net_sched: move the empty tp check from ->destroy() to ->delete()

We could have a race condition where in ->classify() path we
dereference tp->root and meanwhile a parallel ->destroy() makes it
a NULL. Daniel cured this bug in commit d936377414fa
("net, sched: respect rcu grace period on cls destruction").

This happens when ->destroy() is called for deleting a filter to
check if we are the last one in tp, this tp is still linked and
visible at that time. The root cause of this problem is the semantic
of ->destroy(), it does two things (for non-force case):

1) check if tp is empty
2) if tp is empty we could really destroy it

and its caller, if cares, needs to check its return value to see if it
is really destroyed. Therefore we can't unlink tp unless we know it is
empty.

As suggested by Daniel, we could actually move the test logic to ->delete()
so that we can safely unlink tp after ->delete() tells us the last one is
just deleted and before ->destroy().

Fixes: 1e052be69d04 ("net_sched: destroy proto tp when all filters are gone")
Cc: Roi Dayan <roid@mellanox.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 17 insertions, 12 deletions

diff --git a/net/sched/cls_fw.c b/net/sched/cls_fw.c
index 996209083c6b..78ccb4a85a4d 100644
--- a/net/sched/cls_fw.c
+++ b/net/sched/cls_fw.c
@@ -127,20 +127,14 @@ static void fw_delete_filter(struct rcu_head *head)
 	kfree(f);
 }
 
-static bool fw_destroy(struct tcf_proto *tp, bool force)
+static void fw_destroy(struct tcf_proto *tp)
 {
 	struct fw_head *head = rtnl_dereference(tp->root);
 	struct fw_filter *f;
 	int h;
 
 	if (head == NULL)
-		return true;
-
-	if (!force) {
-		for (h = 0; h < HTSIZE; h++)
-			if (rcu_access_pointer(head->ht[h]))
-				return false;
-	}
+		return;
 
 	for (h = 0; h < HTSIZE; h++) {
 		while ((f = rtnl_dereference(head->ht[h])) != NULL) {
@@ -152,15 +146,16 @@ static bool fw_destroy(struct tcf_proto *tp, bool force)
 	}
 	RCU_INIT_POINTER(tp->root, NULL);
 	kfree_rcu(head, rcu);
-	return true;
 }
 
-static int fw_delete(struct tcf_proto *tp, unsigned long arg)
+static int fw_delete(struct tcf_proto *tp, unsigned long arg, bool *last)
 {
 	struct fw_head *head = rtnl_dereference(tp->root);
 	struct fw_filter *f = (struct fw_filter *)arg;
 	struct fw_filter __rcu **fp;
 	struct fw_filter *pfp;
+	int ret = -EINVAL;
+	int h;
 
 	if (head == NULL || f == NULL)
 		goto out;
@@ -173,11 +168,21 @@ static int fw_delete(struct tcf_proto *tp, unsigned long arg)
 			RCU_INIT_POINTER(*fp, rtnl_dereference(f->next));
 			tcf_unbind_filter(tp, &f->res);
 			call_rcu(&f->rcu, fw_delete_filter);
-			return 0;
+			ret = 0;
+			break;
 		}
 	}
+
+	*last = true;
+	for (h = 0; h < HTSIZE; h++) {
+		if (rcu_access_pointer(head->ht[h])) {
+			*last = false;
+			break;
+		}
+	}
+
 out:
-	return -EINVAL;
+	return ret;
 }
 
 static const struct nla_policy fw_policy[TCA_FW_MAX + 1] = {