fix d_walk()/non-delayed __d_free() race

Ascend-to-parent logics in d_walk() depends on all encountered child
dentries not getting freed without an RCU delay.  Unfortunately, in
quite a few cases it is not true, with hard-to-hit oopsable race as
the result.

Fortunately, the fix is simiple; right now the rule is "if it ever
been hashed, freeing must be delayed" and changing it to "if it
ever had a parent, freeing must be delayed" closes that hole and
covers all cases the old rule used to cover.  Moreover, pipes and
sockets remain _not_ covered, so we do not introduce RCU delay in
the cases which are the reason for having that delay conditional
in the first place.

Cc: stable@vger.kernel.org # v3.2+ (and watch out for __d_materialise_dentry())
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

diff --git a/fs/dcache.c b/fs/dcache.c
index ad4a542..817c243 100644 (file)
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1636,7 +1636,7 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
        struct dentry *dentry = __d_alloc(parent->d_sb, name);
        if (!dentry)
                return NULL;
-
+       dentry->d_flags |= DCACHE_RCUACCESS;
        spin_lock(&parent->d_lock);
        /*
         * don't need child lock because it is not subject
@@ -2358,7 +2358,6 @@ static void __d_rehash(struct dentry * entry, struct hlist_bl_head *b)
 {
        BUG_ON(!d_unhashed(entry));
        hlist_bl_lock(b);
-       entry->d_flags |= DCACHE_RCUACCESS;
        hlist_bl_add_head_rcu(&entry->d_hash, b);
        hlist_bl_unlock(b);
 }
@@ -2843,6 +2842,7 @@ static void __d_move(struct dentry *dentry, struct dentry *target,
        /* ... and switch them in the tree */
        if (IS_ROOT(dentry)) {
                /* splicing a tree */
+               dentry->d_flags |= DCACHE_RCUACCESS;
                dentry->d_parent = target->d_parent;
                target->d_parent = target;
                list_del_init(&target->d_child);