Btrfs: use percpu counter for dirty metadata count

Browse source

->dirty_metadata_bytes is accessed very frequently, so use percpu
counter instead of the u64 variant to reduce the contention of
the lock.

This patch also fixed the problem that we access it without
lock protection in __btrfs_btree_balance_dirty(), which may
cause we skip the dirty pages flush.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>

...

This commit is contained in:

Miao Xie 2013-01-29 10:09:20 +00:00 • committed by Josef Bacik

parent c018daecea

commit e2d845211e

3 changed files with 42 additions and 40 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

9

fs/btrfs/ctree.h

View file

64

fs/btrfs/disk-io.c

View file

9

fs/btrfs/extent_io.c

View file