net: hns3: add support for TX hardware checksum offload

For the device that supports TX hardware checksum, the hardware
can calculate the checksum from the start and fill the checksum
to the offset position, which reduces the operations of
calculating the type and header length of L3/L4. So add this
feature for the HNS3 ethernet driver.

The previous simple BD description is unsuitable, rename it as
HW TX CSUM.

Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

1 files changed, 51 insertions, 11 deletions

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 1647877be148..904328e6f9e7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1055,15 +1055,31 @@ static int hns3_handle_vtags(struct hns3_enet_ring *tx_ring,
 	return 0;
 }
 
+/* check if the hardware is capable of checksum offloading */
+static bool hns3_check_hw_tx_csum(struct sk_buff *skb)
+{
+	struct hns3_nic_priv *priv = netdev_priv(skb->dev);
+
+	/* Kindly note, due to backward compatibility of the TX descriptor,
+	 * HW checksum of the non-IP packets and GSO packets is handled at
+	 * different place in the following code
+	 */
+	if (skb->csum_not_inet || skb_is_gso(skb) ||
+	    !test_bit(HNS3_NIC_STATE_HW_TX_CSUM_ENABLE, &priv->state))
+		return false;
+
+	return true;
+}
+
 static int hns3_fill_skb_desc(struct hns3_enet_ring *ring,
 			      struct sk_buff *skb, struct hns3_desc *desc)
 {
 	u32 ol_type_vlan_len_msec = 0;
 	u32 type_cs_vlan_tso = 0;
 	u32 paylen = skb->len;
+	u16 mss_hw_csum = 0;
 	u16 inner_vtag = 0;
 	u16 out_vtag = 0;
-	u16 mss = 0;
 	int ret;
 
 	ret = hns3_handle_vtags(ring, skb);
@@ -1088,6 +1104,17 @@ static int hns3_fill_skb_desc(struct hns3_enet_ring *ring,
 	if (skb->ip_summed == CHECKSUM_PARTIAL) {
 		u8 ol4_proto, il4_proto;
 
+		if (hns3_check_hw_tx_csum(skb)) {
+			/* set checksum start and offset, defined in 2 Bytes */
+			hns3_set_field(type_cs_vlan_tso, HNS3_TXD_CSUM_START_S,
+				       skb_checksum_start_offset(skb) >> 1);
+			hns3_set_field(ol_type_vlan_len_msec,
+				       HNS3_TXD_CSUM_OFFSET_S,
+				       skb->csum_offset >> 1);
+			mss_hw_csum |= BIT(HNS3_TXD_HW_CS_B);
+			goto out_hw_tx_csum;
+		}
+
 		skb_reset_mac_len(skb);
 
 		ret = hns3_get_l4_protocol(skb, &ol4_proto, &il4_proto);
@@ -1108,7 +1135,7 @@ static int hns3_fill_skb_desc(struct hns3_enet_ring *ring,
 			return ret;
 		}
 
-		ret = hns3_set_tso(skb, &paylen, &mss,
+		ret = hns3_set_tso(skb, &paylen, &mss_hw_csum,
 				   &type_cs_vlan_tso);
 		if (unlikely(ret < 0)) {
 			u64_stats_update_begin(&ring->syncp);
@@ -1118,12 +1145,13 @@ static int hns3_fill_skb_desc(struct hns3_enet_ring *ring,
 		}
 	}
 
+out_hw_tx_csum:
 	/* Set txbd */
 	desc->tx.ol_type_vlan_len_msec =
 		cpu_to_le32(ol_type_vlan_len_msec);
 	desc->tx.type_cs_vlan_tso_len = cpu_to_le32(type_cs_vlan_tso);
 	desc->tx.paylen = cpu_to_le32(paylen);
-	desc->tx.mss = cpu_to_le16(mss);
+	desc->tx.mss_hw_csum = cpu_to_le16(mss_hw_csum);
 	desc->tx.vlan_tag = cpu_to_le16(inner_vtag);
 	desc->tx.outer_vlan_tag = cpu_to_le16(out_vtag);
 
@@ -2326,8 +2354,7 @@ static void hns3_set_default_feature(struct net_device *netdev)
 
 	netdev->priv_flags |= IFF_UNICAST_FLT;
 
-	netdev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
-		NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_GSO |
+	netdev->hw_enc_features |= NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_GSO |
 		NETIF_F_GRO | NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_GRE |
 		NETIF_F_GSO_GRE_CSUM | NETIF_F_GSO_UDP_TUNNEL |
 		NETIF_F_GSO_UDP_TUNNEL_CSUM | NETIF_F_SCTP_CRC |
@@ -2335,8 +2362,7 @@ static void hns3_set_default_feature(struct net_device *netdev)
 
 	netdev->gso_partial_features |= NETIF_F_GSO_GRE_CSUM;
 
-	netdev->features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
-		NETIF_F_HW_VLAN_CTAG_FILTER |
+	netdev->features |= NETIF_F_HW_VLAN_CTAG_FILTER |
 		NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX |
 		NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_GSO |
 		NETIF_F_GRO | NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_GRE |
@@ -2344,16 +2370,15 @@ static void hns3_set_default_feature(struct net_device *netdev)
 		NETIF_F_GSO_UDP_TUNNEL_CSUM | NETIF_F_SCTP_CRC |
 		NETIF_F_FRAGLIST;
 
-	netdev->vlan_features |=
-		NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | NETIF_F_RXCSUM |
+	netdev->vlan_features |= NETIF_F_RXCSUM |
 		NETIF_F_SG | NETIF_F_GSO | NETIF_F_GRO |
 		NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_GRE |
 		NETIF_F_GSO_GRE_CSUM | NETIF_F_GSO_UDP_TUNNEL |
 		NETIF_F_GSO_UDP_TUNNEL_CSUM | NETIF_F_SCTP_CRC |
 		NETIF_F_FRAGLIST;
 
-	netdev->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
-		NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX |
+	netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_TX |
+		NETIF_F_HW_VLAN_CTAG_RX |
 		NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_GSO |
 		NETIF_F_GRO | NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_GRE |
 		NETIF_F_GSO_GRE_CSUM | NETIF_F_GSO_UDP_TUNNEL |
@@ -2376,6 +2401,18 @@ static void hns3_set_default_feature(struct net_device *netdev)
 		netdev->vlan_features |= NETIF_F_GSO_UDP_L4;
 		netdev->hw_enc_features |= NETIF_F_GSO_UDP_L4;
 	}
+
+	if (test_bit(HNAE3_DEV_SUPPORT_HW_TX_CSUM_B, ae_dev->caps)) {
+		netdev->hw_features |= NETIF_F_HW_CSUM;
+		netdev->features |= NETIF_F_HW_CSUM;
+		netdev->vlan_features |= NETIF_F_HW_CSUM;
+		netdev->hw_enc_features |= NETIF_F_HW_CSUM;
+	} else {
+		netdev->hw_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
+		netdev->features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
+		netdev->vlan_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
+		netdev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
+	}
 }
 
 static int hns3_alloc_buffer(struct hns3_enet_ring *ring,
@@ -4178,6 +4215,9 @@ static int hns3_client_init(struct hnae3_handle *handle)
 	/* MTU range: (ETH_MIN_MTU(kernel default) - 9702) */
 	netdev->max_mtu = HNS3_MAX_MTU;
 
+	if (test_bit(HNAE3_DEV_SUPPORT_HW_TX_CSUM_B, ae_dev->caps))
+		set_bit(HNS3_NIC_STATE_HW_TX_CSUM_ENABLE, &priv->state);
+
 	set_bit(HNS3_NIC_STATE_INITED, &priv->state);
 
 	if (netif_msg_drv(handle))