diff options
| author | Waiman Long <Waiman.Long@hp.com> | 2013-09-02 20:29:22 +0200 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2013-09-02 20:29:22 +0200 |
| commit | df3d0bbcdb2cafa23a70223d806655bd37e64a9b (patch) | |
| tree | e25e1f06ad304ab4c87293f37961b8e97d7fd242 /fs/dcache.c | |
| parent | lockref: add 'lockref_get_or_lock() helper (diff) | |
| download | linux-df3d0bbcdb2cafa23a70223d806655bd37e64a9b.tar.xz linux-df3d0bbcdb2cafa23a70223d806655bd37e64a9b.zip | |
vfs: use lockref_get_not_zero() for optimistic lockless dget_parent()
A valid parent pointer is always going to have a non-zero reference
count, but if we look up the parent optimistically without locking, we
have to protect against the (very unlikely) race against renaming
changing the parent from under us.
We do that by using lockref_get_not_zero(), and then re-checking the
parent pointer after getting a valid reference.
[ This is a re-implementation of a chunk from the original patch by
Waiman Long: "dcache: Enable lockless update of dentry's refcount".
I've completely rewritten the patch-series and split it up, but I'm
attributing this part to Waiman as it's close enough to his earlier
patch - Linus ]
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'fs/dcache.c')
| -rw-r--r-- | fs/dcache.c | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/fs/dcache.c b/fs/dcache.c index b949af850cd6..2d244227999d 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -611,8 +611,23 @@ static inline void __dget(struct dentry *dentry) struct dentry *dget_parent(struct dentry *dentry) { + int gotref; struct dentry *ret; + /* + * Do optimistic parent lookup without any + * locking. + */ + rcu_read_lock(); + ret = ACCESS_ONCE(dentry->d_parent); + gotref = lockref_get_not_zero(&ret->d_lockref); + rcu_read_unlock(); + if (likely(gotref)) { + if (likely(ret == ACCESS_ONCE(dentry->d_parent))) + return ret; + dput(ret); + } + repeat: /* * Don't need rcu_dereference because we re-check it was correct under |