i40e: Fix tunnel checksum offload with fragmented traffic

commit 2c64820 upstream.

Fix checksum offload on VXLAN tunnels.
In case, when mpls protocol is not used, set l4 header to transport
header of skb. This fixes case, when user tries to offload checksums
of VXLAN tunneled traffic.

Steps for reproduction (requires link partner with tunnels):
ip l s enp130s0f0 up
ip a f enp130s0f0
ip a a 10.10.110.2/24 dev enp130s0f0
ip l s enp130s0f0 mtu 1600
ip link add vxlan12_sut type vxlan id 12 group 238.168.100.100 dev \
enp130s0f0 dstport 4789
ip l s vxlan12_sut up
ip a a 20.10.110.2/24 dev vxlan12_sut
iperf3 -c 20.10.110.1 #should connect

Without this patch, TX descriptor was using wrong data, due to
l4 header pointing wrong address. NIC would then drop those packets
internally, due to incorrect TX descriptor data, which increased
GLV_TEPC register.

Fixes: b4fb2d3 ("i40e: Add support for MPLS + TSO")
Signed-off-by: Przemyslaw Patynowski <przemyslawx.patynowski@intel.com>
Signed-off-by: Mateusz Palczewski <mateusz.palczewski@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/net/ethernet/intel/i40e/i40e_txrx.c

@@ -3204,11 +3204,13 @@ static int i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
 
 	protocol = vlan_get_protocol(skb);
 
-	if (eth_p_mpls(protocol))
+	if (eth_p_mpls(protocol)) {
 		ip.hdr = skb_inner_network_header(skb);
-	else
+		l4.hdr = skb_checksum_start(skb);
+	} else {
 		ip.hdr = skb_network_header(skb);
-	l4.hdr = skb_checksum_start(skb);
+		l4.hdr = skb_transport_header(skb);
+	}
 
 	/* set the tx_flags to indicate the IP protocol type. this is
 	 * required so that checksum header computation below is accurate.