xref: /dpdk/drivers/net/hns3/hns3_flow.c (revision ec169149)

/* SPDX-License-Identifier: BSD-3-Clause
 * Copyright(c) 2018-2021 HiSilicon Limited.
 */

#include <rte_flow_driver.h>
#include <rte_io.h>
#include <rte_malloc.h>

#include "hns3_ethdev.h"
#include "hns3_logs.h"
#include "hns3_flow.h"

/* Default default keys */
static uint8_t hns3_hash_key[] = {
	0x6D, 0x5A, 0x56, 0xDA, 0x25, 0x5B, 0x0E, 0xC2,
	0x41, 0x67, 0x25, 0x3D, 0x43, 0xA3, 0x8F, 0xB0,
	0xD0, 0xCA, 0x2B, 0xCB, 0xAE, 0x7B, 0x30, 0xB4,
	0x77, 0xCB, 0x2D, 0xA3, 0x80, 0x30, 0xF2, 0x0C,
	0x6A, 0x42, 0xB7, 0x3B, 0xBE, 0xAC, 0x01, 0xFA
};

static const uint8_t full_mask[VNI_OR_TNI_LEN] = { 0xFF, 0xFF, 0xFF };
static const uint8_t zero_mask[VNI_OR_TNI_LEN] = { 0x00, 0x00, 0x00 };

/* Special Filter id for non-specific packet flagging. Don't change value */
#define HNS3_MAX_FILTER_ID	0x0FFF

#define ETHER_TYPE_MASK		0xFFFF
#define IPPROTO_MASK		0xFF
#define TUNNEL_TYPE_MASK	0xFFFF

#define HNS3_TUNNEL_TYPE_VXLAN		0x12B5
#define HNS3_TUNNEL_TYPE_VXLAN_GPE	0x12B6
#define HNS3_TUNNEL_TYPE_GENEVE		0x17C1
#define HNS3_TUNNEL_TYPE_NVGRE		0x6558

static enum rte_flow_item_type first_items[] = {
	RTE_FLOW_ITEM_TYPE_ETH,
	RTE_FLOW_ITEM_TYPE_IPV4,
	RTE_FLOW_ITEM_TYPE_IPV6,
	RTE_FLOW_ITEM_TYPE_TCP,
	RTE_FLOW_ITEM_TYPE_UDP,
	RTE_FLOW_ITEM_TYPE_SCTP,
	RTE_FLOW_ITEM_TYPE_ICMP,
	RTE_FLOW_ITEM_TYPE_NVGRE,
	RTE_FLOW_ITEM_TYPE_VXLAN,
	RTE_FLOW_ITEM_TYPE_GENEVE,
	RTE_FLOW_ITEM_TYPE_VXLAN_GPE
};

static enum rte_flow_item_type L2_next_items[] = {
	RTE_FLOW_ITEM_TYPE_VLAN,
	RTE_FLOW_ITEM_TYPE_IPV4,
	RTE_FLOW_ITEM_TYPE_IPV6
};

static enum rte_flow_item_type L3_next_items[] = {
	RTE_FLOW_ITEM_TYPE_TCP,
	RTE_FLOW_ITEM_TYPE_UDP,
	RTE_FLOW_ITEM_TYPE_SCTP,
	RTE_FLOW_ITEM_TYPE_NVGRE,
	RTE_FLOW_ITEM_TYPE_ICMP
};

static enum rte_flow_item_type L4_next_items[] = {
	RTE_FLOW_ITEM_TYPE_VXLAN,
	RTE_FLOW_ITEM_TYPE_GENEVE,
	RTE_FLOW_ITEM_TYPE_VXLAN_GPE
};

static enum rte_flow_item_type tunnel_next_items[] = {
	RTE_FLOW_ITEM_TYPE_ETH,
	RTE_FLOW_ITEM_TYPE_VLAN
};

struct items_step_mngr {
	enum rte_flow_item_type *items;
	int count;
};

static inline void
net_addr_to_host(uint32_t *dst, const rte_be32_t *src, size_t len)
{
	size_t i;

	for (i = 0; i < len; i++)
		dst[i] = rte_be_to_cpu_32(src[i]);
}

/*
 * This function is used to find rss general action.
 * 1. As we know RSS is used to spread packets among several queues, the flow
 *    API provide the struct rte_flow_action_rss, user could config its field
 *    sush as: func/level/types/key/queue to control RSS function.
 * 2. The flow API also supports queue region configuration for hns3. It was
 *    implemented by FDIR + RSS in hns3 hardware, user can create one FDIR rule
 *    which action is RSS queues region.
 * 3. When action is RSS, we use the following rule to distinguish:
 *    Case 1: pattern have ETH and action's queue_num > 0, indicate it is queue
 *            region configuration.
 *    Case other: an rss general action.
 */
static const struct rte_flow_action *
hns3_find_rss_general_action(const struct rte_flow_item pattern[],
			     const struct rte_flow_action actions[])
{
	const struct rte_flow_action *act = NULL;
	const struct hns3_rss_conf *rss;
	bool have_eth = false;

	for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) {
		if (actions->type == RTE_FLOW_ACTION_TYPE_RSS) {
			act = actions;
			break;
		}
	}
	if (!act)
		return NULL;

	for (; pattern->type != RTE_FLOW_ITEM_TYPE_END; pattern++) {
		if (pattern->type == RTE_FLOW_ITEM_TYPE_ETH) {
			have_eth = true;
			break;
		}
	}

	rss = act->conf;
	if (have_eth && rss->conf.queue_num) {
		/*
		 * Pattern have ETH and action's queue_num > 0, indicate this is
		 * queue region configuration.
		 * Because queue region is implemented by FDIR + RSS in hns3
		 * hardware, it needs to enter FDIR process, so here return NULL
		 * to avoid enter RSS process.
		 */
		return NULL;
	}

	return act;
}

static inline struct hns3_flow_counter *
hns3_counter_lookup(struct rte_eth_dev *dev, uint32_t id)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_pf *pf = &hns->pf;
	struct hns3_flow_counter *cnt;

	LIST_FOREACH(cnt, &pf->flow_counters, next) {
		if (cnt->id == id)
			return cnt;
	}
	return NULL;
}

static int
hns3_counter_new(struct rte_eth_dev *dev, uint32_t indirect, uint32_t id,
		 struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_pf *pf = &hns->pf;
	struct hns3_hw *hw = &hns->hw;
	struct hns3_flow_counter *cnt;
	uint64_t value;
	int ret;

	cnt = hns3_counter_lookup(dev, id);
	if (cnt) {
		if (!cnt->indirect || cnt->indirect != indirect)
			return rte_flow_error_set(error, ENOTSUP,
				RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				cnt,
				"Counter id is used, indirect flag not match");
		/* Clear the indirect counter on first use. */
		if (cnt->indirect && cnt->ref_cnt == 1)
			(void)hns3_get_count(hw, id, &value);
		cnt->ref_cnt++;
		return 0;
	}

	/* Clear the counter by read ops because the counter is read-clear */
	ret = hns3_get_count(hw, id, &value);
	if (ret)
		return rte_flow_error_set(error, EIO,
					  RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
					  "Clear counter failed!");

	cnt = rte_zmalloc("hns3 counter", sizeof(*cnt), 0);
	if (cnt == NULL)
		return rte_flow_error_set(error, ENOMEM,
					  RTE_FLOW_ERROR_TYPE_HANDLE, cnt,
					  "Alloc mem for counter failed");
	cnt->id = id;
	cnt->indirect = indirect;
	cnt->ref_cnt = 1;
	cnt->hits = 0;
	LIST_INSERT_HEAD(&pf->flow_counters, cnt, next);
	return 0;
}

static int
hns3_counter_query(struct rte_eth_dev *dev, struct rte_flow *flow,
		   struct rte_flow_query_count *qc,
		   struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_flow_counter *cnt;
	uint64_t value;
	int ret;

	/* FDIR is available only in PF driver */
	if (hns->is_vf)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
					  "Fdir is not supported in VF");
	cnt = hns3_counter_lookup(dev, flow->counter_id);
	if (cnt == NULL)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
					  "Can't find counter id");

	ret = hns3_get_count(&hns->hw, flow->counter_id, &value);
	if (ret) {
		rte_flow_error_set(error, -ret, RTE_FLOW_ERROR_TYPE_HANDLE,
				   NULL, "Read counter fail.");
		return ret;
	}
	qc->hits_set = 1;
	qc->hits = value;
	qc->bytes_set = 0;
	qc->bytes = 0;

	return 0;
}

static int
hns3_counter_release(struct rte_eth_dev *dev, uint32_t id)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;
	struct hns3_flow_counter *cnt;

	cnt = hns3_counter_lookup(dev, id);
	if (cnt == NULL) {
		hns3_err(hw, "Can't find available counter to release");
		return -EINVAL;
	}
	cnt->ref_cnt--;
	if (cnt->ref_cnt == 0) {
		LIST_REMOVE(cnt, next);
		rte_free(cnt);
	}
	return 0;
}

static void
hns3_counter_flush(struct rte_eth_dev *dev)
{
	struct hns3_pf *pf = HNS3_DEV_PRIVATE_TO_PF(dev->data->dev_private);
	LIST_HEAD(counters, hns3_flow_counter) indir_counters;
	struct hns3_flow_counter *cnt_ptr;

	LIST_INIT(&indir_counters);
	cnt_ptr = LIST_FIRST(&pf->flow_counters);
	while (cnt_ptr) {
		LIST_REMOVE(cnt_ptr, next);
		if (cnt_ptr->indirect)
			LIST_INSERT_HEAD(&indir_counters, cnt_ptr, next);
		else
			rte_free(cnt_ptr);
		cnt_ptr = LIST_FIRST(&pf->flow_counters);
	}

	/* Reset the indirect action and add to pf->flow_counters list. */
	cnt_ptr = LIST_FIRST(&indir_counters);
	while (cnt_ptr) {
		LIST_REMOVE(cnt_ptr, next);
		cnt_ptr->ref_cnt = 1;
		cnt_ptr->hits = 0;
		LIST_INSERT_HEAD(&pf->flow_counters, cnt_ptr, next);
		cnt_ptr = LIST_FIRST(&indir_counters);
	}
}

static int
hns3_handle_action_queue(struct rte_eth_dev *dev,
			 const struct rte_flow_action *action,
			 struct hns3_fdir_rule *rule,
			 struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	const struct rte_flow_action_queue *queue;
	struct hns3_hw *hw = &hns->hw;

	queue = (const struct rte_flow_action_queue *)action->conf;
	if (queue->index >= hw->data->nb_rx_queues) {
		hns3_err(hw, "queue ID(%u) is greater than number of "
			  "available queue (%u) in driver.",
			  queue->index, hw->data->nb_rx_queues);
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					  action, "Invalid queue ID in PF");
	}

	rule->queue_id = queue->index;
	rule->nb_queues = 1;
	rule->action = HNS3_FD_ACTION_ACCEPT_PACKET;
	return 0;
}

static int
hns3_handle_action_queue_region(struct rte_eth_dev *dev,
				const struct rte_flow_action *action,
				struct hns3_fdir_rule *rule,
				struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	const struct rte_flow_action_rss *conf = action->conf;
	struct hns3_hw *hw = &hns->hw;
	uint16_t idx;

	if (!hns3_dev_get_support(hw, FD_QUEUE_REGION))
		return rte_flow_error_set(error, ENOTSUP,
			RTE_FLOW_ERROR_TYPE_ACTION, action,
			"Not support config queue region!");

	if ((!rte_is_power_of_2(conf->queue_num)) ||
		conf->queue_num > hw->rss_size_max ||
		conf->queue[0] >= hw->data->nb_rx_queues ||
		conf->queue[0] + conf->queue_num > hw->data->nb_rx_queues) {
		return rte_flow_error_set(error, EINVAL,
			RTE_FLOW_ERROR_TYPE_ACTION_CONF, action,
			"Invalid start queue ID and queue num! the start queue "
			"ID must valid, the queue num must be power of 2 and "
			"<= rss_size_max.");
	}

	for (idx = 1; idx < conf->queue_num; idx++) {
		if (conf->queue[idx] != conf->queue[idx - 1] + 1)
			return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION_CONF, action,
				"Invalid queue ID sequence! the queue ID "
				"must be continuous increment.");
	}

	rule->queue_id = conf->queue[0];
	rule->nb_queues = conf->queue_num;
	rule->action = HNS3_FD_ACTION_ACCEPT_PACKET;
	return 0;
}

static int
hns3_handle_action_indirect(struct rte_eth_dev *dev,
			    const struct rte_flow_action *action,
			    struct hns3_fdir_rule *rule,
			    struct rte_flow_error *error)
{
	const struct rte_flow_action_handle *indir = action->conf;

	if (indir->indirect_type != HNS3_INDIRECT_ACTION_TYPE_COUNT)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				action, "Invalid indirect type");

	if (hns3_counter_lookup(dev, indir->counter_id) == NULL)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				action, "Counter id not exist");

	rule->act_cnt.id = indir->counter_id;
	rule->flags |= (HNS3_RULE_FLAG_COUNTER | HNS3_RULE_FLAG_COUNTER_INDIR);

	return 0;
}

/*
 * Parse actions structure from the provided pattern.
 * The pattern is validated as the items are copied.
 *
 * @param actions[in]
 * @param rule[out]
 *   NIC specific actions derived from the actions.
 * @param error[out]
 */
static int
hns3_handle_actions(struct rte_eth_dev *dev,
		    const struct rte_flow_action actions[],
		    struct hns3_fdir_rule *rule, struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	const struct rte_flow_action_count *act_count;
	const struct rte_flow_action_mark *mark;
	struct hns3_pf *pf = &hns->pf;
	uint32_t counter_num;
	int ret;

	for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) {
		switch (actions->type) {
		case RTE_FLOW_ACTION_TYPE_QUEUE:
			ret = hns3_handle_action_queue(dev, actions, rule,
						       error);
			if (ret)
				return ret;
			break;
		case RTE_FLOW_ACTION_TYPE_DROP:
			rule->action = HNS3_FD_ACTION_DROP_PACKET;
			break;
		/*
		 * Here RSS's real action is queue region.
		 * Queue region is implemented by FDIR + RSS in hns3 hardware,
		 * the FDIR's action is one queue region (start_queue_id and
		 * queue_num), then RSS spread packets to the queue region by
		 * RSS algorithm.
		 */
		case RTE_FLOW_ACTION_TYPE_RSS:
			ret = hns3_handle_action_queue_region(dev, actions,
							      rule, error);
			if (ret)
				return ret;
			break;
		case RTE_FLOW_ACTION_TYPE_MARK:
			mark =
			    (const struct rte_flow_action_mark *)actions->conf;
			if (mark->id >= HNS3_MAX_FILTER_ID)
				return rte_flow_error_set(error, EINVAL,
						RTE_FLOW_ERROR_TYPE_ACTION_CONF,
						actions,
						"Invalid Mark ID");
			rule->fd_id = mark->id;
			rule->flags |= HNS3_RULE_FLAG_FDID;
			break;
		case RTE_FLOW_ACTION_TYPE_FLAG:
			rule->fd_id = HNS3_MAX_FILTER_ID;
			rule->flags |= HNS3_RULE_FLAG_FDID;
			break;
		case RTE_FLOW_ACTION_TYPE_COUNT:
			act_count =
			    (const struct rte_flow_action_count *)actions->conf;
			counter_num = pf->fdir.fd_cfg.cnt_num[HNS3_FD_STAGE_1];
			if (act_count->id >= counter_num)
				return rte_flow_error_set(error, EINVAL,
						RTE_FLOW_ERROR_TYPE_ACTION_CONF,
						actions,
						"Invalid counter id");
			rule->act_cnt = *act_count;
			rule->flags |= HNS3_RULE_FLAG_COUNTER;
			rule->flags &= ~HNS3_RULE_FLAG_COUNTER_INDIR;
			break;
		case RTE_FLOW_ACTION_TYPE_INDIRECT:
			ret = hns3_handle_action_indirect(dev, actions, rule,
							  error);
			if (ret)
				return ret;
			break;
		case RTE_FLOW_ACTION_TYPE_VOID:
			break;
		default:
			return rte_flow_error_set(error, ENOTSUP,
						  RTE_FLOW_ERROR_TYPE_ACTION,
						  NULL, "Unsupported action");
		}
	}

	return 0;
}

static int
hns3_check_attr(const struct rte_flow_attr *attr, struct rte_flow_error *error)
{
	if (!attr->ingress)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
					  attr, "Ingress can't be zero");
	if (attr->egress)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
					  attr, "Not support egress");
	if (attr->transfer)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ATTR_TRANSFER,
					  attr, "No support for transfer");
	if (attr->priority)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
					  attr, "Not support priority");
	if (attr->group)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ATTR_GROUP,
					  attr, "Not support group");
	return 0;
}

static int
hns3_parse_eth(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
	       struct rte_flow_error *error __rte_unused)
{
	const struct rte_flow_item_eth *eth_spec;
	const struct rte_flow_item_eth *eth_mask;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		eth_mask = item->mask;
		if (eth_mask->type) {
			hns3_set_bit(rule->input_set, INNER_ETH_TYPE, 1);
			rule->key_conf.mask.ether_type =
			    rte_be_to_cpu_16(eth_mask->type);
		}
		if (!rte_is_zero_ether_addr(&eth_mask->src)) {
			hns3_set_bit(rule->input_set, INNER_SRC_MAC, 1);
			memcpy(rule->key_conf.mask.src_mac,
			       eth_mask->src.addr_bytes, RTE_ETHER_ADDR_LEN);
		}
		if (!rte_is_zero_ether_addr(&eth_mask->dst)) {
			hns3_set_bit(rule->input_set, INNER_DST_MAC, 1);
			memcpy(rule->key_conf.mask.dst_mac,
			       eth_mask->dst.addr_bytes, RTE_ETHER_ADDR_LEN);
		}
	}

	eth_spec = item->spec;
	rule->key_conf.spec.ether_type = rte_be_to_cpu_16(eth_spec->type);
	memcpy(rule->key_conf.spec.src_mac, eth_spec->src.addr_bytes,
	       RTE_ETHER_ADDR_LEN);
	memcpy(rule->key_conf.spec.dst_mac, eth_spec->dst.addr_bytes,
	       RTE_ETHER_ADDR_LEN);
	return 0;
}

static int
hns3_parse_vlan(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		struct rte_flow_error *error)
{
	const struct rte_flow_item_vlan *vlan_spec;
	const struct rte_flow_item_vlan *vlan_mask;

	rule->key_conf.vlan_num++;
	if (rule->key_conf.vlan_num > VLAN_TAG_NUM_MAX)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM, item,
					  "Vlan_num is more than 2");

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		vlan_mask = item->mask;
		if (vlan_mask->tci) {
			if (rule->key_conf.vlan_num == 1) {
				hns3_set_bit(rule->input_set, INNER_VLAN_TAG1,
					     1);
				rule->key_conf.mask.vlan_tag1 =
				    rte_be_to_cpu_16(vlan_mask->tci);
			} else {
				hns3_set_bit(rule->input_set, INNER_VLAN_TAG2,
					     1);
				rule->key_conf.mask.vlan_tag2 =
				    rte_be_to_cpu_16(vlan_mask->tci);
			}
		}
	}

	vlan_spec = item->spec;
	if (rule->key_conf.vlan_num == 1)
		rule->key_conf.spec.vlan_tag1 =
		    rte_be_to_cpu_16(vlan_spec->tci);
	else
		rule->key_conf.spec.vlan_tag2 =
		    rte_be_to_cpu_16(vlan_spec->tci);
	return 0;
}

static bool
hns3_check_ipv4_mask_supported(const struct rte_flow_item_ipv4 *ipv4_mask)
{
	if (ipv4_mask->hdr.total_length || ipv4_mask->hdr.packet_id ||
	    ipv4_mask->hdr.fragment_offset || ipv4_mask->hdr.time_to_live ||
	    ipv4_mask->hdr.hdr_checksum)
		return false;

	return true;
}

static int
hns3_parse_ipv4(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		struct rte_flow_error *error)
{
	const struct rte_flow_item_ipv4 *ipv4_spec;
	const struct rte_flow_item_ipv4 *ipv4_mask;

	hns3_set_bit(rule->input_set, INNER_ETH_TYPE, 1);
	rule->key_conf.spec.ether_type = RTE_ETHER_TYPE_IPV4;
	rule->key_conf.mask.ether_type = ETHER_TYPE_MASK;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		ipv4_mask = item->mask;
		if (!hns3_check_ipv4_mask_supported(ipv4_mask)) {
			return rte_flow_error_set(error, EINVAL,
						  RTE_FLOW_ERROR_TYPE_ITEM_MASK,
						  item,
						  "Only support src & dst ip,tos,proto in IPV4");
		}

		if (ipv4_mask->hdr.src_addr) {
			hns3_set_bit(rule->input_set, INNER_SRC_IP, 1);
			rule->key_conf.mask.src_ip[IP_ADDR_KEY_ID] =
			    rte_be_to_cpu_32(ipv4_mask->hdr.src_addr);
		}

		if (ipv4_mask->hdr.dst_addr) {
			hns3_set_bit(rule->input_set, INNER_DST_IP, 1);
			rule->key_conf.mask.dst_ip[IP_ADDR_KEY_ID] =
			    rte_be_to_cpu_32(ipv4_mask->hdr.dst_addr);
		}

		if (ipv4_mask->hdr.type_of_service) {
			hns3_set_bit(rule->input_set, INNER_IP_TOS, 1);
			rule->key_conf.mask.ip_tos =
			    ipv4_mask->hdr.type_of_service;
		}

		if (ipv4_mask->hdr.next_proto_id) {
			hns3_set_bit(rule->input_set, INNER_IP_PROTO, 1);
			rule->key_conf.mask.ip_proto =
			    ipv4_mask->hdr.next_proto_id;
		}
	}

	ipv4_spec = item->spec;
	rule->key_conf.spec.src_ip[IP_ADDR_KEY_ID] =
	    rte_be_to_cpu_32(ipv4_spec->hdr.src_addr);
	rule->key_conf.spec.dst_ip[IP_ADDR_KEY_ID] =
	    rte_be_to_cpu_32(ipv4_spec->hdr.dst_addr);
	rule->key_conf.spec.ip_tos = ipv4_spec->hdr.type_of_service;
	rule->key_conf.spec.ip_proto = ipv4_spec->hdr.next_proto_id;
	return 0;
}

static int
hns3_parse_ipv6(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		struct rte_flow_error *error)
{
	const struct rte_flow_item_ipv6 *ipv6_spec;
	const struct rte_flow_item_ipv6 *ipv6_mask;

	hns3_set_bit(rule->input_set, INNER_ETH_TYPE, 1);
	rule->key_conf.spec.ether_type = RTE_ETHER_TYPE_IPV6;
	rule->key_conf.mask.ether_type = ETHER_TYPE_MASK;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		ipv6_mask = item->mask;
		if (ipv6_mask->hdr.vtc_flow || ipv6_mask->hdr.payload_len ||
		    ipv6_mask->hdr.hop_limits) {
			return rte_flow_error_set(error, EINVAL,
						  RTE_FLOW_ERROR_TYPE_ITEM_MASK,
						  item,
						  "Only support src & dst ip,proto in IPV6");
		}
		net_addr_to_host(rule->key_conf.mask.src_ip,
				 (const rte_be32_t *)ipv6_mask->hdr.src_addr,
				 IP_ADDR_LEN);
		net_addr_to_host(rule->key_conf.mask.dst_ip,
				 (const rte_be32_t *)ipv6_mask->hdr.dst_addr,
				 IP_ADDR_LEN);
		rule->key_conf.mask.ip_proto = ipv6_mask->hdr.proto;
		if (rule->key_conf.mask.src_ip[IP_ADDR_KEY_ID])
			hns3_set_bit(rule->input_set, INNER_SRC_IP, 1);
		if (rule->key_conf.mask.dst_ip[IP_ADDR_KEY_ID])
			hns3_set_bit(rule->input_set, INNER_DST_IP, 1);
		if (ipv6_mask->hdr.proto)
			hns3_set_bit(rule->input_set, INNER_IP_PROTO, 1);
	}

	ipv6_spec = item->spec;
	net_addr_to_host(rule->key_conf.spec.src_ip,
			 (const rte_be32_t *)ipv6_spec->hdr.src_addr,
			 IP_ADDR_LEN);
	net_addr_to_host(rule->key_conf.spec.dst_ip,
			 (const rte_be32_t *)ipv6_spec->hdr.dst_addr,
			 IP_ADDR_LEN);
	rule->key_conf.spec.ip_proto = ipv6_spec->hdr.proto;

	return 0;
}

static bool
hns3_check_tcp_mask_supported(const struct rte_flow_item_tcp *tcp_mask)
{
	if (tcp_mask->hdr.sent_seq || tcp_mask->hdr.recv_ack ||
	    tcp_mask->hdr.data_off || tcp_mask->hdr.tcp_flags ||
	    tcp_mask->hdr.rx_win || tcp_mask->hdr.cksum ||
	    tcp_mask->hdr.tcp_urp)
		return false;

	return true;
}

static int
hns3_parse_tcp(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
	       struct rte_flow_error *error)
{
	const struct rte_flow_item_tcp *tcp_spec;
	const struct rte_flow_item_tcp *tcp_mask;

	hns3_set_bit(rule->input_set, INNER_IP_PROTO, 1);
	rule->key_conf.spec.ip_proto = IPPROTO_TCP;
	rule->key_conf.mask.ip_proto = IPPROTO_MASK;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		tcp_mask = item->mask;
		if (!hns3_check_tcp_mask_supported(tcp_mask)) {
			return rte_flow_error_set(error, EINVAL,
						  RTE_FLOW_ERROR_TYPE_ITEM_MASK,
						  item,
						  "Only support src & dst port in TCP");
		}

		if (tcp_mask->hdr.src_port) {
			hns3_set_bit(rule->input_set, INNER_SRC_PORT, 1);
			rule->key_conf.mask.src_port =
			    rte_be_to_cpu_16(tcp_mask->hdr.src_port);
		}
		if (tcp_mask->hdr.dst_port) {
			hns3_set_bit(rule->input_set, INNER_DST_PORT, 1);
			rule->key_conf.mask.dst_port =
			    rte_be_to_cpu_16(tcp_mask->hdr.dst_port);
		}
	}

	tcp_spec = item->spec;
	rule->key_conf.spec.src_port = rte_be_to_cpu_16(tcp_spec->hdr.src_port);
	rule->key_conf.spec.dst_port = rte_be_to_cpu_16(tcp_spec->hdr.dst_port);

	return 0;
}

static int
hns3_parse_udp(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
	       struct rte_flow_error *error)
{
	const struct rte_flow_item_udp *udp_spec;
	const struct rte_flow_item_udp *udp_mask;

	hns3_set_bit(rule->input_set, INNER_IP_PROTO, 1);
	rule->key_conf.spec.ip_proto = IPPROTO_UDP;
	rule->key_conf.mask.ip_proto = IPPROTO_MASK;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		udp_mask = item->mask;
		if (udp_mask->hdr.dgram_len || udp_mask->hdr.dgram_cksum) {
			return rte_flow_error_set(error, EINVAL,
						  RTE_FLOW_ERROR_TYPE_ITEM_MASK,
						  item,
						  "Only support src & dst port in UDP");
		}
		if (udp_mask->hdr.src_port) {
			hns3_set_bit(rule->input_set, INNER_SRC_PORT, 1);
			rule->key_conf.mask.src_port =
			    rte_be_to_cpu_16(udp_mask->hdr.src_port);
		}
		if (udp_mask->hdr.dst_port) {
			hns3_set_bit(rule->input_set, INNER_DST_PORT, 1);
			rule->key_conf.mask.dst_port =
			    rte_be_to_cpu_16(udp_mask->hdr.dst_port);
		}
	}

	udp_spec = item->spec;
	rule->key_conf.spec.src_port = rte_be_to_cpu_16(udp_spec->hdr.src_port);
	rule->key_conf.spec.dst_port = rte_be_to_cpu_16(udp_spec->hdr.dst_port);

	return 0;
}

static int
hns3_parse_sctp(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		struct rte_flow_error *error)
{
	const struct rte_flow_item_sctp *sctp_spec;
	const struct rte_flow_item_sctp *sctp_mask;

	hns3_set_bit(rule->input_set, INNER_IP_PROTO, 1);
	rule->key_conf.spec.ip_proto = IPPROTO_SCTP;
	rule->key_conf.mask.ip_proto = IPPROTO_MASK;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	if (item->mask) {
		sctp_mask = item->mask;
		if (sctp_mask->hdr.cksum)
			return rte_flow_error_set(error, EINVAL,
						  RTE_FLOW_ERROR_TYPE_ITEM_MASK,
						  item,
						  "Only support src & dst port in SCTP");
		if (sctp_mask->hdr.src_port) {
			hns3_set_bit(rule->input_set, INNER_SRC_PORT, 1);
			rule->key_conf.mask.src_port =
			    rte_be_to_cpu_16(sctp_mask->hdr.src_port);
		}
		if (sctp_mask->hdr.dst_port) {
			hns3_set_bit(rule->input_set, INNER_DST_PORT, 1);
			rule->key_conf.mask.dst_port =
			    rte_be_to_cpu_16(sctp_mask->hdr.dst_port);
		}
		if (sctp_mask->hdr.tag) {
			hns3_set_bit(rule->input_set, INNER_SCTP_TAG, 1);
			rule->key_conf.mask.sctp_tag =
			    rte_be_to_cpu_32(sctp_mask->hdr.tag);
		}
	}

	sctp_spec = item->spec;
	rule->key_conf.spec.src_port =
	    rte_be_to_cpu_16(sctp_spec->hdr.src_port);
	rule->key_conf.spec.dst_port =
	    rte_be_to_cpu_16(sctp_spec->hdr.dst_port);
	rule->key_conf.spec.sctp_tag = rte_be_to_cpu_32(sctp_spec->hdr.tag);

	return 0;
}

/*
 * Check items before tunnel, save inner configs to outer configs, and clear
 * inner configs.
 * The key consists of two parts: meta_data and tuple keys.
 * Meta data uses 15 bits, including vlan_num(2bit), des_port(12bit) and tunnel
 * packet(1bit).
 * Tuple keys uses 384bit, including ot_dst-mac(48bit), ot_dst-port(16bit),
 * ot_tun_vni(24bit), ot_flow_id(8bit), src-mac(48bit), dst-mac(48bit),
 * src-ip(32/128bit), dst-ip(32/128bit), src-port(16bit), dst-port(16bit),
 * tos(8bit), ether-proto(16bit), ip-proto(8bit), vlantag1(16bit),
 * Vlantag2(16bit) and sctp-tag(32bit).
 */
static int
hns3_handle_tunnel(const struct rte_flow_item *item,
		   struct hns3_fdir_rule *rule, struct rte_flow_error *error)
{
	/* check eth config */
	if (rule->input_set & (BIT(INNER_SRC_MAC) | BIT(INNER_DST_MAC)))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  item, "Outer eth mac is unsupported");
	if (rule->input_set & BIT(INNER_ETH_TYPE)) {
		hns3_set_bit(rule->input_set, OUTER_ETH_TYPE, 1);
		rule->key_conf.spec.outer_ether_type =
		    rule->key_conf.spec.ether_type;
		rule->key_conf.mask.outer_ether_type =
		    rule->key_conf.mask.ether_type;
		hns3_set_bit(rule->input_set, INNER_ETH_TYPE, 0);
		rule->key_conf.spec.ether_type = 0;
		rule->key_conf.mask.ether_type = 0;
	}

	/* check vlan config */
	if (rule->input_set & (BIT(INNER_VLAN_TAG1) | BIT(INNER_VLAN_TAG2)))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  item,
					  "Outer vlan tags is unsupported");

	/* clear vlan_num for inner vlan select */
	rule->key_conf.outer_vlan_num = rule->key_conf.vlan_num;
	rule->key_conf.vlan_num = 0;

	/* check L3 config */
	if (rule->input_set &
	    (BIT(INNER_SRC_IP) | BIT(INNER_DST_IP) | BIT(INNER_IP_TOS)))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  item, "Outer ip is unsupported");
	if (rule->input_set & BIT(INNER_IP_PROTO)) {
		hns3_set_bit(rule->input_set, OUTER_IP_PROTO, 1);
		rule->key_conf.spec.outer_proto = rule->key_conf.spec.ip_proto;
		rule->key_conf.mask.outer_proto = rule->key_conf.mask.ip_proto;
		hns3_set_bit(rule->input_set, INNER_IP_PROTO, 0);
		rule->key_conf.spec.ip_proto = 0;
		rule->key_conf.mask.ip_proto = 0;
	}

	/* check L4 config */
	if (rule->input_set & BIT(INNER_SCTP_TAG))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM, item,
					  "Outer sctp tag is unsupported");

	if (rule->input_set & BIT(INNER_SRC_PORT)) {
		hns3_set_bit(rule->input_set, OUTER_SRC_PORT, 1);
		rule->key_conf.spec.outer_src_port =
		    rule->key_conf.spec.src_port;
		rule->key_conf.mask.outer_src_port =
		    rule->key_conf.mask.src_port;
		hns3_set_bit(rule->input_set, INNER_SRC_PORT, 0);
		rule->key_conf.spec.src_port = 0;
		rule->key_conf.mask.src_port = 0;
	}
	if (rule->input_set & BIT(INNER_DST_PORT)) {
		hns3_set_bit(rule->input_set, INNER_DST_PORT, 0);
		rule->key_conf.spec.dst_port = 0;
		rule->key_conf.mask.dst_port = 0;
	}
	return 0;
}

static int
hns3_parse_vxlan(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		 struct rte_flow_error *error)
{
	const struct rte_flow_item_vxlan *vxlan_spec;
	const struct rte_flow_item_vxlan *vxlan_mask;

	hns3_set_bit(rule->input_set, OUTER_DST_PORT, 1);
	rule->key_conf.mask.tunnel_type = TUNNEL_TYPE_MASK;
	if (item->type == RTE_FLOW_ITEM_TYPE_VXLAN)
		rule->key_conf.spec.tunnel_type = HNS3_TUNNEL_TYPE_VXLAN;
	else
		rule->key_conf.spec.tunnel_type = HNS3_TUNNEL_TYPE_VXLAN_GPE;

	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	vxlan_mask = item->mask;
	vxlan_spec = item->spec;

	if (vxlan_mask->flags)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "Flags is not supported in VxLAN");

	/* VNI must be totally masked or not. */
	if (memcmp(vxlan_mask->vni, full_mask, VNI_OR_TNI_LEN) &&
	    memcmp(vxlan_mask->vni, zero_mask, VNI_OR_TNI_LEN))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "VNI must be totally masked or not in VxLAN");
	if (vxlan_mask->vni[0]) {
		hns3_set_bit(rule->input_set, OUTER_TUN_VNI, 1);
		memcpy(rule->key_conf.mask.outer_tun_vni, vxlan_mask->vni,
			   VNI_OR_TNI_LEN);
	}
	memcpy(rule->key_conf.spec.outer_tun_vni, vxlan_spec->vni,
		   VNI_OR_TNI_LEN);
	return 0;
}

static int
hns3_parse_nvgre(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		 struct rte_flow_error *error)
{
	const struct rte_flow_item_nvgre *nvgre_spec;
	const struct rte_flow_item_nvgre *nvgre_mask;

	hns3_set_bit(rule->input_set, OUTER_IP_PROTO, 1);
	rule->key_conf.spec.outer_proto = IPPROTO_GRE;
	rule->key_conf.mask.outer_proto = IPPROTO_MASK;

	hns3_set_bit(rule->input_set, OUTER_DST_PORT, 1);
	rule->key_conf.spec.tunnel_type = HNS3_TUNNEL_TYPE_NVGRE;
	rule->key_conf.mask.tunnel_type = ~HNS3_TUNNEL_TYPE_NVGRE;
	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	nvgre_mask = item->mask;
	nvgre_spec = item->spec;

	if (nvgre_mask->protocol || nvgre_mask->c_k_s_rsvd0_ver)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "Ver/protocol is not supported in NVGRE");

	/* TNI must be totally masked or not. */
	if (memcmp(nvgre_mask->tni, full_mask, VNI_OR_TNI_LEN) &&
	    memcmp(nvgre_mask->tni, zero_mask, VNI_OR_TNI_LEN))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "TNI must be totally masked or not in NVGRE");

	if (nvgre_mask->tni[0]) {
		hns3_set_bit(rule->input_set, OUTER_TUN_VNI, 1);
		memcpy(rule->key_conf.mask.outer_tun_vni, nvgre_mask->tni,
			   VNI_OR_TNI_LEN);
	}
	memcpy(rule->key_conf.spec.outer_tun_vni, nvgre_spec->tni,
		   VNI_OR_TNI_LEN);

	if (nvgre_mask->flow_id) {
		hns3_set_bit(rule->input_set, OUTER_TUN_FLOW_ID, 1);
		rule->key_conf.mask.outer_tun_flow_id = nvgre_mask->flow_id;
	}
	rule->key_conf.spec.outer_tun_flow_id = nvgre_spec->flow_id;
	return 0;
}

static int
hns3_parse_geneve(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		  struct rte_flow_error *error)
{
	const struct rte_flow_item_geneve *geneve_spec;
	const struct rte_flow_item_geneve *geneve_mask;

	hns3_set_bit(rule->input_set, OUTER_DST_PORT, 1);
	rule->key_conf.spec.tunnel_type = HNS3_TUNNEL_TYPE_GENEVE;
	rule->key_conf.mask.tunnel_type = TUNNEL_TYPE_MASK;
	/* Only used to describe the protocol stack. */
	if (item->spec == NULL && item->mask == NULL)
		return 0;

	geneve_mask = item->mask;
	geneve_spec = item->spec;

	if (geneve_mask->ver_opt_len_o_c_rsvd0 || geneve_mask->protocol)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "Ver/protocol is not supported in GENEVE");
	/* VNI must be totally masked or not. */
	if (memcmp(geneve_mask->vni, full_mask, VNI_OR_TNI_LEN) &&
	    memcmp(geneve_mask->vni, zero_mask, VNI_OR_TNI_LEN))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_MASK, item,
					  "VNI must be totally masked or not in GENEVE");
	if (geneve_mask->vni[0]) {
		hns3_set_bit(rule->input_set, OUTER_TUN_VNI, 1);
		memcpy(rule->key_conf.mask.outer_tun_vni, geneve_mask->vni,
			   VNI_OR_TNI_LEN);
	}
	memcpy(rule->key_conf.spec.outer_tun_vni, geneve_spec->vni,
		   VNI_OR_TNI_LEN);
	return 0;
}

static int
hns3_parse_tunnel(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		  struct rte_flow_error *error)
{
	int ret;

	if (item->spec == NULL && item->mask)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM, item,
					  "Can't configure FDIR with mask "
					  "but without spec");
	else if (item->spec && (item->mask == NULL))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM, item,
					  "Tunnel packets must configure "
					  "with mask");

	switch (item->type) {
	case RTE_FLOW_ITEM_TYPE_VXLAN:
	case RTE_FLOW_ITEM_TYPE_VXLAN_GPE:
		ret = hns3_parse_vxlan(item, rule, error);
		break;
	case RTE_FLOW_ITEM_TYPE_NVGRE:
		ret = hns3_parse_nvgre(item, rule, error);
		break;
	case RTE_FLOW_ITEM_TYPE_GENEVE:
		ret = hns3_parse_geneve(item, rule, error);
		break;
	default:
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  NULL, "Unsupported tunnel type!");
	}
	if (ret)
		return ret;
	return hns3_handle_tunnel(item, rule, error);
}

static int
hns3_parse_normal(const struct rte_flow_item *item, struct hns3_fdir_rule *rule,
		  struct items_step_mngr *step_mngr,
		  struct rte_flow_error *error)
{
	int ret;

	if (item->spec == NULL && item->mask)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM, item,
					  "Can't configure FDIR with mask "
					  "but without spec");

	switch (item->type) {
	case RTE_FLOW_ITEM_TYPE_ETH:
		ret = hns3_parse_eth(item, rule, error);
		step_mngr->items = L2_next_items;
		step_mngr->count = RTE_DIM(L2_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_VLAN:
		ret = hns3_parse_vlan(item, rule, error);
		step_mngr->items = L2_next_items;
		step_mngr->count = RTE_DIM(L2_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_IPV4:
		ret = hns3_parse_ipv4(item, rule, error);
		step_mngr->items = L3_next_items;
		step_mngr->count = RTE_DIM(L3_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_IPV6:
		ret = hns3_parse_ipv6(item, rule, error);
		step_mngr->items = L3_next_items;
		step_mngr->count = RTE_DIM(L3_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_TCP:
		ret = hns3_parse_tcp(item, rule, error);
		step_mngr->items = L4_next_items;
		step_mngr->count = RTE_DIM(L4_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_UDP:
		ret = hns3_parse_udp(item, rule, error);
		step_mngr->items = L4_next_items;
		step_mngr->count = RTE_DIM(L4_next_items);
		break;
	case RTE_FLOW_ITEM_TYPE_SCTP:
		ret = hns3_parse_sctp(item, rule, error);
		step_mngr->items = L4_next_items;
		step_mngr->count = RTE_DIM(L4_next_items);
		break;
	default:
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  NULL, "Unsupported normal type!");
	}

	return ret;
}

static int
hns3_validate_item(const struct rte_flow_item *item,
		   struct items_step_mngr step_mngr,
		   struct rte_flow_error *error)
{
	int i;

	if (item->last)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ITEM_LAST, item,
					  "Not supported last point for range");

	for (i = 0; i < step_mngr.count; i++) {
		if (item->type == step_mngr.items[i])
			break;
	}

	if (i == step_mngr.count) {
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM,
					  item, "Inval or missing item");
	}
	return 0;
}

static inline bool
is_tunnel_packet(enum rte_flow_item_type type)
{
	if (type == RTE_FLOW_ITEM_TYPE_VXLAN_GPE ||
	    type == RTE_FLOW_ITEM_TYPE_VXLAN ||
	    type == RTE_FLOW_ITEM_TYPE_NVGRE ||
	    type == RTE_FLOW_ITEM_TYPE_GENEVE)
		return true;
	return false;
}

/*
 * Parse the flow director rule.
 * The supported PATTERN:
 *   case: non-tunnel packet:
 *     ETH : src-mac, dst-mac, ethertype
 *     VLAN: tag1, tag2
 *     IPv4: src-ip, dst-ip, tos, proto
 *     IPv6: src-ip(last 32 bit addr), dst-ip(last 32 bit addr), proto
 *     UDP : src-port, dst-port
 *     TCP : src-port, dst-port
 *     SCTP: src-port, dst-port, tag
 *   case: tunnel packet:
 *     OUTER-ETH: ethertype
 *     OUTER-L3 : proto
 *     OUTER-L4 : src-port, dst-port
 *     TUNNEL   : vni, flow-id(only valid when NVGRE)
 *     INNER-ETH/VLAN/IPv4/IPv6/UDP/TCP/SCTP: same as non-tunnel packet
 * The supported ACTION:
 *    QUEUE
 *    DROP
 *    COUNT
 *    MARK: the id range [0, 4094]
 *    FLAG
 *    RSS: only valid if firmware support FD_QUEUE_REGION.
 */
static int
hns3_parse_fdir_filter(struct rte_eth_dev *dev,
		       const struct rte_flow_item pattern[],
		       const struct rte_flow_action actions[],
		       struct hns3_fdir_rule *rule,
		       struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	const struct rte_flow_item *item;
	struct items_step_mngr step_mngr;
	int ret;

	/* FDIR is available only in PF driver */
	if (hns->is_vf)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
					  "Fdir not supported in VF");

	step_mngr.items = first_items;
	step_mngr.count = RTE_DIM(first_items);
	for (item = pattern; item->type != RTE_FLOW_ITEM_TYPE_END; item++) {
		if (item->type == RTE_FLOW_ITEM_TYPE_VOID)
			continue;

		ret = hns3_validate_item(item, step_mngr, error);
		if (ret)
			return ret;

		if (is_tunnel_packet(item->type)) {
			ret = hns3_parse_tunnel(item, rule, error);
			if (ret)
				return ret;
			step_mngr.items = tunnel_next_items;
			step_mngr.count = RTE_DIM(tunnel_next_items);
		} else {
			ret = hns3_parse_normal(item, rule, &step_mngr, error);
			if (ret)
				return ret;
		}
	}

	return hns3_handle_actions(dev, actions, rule, error);
}

static void
hns3_filterlist_flush(struct rte_eth_dev *dev)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct hns3_fdir_rule_ele *fdir_rule_ptr;
	struct hns3_rss_conf_ele *rss_filter_ptr;
	struct hns3_flow_mem *flow_node;

	fdir_rule_ptr = TAILQ_FIRST(&hw->flow_fdir_list);
	while (fdir_rule_ptr) {
		TAILQ_REMOVE(&hw->flow_fdir_list, fdir_rule_ptr, entries);
		rte_free(fdir_rule_ptr);
		fdir_rule_ptr = TAILQ_FIRST(&hw->flow_fdir_list);
	}

	rss_filter_ptr = TAILQ_FIRST(&hw->flow_rss_list);
	while (rss_filter_ptr) {
		TAILQ_REMOVE(&hw->flow_rss_list, rss_filter_ptr, entries);
		rte_free(rss_filter_ptr);
		rss_filter_ptr = TAILQ_FIRST(&hw->flow_rss_list);
	}

	flow_node = TAILQ_FIRST(&hw->flow_list);
	while (flow_node) {
		TAILQ_REMOVE(&hw->flow_list, flow_node, entries);
		rte_free(flow_node->flow);
		rte_free(flow_node);
		flow_node = TAILQ_FIRST(&hw->flow_list);
	}
}

static bool
hns3_action_rss_same(const struct rte_flow_action_rss *comp,
		     const struct rte_flow_action_rss *with)
{
	bool rss_key_is_same;
	bool func_is_same;

	/*
	 * When user flush all RSS rule, RSS func is set invalid with
	 * RTE_ETH_HASH_FUNCTION_MAX. Then the user create a flow after
	 * flushed, any validate RSS func is different with it before
	 * flushed. Others, when user create an action RSS with RSS func
	 * specified RTE_ETH_HASH_FUNCTION_DEFAULT, the func is the same
	 * between continuous RSS flow.
	 */
	if (comp->func == RTE_ETH_HASH_FUNCTION_MAX)
		func_is_same = false;
	else
		func_is_same = (with->func != RTE_ETH_HASH_FUNCTION_DEFAULT) ?
				(comp->func == with->func) : true;

	if (with->key_len == 0 || with->key == NULL)
		rss_key_is_same = 1;
	else
		rss_key_is_same = comp->key_len == with->key_len &&
		!memcmp(comp->key, with->key, with->key_len);

	return (func_is_same && rss_key_is_same &&
		comp->types == (with->types & HNS3_ETH_RSS_SUPPORT) &&
		comp->level == with->level &&
		comp->queue_num == with->queue_num &&
		!memcmp(comp->queue, with->queue,
			sizeof(*with->queue) * with->queue_num));
}

static int
hns3_rss_conf_copy(struct hns3_rss_conf *out,
		   const struct rte_flow_action_rss *in)
{
	if (in->key_len > RTE_DIM(out->key) ||
	    in->queue_num > RTE_DIM(out->queue))
		return -EINVAL;
	if (in->key == NULL && in->key_len)
		return -EINVAL;
	out->conf = (struct rte_flow_action_rss) {
		.func = in->func,
		.level = in->level,
		.types = in->types,
		.key_len = in->key_len,
		.queue_num = in->queue_num,
	};
	out->conf.queue = memcpy(out->queue, in->queue,
				sizeof(*in->queue) * in->queue_num);
	if (in->key)
		out->conf.key = memcpy(out->key, in->key, in->key_len);

	return 0;
}

static bool
hns3_rss_input_tuple_supported(struct hns3_hw *hw,
			       const struct rte_flow_action_rss *rss)
{
	/*
	 * For IP packet, it is not supported to use src/dst port fields to RSS
	 * hash for the following packet types.
	 * - IPV4 FRAG | IPV4 NONFRAG | IPV6 FRAG | IPV6 NONFRAG
	 * Besides, for Kunpeng920, the NIC HW is not supported to use src/dst
	 * port fields to RSS hash for IPV6 SCTP packet type. However, the
	 * Kunpeng930 and future kunpeng series support to use src/dst port
	 * fields to RSS hash for IPv6 SCTP packet type.
	 */
	if (rss->types & (RTE_ETH_RSS_L4_DST_ONLY | RTE_ETH_RSS_L4_SRC_ONLY) &&
	    (rss->types & RTE_ETH_RSS_IP ||
	    (!hw->rss_info.ipv6_sctp_offload_supported &&
	    rss->types & RTE_ETH_RSS_NONFRAG_IPV6_SCTP)))
		return false;

	return true;
}

/*
 * This function is used to parse rss action validation.
 */
static int
hns3_parse_rss_filter(struct rte_eth_dev *dev,
		      const struct rte_flow_action *actions,
		      struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;
	struct hns3_rss_conf *rss_conf = &hw->rss_info;
	const struct rte_flow_action_rss *rss;
	const struct rte_flow_action *act;
	uint32_t act_index = 0;
	uint16_t n;

	NEXT_ITEM_OF_ACTION(act, actions, act_index);
	rss = act->conf;

	if (rss == NULL) {
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					  act, "no valid queues");
	}

	if (rss->queue_num > RTE_DIM(rss_conf->queue))
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF, act,
					  "queue number configured exceeds "
					  "queue buffer size driver supported");

	for (n = 0; n < rss->queue_num; n++) {
		if (rss->queue[n] < hw->alloc_rss_size)
			continue;
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF, act,
					  "queue id must be less than queue number allocated to a TC");
	}

	if (!(rss->types & HNS3_ETH_RSS_SUPPORT) && rss->types)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					  act,
					  "Flow types is unsupported by "
					  "hns3's RSS");
	if (rss->func >= RTE_ETH_HASH_FUNCTION_MAX)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF, act,
					  "RSS hash func are not supported");
	if (rss->level)
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF, act,
					  "a nonzero RSS encapsulation level is not supported");
	if (rss->key_len && rss->key_len != RTE_DIM(rss_conf->key))
		return rte_flow_error_set(error, ENOTSUP,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF, act,
					  "RSS hash key must be exactly 40 bytes");

	if (!hns3_rss_input_tuple_supported(hw, rss))
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					  &rss->types,
					  "input RSS types are not supported");

	act_index++;

	/* Check if the next not void action is END */
	NEXT_ITEM_OF_ACTION(act, actions, act_index);
	if (act->type != RTE_FLOW_ACTION_TYPE_END) {
		memset(rss_conf, 0, sizeof(struct hns3_rss_conf));
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION,
					  act, "Not supported action.");
	}

	return 0;
}

static int
hns3_disable_rss(struct hns3_hw *hw)
{
	int ret;

	ret = hns3_set_rss_tuple_by_rss_hf(hw, 0);
	if (ret)
		return ret;

	return 0;
}

static void
hns3_adjust_rss_key(struct hns3_hw *hw, struct rte_flow_action_rss *rss_conf)
{
	if (rss_conf->key == NULL || rss_conf->key_len < HNS3_RSS_KEY_SIZE) {
		hns3_warn(hw, "Default RSS hash key to be set");
		rss_conf->key = hns3_hash_key;
		rss_conf->key_len = HNS3_RSS_KEY_SIZE;
	}
}

static int
hns3_parse_rss_algorithm(struct hns3_hw *hw, enum rte_eth_hash_function *func,
			 uint8_t *hash_algo)
{
	enum rte_eth_hash_function algo_func = *func;
	switch (algo_func) {
	case RTE_ETH_HASH_FUNCTION_DEFAULT:
		/* Keep *hash_algo as what it used to be */
		algo_func = hw->rss_info.conf.func;
		break;
	case RTE_ETH_HASH_FUNCTION_TOEPLITZ:
		*hash_algo = HNS3_RSS_HASH_ALGO_TOEPLITZ;
		break;
	case RTE_ETH_HASH_FUNCTION_SIMPLE_XOR:
		*hash_algo = HNS3_RSS_HASH_ALGO_SIMPLE;
		break;
	case RTE_ETH_HASH_FUNCTION_SYMMETRIC_TOEPLITZ:
		*hash_algo = HNS3_RSS_HASH_ALGO_SYMMETRIC_TOEP;
		break;
	default:
		hns3_err(hw, "Invalid RSS algorithm configuration(%d)",
			 algo_func);
		return -EINVAL;
	}
	*func = algo_func;

	return 0;
}

static int
hns3_hw_rss_hash_set(struct hns3_hw *hw, struct rte_flow_action_rss *rss_config)
{
	int ret;

	hns3_adjust_rss_key(hw, rss_config);

	ret = hns3_parse_rss_algorithm(hw, &rss_config->func,
				       &hw->rss_info.hash_algo);
	if (ret)
		return ret;

	ret = hns3_rss_set_algo_key(hw, rss_config->key);
	if (ret)
		return ret;

	hw->rss_info.conf.func = rss_config->func;

	ret = hns3_set_rss_tuple_by_rss_hf(hw, rss_config->types);
	if (ret)
		hns3_err(hw, "Update RSS tuples by rss hf failed %d", ret);

	return ret;
}

static int
hns3_update_indir_table(struct rte_eth_dev *dev,
			const struct rte_flow_action_rss *conf, uint16_t num)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;
	uint16_t indir_tbl[HNS3_RSS_IND_TBL_SIZE_MAX];
	uint16_t j;
	uint32_t i;

	/* Fill in redirection table */
	memcpy(indir_tbl, hw->rss_info.rss_indirection_tbl,
	       sizeof(hw->rss_info.rss_indirection_tbl));
	for (i = 0, j = 0; i < hw->rss_ind_tbl_size; i++, j++) {
		j %= num;
		if (conf->queue[j] >= hw->alloc_rss_size) {
			hns3_err(hw, "queue id(%u) set to redirection table "
				 "exceeds queue number(%u) allocated to a TC.",
				 conf->queue[j], hw->alloc_rss_size);
			return -EINVAL;
		}
		indir_tbl[i] = conf->queue[j];
	}

	return hns3_set_rss_indir_table(hw, indir_tbl, hw->rss_ind_tbl_size);
}

static int
hns3_config_rss_filter(struct rte_eth_dev *dev,
		       const struct hns3_rss_conf *conf, bool add)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_rss_conf_ele *rss_filter_ptr;
	struct hns3_hw *hw = &hns->hw;
	struct hns3_rss_conf *rss_info;
	uint64_t flow_types;
	uint16_t num;
	int ret;

	struct rte_flow_action_rss rss_flow_conf = {
		.func = conf->conf.func,
		.level = conf->conf.level,
		.types = conf->conf.types,
		.key_len = conf->conf.key_len,
		.queue_num = conf->conf.queue_num,
		.key = conf->conf.key_len ?
		    (void *)(uintptr_t)conf->conf.key : NULL,
		.queue = conf->conf.queue,
	};

	/* Filter the unsupported flow types */
	flow_types = conf->conf.types ?
		     rss_flow_conf.types & HNS3_ETH_RSS_SUPPORT :
		     hw->rss_info.conf.types;
	if (flow_types != rss_flow_conf.types)
		hns3_warn(hw, "modified RSS types based on hardware support, "
			      "requested:0x%" PRIx64 " configured:0x%" PRIx64,
			  rss_flow_conf.types, flow_types);
	/* Update the useful flow types */
	rss_flow_conf.types = flow_types;

	rss_info = &hw->rss_info;
	if (!add) {
		if (!conf->valid)
			return 0;

		ret = hns3_disable_rss(hw);
		if (ret) {
			hns3_err(hw, "RSS disable failed(%d)", ret);
			return ret;
		}

		if (rss_flow_conf.queue_num) {
			/*
			 * Due the content of queue pointer have been reset to
			 * 0, the rss_info->conf.queue should be set to NULL
			 */
			rss_info->conf.queue = NULL;
			rss_info->conf.queue_num = 0;
		}

		/* set RSS func invalid after flushed */
		rss_info->conf.func = RTE_ETH_HASH_FUNCTION_MAX;
		return 0;
	}

	/* Set rx queues to use */
	num = RTE_MIN(dev->data->nb_rx_queues, rss_flow_conf.queue_num);
	if (rss_flow_conf.queue_num > num)
		hns3_warn(hw, "Config queue numbers %u are beyond the scope of truncated",
			  rss_flow_conf.queue_num);
	hns3_info(hw, "Max of contiguous %u PF queues are configured", num);

	rte_spinlock_lock(&hw->lock);
	if (num) {
		ret = hns3_update_indir_table(dev, &rss_flow_conf, num);
		if (ret)
			goto rss_config_err;
	}

	/* Set hash algorithm and flow types by the user's config */
	ret = hns3_hw_rss_hash_set(hw, &rss_flow_conf);
	if (ret)
		goto rss_config_err;

	ret = hns3_rss_conf_copy(rss_info, &rss_flow_conf);
	if (ret) {
		hns3_err(hw, "RSS config init fail(%d)", ret);
		goto rss_config_err;
	}

	/*
	 * When create a new RSS rule, the old rule will be overlaid and set
	 * invalid.
	 */
	TAILQ_FOREACH(rss_filter_ptr, &hw->flow_rss_list, entries)
		rss_filter_ptr->filter_info.valid = false;

rss_config_err:
	rte_spinlock_unlock(&hw->lock);

	return ret;
}

static int
hns3_clear_rss_filter(struct rte_eth_dev *dev)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_rss_conf_ele *rss_filter_ptr;
	struct hns3_hw *hw = &hns->hw;
	int rss_rule_succ_cnt = 0; /* count for success of clearing RSS rules */
	int rss_rule_fail_cnt = 0; /* count for failure of clearing RSS rules */
	int ret = 0;

	rss_filter_ptr = TAILQ_FIRST(&hw->flow_rss_list);
	while (rss_filter_ptr) {
		TAILQ_REMOVE(&hw->flow_rss_list, rss_filter_ptr, entries);
		ret = hns3_config_rss_filter(dev, &rss_filter_ptr->filter_info,
					     false);
		if (ret)
			rss_rule_fail_cnt++;
		else
			rss_rule_succ_cnt++;
		rte_free(rss_filter_ptr);
		rss_filter_ptr = TAILQ_FIRST(&hw->flow_rss_list);
	}

	if (rss_rule_fail_cnt) {
		hns3_err(hw, "fail to delete all RSS filters, success num = %d "
			     "fail num = %d", rss_rule_succ_cnt,
			     rss_rule_fail_cnt);
		ret = -EIO;
	}

	return ret;
}

int
hns3_restore_rss_filter(struct rte_eth_dev *dev)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;

	/* When user flush all rules, it doesn't need to restore RSS rule */
	if (hw->rss_info.conf.func == RTE_ETH_HASH_FUNCTION_MAX)
		return 0;

	return hns3_config_rss_filter(dev, &hw->rss_info, true);
}

static int
hns3_flow_parse_rss(struct rte_eth_dev *dev,
		    const struct hns3_rss_conf *conf, bool add)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;
	bool ret;

	ret = hns3_action_rss_same(&hw->rss_info.conf, &conf->conf);
	if (ret) {
		hns3_err(hw, "Enter duplicate RSS configuration : %d", ret);
		return -EINVAL;
	}

	return hns3_config_rss_filter(dev, conf, add);
}

static int
hns3_flow_args_check(const struct rte_flow_attr *attr,
		     const struct rte_flow_item pattern[],
		     const struct rte_flow_action actions[],
		     struct rte_flow_error *error)
{
	if (pattern == NULL)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ITEM_NUM,
					  NULL, "NULL pattern.");

	if (actions == NULL)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ACTION_NUM,
					  NULL, "NULL action.");

	if (attr == NULL)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_ATTR,
					  NULL, "NULL attribute.");

	return hns3_check_attr(attr, error);
}

/*
 * Check if the flow rule is supported by hns3.
 * It only checks the format. Don't guarantee the rule can be programmed into
 * the HW. Because there can be no enough room for the rule.
 */
static int
hns3_flow_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
		   const struct rte_flow_item pattern[],
		   const struct rte_flow_action actions[],
		   struct rte_flow_error *error)
{
	struct hns3_fdir_rule fdir_rule;
	int ret;

	ret = hns3_flow_args_check(attr, pattern, actions, error);
	if (ret)
		return ret;

	if (hns3_find_rss_general_action(pattern, actions))
		return hns3_parse_rss_filter(dev, actions, error);

	memset(&fdir_rule, 0, sizeof(struct hns3_fdir_rule));
	return hns3_parse_fdir_filter(dev, pattern, actions, &fdir_rule, error);
}

static int
hns3_flow_create_rss_rule(struct rte_eth_dev *dev,
			  const struct rte_flow_action *act,
			  struct rte_flow *flow)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct hns3_rss_conf_ele *rss_filter_ptr;
	const struct hns3_rss_conf *rss_conf;
	int ret;

	rss_filter_ptr = rte_zmalloc("hns3 rss filter",
				     sizeof(struct hns3_rss_conf_ele), 0);
	if (rss_filter_ptr == NULL) {
		hns3_err(hw, "failed to allocate hns3_rss_filter memory");
		return -ENOMEM;
	}

	/*
	 * After all the preceding tasks are successfully configured, configure
	 * rules to the hardware to simplify the rollback of rules in the
	 * hardware.
	 */
	rss_conf = (const struct hns3_rss_conf *)act->conf;
	ret = hns3_flow_parse_rss(dev, rss_conf, true);
	if (ret != 0) {
		rte_free(rss_filter_ptr);
		return ret;
	}

	hns3_rss_conf_copy(&rss_filter_ptr->filter_info, &rss_conf->conf);
	rss_filter_ptr->filter_info.valid = true;
	TAILQ_INSERT_TAIL(&hw->flow_rss_list, rss_filter_ptr, entries);
	flow->rule = rss_filter_ptr;
	flow->filter_type = RTE_ETH_FILTER_HASH;

	return 0;
}

static int
hns3_flow_create_fdir_rule(struct rte_eth_dev *dev,
			   const struct rte_flow_item pattern[],
			   const struct rte_flow_action actions[],
			   struct rte_flow_error *error,
			   struct rte_flow *flow)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct hns3_adapter *hns = HNS3_DEV_HW_TO_ADAPTER(hw);
	struct hns3_fdir_rule_ele *fdir_rule_ptr;
	struct hns3_fdir_rule fdir_rule;
	bool indir;
	int ret;

	memset(&fdir_rule, 0, sizeof(struct hns3_fdir_rule));
	ret = hns3_parse_fdir_filter(dev, pattern, actions, &fdir_rule, error);
	if (ret != 0)
		return ret;

	indir = !!(fdir_rule.flags & HNS3_RULE_FLAG_COUNTER_INDIR);
	if (fdir_rule.flags & HNS3_RULE_FLAG_COUNTER) {
		ret = hns3_counter_new(dev, indir, fdir_rule.act_cnt.id,
				       error);
		if (ret != 0)
			return ret;

		flow->counter_id = fdir_rule.act_cnt.id;
	}

	fdir_rule_ptr = rte_zmalloc("hns3 fdir rule",
				    sizeof(struct hns3_fdir_rule_ele), 0);
	if (fdir_rule_ptr == NULL) {
		hns3_err(hw, "failed to allocate fdir_rule memory.");
		ret = -ENOMEM;
		goto err_malloc;
	}

	/*
	 * After all the preceding tasks are successfully configured, configure
	 * rules to the hardware to simplify the rollback of rules in the
	 * hardware.
	 */
	ret = hns3_fdir_filter_program(hns, &fdir_rule, false);
	if (ret != 0)
		goto err_fdir_filter;

	memcpy(&fdir_rule_ptr->fdir_conf, &fdir_rule,
		sizeof(struct hns3_fdir_rule));
	TAILQ_INSERT_TAIL(&hw->flow_fdir_list, fdir_rule_ptr, entries);
	flow->rule = fdir_rule_ptr;
	flow->filter_type = RTE_ETH_FILTER_FDIR;

	return 0;

err_fdir_filter:
	rte_free(fdir_rule_ptr);
err_malloc:
	if (fdir_rule.flags & HNS3_RULE_FLAG_COUNTER)
		hns3_counter_release(dev, fdir_rule.act_cnt.id);

	return ret;
}

/*
 * Create or destroy a flow rule.
 * Theorically one rule can match more than one filters.
 * We will let it use the filter which it hit first.
 * So, the sequence matters.
 */
static struct rte_flow *
hns3_flow_create(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
		 const struct rte_flow_item pattern[],
		 const struct rte_flow_action actions[],
		 struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_hw *hw = &hns->hw;
	struct hns3_flow_mem *flow_node;
	const struct rte_flow_action *act;
	struct rte_flow *flow;
	int ret;

	ret = hns3_flow_validate(dev, attr, pattern, actions, error);
	if (ret)
		return NULL;

	flow = rte_zmalloc("hns3 flow", sizeof(struct rte_flow), 0);
	if (flow == NULL) {
		rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
				   NULL, "Failed to allocate flow memory");
		return NULL;
	}
	flow_node = rte_zmalloc("hns3 flow node",
				sizeof(struct hns3_flow_mem), 0);
	if (flow_node == NULL) {
		rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
				   NULL, "Failed to allocate flow list memory");
		rte_free(flow);
		return NULL;
	}

	flow_node->flow = flow;
	TAILQ_INSERT_TAIL(&hw->flow_list, flow_node, entries);

	act = hns3_find_rss_general_action(pattern, actions);
	if (act)
		ret = hns3_flow_create_rss_rule(dev, act, flow);
	else
		ret = hns3_flow_create_fdir_rule(dev, pattern, actions,
						 error, flow);
	if (ret == 0)
		return flow;

	rte_flow_error_set(error, -ret, RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
			   "Failed to create flow");
	TAILQ_REMOVE(&hw->flow_list, flow_node, entries);
	rte_free(flow_node);
	rte_free(flow);

	return NULL;
}

/* Destroy a flow rule on hns3. */
static int
hns3_flow_destroy(struct rte_eth_dev *dev, struct rte_flow *flow,
		  struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	struct hns3_fdir_rule_ele *fdir_rule_ptr;
	struct hns3_rss_conf_ele *rss_filter_ptr;
	struct hns3_flow_mem *flow_node;
	enum rte_filter_type filter_type;
	struct hns3_fdir_rule fdir_rule;
	struct hns3_hw *hw = &hns->hw;
	int ret;

	if (flow == NULL)
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_HANDLE,
					  flow, "Flow is NULL");

	filter_type = flow->filter_type;
	switch (filter_type) {
	case RTE_ETH_FILTER_FDIR:
		fdir_rule_ptr = (struct hns3_fdir_rule_ele *)flow->rule;
		memcpy(&fdir_rule, &fdir_rule_ptr->fdir_conf,
			   sizeof(struct hns3_fdir_rule));

		ret = hns3_fdir_filter_program(hns, &fdir_rule, true);
		if (ret)
			return rte_flow_error_set(error, EIO,
						  RTE_FLOW_ERROR_TYPE_HANDLE,
						  flow,
						  "Destroy FDIR fail.Try again");
		if (fdir_rule.flags & HNS3_RULE_FLAG_COUNTER)
			hns3_counter_release(dev, fdir_rule.act_cnt.id);
		TAILQ_REMOVE(&hw->flow_fdir_list, fdir_rule_ptr, entries);
		rte_free(fdir_rule_ptr);
		fdir_rule_ptr = NULL;
		break;
	case RTE_ETH_FILTER_HASH:
		rss_filter_ptr = (struct hns3_rss_conf_ele *)flow->rule;
		ret = hns3_config_rss_filter(dev, &rss_filter_ptr->filter_info,
					     false);
		if (ret)
			return rte_flow_error_set(error, EIO,
						  RTE_FLOW_ERROR_TYPE_HANDLE,
						  flow,
						  "Destroy RSS fail.Try again");
		TAILQ_REMOVE(&hw->flow_rss_list, rss_filter_ptr, entries);
		rte_free(rss_filter_ptr);
		rss_filter_ptr = NULL;
		break;
	default:
		return rte_flow_error_set(error, EINVAL,
					  RTE_FLOW_ERROR_TYPE_HANDLE, flow,
					  "Unsupported filter type");
	}

	TAILQ_FOREACH(flow_node, &hw->flow_list, entries) {
		if (flow_node->flow == flow) {
			TAILQ_REMOVE(&hw->flow_list, flow_node, entries);
			rte_free(flow_node);
			flow_node = NULL;
			break;
		}
	}
	rte_free(flow);

	return 0;
}

/*  Destroy all flow rules associated with a port on hns3. */
static int
hns3_flow_flush(struct rte_eth_dev *dev, struct rte_flow_error *error)
{
	struct hns3_adapter *hns = dev->data->dev_private;
	int ret;

	/* FDIR is available only in PF driver */
	if (!hns->is_vf) {
		ret = hns3_clear_all_fdir_filter(hns);
		if (ret) {
			rte_flow_error_set(error, ret,
					   RTE_FLOW_ERROR_TYPE_HANDLE,
					   NULL, "Failed to flush rule");
			return ret;
		}
		hns3_counter_flush(dev);
	}

	ret = hns3_clear_rss_filter(dev);
	if (ret) {
		rte_flow_error_set(error, ret, RTE_FLOW_ERROR_TYPE_HANDLE,
				   NULL, "Failed to flush rss filter");
		return ret;
	}

	hns3_filterlist_flush(dev);

	return 0;
}

/* Query an existing flow rule. */
static int
hns3_flow_query(struct rte_eth_dev *dev, struct rte_flow *flow,
		const struct rte_flow_action *actions, void *data,
		struct rte_flow_error *error)
{
	struct rte_flow_action_rss *rss_conf;
	struct hns3_rss_conf_ele *rss_rule;
	struct rte_flow_query_count *qc;
	int ret;

	if (!flow->rule)
		return rte_flow_error_set(error, EINVAL,
			RTE_FLOW_ERROR_TYPE_HANDLE, NULL, "invalid rule");

	for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) {
		switch (actions->type) {
		case RTE_FLOW_ACTION_TYPE_VOID:
			break;
		case RTE_FLOW_ACTION_TYPE_COUNT:
			qc = (struct rte_flow_query_count *)data;
			ret = hns3_counter_query(dev, flow, qc, error);
			if (ret)
				return ret;
			break;
		case RTE_FLOW_ACTION_TYPE_RSS:
			if (flow->filter_type != RTE_ETH_FILTER_HASH) {
				return rte_flow_error_set(error, ENOTSUP,
					RTE_FLOW_ERROR_TYPE_ACTION,
					actions, "action is not supported");
			}
			rss_conf = (struct rte_flow_action_rss *)data;
			rss_rule = (struct hns3_rss_conf_ele *)flow->rule;
			rte_memcpy(rss_conf, &rss_rule->filter_info.conf,
				   sizeof(struct rte_flow_action_rss));
			break;
		default:
			return rte_flow_error_set(error, ENOTSUP,
				RTE_FLOW_ERROR_TYPE_ACTION,
				actions, "action is not supported");
		}
	}

	return 0;
}

static int
hns3_flow_validate_wrap(struct rte_eth_dev *dev,
			const struct rte_flow_attr *attr,
			const struct rte_flow_item pattern[],
			const struct rte_flow_action actions[],
			struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	int ret;

	pthread_mutex_lock(&hw->flows_lock);
	ret = hns3_flow_validate(dev, attr, pattern, actions, error);
	pthread_mutex_unlock(&hw->flows_lock);

	return ret;
}

static struct rte_flow *
hns3_flow_create_wrap(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
		      const struct rte_flow_item pattern[],
		      const struct rte_flow_action actions[],
		      struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct rte_flow *flow;

	pthread_mutex_lock(&hw->flows_lock);
	flow = hns3_flow_create(dev, attr, pattern, actions, error);
	pthread_mutex_unlock(&hw->flows_lock);

	return flow;
}

static int
hns3_flow_destroy_wrap(struct rte_eth_dev *dev, struct rte_flow *flow,
		       struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	int ret;

	pthread_mutex_lock(&hw->flows_lock);
	ret = hns3_flow_destroy(dev, flow, error);
	pthread_mutex_unlock(&hw->flows_lock);

	return ret;
}

static int
hns3_flow_flush_wrap(struct rte_eth_dev *dev, struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	int ret;

	pthread_mutex_lock(&hw->flows_lock);
	ret = hns3_flow_flush(dev, error);
	pthread_mutex_unlock(&hw->flows_lock);

	return ret;
}

static int
hns3_flow_query_wrap(struct rte_eth_dev *dev, struct rte_flow *flow,
		     const struct rte_flow_action *actions, void *data,
		     struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	int ret;

	pthread_mutex_lock(&hw->flows_lock);
	ret = hns3_flow_query(dev, flow, actions, data, error);
	pthread_mutex_unlock(&hw->flows_lock);

	return ret;
}

static int
hns3_check_indir_action(const struct rte_flow_indir_action_conf *conf,
			const struct rte_flow_action *action,
			struct rte_flow_error *error)
{
	if (!conf->ingress)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION,
				NULL, "Indir action ingress can't be zero");

	if (conf->egress)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION,
				NULL, "Indir action not support egress");

	if (conf->transfer)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION,
				NULL, "Indir action not support transfer");

	if (action->type != RTE_FLOW_ACTION_TYPE_COUNT)
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION,
				NULL, "Indir action only support count");

	return 0;
}

static struct rte_flow_action_handle *
hns3_flow_action_create(struct rte_eth_dev *dev,
			const struct rte_flow_indir_action_conf *conf,
			const struct rte_flow_action *action,
			struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct hns3_pf *pf = HNS3_DEV_PRIVATE_TO_PF(dev->data->dev_private);
	const struct rte_flow_action_count *act_count;
	struct rte_flow_action_handle *handle = NULL;
	struct hns3_flow_counter *counter;

	if (hns3_check_indir_action(conf, action, error))
		return NULL;

	handle = rte_zmalloc("hns3 action handle",
			     sizeof(struct rte_flow_action_handle), 0);
	if (handle == NULL) {
		rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
				   NULL, "Failed to allocate action memory");
		return NULL;
	}

	pthread_mutex_lock(&hw->flows_lock);

	act_count = (const struct rte_flow_action_count *)action->conf;
	if (act_count->id >= pf->fdir.fd_cfg.cnt_num[HNS3_FD_STAGE_1]) {
		rte_flow_error_set(error, EINVAL,
				   RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				   action, "Invalid counter id");
		goto err_exit;
	}

	if (hns3_counter_new(dev, false, act_count->id, error))
		goto err_exit;

	counter = hns3_counter_lookup(dev, act_count->id);
	if (counter == NULL) {
		rte_flow_error_set(error, EINVAL,
				   RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				   action, "Counter id not found");
		goto err_exit;
	}

	counter->indirect = true;
	handle->indirect_type = HNS3_INDIRECT_ACTION_TYPE_COUNT;
	handle->counter_id = counter->id;

	pthread_mutex_unlock(&hw->flows_lock);
	return handle;

err_exit:
	pthread_mutex_unlock(&hw->flows_lock);
	rte_free(handle);
	return NULL;
}

static int
hns3_flow_action_destroy(struct rte_eth_dev *dev,
			 struct rte_flow_action_handle *handle,
			 struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct hns3_flow_counter *counter;

	pthread_mutex_lock(&hw->flows_lock);

	if (handle->indirect_type != HNS3_INDIRECT_ACTION_TYPE_COUNT) {
		pthread_mutex_unlock(&hw->flows_lock);
		return rte_flow_error_set(error, EINVAL,
					RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					handle, "Invalid indirect type");
	}

	counter = hns3_counter_lookup(dev, handle->counter_id);
	if (counter == NULL) {
		pthread_mutex_unlock(&hw->flows_lock);
		return rte_flow_error_set(error, EINVAL,
				RTE_FLOW_ERROR_TYPE_ACTION_CONF,
				handle, "Counter id not exist");
	}

	if (counter->ref_cnt > 1) {
		pthread_mutex_unlock(&hw->flows_lock);
		return rte_flow_error_set(error, EBUSY,
				RTE_FLOW_ERROR_TYPE_HANDLE,
				handle, "Counter id in use");
	}

	(void)hns3_counter_release(dev, handle->counter_id);
	rte_free(handle);

	pthread_mutex_unlock(&hw->flows_lock);
	return 0;
}

static int
hns3_flow_action_query(struct rte_eth_dev *dev,
		 const struct rte_flow_action_handle *handle,
		 void *data,
		 struct rte_flow_error *error)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	struct rte_flow flow;
	int ret;

	pthread_mutex_lock(&hw->flows_lock);

	if (handle->indirect_type != HNS3_INDIRECT_ACTION_TYPE_COUNT) {
		pthread_mutex_unlock(&hw->flows_lock);
		return rte_flow_error_set(error, EINVAL,
					RTE_FLOW_ERROR_TYPE_ACTION_CONF,
					handle, "Invalid indirect type");
	}

	memset(&flow, 0, sizeof(flow));
	flow.counter_id = handle->counter_id;
	ret = hns3_counter_query(dev, &flow,
				 (struct rte_flow_query_count *)data, error);
	pthread_mutex_unlock(&hw->flows_lock);
	return ret;
}

static const struct rte_flow_ops hns3_flow_ops = {
	.validate = hns3_flow_validate_wrap,
	.create = hns3_flow_create_wrap,
	.destroy = hns3_flow_destroy_wrap,
	.flush = hns3_flow_flush_wrap,
	.query = hns3_flow_query_wrap,
	.isolate = NULL,
	.action_handle_create = hns3_flow_action_create,
	.action_handle_destroy = hns3_flow_action_destroy,
	.action_handle_query = hns3_flow_action_query,
};

int
hns3_dev_flow_ops_get(struct rte_eth_dev *dev,
		      const struct rte_flow_ops **ops)
{
	struct hns3_hw *hw;

	hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	if (hw->adapter_state >= HNS3_NIC_CLOSED)
		return -ENODEV;

	*ops = &hns3_flow_ops;
	return 0;
}

void
hns3_flow_init(struct rte_eth_dev *dev)
{
	struct hns3_hw *hw = HNS3_DEV_PRIVATE_TO_HW(dev->data->dev_private);
	pthread_mutexattr_t attr;

	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
		return;

	pthread_mutexattr_init(&attr);
	pthread_mutexattr_setpshared(&attr, PTHREAD_PROCESS_SHARED);
	pthread_mutex_init(&hw->flows_lock, &attr);
	dev->data->dev_flags |= RTE_ETH_DEV_FLOW_OPS_THREAD_SAFE;

	TAILQ_INIT(&hw->flow_fdir_list);
	TAILQ_INIT(&hw->flow_rss_list);
	TAILQ_INIT(&hw->flow_list);
}

void
hns3_flow_uninit(struct rte_eth_dev *dev)
{
	struct rte_flow_error error;
	if (rte_eal_process_type() == RTE_PROC_PRIMARY)
		hns3_flow_flush_wrap(dev, &error);
}