net/packet: fix overflow in tpacket_rcv

Using tp_reserve to calculate netoff can overflow as
tp_reserve is unsigned int and netoff is unsigned short.

This may lead to macoff receving a smaller value then
sizeof(struct virtio_net_hdr), and if po->has_vnet_hdr
is set, an out-of-bounds write will occur when
calling virtio_net_hdr_from_skb.

The bug is fixed by converting netoff to unsigned int
and checking if it exceeds USHRT_MAX.

This addresses CVE-2020-14386

Fixes: 8913336a7e8d ("packet: add PACKET_RESERVE sockopt")
Signed-off-by: Or Cohen <>
Signed-off-by: Eric Dumazet <>
(backported from
Signed-off-by: Robert Kolchmeyer <>

TEST=Manually tried the reproducer before and after this fix.
RELEASE_NOTE=Fixed overflow in tpacket_rcv, which caused CVE-2020-14386.

Change-Id: Ifc451bdd97c124917e49023db2d7f3a0356d4801
Tested-by: Robert Kolchmeyer <>
Reviewed-by: Roy Yang <>
(cherry picked from commit c46f8a87c479cd2e964513992d3eab92eff5c160)
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 387589a..ba741d9 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -2199,7 +2199,8 @@
 	int skb_len = skb->len;
 	unsigned int snaplen, res;
 	unsigned long status = TP_STATUS_USER;
-	unsigned short macoff, netoff, hdrlen;
+	unsigned short macoff, hdrlen;
+	unsigned int netoff;
 	struct sk_buff *copy_skb = NULL;
 	struct timespec ts;
 	__u32 ts_status;
@@ -2261,6 +2262,12 @@
 		macoff = netoff - maclen;
+	if (netoff > USHRT_MAX) {
+		spin_lock(&sk->sk_receive_queue.lock);
+		po->stats.stats1.tp_drops++;
+		spin_unlock(&sk->sk_receive_queue.lock);
+		goto drop_n_restore;
+	}
 	if (po->tp_version <= TPACKET_V2) {
 		if (macoff + snaplen > po->rx_ring.frame_size) {
 			if (po->copy_thresh &&