netvsc: fix incorrect receive checksum offloading

PATCH In review Stephen Hemminger 2 participants 2 messages Oct 24, 2016 – Oct 27, 2016

Subsystems (3)

Networking Network drivers Hyper-V/Azure core and drivers

lore

Share & tools

lore

Stephen Hemminger

author

2016-10-24 04:34 UTC #0 lore raw

PATCH 1/1 netvsc: fix incorrect receive checksum offloading

From: Stephen Hemminger <[email protected]>

The Hyper-V netvsc driver was looking at the incorrect status bits in the checksum info. It was setting the receive checksum unnecessary flag based on the IP header checksum being correct. The checksum flag is skb is about TCP and UDP checksum status. Because of this bug, any packet received with bad TCP checksum would be passed up the stack and to the application causing data corruption. The problem is reproducible via netcat and netem.

This had a side effect of not doing receive checksum offload on IPv6. The driver was also also always doing checksum offload independent of the checksum setting done via ethtool.

Signed-off-by Stephen Hemminger <[email protected]>

---
Please apply to -stable as well!

 drivers/net/hyperv/netvsc_drv.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

drivers/net/hyperv/netvsc_drv.c +11 -8

index 5d6e75a..c71d966 100644
@@ -607,15 +607,18 @@ static struct sk_buff *netvsc_alloc_recv_skb(struct net_device *net,
 	       packet->total_data_buflen);
 
 	skb->protocol = eth_type_trans(skb, net);
-	if (csum_info) {
-		/* We only look at the IP checksum here.
-		 * Should we be dropping the packet if checksum
-		 * failed? How do we deal with other checksums - TCP/UDP?
-		 */
-		if (csum_info->receive.ip_checksum_succeeded)
+
+	/* skb is already created with CHECKSUM_NONE */
+	skb_checksum_none_assert(skb);
+
+	/*
+	 * In Linux, the IP checksum is always checked.
+	 * Do L4 checksum offload if enabled and present.
+	 */
+	if (csum_info && (net->features & NETIF_F_RXCSUM)) {
+		if (csum_info->receive.tcp_checksum_succeeded ||
+		    csum_info->receive.udp_checksum_succeeded)
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
-		else
-			skb->ip_summed = CHECKSUM_NONE;
 	}
 
 	if (vlan_tci & VLAN_TAG_PRESENT)

Signature · 1 line

2.9.3

David Miller

maintainer

2016-10-27 03:19 UTC #1 lore raw

From: Stephen Hemminger <[email protected]>
Date: Sun, 23 Oct 2016 21:32:47 -0700

Quoted 10 lines

From: Stephen Hemminger <[email protected]>
The Hyper-V netvsc driver was looking at the incorrect status bits in the checksum info. It was setting the receive checksum unnecessary flag based on the IP header checksum being correct. The checksum flag is skb is about TCP and UDP checksum status. Because of this bug, any packet received with bad TCP checksum would be passed up the stack and to the application causing data corruption. The problem is reproducible via netcat and netem.

Quoted 5 lines

This had a side effect of not doing receive checksum offload on IPv6. The driver was also also always doing checksum offload independent of the checksum setting done via ethtool.
Signed-off-by Stephen Hemminger <[email protected]>

Applied and queued up for -stable.