net/core: move vlan_depth out of while loop in skb_network_protocol()
authorLi RongQing <roy.qing.li@gmail.com>
Mon, 11 Mar 2013 20:30:44 +0000 (20:30 +0000)
committerDavid S. Miller <davem@davemloft.net>
Tue, 12 Mar 2013 15:47:40 +0000 (11:47 -0400)
[ Bug added added in commit 05e8ef4ab2d8087d (net: factor out
  skb_mac_gso_segment() from skb_gso_segment() ) ]

move vlan_depth out of while loop, or else vlan_depth always is ETH_HLEN,
can not be increased, and lead to infinite loop when frame has two vlan headers.

Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/dev.c

index dffbef70cd31253625da50819a171fbc59b386a5..d540ced1f6c66dc5b84a8ef06c24199e747a0f86 100644 (file)
@@ -2219,9 +2219,9 @@ struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb,
        struct sk_buff *segs = ERR_PTR(-EPROTONOSUPPORT);
        struct packet_offload *ptype;
        __be16 type = skb->protocol;
+       int vlan_depth = ETH_HLEN;
 
        while (type == htons(ETH_P_8021Q)) {
-               int vlan_depth = ETH_HLEN;
                struct vlan_hdr *vh;
 
                if (unlikely(!pskb_may_pull(skb, vlan_depth + VLAN_HLEN)))