Skip to contentSkip to conversation

Q: spin_unlock(dentry) after lock_parent(dentry)

DISCUSSION J. R. Okajima 2 participants 4 messages Jun 3, 2014 – Jun 7, 2014 VFS

lore
Share & tools
lore
#0 lore raw

Hello Al Viro,

I have a question about spin_unlock(dentry) after lock_parent(dentry). In lock_parent(dentry), spin_unlock(dentry) is called. And spin_lock(dentry) is called again when (parent != dentry) is true. Otherwise, dentry left spin_unlock-ed and lock_parent() returns NULL.

Even in the case of lock_parent() returns NULL, shrink_dentry_list() calls spin_unlock(dentry). Is it balanced and correct?

J. R. Okajima

maintainer
#1 lore raw
Quoted 8 lines · J. R. Okajima

Hello Al Viro,

I have a question about spin_unlock(dentry) after lock_parent(dentry). In lock_parent(dentry), spin_unlock(dentry) is called. And spin_lock(dentry) is called again when (parent != dentry) is true. Otherwise, dentry left spin_unlock-ed and lock_parent() returns NULL.

Nope. You are misreading it. Note that in *all* cases the parent is locked. And NULL is returned exactly in the case when that parent is equal to dentry itself. IOW, in all cases, lock_parent(d) returns with d locked and after p = lock_parent(d) either
1) d->d_parent == d and p == NULL, or
2) d->d_parent == p, p != d and p is locked as well.
That holds whether we had succeeded with trylock or had to drop and regain the lock on d.

Quoted 2 lines

Even in the case of lock_parent() returns NULL, shrink_dentry_list() calls spin_unlock(dentry). Is it balanced and correct?

Yes.

#2 lore raw

Al Viro:

Quoted 2 lines

Nope. You are misreading it. Note that in *all* cases the parent is locked. And NULL is returned exactly in the case when that parent is

	:::

Thanks.
Please allow me another question.

In the case of concurrent dentry_kill() and shrink_dentry_list(), dentry_kill() kills the dentry and calls list_del(&dentry->d_u.d_child). lock_parent() (called by shrink_dentry_list()) doesn't check d_child nor DCACHE_DENTRY_KILLED. So shrink_dentry_list() may get a parent of a dead dentry and tries traversing via d_parent.
Is it correct to track d_parent after list_del(&dentry->d_u.d_child)? Should lock_parent() consider DCACHE_DENTRY_KILLED too?

fs/dcache.c +2 -1
@@ -533,7 +533,8 @@ failed:
 static inline struct dentry *lock_parent(struct dentry *dentry)
 {
 	struct dentry *parent = dentry->d_parent;
-	if (IS_ROOT(dentry))
+	if (IS_ROOT(dentry)
+	    || (dentry->d_flags & DCACHE_DENTRY_KILLED))
 		return NULL;
 	if (likely(spin_trylock(&parent->d_lock)))
 		return parent;

J. R. Okajima

#3 lore raw

Al Viro,

I think I could find a clearer problematic scenario. Can it happen in real world?

- /.../dirA/dirB/fileC exists.
- shrink_dentry_list() gets a list containing fileC and dirB, but not dirA
+ Round 1: fileC
- shrink_dentry_list() gets d_parent which is dirB
- dirB is DCACHE_SHRINK_LIST set, so __dentry_kill() is called but not dentry_free()
- shrink_dentry_list() gets higher d_parent which is dirA
- dirA is not DCACHE_SHRINK_LIST set, so both of __dentry_kill() and dentry_free() are is called
+ Round 2: dirB
- shrink_dentry_list() tries getting d_parent which is dirA
- dirA is freed by RCU. if RCU already activated __d_free(), BANG!

Quoted 11 lines

In the case of concurrent dentry_kill() and shrink_dentry_list(), dentry_kill() kills the dentry and calls list_del(&dentry->d_u.d_child). lock_parent() (called by shrink_dentry_list()) doesn't check d_child nor DCACHE_DENTRY_KILLED. So shrink_dentry_list() may get a parent of a dead dentry and tries traversing via d_parent.
Is it correct to track d_parent after list_del(&dentry->d_u.d_child)? Should lock_parent() consider DCACHE_DENTRY_KILLED too?

fs/dcache.c failed: · line 540

Quoted 9 lines
fs/dcache.c +2 -1
@@ -533,7 +533,8 @@ failed:
 static inline struct dentry *lock_parent(struct dentry *dentry)
 {
 	struct dentry *parent = dentry->d_parent;
-	if (IS_ROOT(dentry))
+	if (IS_ROOT(dentry)
+	    || (dentry->d_flags & DCACHE_DENTRY_KILLED))
 		return NULL;
 	if (likely(spin_trylock(&parent->d_lock)))
 		return parent;

J. R. Okajima