#include "en.h"
#include "en_rep.h"
#include "en_tc.h"
+#include "fs_core.h"
static const char mlx5e_rep_driver_name[] = "mlx5e_rep";
mlx5_eswitch_sqs2vport_stop(esw, rep);
}
+static void mlx5e_rep_neigh_update_init_interval(struct mlx5e_rep_priv *rpriv)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+ unsigned long ipv6_interval = NEIGH_VAR(&ipv6_stub->nd_tbl->parms,
+ DELAY_PROBE_TIME);
+#else
+ unsigned long ipv6_interval = ~0UL;
+#endif
+ unsigned long ipv4_interval = NEIGH_VAR(&arp_tbl.parms,
+ DELAY_PROBE_TIME);
+ struct net_device *netdev = rpriv->rep->netdev;
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+
+ rpriv->neigh_update.min_interval = min_t(unsigned long, ipv6_interval, ipv4_interval);
+ mlx5_fc_update_sampling_interval(priv->mdev, rpriv->neigh_update.min_interval);
+}
+
+void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv)
+{
+ struct mlx5e_rep_priv *rpriv = priv->ppriv;
+ struct mlx5e_neigh_update_table *neigh_update = &rpriv->neigh_update;
+
+ mlx5_fc_queue_stats_work(priv->mdev,
+ &neigh_update->neigh_stats_work,
+ neigh_update->min_interval);
+}
+
+static void mlx5e_rep_neigh_stats_work(struct work_struct *work)
+{
+ struct mlx5e_rep_priv *rpriv = container_of(work, struct mlx5e_rep_priv,
+ neigh_update.neigh_stats_work.work);
+ struct net_device *netdev = rpriv->rep->netdev;
+ struct mlx5e_priv *priv = netdev_priv(netdev);
+ struct mlx5e_neigh_hash_entry *nhe;
+
+ rtnl_lock();
+ if (!list_empty(&rpriv->neigh_update.neigh_list))
+ mlx5e_rep_queue_neigh_stats_work(priv);
+
+ list_for_each_entry(nhe, &rpriv->neigh_update.neigh_list, neigh_list)
+ mlx5e_tc_update_neigh_used_value(nhe);
+
+ rtnl_unlock();
+}
+
static void mlx5e_rep_neigh_entry_hold(struct mlx5e_neigh_hash_entry *nhe)
{
refcount_inc(&nhe->refcnt);
return NOTIFY_DONE;
m_neigh.dev = n->dev;
+ m_neigh.family = n->ops->family;
memcpy(&m_neigh.dst_ip, n->primary_key, n->tbl->key_len);
/* We are in atomic context and can't take RTNL mutex, so use
INIT_LIST_HEAD(&neigh_update->neigh_list);
spin_lock_init(&neigh_update->encap_lock);
+ INIT_DELAYED_WORK(&neigh_update->neigh_stats_work,
+ mlx5e_rep_neigh_stats_work);
+ mlx5e_rep_neigh_update_init_interval(rpriv);
rpriv->neigh_update.netevent_nb.notifier_call = mlx5e_rep_netevent_event;
err = register_netevent_notifier(&rpriv->neigh_update.netevent_nb);
flush_workqueue(priv->wq); /* flush neigh update works */
+ cancel_delayed_work_sync(&rpriv->neigh_update.neigh_stats_work);
+
rhashtable_destroy(&neigh_update->neigh_ht);
}
/* protect lookup/remove operations */
spinlock_t encap_lock;
struct notifier_block netevent_nb;
+ struct delayed_work neigh_stats_work;
+ unsigned long min_interval; /* jiffies */
};
struct mlx5e_rep_priv {
__be32 v4;
struct in6_addr v6;
} dst_ip;
+ int family;
};
struct mlx5e_neigh_hash_entry {
* it's used by the neigh notification call.
*/
refcount_t refcnt;
+
+ /* Save the last reported time offloaded trafic pass over one of the
+ * neigh hash entry flows. Use it to periodically update the neigh
+ * 'used' value and avoid neigh deleting by the kernel.
+ */
+ unsigned long reported_lastuse;
};
enum {
void mlx5e_rep_encap_entry_detach(struct mlx5e_priv *priv,
struct mlx5e_encap_entry *e);
+void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv);
+
#endif /* __MLX5E_REP_H__ */
#include <net/tc_act/tc_tunnel_key.h>
#include <net/tc_act/tc_pedit.h>
#include <net/vxlan.h>
+#include <net/arp.h>
#include "en.h"
#include "en_rep.h"
#include "en_tc.h"
return;
}
e->flags |= MLX5_ENCAP_ENTRY_VALID;
+ mlx5e_rep_queue_neigh_stats_work(priv);
list_for_each_entry(flow, &e->flows, encap) {
flow->esw_attr->encap_id = e->encap_id;
}
}
+void mlx5e_tc_update_neigh_used_value(struct mlx5e_neigh_hash_entry *nhe)
+{
+ struct mlx5e_neigh *m_neigh = &nhe->m_neigh;
+ u64 bytes, packets, lastuse = 0;
+ struct mlx5e_tc_flow *flow;
+ struct mlx5e_encap_entry *e;
+ struct mlx5_fc *counter;
+ struct neigh_table *tbl;
+ bool neigh_used = false;
+ struct neighbour *n;
+
+ if (m_neigh->family == AF_INET)
+ tbl = &arp_tbl;
+#if IS_ENABLED(CONFIG_IPV6)
+ else if (m_neigh->family == AF_INET6)
+ tbl = ipv6_stub->nd_tbl;
+#endif
+ else
+ return;
+
+ list_for_each_entry(e, &nhe->encap_list, encap_list) {
+ if (!(e->flags & MLX5_ENCAP_ENTRY_VALID))
+ continue;
+ list_for_each_entry(flow, &e->flows, encap) {
+ if (flow->flags & MLX5E_TC_FLOW_OFFLOADED) {
+ counter = mlx5_flow_rule_counter(flow->rule);
+ mlx5_fc_query_cached(counter, &bytes, &packets, &lastuse);
+ if (time_after((unsigned long)lastuse, nhe->reported_lastuse)) {
+ neigh_used = true;
+ break;
+ }
+ }
+ }
+ }
+
+ if (neigh_used) {
+ nhe->reported_lastuse = jiffies;
+
+ /* find the relevant neigh according to the cached device and
+ * dst ip pair
+ */
+ n = neigh_lookup(tbl, &m_neigh->dst_ip, m_neigh->dev);
+ if (!n) {
+ WARN(1, "The neighbour already freed\n");
+ return;
+ }
+
+ neigh_event_send(n, NULL);
+ neigh_release(n);
+ }
+}
+
static void mlx5e_detach_encap(struct mlx5e_priv *priv,
struct mlx5e_tc_flow *flow)
{
* entry in the neigh hash table when a user deletes a rule
*/
e->m_neigh.dev = n->dev;
+ e->m_neigh.family = n->ops->family;
memcpy(&e->m_neigh.dst_ip, n->primary_key, n->tbl->key_len);
e->out_dev = out_dev;
goto destroy_neigh_entry;
e->flags |= MLX5_ENCAP_ENTRY_VALID;
+ mlx5e_rep_queue_neigh_stats_work(netdev_priv(out_dev));
neigh_release(n);
return err;
* entry in the neigh hash table when a user deletes a rule
*/
e->m_neigh.dev = n->dev;
+ e->m_neigh.family = n->ops->family;
memcpy(&e->m_neigh.dst_ip, n->primary_key, n->tbl->key_len);
e->out_dev = out_dev;
goto destroy_neigh_entry;
e->flags |= MLX5_ENCAP_ENTRY_VALID;
+ mlx5e_rep_queue_neigh_stats_work(netdev_priv(out_dev));
neigh_release(n);
return err;
void mlx5e_tc_encap_flows_del(struct mlx5e_priv *priv,
struct mlx5e_encap_entry *e);
+struct mlx5e_neigh_hash_entry;
+void mlx5e_tc_update_neigh_used_value(struct mlx5e_neigh_hash_entry *nhe);
+
static inline int mlx5e_tc_num_filters(struct mlx5e_priv *priv)
{
return atomic_read(&priv->fs.tc.ht.nelems);
int mlx5_init_fc_stats(struct mlx5_core_dev *dev);
void mlx5_cleanup_fc_stats(struct mlx5_core_dev *dev);
+void mlx5_fc_queue_stats_work(struct mlx5_core_dev *dev,
+ struct delayed_work *dwork,
+ unsigned long delay);
+void mlx5_fc_update_sampling_interval(struct mlx5_core_dev *dev,
+ unsigned long interval);
int mlx5_init_fs(struct mlx5_core_dev *dev);
void mlx5_cleanup_fs(struct mlx5_core_dev *dev);
list_splice_tail_init(&fc_stats->addlist, &tmplist);
if (!list_empty(&tmplist) || !RB_EMPTY_ROOT(&fc_stats->counters))
- queue_delayed_work(fc_stats->wq, &fc_stats->work, MLX5_FC_STATS_PERIOD);
+ queue_delayed_work(fc_stats->wq, &fc_stats->work,
+ fc_stats->sampling_interval);
spin_unlock(&fc_stats->addlist_lock);
node = mlx5_fc_stats_query(dev, counter, last->id);
}
- fc_stats->next_query = now + MLX5_FC_STATS_PERIOD;
+ fc_stats->next_query = now + fc_stats->sampling_interval;
}
struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging)
if (!fc_stats->wq)
return -ENOMEM;
+ fc_stats->sampling_interval = MLX5_FC_STATS_PERIOD;
INIT_DELAYED_WORK(&fc_stats->work, mlx5_fc_stats_work);
return 0;
counter->lastbytes = c.bytes;
counter->lastpackets = c.packets;
}
+
+void mlx5_fc_queue_stats_work(struct mlx5_core_dev *dev,
+ struct delayed_work *dwork,
+ unsigned long delay)
+{
+ struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats;
+
+ queue_delayed_work(fc_stats->wq, dwork, delay);
+}
+
+void mlx5_fc_update_sampling_interval(struct mlx5_core_dev *dev,
+ unsigned long interval)
+{
+ struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats;
+
+ fc_stats->sampling_interval = min_t(unsigned long, interval,
+ fc_stats->sampling_interval);
+}
struct workqueue_struct *wq;
struct delayed_work work;
unsigned long next_query;
+ unsigned long sampling_interval; /* jiffies */
};
struct mlx5_eswitch;