[PATCH] ipv6: do not increment mac header when it's unset

From: Jason A. Donenfeld
Date: Fri Oct 21 2016 - 05:28:54 EST


Otherwise we'll overflow the integer. This occurs when layer 3 tunneled
packets are handed off to the IPv6 layer.

Signed-off-by: Jason A. Donenfeld <Jason@xxxxxxxxx>
---
net/ipv6/reassembly.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/ipv6/reassembly.c b/net/ipv6/reassembly.c
index e2ea311..a943203 100644
--- a/net/ipv6/reassembly.c
+++ b/net/ipv6/reassembly.c
@@ -456,7 +456,8 @@ static int ip6_frag_reasm(struct frag_queue *fq, struct sk_buff *prev,
skb_network_header(head)[nhoff] = skb_transport_header(head)[0];
memmove(head->head + sizeof(struct frag_hdr), head->head,
(head->data - head->head) - sizeof(struct frag_hdr));
- head->mac_header += sizeof(struct frag_hdr);
+ if (skb_mac_header_was_set(head))
+ head->mac_header += sizeof(struct frag_hdr);
head->network_header += sizeof(struct frag_hdr);

skb_reset_transport_header(head);
--
2.10.1