]> git.baikalelectronics.ru Git - kernel.git/commitdiff
mlxsw: spectrum_router: Generalize __mlxsw_sp_ipip_entry_update_tunnel()
authorPetr Machata <petrm@mellanox.com>
Fri, 3 Nov 2017 09:03:38 +0000 (10:03 +0100)
committerDavid S. Miller <davem@davemloft.net>
Sat, 4 Nov 2017 00:15:17 +0000 (09:15 +0900)
The work that needs to be done to update HW configuration in response to
changes is similar to what __mlxsw_sp_ipip_entry_update_tunnel() already
does, but with a number of twists: each change requires a different
subset of things to happen. Extend the function to support all these
uses, and allow finely-grained configuration of what should happen at
each call through a suite of function arguments.

Publish the updated function to allow use from the spectrum_ipip module.

Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h

index e2795b8890686db058ca07d5e045598da488ba7f..1376a9738b3caf21b4644c1a06a7cd9bef3847f9 100644 (file)
@@ -1355,9 +1355,12 @@ static void mlxsw_sp_netdevice_ipip_ol_down_event(struct mlxsw_sp *mlxsw_sp,
                mlxsw_sp_ipip_entry_ol_down_event(mlxsw_sp, ipip_entry);
 }
 
+static void mlxsw_sp_nexthop_rif_update(struct mlxsw_sp *mlxsw_sp,
+                                       struct mlxsw_sp_rif *rif);
 static int
 mlxsw_sp_ipip_entry_ol_lb_update(struct mlxsw_sp *mlxsw_sp,
                                 struct mlxsw_sp_ipip_entry *ipip_entry,
+                                bool keep_encap,
                                 struct netlink_ext_ack *extack)
 {
        struct mlxsw_sp_rif_ipip_lb *old_lb_rif = ipip_entry->ol_lb;
@@ -1370,13 +1373,32 @@ mlxsw_sp_ipip_entry_ol_lb_update(struct mlxsw_sp *mlxsw_sp,
        if (IS_ERR(new_lb_rif))
                return PTR_ERR(new_lb_rif);
        ipip_entry->ol_lb = new_lb_rif;
+
+       if (keep_encap) {
+               list_splice_init(&old_lb_rif->common.nexthop_list,
+                                &new_lb_rif->common.nexthop_list);
+               mlxsw_sp_nexthop_rif_update(mlxsw_sp, &new_lb_rif->common);
+       }
+
        mlxsw_sp_rif_destroy(&old_lb_rif->common);
 
        return 0;
 }
 
+/**
+ * Update the offload related to an IPIP entry. This always updates decap, and
+ * in addition to that it also:
+ * @recreate_loopback: recreates the associated loopback RIF
+ * @keep_encap: updates next hops that use the tunnel netdevice. This is only
+ *              relevant when recreate_loopback is true.
+ * @update_nexthops: updates next hops, keeping the current loopback RIF. This
+ *                   is only relevant when recreate_loopback is false.
+ */
 int __mlxsw_sp_ipip_entry_update_tunnel(struct mlxsw_sp *mlxsw_sp,
                                        struct mlxsw_sp_ipip_entry *ipip_entry,
+                                       bool recreate_loopback,
+                                       bool keep_encap,
+                                       bool update_nexthops,
                                        struct netlink_ext_ack *extack)
 {
        int err;
@@ -1390,9 +1412,15 @@ int __mlxsw_sp_ipip_entry_update_tunnel(struct mlxsw_sp *mlxsw_sp,
        if (ipip_entry->decap_fib_entry)
                mlxsw_sp_ipip_entry_demote_decap(mlxsw_sp, ipip_entry);
 
-       err = mlxsw_sp_ipip_entry_ol_lb_update(mlxsw_sp, ipip_entry, extack);
-       if (err)
-               return err;
+       if (recreate_loopback) {
+               err = mlxsw_sp_ipip_entry_ol_lb_update(mlxsw_sp, ipip_entry,
+                                                      keep_encap, extack);
+               if (err)
+                       return err;
+       } else if (update_nexthops) {
+               mlxsw_sp_nexthop_rif_update(mlxsw_sp,
+                                           &ipip_entry->ol_lb->common);
+       }
 
        if (ipip_entry->ol_dev->flags & IFF_UP)
                mlxsw_sp_ipip_entry_ol_up_event(mlxsw_sp, ipip_entry);
@@ -1410,7 +1438,7 @@ static int mlxsw_sp_netdevice_ipip_ol_vrf_event(struct mlxsw_sp *mlxsw_sp,
        if (!ipip_entry)
                return 0;
        return __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry,
-                                                  extack);
+                                                  true, false, false, extack);
 }
 
 int mlxsw_sp_netdevice_ipip_ol_event(struct mlxsw_sp *mlxsw_sp,
@@ -3285,6 +3313,17 @@ static void mlxsw_sp_nexthop4_event(struct mlxsw_sp *mlxsw_sp,
        mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nh_grp);
 }
 
+static void mlxsw_sp_nexthop_rif_update(struct mlxsw_sp *mlxsw_sp,
+                                       struct mlxsw_sp_rif *rif)
+{
+       struct mlxsw_sp_nexthop *nh;
+
+       list_for_each_entry(nh, &rif->nexthop_list, rif_list_node) {
+               __mlxsw_sp_nexthop_neigh_update(nh, false);
+               mlxsw_sp_nexthop_group_refresh(mlxsw_sp, nh->nh_grp);
+       }
+}
+
 static void mlxsw_sp_nexthop_rif_gone_sync(struct mlxsw_sp *mlxsw_sp,
                                           struct mlxsw_sp_rif *rif)
 {
index 8120b01a9c366d64ce05d8d16ae70adefdfce12b..4b8a12a4f49396c3276622dd5d5f95f21974e62b 100644 (file)
@@ -63,6 +63,7 @@ enum mlxsw_sp_rif_counter_dir {
 
 struct mlxsw_sp_neigh_entry;
 struct mlxsw_sp_nexthop;
+struct mlxsw_sp_ipip_entry;
 
 struct mlxsw_sp_rif *mlxsw_sp_rif_by_index(const struct mlxsw_sp *mlxsw_sp,
                                           u16 rif_index);
@@ -103,6 +104,12 @@ mlxsw_sp_neigh_entry_counter_update(struct mlxsw_sp *mlxsw_sp,
                                    struct mlxsw_sp_neigh_entry *neigh_entry,
                                    bool adding);
 bool mlxsw_sp_neigh_ipv6_ignore(struct mlxsw_sp_neigh_entry *neigh_entry);
+int __mlxsw_sp_ipip_entry_update_tunnel(struct mlxsw_sp *mlxsw_sp,
+                                       struct mlxsw_sp_ipip_entry *ipip_entry,
+                                       bool recreate_loopback,
+                                       bool keep_encap,
+                                       bool update_nexthops,
+                                       struct netlink_ext_ack *extack);
 struct mlxsw_sp_nexthop *mlxsw_sp_nexthop_next(struct mlxsw_sp_router *router,
                                               struct mlxsw_sp_nexthop *nh);
 bool mlxsw_sp_nexthop_offload(struct mlxsw_sp_nexthop *nh);