| Message ID | 871txnqn7j.fsf_-_@x220.int.ebiederm.org |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
diff --git a/net/core/netpoll.c b/net/core/netpoll.c index f3012800a81b..2136c9aacdbf 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -112,7 +112,7 @@ static void queue_process(struct work_struct *work) struct netdev_queue *txq; if (!netif_device_present(dev) || !netif_running(dev)) { - __kfree_skb(skb); + kfree_skb(skb); continue; } @@ -332,7 +332,7 @@ void netpoll_send_skb_on_dev(struct netpoll *np, struct sk_buff *skb, npinfo = rcu_dereference_bh(np->dev->npinfo); if (!npinfo || !netif_running(dev) || !netif_device_present(dev)) { - __kfree_skb(skb); + dev_kfree_skb_irq(skb); return; }
Remove the assumption that the skbs that make it to netpoll_send_skb_on_dev are allocated with find_skb, such that skb->users == 1 and nothing is attached that would prevent the skbs from being freed from hard irq context. Remove this assumption by replacing __kfree_skb on error paths with dev_kfree_skb_irq (in hard irq context) and kfree_skb (in process context). Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> --- net/core/netpoll.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)