cxgb4: don't offload Rx checksums for IPv6 fragments

The checksum provided by the device doesn't include the L3 headers,
as IPv6 expects.

Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 4 insertions, 3 deletions

diff --git a/drivers/net/cxgb4/sge.c b/drivers/net/cxgb4/sge.c
index 4388f72d586a..304302a71dbc 100644
--- a/drivers/net/cxgb4/sge.c
+++ b/drivers/net/cxgb4/sge.c
@@ -1593,14 +1593,15 @@ int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
 
         if (csum_ok && (pi->rx_offload & RX_CSO) &&
             (pkt->l2info & htonl(RXF_UDP | RXF_TCP))) {
-                if (!pkt->ip_frag)
+                if (!pkt->ip_frag) {
                         skb->ip_summed = CHECKSUM_UNNECESSARY;
-                else {
+                        rxq->stats.rx_cso++;
+                } else if (pkt->l2info & htonl(RXF_IP)) {
                         __sum16 c = (__force __sum16)pkt->csum;
                         skb->csum = csum_unfold(c);
                         skb->ip_summed = CHECKSUM_COMPLETE;
+                        rxq->stats.rx_cso++;
                 }
-                rxq->stats.rx_cso++;
         } else
                 skb->ip_summed = CHECKSUM_NONE;