mac80211: don't leave skb->next/prev pointing to stack
authorJohannes Berg <johannes.berg@intel.com>
Fri, 20 Mar 2020 09:20:23 +0000 (10:20 +0100)
committerJohannes Berg <johannes.berg@intel.com>
Fri, 20 Mar 2020 13:42:19 +0000 (14:42 +0100)
In beacon protection, don't leave skb->next/prev pointing to the
on-stack list, even if that's actually harmless since we don't use
them again afterwards.

While at it, check that the SKB on the list is still the same, as
that's required here. If not, the encryption (protection) code is
buggy.

Fixes: 0a3a84360b37 ("mac80211: Beacon protection using the new BIGTK (AP)")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Link: https://lore.kernel.org/r/20200320102021.1be7823fc05e.Ia89fb79a0469d32137c9a04315a1d2dfc7b7d6f5@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
net/mac80211/tx.c

index 8314738..49d3593 100644 (file)
@@ -4670,6 +4670,7 @@ static int ieee80211_beacon_protect(struct sk_buff *skb,
 {
        ieee80211_tx_result res;
        struct ieee80211_tx_data tx;
+       struct sk_buff *check_skb;
 
        memset(&tx, 0, sizeof(tx));
        tx.key = rcu_dereference(sdata->default_beacon_key);
@@ -4680,8 +4681,11 @@ static int ieee80211_beacon_protect(struct sk_buff *skb,
        __skb_queue_head_init(&tx.skbs);
        __skb_queue_tail(&tx.skbs, skb);
        res = ieee80211_tx_h_encrypt(&tx);
+       check_skb = __skb_dequeue(&tx.skbs);
+       /* we may crash after this, but it'd be a bug in crypto */
+       WARN_ON(check_skb != skb);
        if (WARN_ON_ONCE(res != TX_CONTINUE))
-               return -1;
+               return -EINVAL;
 
        return 0;
 }