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, 3 insertions, 7 deletions

diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c
index 3e7bd7801aa8..31ee3404aeb4 100644
--- a/net/sched/cls_flower.c
+++ b/net/sched/cls_flower.c
@@ -328,21 +328,16 @@ static void fl_destroy_rcu(struct rcu_head *rcu)
 	schedule_work(&head->work);
 }
 
-static bool fl_destroy(struct tcf_proto *tp, bool force)
+static void fl_destroy(struct tcf_proto *tp)
 {
 	struct cls_fl_head *head = rtnl_dereference(tp->root);
 	struct cls_fl_filter *f, *next;
 
-	if (!force && !list_empty(&head->filters))
-		return false;
-
 	list_for_each_entry_safe(f, next, &head->filters, list)
 		__fl_delete(tp, f);
 
 	__module_get(THIS_MODULE);
 	call_rcu(&head->rcu, fl_destroy_rcu);
-
-	return true;
 }
 
 static unsigned long fl_get(struct tcf_proto *tp, u32 handle)
@@ -947,7 +942,7 @@ errout_tb:
 	return err;
 }
 
-static int fl_delete(struct tcf_proto *tp, unsigned long arg)
+static int fl_delete(struct tcf_proto *tp, unsigned long arg, bool *last)
 {
 	struct cls_fl_head *head = rtnl_dereference(tp->root);
 	struct cls_fl_filter *f = (struct cls_fl_filter *) arg;
@@ -956,6 +951,7 @@ static int fl_delete(struct tcf_proto *tp, unsigned long arg)
 		rhashtable_remove_fast(&head->ht, &f->ht_node,
 				       head->ht_params);
 	__fl_delete(tp, f);
+	*last = list_empty(&head->filters);
 	return 0;
 }