cxgb4: add ethtool n-tuple filter deletion
authorVishal Kulkarni <vishal@chelsio.com>
Fri, 19 Jun 2020 14:21:37 +0000 (19:51 +0530)
committerDavid S. Miller <davem@davemloft.net>
Fri, 19 Jun 2020 20:17:31 +0000 (13:17 -0700)
Add support to delete ethtool n-tuple filter. Fetch the appropriate
filter region (HPFILTER, HASH, NORMAL) in which the filter exists,
and delete it from the respective region, accordingly.

Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Vishal Kulkarni <vishal@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c
drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h

index 82fc09b..5c58867 100644 (file)
@@ -1573,6 +1573,22 @@ static int set_rss_table(struct net_device *dev, const u32 *p, const u8 *key,
        return -EPERM;
 }
 
+static struct filter_entry *cxgb4_get_filter_entry(struct adapter *adap,
+                                                  u32 ftid)
+{
+       struct tid_info *t = &adap->tids;
+       struct filter_entry *f;
+
+       if (ftid < t->nhpftids)
+               f = &adap->tids.hpftid_tab[ftid];
+       else if (ftid < t->nftids)
+               f = &adap->tids.ftid_tab[ftid - t->nhpftids];
+       else
+               f = lookup_tid(&adap->tids, ftid);
+
+       return f;
+}
+
 static int get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
                     u32 *rules)
 {
@@ -1636,6 +1652,48 @@ static int get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
        return -EOPNOTSUPP;
 }
 
+static int cxgb4_ntuple_del_filter(struct net_device *dev,
+                                  struct ethtool_rxnfc *cmd)
+{
+       struct cxgb4_ethtool_filter_info *filter_info;
+       struct adapter *adapter = netdev2adap(dev);
+       struct port_info *pi = netdev_priv(dev);
+       struct filter_entry *f;
+       u32 filter_id;
+       int ret;
+
+       if (!(adapter->flags & CXGB4_FULL_INIT_DONE))
+               return -EAGAIN;  /* can still change nfilters */
+
+       if (!adapter->ethtool_filters)
+               return -EOPNOTSUPP;
+
+       if (cmd->fs.location >= adapter->ethtool_filters->nentries) {
+               dev_err(adapter->pdev_dev,
+                       "Location must be < %u",
+                       adapter->ethtool_filters->nentries);
+               return -ERANGE;
+       }
+
+       filter_info = &adapter->ethtool_filters->port[pi->port_id];
+
+       if (!test_bit(cmd->fs.location, filter_info->bmap))
+               return -ENOENT;
+
+       filter_id = filter_info->loc_array[cmd->fs.location];
+       f = cxgb4_get_filter_entry(adapter, filter_id);
+
+       ret = cxgb4_flow_rule_destroy(dev, f->fs.tc_prio, &f->fs, filter_id);
+       if (ret)
+               goto err;
+
+       clear_bit(cmd->fs.location, filter_info->bmap);
+       filter_info->in_use--;
+
+err:
+       return ret;
+}
+
 /* Add Ethtool n-tuple filters. */
 static int cxgb4_ntuple_set_filter(struct net_device *netdev,
                                   struct ethtool_rxnfc *cmd)
@@ -1702,6 +1760,9 @@ static int set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
        case ETHTOOL_SRXCLSRLINS:
                ret = cxgb4_ntuple_set_filter(dev, cmd);
                break;
+       case ETHTOOL_SRXCLSRLDEL:
+               ret = cxgb4_ntuple_del_filter(dev, cmd);
+               break;
        default:
                break;
        }
index 260d121..3ded4fe 100644 (file)
@@ -797,23 +797,38 @@ free_entry:
        return ret;
 }
 
+int cxgb4_flow_rule_destroy(struct net_device *dev, u32 tc_prio,
+                           struct ch_filter_specification *fs, int tid)
+{
+       struct adapter *adap = netdev2adap(dev);
+       u8 hash;
+       int ret;
+
+       hash = fs->hash;
+
+       ret = cxgb4_del_filter(dev, tid, fs);
+       if (ret)
+               return ret;
+
+       if (hash)
+               cxgb4_tc_flower_hash_prio_del(adap, tc_prio);
+
+       return ret;
+}
+
 int cxgb4_tc_flower_destroy(struct net_device *dev,
                            struct flow_cls_offload *cls)
 {
        struct adapter *adap = netdev2adap(dev);
        struct ch_tc_flower_entry *ch_flower;
-       u32 tc_prio;
-       bool hash;
        int ret;
 
        ch_flower = ch_flower_lookup(adap, cls->cookie);
        if (!ch_flower)
                return -ENOENT;
 
-       hash = ch_flower->fs.hash;
-       tc_prio = ch_flower->fs.tc_prio;
-
-       ret = cxgb4_del_filter(dev, ch_flower->filter_id, &ch_flower->fs);
+       ret = cxgb4_flow_rule_destroy(dev, ch_flower->fs.tc_prio,
+                                     &ch_flower->fs, ch_flower->filter_id);
        if (ret)
                goto err;
 
@@ -825,9 +840,6 @@ int cxgb4_tc_flower_destroy(struct net_device *dev,
        }
        kfree_rcu(ch_flower, rcu);
 
-       if (hash)
-               cxgb4_tc_flower_hash_prio_del(adap, tc_prio);
-
 err:
        return ret;
 }
index 7fa3797..befa459 100644 (file)
@@ -124,6 +124,8 @@ int cxgb4_tc_flower_stats(struct net_device *dev,
 int cxgb4_flow_rule_replace(struct net_device *dev, struct flow_rule *rule,
                            u32 tc_prio, struct netlink_ext_ack *extack,
                            struct ch_filter_specification *fs, u32 *tid);
+int cxgb4_flow_rule_destroy(struct net_device *dev, u32 tc_prio,
+                           struct ch_filter_specification *fs, int tid);
 
 int cxgb4_init_tc_flower(struct adapter *adap);
 void cxgb4_cleanup_tc_flower(struct adapter *adap);