ice: Add support for VLANs and offloads

This patch adds support for VLANs. When a VLAN is created a switch filter
is added to direct the VLAN traffic to the corresponding VSI. When a VLAN
is deleted, the filter is deleted as well.

This patch also adds support for the following hardware offloads.
    1) VLAN tag insertion/stripping
    2) Receive Side Scaling (RSS)
    3) Tx checksum and TCP segmentation
    4) Rx checksum
Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: NTony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

drivers/net/ethernet/intel/ice/ice.h

@@ -23,7 +23,10 @@
 #include <linux/delay.h>
 #include <linux/bitmap.h>
 #include <linux/log2.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
 #include <linux/if_bridge.h>
+#include <net/ipv6.h>
 #include "ice_devids.h"
 #include "ice_type.h"
 #include "ice_txrx.h"
@@ -47,6 +50,8 @@
 #define ICE_MAX_SCATTER_RXQS	16
 #define ICE_Q_WAIT_RETRY_LIMIT	10
 #define ICE_Q_WAIT_MAX_RETRY	(5 * ICE_Q_WAIT_RETRY_LIMIT)
+#define ICE_MAX_LG_RSS_QS	256
+#define ICE_MAX_SMALL_RSS_QS	8
 #define ICE_RES_VALID_BIT	0x8000
 #define ICE_RES_MISC_VEC_ID	(ICE_RES_VALID_BIT - 1)
 #define ICE_INVAL_Q_INDEX	0xffff
@@ -62,6 +67,7 @@
 
 #define ICE_TX_DESC(R, i) (&(((struct ice_tx_desc *)((R)->desc))[i]))
 #define ICE_RX_DESC(R, i) (&(((union ice_32b_rx_flex_desc *)((R)->desc))[i]))
+#define ICE_TX_CTX_DESC(R, i) (&(((struct ice_tx_ctx_desc *)((R)->desc))[i]))
 
 #define ice_for_each_txq(vsi, i) \
 	for ((i) = 0; (i) < (vsi)->num_txq; (i)++)
@@ -113,6 +119,7 @@ struct ice_vsi {
 	irqreturn_t (*irq_handler)(int irq, void *data);
 
 	DECLARE_BITMAP(state, __ICE_STATE_NBITS);
+	unsigned long active_vlans[BITS_TO_LONGS(VLAN_N_VID)];
 	int num_q_vectors;
 	int base_vector;
 	enum ice_vsi_type type;
@@ -122,6 +129,13 @@ struct ice_vsi {
 	/* Interrupt thresholds */
 	u16 work_lmt;
 
+	/* RSS config */
+	u16 rss_table_size;	/* HW RSS table size */
+	u16 rss_size;		/* Allocated RSS queues */
+	u8 *rss_hkey_user;	/* User configured hash keys */
+	u8 *rss_lut_user;	/* User configured lookup table entries */
+	u8 rss_lut_type;	/* used to configure Get/Set RSS LUT AQ call */
+
 	u16 max_frame;
 	u16 rx_buf_len;
 
@@ -181,6 +195,7 @@ struct ice_pf {
 	struct mutex avail_q_mutex;	/* protects access to avail_[rx|tx]qs */
 	struct mutex sw_mutex;		/* lock for protecting VSI alloc flow */
 	u32 msg_enable;
+	u32 hw_csum_rx_error;
 	u32 oicr_idx;		/* Other interrupt cause vector index */
 	u32 num_lan_msix;	/* Total MSIX vectors for base driver */
 	u32 num_avail_msix;	/* remaining MSIX vectors left unclaimed */
@@ -224,4 +239,8 @@ static inline void ice_irq_dynamic_ena(struct ice_hw *hw, struct ice_vsi *vsi,
 	wr32(hw, GLINT_DYN_CTL(vector), val);
 }
 
+int ice_set_rss(struct ice_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size);
+int ice_get_rss(struct ice_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size);
+void ice_fill_rss_lut(u8 *lut, u16 rss_table_size, u16 rss_size);
+
 #endif /* _ICE_H_ */

drivers/net/ethernet/intel/ice/ice_adminq_cmd.h

@@ -968,6 +968,60 @@ struct ice_aqc_nvm {
 	__le32	addr_low;
 };
 
+/* Get/Set RSS key (indirect 0x0B04/0x0B02) */
+struct ice_aqc_get_set_rss_key {
+#define ICE_AQC_GSET_RSS_KEY_VSI_VALID	BIT(15)
+#define ICE_AQC_GSET_RSS_KEY_VSI_ID_S	0
+#define ICE_AQC_GSET_RSS_KEY_VSI_ID_M	(0x3FF << ICE_AQC_GSET_RSS_KEY_VSI_ID_S)
+	__le16 vsi_id;
+	u8 reserved[6];
+	__le32 addr_high;
+	__le32 addr_low;
+};
+
+#define ICE_AQC_GET_SET_RSS_KEY_DATA_RSS_KEY_SIZE	0x28
+#define ICE_AQC_GET_SET_RSS_KEY_DATA_HASH_KEY_SIZE	0xC
+
+struct ice_aqc_get_set_rss_keys {
+	u8 standard_rss_key[ICE_AQC_GET_SET_RSS_KEY_DATA_RSS_KEY_SIZE];
+	u8 extended_hash_key[ICE_AQC_GET_SET_RSS_KEY_DATA_HASH_KEY_SIZE];
+};
+
+/* Get/Set RSS LUT (indirect 0x0B05/0x0B03) */
+struct  ice_aqc_get_set_rss_lut {
+#define ICE_AQC_GSET_RSS_LUT_VSI_VALID	BIT(15)
+#define ICE_AQC_GSET_RSS_LUT_VSI_ID_S	0
+#define ICE_AQC_GSET_RSS_LUT_VSI_ID_M	(0x1FF << ICE_AQC_GSET_RSS_LUT_VSI_ID_S)
+	__le16 vsi_id;
+#define ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_S	0
+#define ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_M	\
+				(0x3 << ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_S)
+
+#define ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_VSI	 0
+#define ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_PF	 1
+#define ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_GLOBAL	 2
+
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_S	 2
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_M	 \
+				(0x3 << ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_S)
+
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_128	 128
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_128_FLAG 0
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_512	 512
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_512_FLAG 1
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_2K	 2048
+#define ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_2K_FLAG	 2
+
+#define ICE_AQC_GSET_RSS_LUT_GLOBAL_IDX_S	 4
+#define ICE_AQC_GSET_RSS_LUT_GLOBAL_IDX_M	 \
+				(0xF << ICE_AQC_GSET_RSS_LUT_GLOBAL_IDX_S)
+
+	__le16 flags;
+	__le32 reserved;
+	__le32 addr_high;
+	__le32 addr_low;
+};
+
 /* Add TX LAN Queues (indirect 0x0C30) */
 struct ice_aqc_add_txqs {
 	u8 num_qgrps;
@@ -1089,6 +1143,8 @@ struct ice_aq_desc {
 		struct ice_aqc_query_txsched_res query_sched_res;
 		struct ice_aqc_add_move_delete_elem add_move_delete_elem;
 		struct ice_aqc_nvm nvm;
+		struct ice_aqc_get_set_rss_lut get_set_rss_lut;
+		struct ice_aqc_get_set_rss_key get_set_rss_key;
 		struct ice_aqc_add_txqs add_txqs;
 		struct ice_aqc_dis_txqs dis_txqs;
 		struct ice_aqc_add_get_update_free_vsi vsi_cmd;
@@ -1171,6 +1227,12 @@ enum ice_adminq_opc {
 	/* NVM commands */
 	ice_aqc_opc_nvm_read				= 0x0701,
 
+	/* RSS commands */
+	ice_aqc_opc_set_rss_key				= 0x0B02,
+	ice_aqc_opc_set_rss_lut				= 0x0B03,
+	ice_aqc_opc_get_rss_key				= 0x0B04,
+	ice_aqc_opc_get_rss_lut				= 0x0B05,
+
 	/* TX queue handling commands/events */
 	ice_aqc_opc_add_txqs				= 0x0C30,
 	ice_aqc_opc_dis_txqs				= 0x0C31,

drivers/net/ethernet/intel/ice/ice_common.c

@@ -1261,6 +1261,194 @@ void ice_clear_pxe_mode(struct ice_hw *hw)
 		ice_aq_clear_pxe_mode(hw);
 }
 
+/**
+ * __ice_aq_get_set_rss_lut
+ * @hw: pointer to the hardware structure
+ * @vsi_id: VSI FW index
+ * @lut_type: LUT table type
+ * @lut: pointer to the LUT buffer provided by the caller
+ * @lut_size: size of the LUT buffer
+ * @glob_lut_idx: global LUT index
+ * @set: set true to set the table, false to get the table
+ *
+ * Internal function to get (0x0B05) or set (0x0B03) RSS look up table
+ */
+static enum ice_status
+__ice_aq_get_set_rss_lut(struct ice_hw *hw, u16 vsi_id, u8 lut_type, u8 *lut,
+			 u16 lut_size, u8 glob_lut_idx, bool set)
+{
+	struct ice_aqc_get_set_rss_lut *cmd_resp;
+	struct ice_aq_desc desc;
+	enum ice_status status;
+	u16 flags = 0;
+
+	cmd_resp = &desc.params.get_set_rss_lut;
+
+	if (set) {
+		ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_set_rss_lut);
+		desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD);
+	} else {
+		ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_get_rss_lut);
+	}
+
+	cmd_resp->vsi_id = cpu_to_le16(((vsi_id <<
+					 ICE_AQC_GSET_RSS_LUT_VSI_ID_S) &
+					ICE_AQC_GSET_RSS_LUT_VSI_ID_M) |
+				       ICE_AQC_GSET_RSS_LUT_VSI_VALID);
+
+	switch (lut_type) {
+	case ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_VSI:
+	case ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_PF:
+	case ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_GLOBAL:
+		flags |= ((lut_type << ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_S) &
+			  ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_M);
+		break;
+	default:
+		status = ICE_ERR_PARAM;
+		goto ice_aq_get_set_rss_lut_exit;
+	}
+
+	if (lut_type == ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_GLOBAL) {
+		flags |= ((glob_lut_idx << ICE_AQC_GSET_RSS_LUT_GLOBAL_IDX_S) &
+			  ICE_AQC_GSET_RSS_LUT_GLOBAL_IDX_M);
+
+		if (!set)
+			goto ice_aq_get_set_rss_lut_send;
+	} else if (lut_type == ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_PF) {
+		if (!set)
+			goto ice_aq_get_set_rss_lut_send;
+	} else {
+		goto ice_aq_get_set_rss_lut_send;
+	}
+
+	/* LUT size is only valid for Global and PF table types */
+	if (lut_size == ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_128) {
+		flags |= (ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_128_FLAG <<
+			  ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_S) &
+			 ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_M;
+	} else if (lut_size == ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_512) {
+		flags |= (ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_512_FLAG <<
+			  ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_S) &
+			 ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_M;
+	} else if ((lut_size == ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_2K) &&
+		   (lut_type == ICE_AQC_GSET_RSS_LUT_TABLE_TYPE_PF)) {
+		flags |= (ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_2K_FLAG <<
+			  ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_S) &
+			 ICE_AQC_GSET_RSS_LUT_TABLE_SIZE_M;
+	} else {
+		status = ICE_ERR_PARAM;
+		goto ice_aq_get_set_rss_lut_exit;
+	}
+
+ice_aq_get_set_rss_lut_send:
+	cmd_resp->flags = cpu_to_le16(flags);
+	status = ice_aq_send_cmd(hw, &desc, lut, lut_size, NULL);
+
+ice_aq_get_set_rss_lut_exit:
+	return status;
+}
+
+/**
+ * ice_aq_get_rss_lut
+ * @hw: pointer to the hardware structure
+ * @vsi_id: VSI FW index
+ * @lut_type: LUT table type
+ * @lut: pointer to the LUT buffer provided by the caller
+ * @lut_size: size of the LUT buffer
+ *
+ * get the RSS lookup table, PF or VSI type
+ */
+enum ice_status
+ice_aq_get_rss_lut(struct ice_hw *hw, u16 vsi_id, u8 lut_type, u8 *lut,
+		   u16 lut_size)
+{
+	return __ice_aq_get_set_rss_lut(hw, vsi_id, lut_type, lut, lut_size, 0,
+					false);
+}
+
+/**
+ * ice_aq_set_rss_lut
+ * @hw: pointer to the hardware structure
+ * @vsi_id: VSI FW index
+ * @lut_type: LUT table type
+ * @lut: pointer to the LUT buffer provided by the caller
+ * @lut_size: size of the LUT buffer
+ *
+ * set the RSS lookup table, PF or VSI type
+ */
+enum ice_status
+ice_aq_set_rss_lut(struct ice_hw *hw, u16 vsi_id, u8 lut_type, u8 *lut,
+		   u16 lut_size)
+{
+	return __ice_aq_get_set_rss_lut(hw, vsi_id, lut_type, lut, lut_size, 0,
+					true);
+}
+
+/**
+ * __ice_aq_get_set_rss_key
+ * @hw: pointer to the hw struct
+ * @vsi_id: VSI FW index
+ * @key: pointer to key info struct
+ * @set: set true to set the key, false to get the key
+ *
+ * get (0x0B04) or set (0x0B02) the RSS key per VSI
+ */
+static enum
+ice_status __ice_aq_get_set_rss_key(struct ice_hw *hw, u16 vsi_id,
+				    struct ice_aqc_get_set_rss_keys *key,
+				    bool set)
+{
+	struct ice_aqc_get_set_rss_key *cmd_resp;
+	u16 key_size = sizeof(*key);
+	struct ice_aq_desc desc;
+
+	cmd_resp = &desc.params.get_set_rss_key;
+
+	if (set) {
+		ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_set_rss_key);
+		desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD);
+	} else {
+		ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_get_rss_key);
+	}
+
+	cmd_resp->vsi_id = cpu_to_le16(((vsi_id <<
+					 ICE_AQC_GSET_RSS_KEY_VSI_ID_S) &
+					ICE_AQC_GSET_RSS_KEY_VSI_ID_M) |
+				       ICE_AQC_GSET_RSS_KEY_VSI_VALID);
+
+	return ice_aq_send_cmd(hw, &desc, key, key_size, NULL);
+}
+
+/**
+ * ice_aq_get_rss_key
+ * @hw: pointer to the hw struct
+ * @vsi_id: VSI FW index
+ * @key: pointer to key info struct
+ *
+ * get the RSS key per VSI
+ */
+enum ice_status
+ice_aq_get_rss_key(struct ice_hw *hw, u16 vsi_id,
+		   struct ice_aqc_get_set_rss_keys *key)
+{
+	return __ice_aq_get_set_rss_key(hw, vsi_id, key, false);
+}
+
+/**
+ * ice_aq_set_rss_key
+ * @hw: pointer to the hw struct
+ * @vsi_id: VSI FW index
+ * @keys: pointer to key info struct
+ *
+ * set the RSS key per VSI
+ */
+enum ice_status
+ice_aq_set_rss_key(struct ice_hw *hw, u16 vsi_id,
+		   struct ice_aqc_get_set_rss_keys *keys)
+{
+	return __ice_aq_get_set_rss_key(hw, vsi_id, keys, true);
+}
+
 /**
  * ice_aq_add_lan_txq
  * @hw: pointer to the hardware structure

drivers/net/ethernet/intel/ice/ice_common.h

@@ -33,6 +33,19 @@ enum ice_status ice_get_caps(struct ice_hw *hw);
 enum ice_status
 ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
 		  u32 rxq_index);
+
+enum ice_status
+ice_aq_get_rss_lut(struct ice_hw *hw, u16 vsi_id, u8 lut_type, u8 *lut,
+		   u16 lut_size);
+enum ice_status
+ice_aq_set_rss_lut(struct ice_hw *hw, u16 vsi_id, u8 lut_type, u8 *lut,
+		   u16 lut_size);
+enum ice_status
+ice_aq_get_rss_key(struct ice_hw *hw, u16 vsi_id,
+		   struct ice_aqc_get_set_rss_keys *keys);
+enum ice_status
+ice_aq_set_rss_key(struct ice_hw *hw, u16 vsi_id,
+		   struct ice_aqc_get_set_rss_keys *keys);
 bool ice_check_sq_alive(struct ice_hw *hw, struct ice_ctl_q_info *cq);
 enum ice_status ice_aq_q_shutdown(struct ice_hw *hw, bool unloading);
 void ice_fill_dflt_direct_cmd_desc(struct ice_aq_desc *desc, u16 opcode);

drivers/net/ethernet/intel/ice/ice_lan_tx_rx.h

@@ -40,6 +40,65 @@ union ice_32byte_rx_desc {
 	} wb; /* writeback */
 };
 
+struct ice_rx_ptype_decoded {
+	u32 ptype:10;
+	u32 known:1;
+	u32 outer_ip:1;
+	u32 outer_ip_ver:2;
+	u32 outer_frag:1;
+	u32 tunnel_type:3;
+	u32 tunnel_end_prot:2;
+	u32 tunnel_end_frag:1;
+	u32 inner_prot:4;
+	u32 payload_layer:3;
+};
+
+enum ice_rx_ptype_outer_ip {
+	ICE_RX_PTYPE_OUTER_L2	= 0,
+	ICE_RX_PTYPE_OUTER_IP	= 1,
+};
+
+enum ice_rx_ptype_outer_ip_ver {
+	ICE_RX_PTYPE_OUTER_NONE	= 0,
+	ICE_RX_PTYPE_OUTER_IPV4	= 1,
+	ICE_RX_PTYPE_OUTER_IPV6	= 2,
+};
+
+enum ice_rx_ptype_outer_fragmented {
+	ICE_RX_PTYPE_NOT_FRAG	= 0,
+	ICE_RX_PTYPE_FRAG	= 1,
+};
+
+enum ice_rx_ptype_tunnel_type {
+	ICE_RX_PTYPE_TUNNEL_NONE		= 0,
+	ICE_RX_PTYPE_TUNNEL_IP_IP		= 1,
+	ICE_RX_PTYPE_TUNNEL_IP_GRENAT		= 2,
+	ICE_RX_PTYPE_TUNNEL_IP_GRENAT_MAC	= 3,
+	ICE_RX_PTYPE_TUNNEL_IP_GRENAT_MAC_VLAN	= 4,
+};
+
+enum ice_rx_ptype_tunnel_end_prot {
+	ICE_RX_PTYPE_TUNNEL_END_NONE	= 0,
+	ICE_RX_PTYPE_TUNNEL_END_IPV4	= 1,
+	ICE_RX_PTYPE_TUNNEL_END_IPV6	= 2,
+};
+
+enum ice_rx_ptype_inner_prot {
+	ICE_RX_PTYPE_INNER_PROT_NONE		= 0,
+	ICE_RX_PTYPE_INNER_PROT_UDP		= 1,
+	ICE_RX_PTYPE_INNER_PROT_TCP		= 2,
+	ICE_RX_PTYPE_INNER_PROT_SCTP		= 3,
+	ICE_RX_PTYPE_INNER_PROT_ICMP		= 4,
+	ICE_RX_PTYPE_INNER_PROT_TIMESYNC	= 5,
+};
+
+enum ice_rx_ptype_payload_layer {
+	ICE_RX_PTYPE_PAYLOAD_LAYER_NONE	= 0,
+	ICE_RX_PTYPE_PAYLOAD_LAYER_PAY2	= 1,
+	ICE_RX_PTYPE_PAYLOAD_LAYER_PAY3	= 2,
+	ICE_RX_PTYPE_PAYLOAD_LAYER_PAY4	= 3,
+};
+
 /* RX Flex Descriptor
  * This descriptor is used instead of the legacy version descriptor when
  * ice_rlan_ctx.adv_desc is set
@@ -88,6 +147,41 @@ union ice_32b_rx_flex_desc {
 	} wb; /* writeback */
 };
 
+/* Rx Flex Descriptor NIC Profile
+ * This descriptor corresponds to RxDID 2 which contains
+ * metadata fields for RSS, flow id and timestamp info
+ */
+struct ice_32b_rx_flex_desc_nic {
+	/* Qword 0 */
+	u8 rxdid;
+	u8 mir_id_umb_cast;
+	__le16 ptype_flexi_flags0;
+	__le16 pkt_len;
+	__le16 hdr_len_sph_flex_flags1;
+
+	/* Qword 1 */
+	__le16 status_error0;
+	__le16 l2tag1;
+	__le32 rss_hash;
+
+	/* Qword 2 */
+	__le16 status_error1;
+	u8 flexi_flags2;
+	u8 ts_low;
+	__le16 l2tag2_1st;
+	__le16 l2tag2_2nd;
+
+	/* Qword 3 */
+	__le32 flow_id;
+	union {
+		struct {
+			__le16 vlan_id;
+			__le16 flow_id_ipv6;
+		} flex;
+		__le32 ts_high;
+	} flex_ts;
+};
+
 /* Receive Flex Descriptor profile IDs: There are a total
  * of 64 profiles where profile IDs 0/1 are for legacy; and
  * profiles 2-63 are flex profiles that can be programmed
@@ -241,12 +335,56 @@ enum ice_tx_desc_dtype_value {
 enum ice_tx_desc_cmd_bits {
 	ICE_TX_DESC_CMD_EOP			= 0x0001,
 	ICE_TX_DESC_CMD_RS			= 0x0002,
+	ICE_TX_DESC_CMD_IL2TAG1			= 0x0008,
+	ICE_TX_DESC_CMD_IIPT_IPV6		= 0x0020, /* 2 BITS */
+	ICE_TX_DESC_CMD_IIPT_IPV4		= 0x0040, /* 2 BITS */
+	ICE_TX_DESC_CMD_IIPT_IPV4_CSUM		= 0x0060, /* 2 BITS */
+	ICE_TX_DESC_CMD_L4T_EOFT_TCP		= 0x0100, /* 2 BITS */
+	ICE_TX_DESC_CMD_L4T_EOFT_UDP		= 0x0300, /* 2 BITS */
 };
 
 #define ICE_TXD_QW1_OFFSET_S	16
+#define ICE_TXD_QW1_OFFSET_M	(0x3FFFFULL << ICE_TXD_QW1_OFFSET_S)
+
+enum ice_tx_desc_len_fields {
+	/* Note: These are predefined bit offsets */
+	ICE_TX_DESC_LEN_MACLEN_S	= 0, /* 7 BITS */
+	ICE_TX_DESC_LEN_IPLEN_S	= 7, /* 7 BITS */
+	ICE_TX_DESC_LEN_L4_LEN_S	= 14 /* 4 BITS */
+};
+
 #define ICE_TXD_QW1_TX_BUF_SZ_S	34
 #define ICE_TXD_QW1_L2TAG1_S	48
 
+/* Context descriptors */
+struct ice_tx_ctx_desc {
+	__le32 tunneling_params;
+	__le16 l2tag2;
+	__le16 rsvd;
+	__le64 qw1;
+};
+
+#define ICE_TXD_CTX_QW1_CMD_S	4
+#define ICE_TXD_CTX_QW1_CMD_M	(0x7FUL << ICE_TXD_CTX_QW1_CMD_S)
+
+#define ICE_TXD_CTX_QW1_TSO_LEN_S	30
+#define ICE_TXD_CTX_QW1_TSO_LEN_M	\
+			(0x3FFFFULL << ICE_TXD_CTX_QW1_TSO_LEN_S)
+
+#define ICE_TXD_CTX_QW1_MSS_S	50
+
+enum ice_tx_ctx_desc_cmd_bits {
+	ICE_TX_CTX_DESC_TSO		= 0x01,
+	ICE_TX_CTX_DESC_TSYN		= 0x02,
+	ICE_TX_CTX_DESC_IL2TAG2		= 0x04,
+	ICE_TX_CTX_DESC_IL2TAG2_IL2H	= 0x08,
+	ICE_TX_CTX_DESC_SWTCH_NOTAG	= 0x00,
+	ICE_TX_CTX_DESC_SWTCH_UPLINK	= 0x10,
+	ICE_TX_CTX_DESC_SWTCH_LOCAL	= 0x20,
+	ICE_TX_CTX_DESC_SWTCH_VSI	= 0x30,
+	ICE_TX_CTX_DESC_RESERVED	= 0x40
+};
+
 #define ICE_LAN_TXQ_MAX_QGRPS	127
 #define ICE_LAN_TXQ_MAX_QDIS	1023
 
@@ -289,4 +427,35 @@ struct ice_tlan_ctx {
 	u8  pkt_shaper_prof_idx;
 	u8  int_q_state;	/* width not needed - internal do not write */
 };
+
+/* macro to make the table lines short */
+#define ICE_PTT(PTYPE, OUTER_IP, OUTER_IP_VER, OUTER_FRAG, T, TE, TEF, I, PL)\
+	{	PTYPE, \
+		1, \
+		ICE_RX_PTYPE_OUTER_##OUTER_IP, \
+		ICE_RX_PTYPE_OUTER_##OUTER_IP_VER, \
+		ICE_RX_PTYPE_##OUTER_FRAG, \
+		ICE_RX_PTYPE_TUNNEL_##T, \
+		ICE_RX_PTYPE_TUNNEL_END_##TE, \
+		ICE_RX_PTYPE_##TEF, \
+		ICE_RX_PTYPE_INNER_PROT_##I, \
+		ICE_RX_PTYPE_PAYLOAD_LAYER_##PL }
+
+#define ICE_PTT_UNUSED_ENTRY(PTYPE) { PTYPE, 0, 0, 0, 0, 0, 0, 0, 0, 0 }
+
+/* shorter macros makes the table fit but are terse */
+#define ICE_RX_PTYPE_NOF		ICE_RX_PTYPE_NOT_FRAG
+
+/* Lookup table mapping the HW PTYPE to the bit field for decoding */
+static const struct ice_rx_ptype_decoded ice_ptype_lkup[] = {
+	/* L2 Packet types */
+	ICE_PTT_UNUSED_ENTRY(0),
+	ICE_PTT(1, L2, NONE, NOF, NONE, NONE, NOF, NONE, PAY2),
+	ICE_PTT(2, L2, NONE, NOF, NONE, NONE, NOF, NONE, NONE),
+};
+
+static inline struct ice_rx_ptype_decoded ice_decode_rx_desc_ptype(u16 ptype)
+{
+	return ice_ptype_lkup[ptype];
+}
 #endif /* _ICE_LAN_TX_RX_H_ */

drivers/net/ethernet/intel/ice/ice_switch.c

@@ -1217,6 +1217,117 @@ ice_add_mac(struct ice_hw *hw, struct list_head *m_list)
 	return status;
 }
 
+/**
+ * ice_find_vlan_entry
+ * @hw: pointer to the hardware structure
+ * @vlan_id: VLAN id to search for
+ *
+ * Helper function to search for a VLAN entry using a given VLAN id
+ * Returns pointer to the entry if found.
+ */
+static struct ice_fltr_mgmt_list_entry *
+ice_find_vlan_entry(struct ice_hw *hw, u16 vlan_id)
+{
+	struct ice_fltr_mgmt_list_entry *vlan_list_itr, *vlan_ret = NULL;
+	struct ice_switch_info *sw = hw->switch_info;
+
+	mutex_lock(&sw->vlan_list_lock);
+	list_for_each_entry(vlan_list_itr, &sw->vlan_list_head, list_entry)
+		if (vlan_list_itr->fltr_info.l_data.vlan.vlan_id == vlan_id) {
+			vlan_ret = vlan_list_itr;
+			break;
+		}
+
+	mutex_unlock(&sw->vlan_list_lock);
+	return vlan_ret;
+}
+
+/**
+ * ice_add_vlan_internal - Add one VLAN based filter rule
+ * @hw: pointer to the hardware structure
+ * @f_entry: filter entry containing one VLAN information
+ */
+static enum ice_status
+ice_add_vlan_internal(struct ice_hw *hw, struct ice_fltr_list_entry *f_entry)
+{
+	struct ice_fltr_info *new_fltr, *cur_fltr;
+	struct ice_fltr_mgmt_list_entry *v_list_itr;
+	u16 vlan_id;
+
+	new_fltr = &f_entry->fltr_info;
+	/* VLAN id should only be 12 bits */
+	if (new_fltr->l_data.vlan.vlan_id > ICE_MAX_VLAN_ID)
+		return ICE_ERR_PARAM;
+
+	vlan_id = new_fltr->l_data.vlan.vlan_id;
+	v_list_itr = ice_find_vlan_entry(hw, vlan_id);
+	if (!v_list_itr) {
+		u16 vsi_id = ICE_VSI_INVAL_ID;
+		enum ice_status status;
+		u16 vsi_list_id = 0;
+
+		if (new_fltr->fltr_act == ICE_FWD_TO_VSI) {
+			enum ice_sw_lkup_type lkup_type = new_fltr->lkup_type;
+
+			/* All VLAN pruning rules use a VSI list.
+			 * Convert the action to forwarding to a VSI list.
+			 */
+			vsi_id = new_fltr->fwd_id.vsi_id;
+			status = ice_create_vsi_list_rule(hw, &vsi_id, 1,
+							  &vsi_list_id,
+							  lkup_type);
+			if (status)
+				return status;
+			new_fltr->fltr_act = ICE_FWD_TO_VSI_LIST;
+			new_fltr->fwd_id.vsi_list_id = vsi_list_id;
+		}
+
+		status = ice_create_pkt_fwd_rule(hw, f_entry);
+		if (!status && vsi_id != ICE_VSI_INVAL_ID) {
+			v_list_itr = ice_find_vlan_entry(hw, vlan_id);
+			if (!v_list_itr)
+				return ICE_ERR_DOES_NOT_EXIST;
+			v_list_itr->vsi_list_info =
+				ice_create_vsi_list_map(hw, &vsi_id, 1,
+							vsi_list_id);
+		}
+
+		return status;
+	}
+
+	cur_fltr = &v_list_itr->fltr_info;
+	return ice_handle_vsi_list_mgmt(hw, v_list_itr, cur_fltr, new_fltr);
+}
+
+/**
+ * ice_add_vlan - Add VLAN based filter rule
+ * @hw: pointer to the hardware structure
+ * @v_list: list of VLAN entries and forwarding information
+ */
+enum ice_status
+ice_add_vlan(struct ice_hw *hw, struct list_head *v_list)
+{
+	struct ice_fltr_list_entry *v_list_itr;
+
+	if (!v_list || !hw)
+		return ICE_ERR_PARAM;
+
+	list_for_each_entry(v_list_itr, v_list, list_entry) {
+		enum ice_status status;
+
+		if (v_list_itr->fltr_info.lkup_type != ICE_SW_LKUP_VLAN)
+			return ICE_ERR_PARAM;
+
+		status = ice_add_vlan_internal(hw, v_list_itr);
+		if (status) {
+			v_list_itr->status = ICE_FLTR_STATUS_FW_FAIL;
+			return status;
+		}
+		v_list_itr->status = ICE_FLTR_STATUS_FW_SUCCESS;
+	}
+	return 0;
+}
+
 /**
  * ice_remove_vsi_list_rule
  * @hw: pointer to the hardware structure
@@ -1515,6 +1626,54 @@ ice_remove_mac(struct ice_hw *hw, struct list_head *m_list)
 	return status;
 }
 
+/**
+ * ice_remove_vlan_internal - Remove one VLAN based filter rule
+ * @hw: pointer to the hardware structure
+ * @f_entry: filter entry containing one VLAN information
+ */
+static enum ice_status
+ice_remove_vlan_internal(struct ice_hw *hw,
+			 struct ice_fltr_list_entry *f_entry)
+{
+	struct ice_fltr_info *new_fltr;
+	struct ice_fltr_mgmt_list_entry *v_list_elem;
+	u16 vsi_id;
+
+	new_fltr = &f_entry->fltr_info;
+
+	v_list_elem = ice_find_vlan_entry(hw, new_fltr->l_data.vlan.vlan_id);
+	if (!v_list_elem)
+		return ICE_ERR_PARAM;
+
+	vsi_id = f_entry->fltr_info.fwd_id.vsi_id;
+	return ice_handle_rem_vsi_list_mgmt(hw, vsi_id, v_list_elem);
+}
+
+/**
+ * ice_remove_vlan - Remove VLAN based filter rule
+ * @hw: pointer to the hardware structure
+ * @v_list: list of VLAN entries and forwarding information
+ */
+enum ice_status
+ice_remove_vlan(struct ice_hw *hw, struct list_head *v_list)
+{
+	struct ice_fltr_list_entry *v_list_itr;
+	enum ice_status status = 0;
+
+	if (!v_list || !hw)
+		return ICE_ERR_PARAM;
+
+	list_for_each_entry(v_list_itr, v_list, list_entry) {
+		status = ice_remove_vlan_internal(hw, v_list_itr);
+		if (status) {
+			v_list_itr->status = ICE_FLTR_STATUS_FW_FAIL;
+			return status;
+		}
+		v_list_itr->status = ICE_FLTR_STATUS_FW_SUCCESS;
+	}
+	return status;
+}
+
 /**
  * ice_add_to_vsi_fltr_list - Add VSI filters to the list
  * @hw: pointer to the hardware structure
@@ -1600,6 +1759,16 @@ ice_remove_vsi_lkup_fltr(struct ice_hw *hw, u16 vsi_id,
 		}
 		break;
 	case ICE_SW_LKUP_VLAN:
+		mutex_lock(&sw->vlan_list_lock);
+		status = ice_add_to_vsi_fltr_list(hw, vsi_id,
+						  &sw->vlan_list_head,
+						  &remove_list_head);
+		mutex_unlock(&sw->vlan_list_lock);
+		if (!status) {
+			ice_remove_vlan(hw, &remove_list_head);
+			goto free_fltr_list;
+		}
+		break;
 	case ICE_SW_LKUP_MAC_VLAN:
 	case ICE_SW_LKUP_ETHERTYPE:
 	case ICE_SW_LKUP_ETHERTYPE_MAC:

drivers/net/ethernet/intel/ice/ice_switch.h

@@ -8,6 +8,7 @@
 
 #define ICE_SW_CFG_MAX_BUF_LEN 2048
 #define ICE_DFLT_VSI_INVAL 0xff
+#define ICE_VSI_INVAL_ID 0xffff
 
 /* VSI context structure for add/get/update/free operations */
 struct ice_vsi_ctx {
@@ -152,4 +153,7 @@ enum ice_status ice_get_initial_sw_cfg(struct ice_hw *hw);
 enum ice_status ice_add_mac(struct ice_hw *hw, struct list_head *m_lst);
 enum ice_status ice_remove_mac(struct ice_hw *hw, struct list_head *m_lst);
 void ice_remove_vsi_fltr(struct ice_hw *hw, u16 vsi_id);
+enum ice_status ice_add_vlan(struct ice_hw *hw, struct list_head *m_list);
+enum ice_status ice_remove_vlan(struct ice_hw *hw, struct list_head *v_list);
+
 #endif /* _ICE_SWITCH_H_ */

drivers/net/ethernet/intel/ice/ice_txrx.c

@@ -796,6 +796,134 @@ static bool ice_is_non_eop(struct ice_ring *rx_ring,
 	return true;
 }
 
+/**
+ * ice_ptype_to_htype - get a hash type
+ * @ptype: the ptype value from the descriptor
+ *
+ * Returns a hash type to be used by skb_set_hash
+ */
+static enum pkt_hash_types ice_ptype_to_htype(u8 __always_unused ptype)
+{
+	return PKT_HASH_TYPE_NONE;
+}
+
+/**
+ * ice_rx_hash - set the hash value in the skb
+ * @rx_ring: descriptor ring
+ * @rx_desc: specific descriptor
+ * @skb: pointer to current skb
+ * @rx_ptype: the ptype value from the descriptor
+ */
+static void
+ice_rx_hash(struct ice_ring *rx_ring, union ice_32b_rx_flex_desc *rx_desc,
+	    struct sk_buff *skb, u8 rx_ptype)
+{
+	struct ice_32b_rx_flex_desc_nic *nic_mdid;
+	u32 hash;
+
+	if (!(rx_ring->netdev->features & NETIF_F_RXHASH))
+		return;
+
+	if (rx_desc->wb.rxdid != ICE_RXDID_FLEX_NIC)
+		return;
+
+	nic_mdid = (struct ice_32b_rx_flex_desc_nic *)rx_desc;
+	hash = le32_to_cpu(nic_mdid->rss_hash);
+	skb_set_hash(skb, hash, ice_ptype_to_htype(rx_ptype));
+}
+
+/**
+ * ice_rx_csum - Indicate in skb if checksum is good
+ * @vsi: the VSI we care about
+ * @skb: skb currently being received and modified
+ * @rx_desc: the receive descriptor
+ * @ptype: the packet type decoded by hardware
+ *
+ * skb->protocol must be set before this function is called
+ */
+static void ice_rx_csum(struct ice_vsi *vsi, struct sk_buff *skb,
+			union ice_32b_rx_flex_desc *rx_desc, u8 ptype)
+{
+	struct ice_rx_ptype_decoded decoded;
+	u32 rx_error, rx_status;
+	bool ipv4, ipv6;
+
+	rx_status = le16_to_cpu(rx_desc->wb.status_error0);
+	rx_error = rx_status;
+
+	decoded = ice_decode_rx_desc_ptype(ptype);
+
+	/* Start with CHECKSUM_NONE and by default csum_level = 0 */
+	skb->ip_summed = CHECKSUM_NONE;
+	skb_checksum_none_assert(skb);
+
+	/* check if Rx checksum is enabled */
+	if (!(vsi->netdev->features & NETIF_F_RXCSUM))
+		return;
+
+	/* check if HW has decoded the packet and checksum */
+	if (!(rx_status & BIT(ICE_RX_FLEX_DESC_STATUS0_L3L4P_S)))
+		return;
+
+	if (!(decoded.known && decoded.outer_ip))
+		return;
+
+	ipv4 = (decoded.outer_ip == ICE_RX_PTYPE_OUTER_IP) &&
+	       (decoded.outer_ip_ver == ICE_RX_PTYPE_OUTER_IPV4);
+	ipv6 = (decoded.outer_ip == ICE_RX_PTYPE_OUTER_IP) &&
+	       (decoded.outer_ip_ver == ICE_RX_PTYPE_OUTER_IPV6);
+
+	if (ipv4 && (rx_error & (BIT(ICE_RX_FLEX_DESC_STATUS0_XSUM_IPE_S) |
+				 BIT(ICE_RX_FLEX_DESC_STATUS0_XSUM_EIPE_S))))
+		goto checksum_fail;
+	else if (ipv6 && (rx_status &
+		 (BIT(ICE_RX_FLEX_DESC_STATUS0_IPV6EXADD_S))))
+		goto checksum_fail;
+
+	/* check for L4 errors and handle packets that were not able to be
+	 * checksummed due to arrival speed
+	 */
+	if (rx_error & BIT(ICE_RX_FLEX_DESC_STATUS0_XSUM_L4E_S))
+		goto checksum_fail;
+
+	/* Only report checksum unnecessary for TCP, UDP, or SCTP */
+	switch (decoded.inner_prot) {
+	case ICE_RX_PTYPE_INNER_PROT_TCP:
+	case ICE_RX_PTYPE_INNER_PROT_UDP:
+	case ICE_RX_PTYPE_INNER_PROT_SCTP:
+		skb->ip_summed = CHECKSUM_UNNECESSARY;
+	default:
+		break;
+	}
+	return;
+
+checksum_fail:
+	vsi->back->hw_csum_rx_error++;
+}
+
+/**
+ * ice_process_skb_fields - Populate skb header fields from Rx descriptor
+ * @rx_ring: rx descriptor ring packet is being transacted on
+ * @rx_desc: pointer to the EOP Rx descriptor
+ * @skb: pointer to current skb being populated
+ * @ptype: the packet type decoded by hardware
+ *
+ * This function checks the ring, descriptor, and packet information in
+ * order to populate the hash, checksum, VLAN, protocol, and
+ * other fields within the skb.
+ */
+static void ice_process_skb_fields(struct ice_ring *rx_ring,
+				   union ice_32b_rx_flex_desc *rx_desc,
+				   struct sk_buff *skb, u8 ptype)
+{
+	ice_rx_hash(rx_ring, rx_desc, skb, ptype);
+
+	/* modifies the skb - consumes the enet header */
+	skb->protocol = eth_type_trans(skb, rx_ring->netdev);
+
+	ice_rx_csum(rx_ring->vsi, skb, rx_desc, ptype);
+}
+
 /**
  * ice_receive_skb - Send a completed packet up the stack
  * @rx_ring: rx ring in play
@@ -839,6 +967,7 @@ static int ice_clean_rx_irq(struct ice_ring *rx_ring, int budget)
 		struct sk_buff *skb;
 		u16 stat_err_bits;
 		u16 vlan_tag = 0;
+		u8 rx_ptype;
 
 		/* return some buffers to hardware, one at a time is too slow */
 		if (cleaned_count >= ICE_RX_BUF_WRITE) {
@@ -882,6 +1011,9 @@ static int ice_clean_rx_irq(struct ice_ring *rx_ring, int budget)
 			continue;
 		}
 
+		rx_ptype = le16_to_cpu(rx_desc->wb.ptype_flex_flags0) &
+			ICE_RX_FLEX_DESC_PTYPE_M;
+
 		stat_err_bits = BIT(ICE_RX_FLEX_DESC_STATUS0_L2TAG1P_S);
 		if (ice_test_staterr(rx_desc, stat_err_bits))
 			vlan_tag = le16_to_cpu(rx_desc->wb.l2tag1);
@@ -897,6 +1029,9 @@ static int ice_clean_rx_irq(struct ice_ring *rx_ring, int budget)
 		/* probably a little skewed due to removing CRC */
 		total_rx_bytes += skb->len;
 
+		/* populate checksum, VLAN, and protocol */
+		ice_process_skb_fields(rx_ring, rx_desc, skb, rx_ptype);
+
 		/* send completed skb up the stack */
 		ice_receive_skb(rx_ring, skb, vlan_tag);
 
@@ -1026,14 +1161,17 @@ static int ice_maybe_stop_tx(struct ice_ring *tx_ring, unsigned int size)
  * ice_tx_map - Build the Tx descriptor
  * @tx_ring: ring to send buffer on
  * @first: first buffer info buffer to use
+ * @off: pointer to struct that holds offload parameters
  *
  * This function loops over the skb data pointed to by *first
  * and gets a physical address for each memory location and programs
  * it and the length into the transmit descriptor.
  */
-static void ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first)
+static void
+ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first,
+	   struct ice_tx_offload_params *off)
 {
-	u64 td_offset = 0, td_tag = 0, td_cmd = 0;
+	u64 td_offset, td_tag, td_cmd;
 	u16 i = tx_ring->next_to_use;
 	struct skb_frag_struct *frag;
 	unsigned int data_len, size;
@@ -1042,6 +1180,9 @@ static void ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first)
 	struct sk_buff *skb;
 	dma_addr_t dma;
 
+	td_tag = off->td_l2tag1;
+	td_cmd = off->td_cmd;
+	td_offset = off->td_offset;
 	skb = first->skb;
 
 	data_len = skb->data_len;
@@ -1049,6 +1190,12 @@ static void ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first)
 
 	tx_desc = ICE_TX_DESC(tx_ring, i);
 
+	if (first->tx_flags & ICE_TX_FLAGS_HW_VLAN) {
+		td_cmd |= (u64)ICE_TX_DESC_CMD_IL2TAG1;
+		td_tag = (first->tx_flags & ICE_TX_FLAGS_VLAN_M) >>
+			  ICE_TX_FLAGS_VLAN_S;
+	}
+
 	dma = dma_map_single(tx_ring->dev, skb->data, size, DMA_TO_DEVICE);
 
 	tx_buf = first;
@@ -1169,6 +1316,223 @@ static void ice_tx_map(struct ice_ring *tx_ring, struct ice_tx_buf *first)
 	tx_ring->next_to_use = i;
 }
 
+/**
+ * ice_tx_csum - Enable Tx checksum offloads
+ * @first: pointer to the first descriptor
+ * @off: pointer to struct that holds offload parameters
+ *
+ * Returns 0 or error (negative) if checksum offload can't happen, 1 otherwise.
+ */
+static
+int ice_tx_csum(struct ice_tx_buf *first, struct ice_tx_offload_params *off)
+{
+	u32 l4_len = 0, l3_len = 0, l2_len = 0;
+	struct sk_buff *skb = first->skb;
+	union {
+		struct iphdr *v4;
+		struct ipv6hdr *v6;
+		unsigned char *hdr;
+	} ip;
+	union {
+		struct tcphdr *tcp;
+		unsigned char *hdr;
+	} l4;
+	__be16 frag_off, protocol;
+	unsigned char *exthdr;
+	u32 offset, cmd = 0;
+	u8 l4_proto = 0;
+
+	if (skb->ip_summed != CHECKSUM_PARTIAL)
+		return 0;
+
+	ip.hdr = skb_network_header(skb);
+	l4.hdr = skb_transport_header(skb);
+
+	/* compute outer L2 header size */
+	l2_len = ip.hdr - skb->data;
+	offset = (l2_len / 2) << ICE_TX_DESC_LEN_MACLEN_S;
+
+	if (skb->encapsulation)
+		return -1;
+
+	/* Enable IP checksum offloads */
+	protocol = vlan_get_protocol(skb);
+	if (protocol == htons(ETH_P_IP)) {
+		l4_proto = ip.v4->protocol;
+		/* the stack computes the IP header already, the only time we
+		 * need the hardware to recompute it is in the case of TSO.
+		 */
+		if (first->tx_flags & ICE_TX_FLAGS_TSO)
+			cmd |= ICE_TX_DESC_CMD_IIPT_IPV4_CSUM;
+		else
+			cmd |= ICE_TX_DESC_CMD_IIPT_IPV4;
+
+	} else if (protocol == htons(ETH_P_IPV6)) {
+		cmd |= ICE_TX_DESC_CMD_IIPT_IPV6;
+		exthdr = ip.hdr + sizeof(*ip.v6);
+		l4_proto = ip.v6->nexthdr;
+		if (l4.hdr != exthdr)
+			ipv6_skip_exthdr(skb, exthdr - skb->data, &l4_proto,
+					 &frag_off);
+	} else {
+		return -1;
+	}
+
+	/* compute inner L3 header size */
+	l3_len = l4.hdr - ip.hdr;
+	offset |= (l3_len / 4) << ICE_TX_DESC_LEN_IPLEN_S;
+
+	/* Enable L4 checksum offloads */
+	switch (l4_proto) {
+	case IPPROTO_TCP:
+		/* enable checksum offloads */
+		cmd |= ICE_TX_DESC_CMD_L4T_EOFT_TCP;
+		l4_len = l4.tcp->doff;
+		offset |= l4_len << ICE_TX_DESC_LEN_L4_LEN_S;
+		break;
+	case IPPROTO_UDP:
+		/* enable UDP checksum offload */
+		cmd |= ICE_TX_DESC_CMD_L4T_EOFT_UDP;
+		l4_len = (sizeof(struct udphdr) >> 2);
+		offset |= l4_len << ICE_TX_DESC_LEN_L4_LEN_S;
+		break;
+	case IPPROTO_SCTP:
+	default:
+		if (first->tx_flags & ICE_TX_FLAGS_TSO)
+			return -1;
+		skb_checksum_help(skb);
+		return 0;
+	}
+
+	off->td_cmd |= cmd;
+	off->td_offset |= offset;
+	return 1;
+}
+
+/**
+ * ice_tx_prepare_vlan_flags - prepare generic TX VLAN tagging flags for HW
+ * @tx_ring: ring to send buffer on
+ * @first: pointer to struct ice_tx_buf
+ *
+ * Checks the skb and set up correspondingly several generic transmit flags
+ * related to VLAN tagging for the HW, such as VLAN, DCB, etc.
+ *
+ * Returns error code indicate the frame should be dropped upon error and the
+ * otherwise returns 0 to indicate the flags has been set properly.
+ */
+static int
+ice_tx_prepare_vlan_flags(struct ice_ring *tx_ring, struct ice_tx_buf *first)
+{
+	struct sk_buff *skb = first->skb;
+	__be16 protocol = skb->protocol;
+
+	if (protocol == htons(ETH_P_8021Q) &&
+	    !(tx_ring->netdev->features & NETIF_F_HW_VLAN_CTAG_TX)) {
+		/* when HW VLAN acceleration is turned off by the user the
+		 * stack sets the protocol to 8021q so that the driver
+		 * can take any steps required to support the SW only
+		 * VLAN handling. In our case the driver doesn't need
+		 * to take any further steps so just set the protocol
+		 * to the encapsulated ethertype.
+		 */
+		skb->protocol = vlan_get_protocol(skb);
+		goto out;
+	}
+
+	/* if we have a HW VLAN tag being added, default to the HW one */
+	if (skb_vlan_tag_present(skb)) {
+		first->tx_flags |= skb_vlan_tag_get(skb) << ICE_TX_FLAGS_VLAN_S;
+		first->tx_flags |= ICE_TX_FLAGS_HW_VLAN;
+	} else if (protocol == htons(ETH_P_8021Q)) {
+		struct vlan_hdr *vhdr, _vhdr;
+
+		/* for SW VLAN, check the next protocol and store the tag */
+		vhdr = (struct vlan_hdr *)skb_header_pointer(skb, ETH_HLEN,
+							     sizeof(_vhdr),
+							     &_vhdr);
+		if (!vhdr)
+			return -EINVAL;
+
+		first->tx_flags |= ntohs(vhdr->h_vlan_TCI) <<
+				   ICE_TX_FLAGS_VLAN_S;
+		first->tx_flags |= ICE_TX_FLAGS_SW_VLAN;
+	}
+
+out:
+	return 0;
+}
+
+/**
+ * ice_tso - computes mss and TSO length to prepare for TSO
+ * @first: pointer to struct ice_tx_buf
+ * @off: pointer to struct that holds offload parameters
+ *
+ * Returns 0 or error (negative) if TSO can't happen, 1 otherwise.
+ */
+static
+int ice_tso(struct ice_tx_buf *first, struct ice_tx_offload_params *off)
+{
+	struct sk_buff *skb = first->skb;
+	union {
+		struct iphdr *v4;
+		struct ipv6hdr *v6;
+		unsigned char *hdr;
+	} ip;
+	union {
+		struct tcphdr *tcp;
+		unsigned char *hdr;
+	} l4;
+	u64 cd_mss, cd_tso_len;
+	u32 paylen, l4_start;
+	int err;
+
+	if (skb->ip_summed != CHECKSUM_PARTIAL)
+		return 0;
+
+	if (!skb_is_gso(skb))
+		return 0;
+
+	err = skb_cow_head(skb, 0);
+	if (err < 0)
+		return err;
+
+	ip.hdr = skb_network_header(skb);
+	l4.hdr = skb_transport_header(skb);
+
+	/* initialize outer IP header fields */
+	if (ip.v4->version == 4) {
+		ip.v4->tot_len = 0;
+		ip.v4->check = 0;
+	} else {
+		ip.v6->payload_len = 0;
+	}
+
+	/* determine offset of transport header */
+	l4_start = l4.hdr - skb->data;
+
+	/* remove payload length from checksum */
+	paylen = skb->len - l4_start;
+	csum_replace_by_diff(&l4.tcp->check, (__force __wsum)htonl(paylen));
+
+	/* compute length of segmentation header */
+	off->header_len = (l4.tcp->doff * 4) + l4_start;
+
+	/* update gso_segs and bytecount */
+	first->gso_segs = skb_shinfo(skb)->gso_segs;
+	first->bytecount = (first->gso_segs - 1) * off->header_len;
+
+	cd_tso_len = skb->len - off->header_len;
+	cd_mss = skb_shinfo(skb)->gso_size;
+
+	/* record cdesc_qw1 with TSO parameters */
+	off->cd_qw1 |= ICE_TX_DESC_DTYPE_CTX |
+			 (ICE_TX_CTX_DESC_TSO << ICE_TXD_CTX_QW1_CMD_S) |
+			 (cd_tso_len << ICE_TXD_CTX_QW1_TSO_LEN_S) |
+			 (cd_mss << ICE_TXD_CTX_QW1_MSS_S);
+	first->tx_flags |= ICE_TX_FLAGS_TSO;
+	return 1;
+}
+
 /**
  * ice_txd_use_count  - estimate the number of descriptors needed for Tx
  * @size: transmit request size in bytes
@@ -1322,8 +1686,10 @@ static bool ice_chk_linearize(struct sk_buff *skb, unsigned int count)
 static netdev_tx_t
 ice_xmit_frame_ring(struct sk_buff *skb, struct ice_ring *tx_ring)
 {
+	struct ice_tx_offload_params offload = { 0 };
 	struct ice_tx_buf *first;
 	unsigned int count;
+	int tso, csum;
 
 	count = ice_xmit_desc_count(skb);
 	if (ice_chk_linearize(skb, count)) {
@@ -1344,13 +1710,46 @@ ice_xmit_frame_ring(struct sk_buff *skb, struct ice_ring *tx_ring)
 		return NETDEV_TX_BUSY;
 	}
 
+	offload.tx_ring = tx_ring;
+
 	/* record the location of the first descriptor for this packet */
 	first = &tx_ring->tx_buf[tx_ring->next_to_use];
 	first->skb = skb;
 	first->bytecount = max_t(unsigned int, skb->len, ETH_ZLEN);
 	first->gso_segs = 1;
+	first->tx_flags = 0;
+
+	/* prepare the VLAN tagging flags for Tx */
+	if (ice_tx_prepare_vlan_flags(tx_ring, first))
+		goto out_drop;
+
+	/* set up TSO offload */
+	tso = ice_tso(first, &offload);
+	if (tso < 0)
+		goto out_drop;
+
+	/* always set up Tx checksum offload */
+	csum = ice_tx_csum(first, &offload);
+	if (csum < 0)
+		goto out_drop;
+
+	if (tso || offload.cd_tunnel_params) {
+		struct ice_tx_ctx_desc *cdesc;
+		int i = tx_ring->next_to_use;
+
+		/* grab the next descriptor */
+		cdesc = ICE_TX_CTX_DESC(tx_ring, i);
+		i++;
+		tx_ring->next_to_use = (i < tx_ring->count) ? i : 0;
+
+		/* setup context descriptor */
+		cdesc->tunneling_params = cpu_to_le32(offload.cd_tunnel_params);
+		cdesc->l2tag2 = cpu_to_le16(offload.cd_l2tag2);
+		cdesc->rsvd = cpu_to_le16(0);
+		cdesc->qw1 = cpu_to_le64(offload.cd_qw1);
+	}
 
-	ice_tx_map(tx_ring, first);
+	ice_tx_map(tx_ring, first, &offload);
 	return NETDEV_TX_OK;
 
 out_drop:

drivers/net/ethernet/intel/ice/ice_txrx.h

@@ -28,6 +28,12 @@
 	((((R)->next_to_clean > (R)->next_to_use) ? 0 : (R)->count) + \
 	(R)->next_to_clean - (R)->next_to_use - 1)
 
+#define ICE_TX_FLAGS_TSO	BIT(0)
+#define ICE_TX_FLAGS_HW_VLAN	BIT(1)
+#define ICE_TX_FLAGS_SW_VLAN	BIT(2)
+#define ICE_TX_FLAGS_VLAN_M	0xffff0000
+#define ICE_TX_FLAGS_VLAN_S	16
+
 struct ice_tx_buf {
 	struct ice_tx_desc *next_to_watch;
 	struct sk_buff *skb;
@@ -38,6 +44,17 @@ struct ice_tx_buf {
 	DEFINE_DMA_UNMAP_LEN(len);
 };
 
+struct ice_tx_offload_params {
+	u8 header_len;
+	u32 td_cmd;
+	u32 td_offset;
+	u32 td_l2tag1;
+	u16 cd_l2tag2;
+	u32 cd_tunnel_params;
+	u64 cd_qw1;
+	struct ice_ring *tx_ring;
+};
+
 struct ice_rx_buf {
 	struct sk_buff *skb;
 	dma_addr_t dma;