|
* 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.
|
|
*/
|
|
u64 disk_i_size;
|
|
|
|
/*
|
|
* if this is a directory then index_cnt is the counter for the index
|
|
* number for new files that are created
|
|
*/
|
|
u64 index_cnt;
|
|
|
|
/* 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;
|
|
|
|
/*
|
|
* Number of bytes outstanding that are going to need csums. This is
|
|
* used in ENOSPC accounting.
|
|
*/
|
|
u64 csum_bytes;
|
|
|
|
/* flags field from the on disk inode */
|
|
u32 flags;
|
|
|
|
/*
|
|
* 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.
|
|
*/
|
|
unsigned outstanding_extents;
|
|
unsigned reserved_extents;
|
|
|
|
/*
|
|
* always compress this one file
|
|
*/
|
|
unsigned force_compress;
|
|
|
|
struct btrfs_delayed_node *delayed_node;
|
|
|
|
struct inode vfs_inode;
|
|
};
|
|
|
|
extern unsigned char btrfs_filetype_table[];
|
|
|
|
static inline struct btrfs_inode *BTRFS_I(struct inode *inode)
|
|
{
|
|
return container_of(inode, struct btrfs_inode, vfs_inode);
|
|
}
|
|
|
|
static inline u64 btrfs_ino(struct inode *inode)
|
|
{
|
|
u64 ino = BTRFS_I(inode)->location.objectid;
|
|
|
|
/*
|
|
* !ino: btree_inode
|
|
* type == BTRFS_ROOT_ITEM_KEY: subvol dir
|
|
*/
|
|
if (!ino || BTRFS_I(inode)->location.type == BTRFS_ROOT_ITEM_KEY)
|
|
ino = inode->i_ino;
|
|
return ino;
|
|
}
|
|
|
|
static inline void btrfs_i_size_write(struct inode *inode, u64 size)
|
|
{
|
|
i_size_write(inode, size);
|
|
BTRFS_I(inode)->disk_i_size = size;
|
|
}
|
|
|
|
static inline bool btrfs_is_free_space_inode(struct inode *inode)
|
|
{
|
|
struct btrfs_root *root = BTRFS_I(inode)->root;
|
|
|
|
if (root == root->fs_info->tree_root &&
|
|
btrfs_ino(inode) != BTRFS_BTREE_INODE_OBJECTID)
|
|
return true;
|
|
if (BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID)
|
|
return true;
|
|
return false;
|
|
}
|
|
|
|
static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
|
|
{
|
|
struct btrfs_root *root = BTRFS_I(inode)->root;
|
|
int ret = 0;
|
|
|
|
mutex_lock(&root->log_mutex);
|
|
if (BTRFS_I(inode)->logged_trans == generation &&
|
|
BTRFS_I(inode)->last_sub_trans <= root->last_log_commit)
|
|
ret = 1;
|
|
mutex_unlock(&root->log_mutex);
|
|
return ret;
|
|
}
|
|
|
|
#endif
|