diff options
author | Christoph Hellwig <hch@infradead.org> | 2011-12-19 00:00:09 +0400 |
---|---|---|
committer | Ben Myers <bpm@sgi.com> | 2012-01-18 01:06:45 +0400 |
commit | 474fce067521a40dbacc722e8ba119e81c2d31bf (patch) | |
tree | fd923aa42a5304182e8a8c64ca5d130f9afef286 /fs/xfs/xfs_super.c | |
parent | 49e4c70e52a2bc2090e5a4e003e2888af21d6a2b (diff) | |
download | linux-474fce067521a40dbacc722e8ba119e81c2d31bf.tar.xz |
xfs: replace i_flock with a sleeping bitlock
We almost never block on i_flock, the exception is synchronous inode
flushing. Instead of bloating the inode with a 16/24-byte completion
that we abuse as a semaphore just implement it as a bitlock that uses
a bit waitqueue for the rare sleeping path. This primarily is a
tradeoff between a much smaller inode and a faster non-blocking
path vs faster wakeups, and we are much better off with the former.
A small downside is that we will lose lockdep checking for i_flock, but
given that it's always taken inside the ilock that should be acceptable.
Note that for example the inode writeback locking is implemented in a
very similar way.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
Diffstat (limited to 'fs/xfs/xfs_super.c')
-rw-r--r-- | fs/xfs/xfs_super.c | 7 |
1 files changed, 0 insertions, 7 deletions
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 281961c1d81a..6851fa7b1afa 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -829,13 +829,6 @@ xfs_fs_inode_init_once( atomic_set(&ip->i_pincount, 0); spin_lock_init(&ip->i_flags_lock); init_waitqueue_head(&ip->i_ipin_wait); - /* - * Because we want to use a counting completion, complete - * the flush completion once to allow a single access to - * the flush completion without blocking. - */ - init_completion(&ip->i_flush); - complete(&ip->i_flush); mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, "xfsino", ip->i_ino); |