From: Tao Ma <boyu...@taobao.com>
When we are using netpoll, we don't go through the normal
transmit process. In this case, bond_select_queue is not called
and qdisc_skb_cb(skb)->slave_dev_queue_mapping isn't set.
So when netpoll_send_skb_on_dev calls ndo_start_xmit and we
enter bond_dev_queue_xmit, we will set skb->queue_mapping to
an invalid value and in some cases cause the driver panic the
kernel(We meet with bnx2 panic because of a very large queue_mapping).
This patch skip skb->queue_mapping if we find we are in netpoll.
CC: "David S. Miller" <da...@davemloft.net>
CC: Cong Wang <amw...@redhat.com>
CC: Eric Dumazet <eric.duma...@gmail.com>
Signed-off-by: Tao Ma <boyu...@taobao.com>
---
drivers/net/bonding/bond_main.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 07f257d..97b2f52 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -405,12 +405,13 @@ int bond_dev_queue_xmit(struct bonding *bond, struct
sk_buff *skb,
BUILD_BUG_ON(sizeof(skb->queue_mapping) !=
sizeof(qdisc_skb_cb(skb)->slave_dev_queue_mapping));
- skb->queue_mapping = qdisc_skb_cb(skb)->slave_dev_queue_mapping;
if (unlikely(netpoll_tx_running(bond->dev)))
bond_netpoll_send_skb(bond_get_slave_by_dev(bond, slave_dev),
skb);
- else
+ else {
+ skb->queue_mapping = qdisc_skb_cb(skb)->slave_dev_queue_mapping;
dev_queue_xmit(skb);
+ }
return 0;
}
--
1.7.0.4
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
