[RFC net-next 2/4] net: sched: add tcf_proto_op to offload a rule

[Jakub Kicinski <jakub.kicinski@netronome.com>]

From: John Hurley <john.hurley@netronome.com>

Create a new tcf_proto_op called offload that generates a new offload
message for each node in a tcf_proto. Pointers to the tcf_proto and
whether the offload request is to add or delete the node are included.
Also included is a callback function to send the offload message to and
the option of priv data to go with the cb.

The function is called to offload all tcf_proto nodes in all chains of a
block when a callback tries to register to a port that already has
offloaded rules. If all existing rules cannot be offloaded then the
registration is rejected. This replaces the previous policy of rejecting
such callback registration outright.

On unregistration of a callback, the rules are flushed for that given cb.
The implementation of block sharing in the NFP driver, for example,
dublicates shared rules to all devs bound to a block. This meant that
rules could still exist in hw even after a device is unbound from a block
(assuming the block still remains active).

Signed-off-by: John Hurley <john.hurley@netronome.com>
---
 .../net/ethernet/mellanox/mlxsw/spectrum.c    |  4 +-
 include/net/act_api.h                         |  3 --
 include/net/pkt_cls.h                         |  6 ++-
 include/net/sch_generic.h                     |  6 +++
 net/sched/cls_api.c                           | 50 ++++++++++++++++---
 5 files changed, 54 insertions(+), 15 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
index 6887c7faaaba..67ff1934fc38 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
@@ -1542,7 +1542,7 @@ mlxsw_sp_setup_tc_block_flower_bind(struct mlxsw_sp_port *mlxsw_sp_port,
 
 err_block_bind:
 	if (!tcf_block_cb_decref(block_cb)) {
-		__tcf_block_cb_unregister(block_cb);
+		__tcf_block_cb_unregister(block, block_cb);
 err_cb_register:
 		mlxsw_sp_acl_block_destroy(acl_block);
 	}
@@ -1572,7 +1572,7 @@ mlxsw_sp_setup_tc_block_flower_unbind(struct mlxsw_sp_port *mlxsw_sp_port,
 	err = mlxsw_sp_acl_block_unbind(mlxsw_sp, acl_block,
 					mlxsw_sp_port, ingress);
 	if (!err && !tcf_block_cb_decref(block_cb)) {
-		__tcf_block_cb_unregister(block_cb);
+		__tcf_block_cb_unregister(block, block_cb);
 		mlxsw_sp_acl_block_destroy(acl_block);
 	}
 }
diff --git a/include/net/act_api.h b/include/net/act_api.h
index 9e59ebfded62..5ff11adbe2a6 100644
--- a/include/net/act_api.h
+++ b/include/net/act_api.h
@@ -190,9 +190,6 @@ static inline void tcf_action_stats_update(struct tc_action *a, u64 bytes,
 #endif
 }
 
-typedef int tc_setup_cb_t(enum tc_setup_type type,
-			  void *type_data, void *cb_priv);
-
 #ifdef CONFIG_NET_CLS_ACT
 int tc_setup_cb_egdev_register(const struct net_device *dev,
 			       tc_setup_cb_t *cb, void *cb_priv);
diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h
index b2b5cbe13086..74dd9784bf88 100644
--- a/include/net/pkt_cls.h
+++ b/include/net/pkt_cls.h
@@ -78,7 +78,8 @@ struct tcf_block_cb *__tcf_block_cb_register(struct tcf_block *block,
 int tcf_block_cb_register(struct tcf_block *block,
 			  tc_setup_cb_t *cb, void *cb_ident,
 			  void *cb_priv, struct netlink_ext_ack *extack);
-void __tcf_block_cb_unregister(struct tcf_block_cb *block_cb);
+void __tcf_block_cb_unregister(struct tcf_block *block,
+			       struct tcf_block_cb *block_cb);
 void tcf_block_cb_unregister(struct tcf_block *block,
 			     tc_setup_cb_t *cb, void *cb_ident);
 
@@ -176,7 +177,8 @@ int tcf_block_cb_register(struct tcf_block *block,
 }
 
 static inline
-void __tcf_block_cb_unregister(struct tcf_block_cb *block_cb)
+void __tcf_block_cb_unregister(struct tcf_block *block,
+			       struct tcf_block_cb *block_cb)
 {
 }
 
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 98c10a28cd01..7b0d62870e82 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -20,6 +20,9 @@ struct qdisc_walker;
 struct tcf_walker;
 struct module;
 
+typedef int tc_setup_cb_t(enum tc_setup_type type,
+			  void *type_data, void *cb_priv);
+
 struct qdisc_rate_table {
 	struct tc_ratespec rate;
 	u32		data[256];
@@ -256,6 +259,9 @@ struct tcf_proto_ops {
 					  bool *last,
 					  struct netlink_ext_ack *);
 	void			(*walk)(struct tcf_proto*, struct tcf_walker *arg);
+	int			(*offload)(struct tcf_proto *, bool,
+					   tc_setup_cb_t *, void *,
+					   struct netlink_ext_ack *);
 	void			(*bind_class)(void *, u32, unsigned long);
 
 	/* rtnetlink specific */
diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c
index 95a83aa798d1..0aba7d7db099 100644
--- a/net/sched/cls_api.c
+++ b/net/sched/cls_api.c
@@ -677,19 +677,50 @@ unsigned int tcf_block_cb_decref(struct tcf_block_cb *block_cb)
 }
 EXPORT_SYMBOL(tcf_block_cb_decref);
 
+static int
+tcf_block_playback_offloads(struct tcf_block *block, tc_setup_cb_t *cb,
+			    void *cb_priv, bool add,
+			    struct netlink_ext_ack *extack)
+{
+	struct tcf_chain *chain;
+	struct tcf_proto *tp;
+	int err;
+
+	list_for_each_entry(chain, &block->chain_list, list) {
+		for (tp = rtnl_dereference(chain->filter_chain); tp;
+		     tp = rtnl_dereference(tp->next)) {
+			if (tp->ops->offload) {
+				err = tp->ops->offload(tp, add, cb, cb_priv,
+						       extack);
+				if (err && add)
+					goto err_playback_remove;
+			} else if (add) {
+				err = -EOPNOTSUPP;
+				NL_SET_ERR_MSG(extack, "Block rule replay failed - no tp offload op");
+				goto err_playback_remove;
+			}
+		}
+	}
+
+	return 0;
+
+err_playback_remove:
+	tcf_block_playback_offloads(block, cb, cb_priv, false, extack);
+	return err;
+}
+
 struct tcf_block_cb *__tcf_block_cb_register(struct tcf_block *block,
 					     tc_setup_cb_t *cb, void *cb_ident,
 					     void *cb_priv,
 					     struct netlink_ext_ack *extack)
 {
 	struct tcf_block_cb *block_cb;
+	int err;
 
-	/* At this point, playback of previous block cb calls is not supported,
-	 * so forbid to register to block which already has some offloaded
-	 * filters present.
-	 */
-	if (tcf_block_offload_in_use(block))
-		return ERR_PTR(-EOPNOTSUPP);
+	/* Replay any already present rules */
+	err = tcf_block_playback_offloads(block, cb, cb_priv, true, extack);
+	if (err)
+		return ERR_PTR(err);
 
 	block_cb = kzalloc(sizeof(*block_cb), GFP_KERNEL);
 	if (!block_cb)
@@ -714,8 +745,11 @@ int tcf_block_cb_register(struct tcf_block *block,
 }
 EXPORT_SYMBOL(tcf_block_cb_register);
 
-void __tcf_block_cb_unregister(struct tcf_block_cb *block_cb)
+void __tcf_block_cb_unregister(struct tcf_block *block,
+			       struct tcf_block_cb *block_cb)
 {
+	tcf_block_playback_offloads(block, block_cb->cb, block_cb->cb_priv,
+				    false, NULL);
 	list_del(&block_cb->list);
 	kfree(block_cb);
 }
@@ -729,7 +763,7 @@ void tcf_block_cb_unregister(struct tcf_block *block,
 	block_cb = tcf_block_cb_lookup(block, cb, cb_ident);
 	if (!block_cb)
 		return;
-	__tcf_block_cb_unregister(block_cb);
+	__tcf_block_cb_unregister(block, block_cb);
 }
 EXPORT_SYMBOL(tcf_block_cb_unregister);
 
-- 
2.17.0