btrfs: make inode ref log recovery faster

When we recover from crash via write-ahead log tree and process
the inode refs, for each btrfs_inode_ref item, we will
1) check if we already have a perfect match in fs/file tree, if
   we have, then we're done.
2) search the corresponding back reference in fs/file tree, and
   check all the names in this back reference to see if they are
   also in the log to avoid conflict corners.
3) recover the logged inode refs to fs/file tree.

In current btrfs, however,
- for 2)'s check, once is enough, since the checked back reference
  will remain unchanged after processing all the inode refs belonged
  to the key.
- it has no need to do another 1) between 2) and 3).

I've made a small test to show how it improves,

$dd if=/dev/zero of=foobar bs=4K count=1
$sync
$make 100 hard links continuously, like ln foobar link_i
$fsync foobar
$echo b > /proc/sysrq-trigger
after reboot
$time mount DEV PATH

without patch:
real    0m0.285s
user    0m0.001s
sys     0m0.009s

with patch:
real    0m0.123s
user    0m0.000s
sys     0m0.010s

Changelog v1->v2:
- fix double free - pointed by David Sterba
Changelog v2->v3:
- adjust free order

Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This commit is contained in:
liubo 2011-03-26 08:01:12 -04:00 committed by root
parent f7039b1d5c
commit c622ae6085

View File

@ -799,12 +799,12 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
struct inode *dir;
int ret;
struct btrfs_inode_ref *ref;
struct btrfs_dir_item *di;
struct inode *inode;
char *name;
int namelen;
unsigned long ref_ptr;
unsigned long ref_end;
int search_done = 0;
/*
* it is possible that we didn't log all the parent directories
@ -845,7 +845,10 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
* existing back reference, and we don't want to create
* dangling pointers in the directory.
*/
conflict_again:
if (search_done)
goto insert;
ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
if (ret == 0) {
char *victim_name;
@ -886,37 +889,21 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
ret = btrfs_unlink_inode(trans, root, dir,
inode, victim_name,
victim_name_len);
kfree(victim_name);
btrfs_release_path(root, path);
goto conflict_again;
}
kfree(victim_name);
ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
}
BUG_ON(ret);
/*
* NOTE: we have searched root tree and checked the
* coresponding ref, it does not need to check again.
*/
search_done = 1;
}
btrfs_release_path(root, path);
/* look for a conflicting sequence number */
di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino,
btrfs_inode_ref_index(eb, ref),
name, namelen, 0);
if (di && !IS_ERR(di)) {
ret = drop_one_dir_item(trans, root, path, dir, di);
BUG_ON(ret);
}
btrfs_release_path(root, path);
/* look for a conflicting name */
di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
name, namelen, 0);
if (di && !IS_ERR(di)) {
ret = drop_one_dir_item(trans, root, path, dir, di);
BUG_ON(ret);
}
btrfs_release_path(root, path);
insert:
/* insert our name */
ret = btrfs_add_link(trans, dir, inode, name, namelen, 0,
btrfs_inode_ref_index(eb, ref));