OSDN Git Service

ice: Add tc-flower filter support for channel
authorKiran Patil <kiran.patil@intel.com>
Fri, 15 Oct 2021 23:35:17 +0000 (16:35 -0700)
committerTony Nguyen <anthony.l.nguyen@intel.com>
Wed, 20 Oct 2021 22:58:11 +0000 (15:58 -0700)
Add support to add/delete channel specific filter using tc-flower.
For now, only supported action is "skip_sw hw_tc <tc_num>"

Filter criteria is specific to channel and it can be
combination of L3, L3+L4, L2+L4.

Example:
MATCH criteria       Action
---------------------------
src and/or dest IPv4[6]/mask -> Forward to "hw_tc <tc_num>"
dest IPv4[6]/mask + dest L4 port -> Forward to "hw_tc <tc_num>"
dest MAC + dest L4 port -> Forward to "hw_tc <tc_num>"
src IPv4[6]/mask + src L4 port -> Forward to "hw_tc <tc_num>"
src MAC + src L4 port -> Forward to "hw_tc <tc_num>"

Adding tc-flower filter for channel using "hw_tc"
-------------------------------------------------
tc qdisc add dev <ethX> clsact

Above two steps are only needed the first time when adding
tc-flower filter.

tc filter add dev <ethX> protocol ip ingress prio 1 flower \
     dst_ip 192.168.0.1/32 ip_proto tcp dst_port 5001 \
     skip_sw hw_tc 1

tc filter show dev <ethX> ingress
filter protocol ip pref 1 flower chain 0
filter protocol ip pref 1 flower chain 0 handle 0x1 hw_tc 1
  eth_type ipv4
  ip_proto tcp
  dst_ip 192.168.0.1
  dst_port 5001
  skip_sw
  in_hw in_hw_count 1

Delete specific filter:
-------------------------
tc filter del  dev <ethx> ingress pref 1 handle 0x1 flower

Delete All filters:
------------------
tc filter del dev <ethX> ingress

Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Kiran Patil <kiran.patil@intel.com>
Signed-off-by: Sudheer Mogilappagari <sudheer.mogilappagari@intel.com>
Tested-by: Bharathi Sreenivas <bharathi.sreenivas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
drivers/net/ethernet/intel/ice/ice.h
drivers/net/ethernet/intel/ice/ice_main.c
drivers/net/ethernet/intel/ice/ice_switch.c
drivers/net/ethernet/intel/ice/ice_switch.h
drivers/net/ethernet/intel/ice/ice_tc_lib.c
drivers/net/ethernet/intel/ice/ice_tc_lib.h

index 4f94ecf..967a90e 100644 (file)
@@ -39,6 +39,9 @@
 #include <linux/cpu_rmap.h>
 #include <linux/dim.h>
 #include <net/pkt_cls.h>
+#include <net/tc_act/tc_mirred.h>
+#include <net/tc_act/tc_gact.h>
+#include <net/ip.h>
 #include <net/devlink.h>
 #include <net/ipv6.h>
 #include <net/xdp_sock.h>
@@ -402,6 +405,7 @@ struct ice_vsi {
        u16 num_chnl_rxq;
        u16 num_chnl_txq;
        u16 ch_rss_size;
+       u16 num_chnl_fltr;
        /* store away rss size info before configuring ADQ channels so that,
         * it can be used after tc-qdisc delete, to get back RSS setting as
         * they were before
@@ -581,7 +585,10 @@ struct ice_pf {
        struct auxiliary_device *adev;
        int aux_idx;
        u32 sw_int_count;
-
+       /* count of tc_flower filters specific to channel (aka where filter
+        * action is "hw_tc <tc_num>")
+        */
+       u16 num_dmac_chnl_fltrs;
        struct hlist_head tc_flower_fltr_list;
 
        __le64 nvm_phy_type_lo; /* NVM PHY type low */
index ff2d000..cb82abd 100644 (file)
@@ -5255,6 +5255,12 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi)
                return -EBUSY;
        }
 
+       if (ice_chnl_dmac_fltr_cnt(pf)) {
+               netdev_err(netdev, "can't set mac %pM. Device has tc-flower filters, delete all of them and try again\n",
+                          mac);
+               return -EAGAIN;
+       }
+
        netif_addr_lock_bh(netdev);
        ether_addr_copy(old_mac, netdev->dev_addr);
        /* change the netdev's MAC address */
@@ -5511,6 +5517,13 @@ ice_set_features(struct net_device *netdev, netdev_features_t features)
                dev_err(ice_pf_to_dev(pf), "ADQ is active, can't turn hw_tc_offload off\n");
                return -EACCES;
        }
+
+       if ((features & NETIF_F_HW_TC) &&
+           !(netdev->features & NETIF_F_HW_TC))
+               set_bit(ICE_FLAG_CLS_FLOWER, pf->flags);
+       else
+               clear_bit(ICE_FLAG_CLS_FLOWER, pf->flags);
+
        return ret;
 }
 
@@ -7650,17 +7663,73 @@ static int ice_create_q_channel(struct ice_vsi *vsi, struct ice_channel *ch)
 }
 
 /**
+ * ice_rem_all_chnl_fltrs - removes all channel filters
+ * @pf: ptr to PF, TC-flower based filter are tracked at PF level
+ *
+ * Remove all advanced switch filters only if they are channel specific
+ * tc-flower based filter
+ */
+static void ice_rem_all_chnl_fltrs(struct ice_pf *pf)
+{
+       struct ice_tc_flower_fltr *fltr;
+       struct hlist_node *node;
+
+       /* to remove all channel filters, iterate an ordered list of filters */
+       hlist_for_each_entry_safe(fltr, node,
+                                 &pf->tc_flower_fltr_list,
+                                 tc_flower_node) {
+               struct ice_rule_query_data rule;
+               int status;
+
+               /* for now process only channel specific filters */
+               if (!ice_is_chnl_fltr(fltr))
+                       continue;
+
+               rule.rid = fltr->rid;
+               rule.rule_id = fltr->rule_id;
+               rule.vsi_handle = fltr->dest_id;
+               status = ice_rem_adv_rule_by_id(&pf->hw, &rule);
+               if (status) {
+                       if (status == -ENOENT)
+                               dev_dbg(ice_pf_to_dev(pf), "TC flower filter (rule_id %u) does not exist\n",
+                                       rule.rule_id);
+                       else
+                               dev_err(ice_pf_to_dev(pf), "failed to delete TC flower filter, status %d\n",
+                                       status);
+               } else if (fltr->dest_vsi) {
+                       /* update advanced switch filter count */
+                       if (fltr->dest_vsi->type == ICE_VSI_CHNL) {
+                               u32 flags = fltr->flags;
+
+                               fltr->dest_vsi->num_chnl_fltr--;
+                               if (flags & (ICE_TC_FLWR_FIELD_DST_MAC |
+                                            ICE_TC_FLWR_FIELD_ENC_DST_MAC))
+                                       pf->num_dmac_chnl_fltrs--;
+                       }
+               }
+
+               hlist_del(&fltr->tc_flower_node);
+               kfree(fltr);
+       }
+}
+
+/**
  * ice_remove_q_channels - Remove queue channels for the TCs
  * @vsi: VSI to be configured
  * @rem_fltr: delete advanced switch filter or not
  *
  * Remove queue channels for the TCs
  */
-static void ice_remove_q_channels(struct ice_vsi *vsi, bool __maybe_unused rem_fltr)
+static void ice_remove_q_channels(struct ice_vsi *vsi, bool rem_fltr)
 {
        struct ice_channel *ch, *ch_tmp;
+       struct ice_pf *pf = vsi->back;
        int i;
 
+       /* remove all tc-flower based filter if they are channel filters only */
+       if (rem_fltr)
+               ice_rem_all_chnl_fltrs(pf);
+
        /* perform cleanup for channels if they exist */
        list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) {
                struct ice_vsi *ch_vsi;
@@ -7926,6 +7995,12 @@ static int ice_setup_tc_mqprio_qdisc(struct net_device *netdev, void *type_data)
                }
                memcpy(&vsi->mqprio_qopt, mqprio_qopt, sizeof(*mqprio_qopt));
                set_bit(ICE_FLAG_TC_MQPRIO, pf->flags);
+               /* don't assume state of hw_tc_offload during driver load
+                * and set the flag for TC flower filter if hw_tc_offload
+                * already ON
+                */
+               if (vsi->netdev->features & NETIF_F_HW_TC)
+                       set_bit(ICE_FLAG_CLS_FLOWER, pf->flags);
                break;
        default:
                return -EINVAL;
index a4a2990..2742e1c 100644 (file)
@@ -2273,6 +2273,125 @@ exit:
 }
 
 /**
+ * ice_mac_fltr_exist - does this MAC filter exist for given VSI
+ * @hw: pointer to the hardware structure
+ * @mac: MAC address to be checked (for MAC filter)
+ * @vsi_handle: check MAC filter for this VSI
+ */
+bool ice_mac_fltr_exist(struct ice_hw *hw, u8 *mac, u16 vsi_handle)
+{
+       struct ice_fltr_mgmt_list_entry *entry;
+       struct list_head *rule_head;
+       struct ice_switch_info *sw;
+       struct mutex *rule_lock; /* Lock to protect filter rule list */
+       u16 hw_vsi_id;
+
+       if (!ice_is_vsi_valid(hw, vsi_handle))
+               return false;
+
+       hw_vsi_id = ice_get_hw_vsi_num(hw, vsi_handle);
+       sw = hw->switch_info;
+       rule_head = &sw->recp_list[ICE_SW_LKUP_MAC].filt_rules;
+       if (!rule_head)
+               return false;
+
+       rule_lock = &sw->recp_list[ICE_SW_LKUP_MAC].filt_rule_lock;
+       mutex_lock(rule_lock);
+       list_for_each_entry(entry, rule_head, list_entry) {
+               struct ice_fltr_info *f_info = &entry->fltr_info;
+               u8 *mac_addr = &f_info->l_data.mac.mac_addr[0];
+
+               if (is_zero_ether_addr(mac_addr))
+                       continue;
+
+               if (f_info->flag != ICE_FLTR_TX ||
+                   f_info->src_id != ICE_SRC_ID_VSI ||
+                   f_info->lkup_type != ICE_SW_LKUP_MAC ||
+                   f_info->fltr_act != ICE_FWD_TO_VSI ||
+                   hw_vsi_id != f_info->fwd_id.hw_vsi_id)
+                       continue;
+
+               if (ether_addr_equal(mac, mac_addr)) {
+                       mutex_unlock(rule_lock);
+                       return true;
+               }
+       }
+       mutex_unlock(rule_lock);
+       return false;
+}
+
+/**
+ * ice_vlan_fltr_exist - does this VLAN filter exist for given VSI
+ * @hw: pointer to the hardware structure
+ * @vlan_id: VLAN ID
+ * @vsi_handle: check MAC filter for this VSI
+ */
+bool ice_vlan_fltr_exist(struct ice_hw *hw, u16 vlan_id, u16 vsi_handle)
+{
+       struct ice_fltr_mgmt_list_entry *entry;
+       struct list_head *rule_head;
+       struct ice_switch_info *sw;
+       struct mutex *rule_lock; /* Lock to protect filter rule list */
+       u16 hw_vsi_id;
+
+       if (vlan_id > ICE_MAX_VLAN_ID)
+               return false;
+
+       if (!ice_is_vsi_valid(hw, vsi_handle))
+               return false;
+
+       hw_vsi_id = ice_get_hw_vsi_num(hw, vsi_handle);
+       sw = hw->switch_info;
+       rule_head = &sw->recp_list[ICE_SW_LKUP_VLAN].filt_rules;
+       if (!rule_head)
+               return false;
+
+       rule_lock = &sw->recp_list[ICE_SW_LKUP_VLAN].filt_rule_lock;
+       mutex_lock(rule_lock);
+       list_for_each_entry(entry, rule_head, list_entry) {
+               struct ice_fltr_info *f_info = &entry->fltr_info;
+               u16 entry_vlan_id = f_info->l_data.vlan.vlan_id;
+               struct ice_vsi_list_map_info *map_info;
+
+               if (entry_vlan_id > ICE_MAX_VLAN_ID)
+                       continue;
+
+               if (f_info->flag != ICE_FLTR_TX ||
+                   f_info->src_id != ICE_SRC_ID_VSI ||
+                   f_info->lkup_type != ICE_SW_LKUP_VLAN)
+                       continue;
+
+               /* Only allowed filter action are FWD_TO_VSI/_VSI_LIST */
+               if (f_info->fltr_act != ICE_FWD_TO_VSI &&
+                   f_info->fltr_act != ICE_FWD_TO_VSI_LIST)
+                       continue;
+
+               if (f_info->fltr_act == ICE_FWD_TO_VSI) {
+                       if (hw_vsi_id != f_info->fwd_id.hw_vsi_id)
+                               continue;
+               } else if (f_info->fltr_act == ICE_FWD_TO_VSI_LIST) {
+                       /* If filter_action is FWD_TO_VSI_LIST, make sure
+                        * that VSI being checked is part of VSI list
+                        */
+                       if (entry->vsi_count == 1 &&
+                           entry->vsi_list_info) {
+                               map_info = entry->vsi_list_info;
+                               if (!test_bit(vsi_handle, map_info->vsi_map))
+                                       continue;
+                       }
+               }
+
+               if (vlan_id == entry_vlan_id) {
+                       mutex_unlock(rule_lock);
+                       return true;
+               }
+       }
+       mutex_unlock(rule_lock);
+
+       return false;
+}
+
+/**
  * ice_add_mac - Add a MAC address based filter rule
  * @hw: pointer to the hardware structure
  * @m_list: list of MAC addresses and forwarding information
index d4c0a3b..c4dd206 100644 (file)
@@ -335,6 +335,8 @@ enum ice_status
 ice_remove_eth_mac(struct ice_hw *hw, struct list_head *em_list);
 int
 ice_cfg_rdma_fltr(struct ice_hw *hw, u16 vsi_handle, bool enable);
+bool ice_mac_fltr_exist(struct ice_hw *hw, u8 *mac, u16 vsi_handle);
+bool ice_vlan_fltr_exist(struct ice_hw *hw, u16 vlan_id, u16 vsi_handle);
 void ice_remove_vsi_fltr(struct ice_hw *hw, u16 vsi_handle);
 enum ice_status
 ice_add_vlan(struct ice_hw *hw, struct list_head *m_list);
index 1dccfd1..725caa1 100644 (file)
@@ -304,6 +304,136 @@ exit:
 }
 
 /**
+ * ice_add_tc_flower_adv_fltr - add appropriate filter rules
+ * @vsi: Pointer to VSI
+ * @tc_fltr: Pointer to TC flower filter structure
+ *
+ * based on filter parameters using Advance recipes supported
+ * by OS package.
+ */
+static int
+ice_add_tc_flower_adv_fltr(struct ice_vsi *vsi,
+                          struct ice_tc_flower_fltr *tc_fltr)
+{
+       struct ice_tc_flower_lyr_2_4_hdrs *headers = &tc_fltr->outer_headers;
+       struct ice_adv_rule_info rule_info = {0};
+       struct ice_rule_query_data rule_added;
+       struct ice_adv_lkup_elem *list;
+       struct ice_pf *pf = vsi->back;
+       struct ice_hw *hw = &pf->hw;
+       u32 flags = tc_fltr->flags;
+       struct ice_vsi *ch_vsi;
+       struct device *dev;
+       u16 lkups_cnt = 0;
+       u16 l4_proto = 0;
+       int ret = 0;
+       u16 i = 0;
+
+       dev = ice_pf_to_dev(pf);
+       if (ice_is_safe_mode(pf)) {
+               NL_SET_ERR_MSG_MOD(tc_fltr->extack, "Unable to add filter because driver is in safe mode");
+               return -EOPNOTSUPP;
+       }
+
+       if (!flags || (flags & (ICE_TC_FLWR_FIELD_ENC_DEST_IPV4 |
+                               ICE_TC_FLWR_FIELD_ENC_SRC_IPV4 |
+                               ICE_TC_FLWR_FIELD_ENC_DEST_IPV6 |
+                               ICE_TC_FLWR_FIELD_ENC_SRC_IPV6 |
+                               ICE_TC_FLWR_FIELD_ENC_SRC_L4_PORT))) {
+               NL_SET_ERR_MSG_MOD(tc_fltr->extack, "Unsupported encap field(s)");
+               return -EOPNOTSUPP;
+       }
+
+       /* get the channel (aka ADQ VSI) */
+       if (tc_fltr->dest_vsi)
+               ch_vsi = tc_fltr->dest_vsi;
+       else
+               ch_vsi = vsi->tc_map_vsi[tc_fltr->action.tc_class];
+
+       lkups_cnt = ice_tc_count_lkups(flags, headers, tc_fltr);
+       list = kcalloc(lkups_cnt, sizeof(*list), GFP_ATOMIC);
+       if (!list)
+               return -ENOMEM;
+
+       i = ice_tc_fill_rules(hw, flags, tc_fltr, list, &rule_info, &l4_proto);
+       if (i != lkups_cnt) {
+               ret = -EINVAL;
+               goto exit;
+       }
+
+       rule_info.sw_act.fltr_act = tc_fltr->action.fltr_act;
+       if (tc_fltr->action.tc_class >= ICE_CHNL_START_TC) {
+               if (!ch_vsi) {
+                       NL_SET_ERR_MSG_MOD(tc_fltr->extack, "Unable to add filter because specified destination doesn't exist");
+                       ret = -EINVAL;
+                       goto exit;
+               }
+
+               rule_info.sw_act.fltr_act = ICE_FWD_TO_VSI;
+               rule_info.sw_act.vsi_handle = ch_vsi->idx;
+               rule_info.priority = 7;
+               rule_info.sw_act.src = hw->pf_id;
+               rule_info.rx = true;
+               dev_dbg(dev, "add switch rule for TC:%u vsi_idx:%u, lkups_cnt:%u\n",
+                       tc_fltr->action.tc_class,
+                       rule_info.sw_act.vsi_handle, lkups_cnt);
+       } else {
+               rule_info.sw_act.flag |= ICE_FLTR_TX;
+               rule_info.sw_act.src = vsi->idx;
+               rule_info.rx = false;
+       }
+
+       /* specify the cookie as filter_rule_id */
+       rule_info.fltr_rule_id = tc_fltr->cookie;
+
+       ret = ice_add_adv_rule(hw, list, lkups_cnt, &rule_info, &rule_added);
+       if (ret == -EEXIST) {
+               NL_SET_ERR_MSG_MOD(tc_fltr->extack,
+                                  "Unable to add filter because it already exist");
+               ret = -EINVAL;
+               goto exit;
+       } else if (ret) {
+               NL_SET_ERR_MSG_MOD(tc_fltr->extack,
+                                  "Unable to add filter due to error");
+               ret = -EIO;
+               goto exit;
+       }
+
+       /* store the output params, which are needed later for removing
+        * advanced switch filter
+        */
+       tc_fltr->rid = rule_added.rid;
+       tc_fltr->rule_id = rule_added.rule_id;
+       if (tc_fltr->action.tc_class > 0 && ch_vsi) {
+               /* For PF ADQ, VSI type is set as ICE_VSI_CHNL, and
+                * for PF ADQ filter, it is not yet set in tc_fltr,
+                * hence store the dest_vsi ptr in tc_fltr
+                */
+               if (ch_vsi->type == ICE_VSI_CHNL)
+                       tc_fltr->dest_vsi = ch_vsi;
+               /* keep track of advanced switch filter for
+                * destination VSI (channel VSI)
+                */
+               ch_vsi->num_chnl_fltr++;
+               /* in this case, dest_id is VSI handle (sw handle) */
+               tc_fltr->dest_id = rule_added.vsi_handle;
+
+               /* keeps track of channel filters for PF VSI */
+               if (vsi->type == ICE_VSI_PF &&
+                   (flags & (ICE_TC_FLWR_FIELD_DST_MAC |
+                             ICE_TC_FLWR_FIELD_ENC_DST_MAC)))
+                       pf->num_dmac_chnl_fltrs++;
+       }
+       dev_dbg(dev, "added switch rule (lkups_cnt %u, flags 0x%x) for TC %u, rid %u, rule_id %u, vsi_idx %u\n",
+               lkups_cnt, flags,
+               tc_fltr->action.tc_class, rule_added.rid,
+               rule_added.rule_id, rule_added.vsi_handle);
+exit:
+       kfree(list);
+       return ret;
+}
+
+/**
  * ice_tc_set_ipv4 - Parse IPv4 addresses from TC flower filter
  * @match: Pointer to flow match structure
  * @fltr: Pointer to filter structure
@@ -561,10 +691,13 @@ ice_parse_cls_flower(struct net_device *filter_dev, struct ice_vsi *vsi,
 static int
 ice_add_switch_fltr(struct ice_vsi *vsi, struct ice_tc_flower_fltr *fltr)
 {
+       if (fltr->action.fltr_act == ICE_FWD_TO_QGRP)
+               return -EOPNOTSUPP;
+
        if (ice_is_eswitch_mode_switchdev(vsi->back))
                return ice_eswitch_add_tc_fltr(vsi, fltr);
 
-       return -EOPNOTSUPP;
+       return ice_add_tc_flower_adv_fltr(vsi, fltr);
 }
 
 /**
@@ -581,6 +714,7 @@ ice_handle_tclass_action(struct ice_vsi *vsi,
                         struct ice_tc_flower_fltr *fltr)
 {
        int tc = tc_classid_to_hwtc(vsi->netdev, cls_flower->classid);
+       struct ice_vsi *main_vsi;
 
        if (tc < 0) {
                NL_SET_ERR_MSG_MOD(fltr->extack, "Unable to add filter because specified destination is invalid");
@@ -591,13 +725,69 @@ ice_handle_tclass_action(struct ice_vsi *vsi,
                return -EINVAL;
        }
 
-       if (!(vsi->tc_cfg.ena_tc & BIT(tc))) {
+       if (!(vsi->all_enatc & BIT(tc))) {
                NL_SET_ERR_MSG_MOD(fltr->extack, "Unable to add filter because of non-existence destination");
                return -EINVAL;
        }
 
        /* Redirect to a TC class or Queue Group */
-       fltr->action.fltr_act = ICE_FWD_TO_QGRP;
+       main_vsi = ice_get_main_vsi(vsi->back);
+       if (!main_vsi || !main_vsi->netdev) {
+               NL_SET_ERR_MSG_MOD(fltr->extack,
+                                  "Unable to add filter because of invalid netdevice");
+               return -EINVAL;
+       }
+
+       if ((fltr->flags & ICE_TC_FLWR_FIELD_TENANT_ID) &&
+           (fltr->flags & (ICE_TC_FLWR_FIELD_DST_MAC |
+                          ICE_TC_FLWR_FIELD_SRC_MAC))) {
+               NL_SET_ERR_MSG_MOD(fltr->extack,
+                                  "Unable to add filter because filter using tunnel key and inner MAC is unsupported combination");
+               return -EOPNOTSUPP;
+       }
+
+       /* For ADQ, filter must include dest MAC address, otherwise unwanted
+        * packets with unrelated MAC address get delivered to ADQ VSIs as long
+        * as remaining filter criteria is satisfied such as dest IP address
+        * and dest/src L4 port. Following code is trying to handle:
+        * 1. For non-tunnel, if user specify MAC addresses, use them (means
+        * this code won't do anything
+        * 2. For non-tunnel, if user didn't specify MAC address, add implicit
+        * dest MAC to be lower netdev's active unicast MAC address
+        */
+       if (!(fltr->flags & ICE_TC_FLWR_FIELD_DST_MAC)) {
+               ether_addr_copy(fltr->outer_headers.l2_key.dst_mac,
+                               main_vsi->netdev->dev_addr);
+               eth_broadcast_addr(fltr->outer_headers.l2_mask.dst_mac);
+               fltr->flags |= ICE_TC_FLWR_FIELD_DST_MAC;
+       }
+
+       /* validate specified dest MAC address, make sure either it belongs to
+        * lower netdev or any of MACVLAN. MACVLANs MAC address are added as
+        * unicast MAC filter destined to main VSI.
+        */
+       if (!ice_mac_fltr_exist(&main_vsi->back->hw,
+                               fltr->outer_headers.l2_key.dst_mac,
+                               main_vsi->idx)) {
+               NL_SET_ERR_MSG_MOD(fltr->extack,
+                                  "Unable to add filter because legacy MAC filter for specified destination doesn't exist");
+               return -EINVAL;
+       }
+
+       /* Make sure VLAN is already added to main VSI, before allowing ADQ to
+        * add a VLAN based filter such as MAC + VLAN + L4 port.
+        */
+       if (fltr->flags & ICE_TC_FLWR_FIELD_VLAN) {
+               u16 vlan_id = be16_to_cpu(fltr->outer_headers.vlan_hdr.vlan_id);
+
+               if (!ice_vlan_fltr_exist(&main_vsi->back->hw, vlan_id,
+                                        main_vsi->idx)) {
+                       NL_SET_ERR_MSG_MOD(fltr->extack,
+                                          "Unable to add filter because legacy VLAN filter for specified destination doesn't exist");
+                       return -EINVAL;
+               }
+       }
+       fltr->action.fltr_act = ICE_FWD_TO_VSI;
        fltr->action.tc_class = tc;
 
        return 0;
@@ -639,8 +829,8 @@ ice_parse_tc_flower_actions(struct ice_vsi *vsi,
 
                /* Drop action */
                if (act->id == FLOW_ACTION_DROP) {
-                       fltr->action.fltr_act = ICE_DROP_PACKET;
-                       return 0;
+                       NL_SET_ERR_MSG_MOD(fltr->extack, "Unsupported action DROP");
+                       return -EINVAL;
                }
                fltr->action.fltr_act = ICE_FWD_TO_VSI;
        }
@@ -673,6 +863,20 @@ static int ice_del_tc_fltr(struct ice_vsi *vsi, struct ice_tc_flower_fltr *fltr)
                return -EIO;
        }
 
+       /* update advanced switch filter count for destination
+        * VSI if filter destination was VSI
+        */
+       if (fltr->dest_vsi) {
+               if (fltr->dest_vsi->type == ICE_VSI_CHNL) {
+                       fltr->dest_vsi->num_chnl_fltr--;
+
+                       /* keeps track of channel filters for PF VSI */
+                       if (vsi->type == ICE_VSI_PF &&
+                           (fltr->flags & (ICE_TC_FLWR_FIELD_DST_MAC |
+                                           ICE_TC_FLWR_FIELD_ENC_DST_MAC)))
+                               pf->num_dmac_chnl_fltrs--;
+               }
+       }
        return 0;
 }
 
@@ -811,7 +1015,8 @@ ice_del_cls_flower(struct ice_vsi *vsi, struct flow_cls_offload *cls_flower)
        /* find filter */
        fltr = ice_find_tc_flower_fltr(pf, cls_flower->cookie);
        if (!fltr) {
-               if (hlist_empty(&pf->tc_flower_fltr_list))
+               if (!test_bit(ICE_FLAG_TC_MQPRIO, pf->flags) &&
+                   hlist_empty(&pf->tc_flower_fltr_list))
                        return 0;
 
                NL_SET_ERR_MSG_MOD(cls_flower->common.extack, "failed to delete TC flower filter because unable to find it");
index d90e9e3..ee9b284 100644 (file)
@@ -120,6 +120,28 @@ struct ice_tc_flower_fltr {
        struct netlink_ext_ack *extack;
 };
 
+/**
+ * ice_is_chnl_fltr - is this a valid channel filter
+ * @f: Pointer to tc-flower filter
+ *
+ * Criteria to determine of given filter is valid channel filter
+ * or not is based on its "destination". If destination is hw_tc (aka tc_class)
+ * and it is non-zero, then it is valid channel (aka ADQ) filter
+ */
+static inline bool ice_is_chnl_fltr(struct ice_tc_flower_fltr *f)
+{
+       return !!f->action.tc_class;
+}
+
+/**
+ * ice_chnl_dmac_fltr_cnt - DMAC based CHNL filter count
+ * @pf: Pointer to PF
+ */
+static inline int ice_chnl_dmac_fltr_cnt(struct ice_pf *pf)
+{
+       return pf->num_dmac_chnl_fltrs;
+}
+
 int
 ice_add_cls_flower(struct net_device *netdev, struct ice_vsi *vsi,
                   struct flow_cls_offload *cls_flower);