1 /*
2 * Copyright (c) 2017, Mellanox Technologies. All rights reserved.
3 *
4 * This software is available to you under a choice of one of two
5 * licenses. You may choose to be licensed under the terms of the GNU
6 * General Public License (GPL) Version 2, available from the file
7 * COPYING in the main directory of this source tree, or the
8 * OpenIB.org BSD license below:
9 *
10 * Redistribution and use in source and binary forms, with or
11 * without modification, are permitted provided that the following
12 * conditions are met:
13 *
14 * - Redistributions of source code must retain the above
15 * copyright notice, this list of conditions and the following
16 * disclaimer.
17 *
18 * - Redistributions in binary form must reproduce the above
19 * copyright notice, this list of conditions and the following
20 * disclaimer in the documentation and/or other materials
21 * provided with the distribution.
22 *
23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
24 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
25 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
26 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
27 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
28 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
29 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
30 * SOFTWARE.
31 */
32
33 #ifndef __MLX5E_REP_H__
34 #define __MLX5E_REP_H__
35
36 #include <net/ip_tunnels.h>
37 #include <linux/rhashtable.h>
38 #include <linux/mutex.h>
39 #include "eswitch.h"
40 #include "en.h"
41 #include "lib/port_tun.h"
42
43 #ifdef CONFIG_MLX5_ESWITCH
44 extern const struct mlx5e_rx_handlers mlx5e_rx_handlers_rep;
45
46 struct mlx5e_neigh_update_table {
47 struct rhashtable neigh_ht;
48 /* Save the neigh hash entries in a list in addition to the hash table
49 * (neigh_ht). In order to iterate easily over the neigh entries.
50 * Used for stats query.
51 */
52 struct list_head neigh_list;
53 /* protect lookup/remove operations */
54 struct mutex encap_lock;
55 struct notifier_block netevent_nb;
56 struct delayed_work neigh_stats_work;
57 unsigned long min_interval; /* jiffies */
58 };
59
60 struct mlx5_tc_ct_priv;
61 struct mlx5_tc_int_port_priv;
62 struct mlx5e_rep_bond;
63 struct mlx5e_tc_tun_encap;
64 struct mlx5e_post_act;
65 struct mlx5e_flow_meters;
66
67 struct mlx5_rep_uplink_priv {
68 /* indirect block callbacks are invoked on bind/unbind events
69 * on registered higher level devices (e.g. tunnel devices)
70 *
71 * tc_indr_block_cb_priv_list is used to lookup indirect callback
72 * private data
73 *
74 */
75 struct list_head tc_indr_block_priv_list;
76
77 struct mlx5_tun_entropy tun_entropy;
78
79 /* protects unready_flows */
80 struct mutex unready_flows_lock;
81 struct list_head unready_flows;
82 struct work_struct reoffload_flows_work;
83
84 /* maps tun_info to a unique id*/
85 struct mapping_ctx *tunnel_mapping;
86 /* maps tun_enc_opts to a unique id*/
87 struct mapping_ctx *tunnel_enc_opts_mapping;
88
89 struct mlx5e_post_act *post_act;
90 struct mlx5_tc_ct_priv *ct_priv;
91 struct mlx5e_tc_psample *tc_psample;
92
93 /* support eswitch vports bonding */
94 struct mlx5e_rep_bond *bond;
95
96 /* tc tunneling encapsulation private data */
97 struct mlx5e_tc_tun_encap *encap;
98
99 /* OVS internal port support */
100 struct mlx5e_tc_int_port_priv *int_port_priv;
101
102 struct mlx5e_flow_meters *flow_meters;
103
104 /* tc action stats */
105 struct mlx5e_tc_act_stats_handle *action_stats_handle;
106
107 struct work_struct mpesw_work;
108 };
109
110 struct mlx5e_rep_priv {
111 struct mlx5_eswitch_rep *rep;
112 struct mlx5e_neigh_update_table neigh_update;
113 struct net_device *netdev;
114 struct mlx5_flow_table *root_ft;
115 struct mlx5_flow_handle *vport_rx_rule;
116 struct list_head vport_sqs_list;
117 struct mlx5_rep_uplink_priv uplink_priv; /* valid for uplink rep */
118 struct rtnl_link_stats64 prev_vf_vport_stats;
119 struct mlx5_flow_handle *send_to_vport_meta_rule;
120 struct rhashtable tc_ht;
121 };
122
123 static inline
mlx5e_rep_to_rep_priv(struct mlx5_eswitch_rep * rep)124 struct mlx5e_rep_priv *mlx5e_rep_to_rep_priv(struct mlx5_eswitch_rep *rep)
125 {
126 return rep->rep_data[REP_ETH].priv;
127 }
128
129 struct mlx5e_neigh {
130 union {
131 __be32 v4;
132 struct in6_addr v6;
133 } dst_ip;
134 int family;
135 };
136
137 struct mlx5e_neigh_hash_entry {
138 struct rhash_head rhash_node;
139 struct mlx5e_neigh m_neigh;
140 struct mlx5e_priv *priv;
141 struct net_device *neigh_dev;
142
143 /* Save the neigh hash entry in a list on the representor in
144 * addition to the hash table. In order to iterate easily over the
145 * neighbour entries. Used for stats query.
146 */
147 struct list_head neigh_list;
148
149 /* protects encap list */
150 spinlock_t encap_list_lock;
151 /* encap list sharing the same neigh */
152 struct list_head encap_list;
153
154 /* neigh hash entry can be deleted only when the refcount is zero.
155 * refcount is needed to avoid neigh hash entry removal by TC, while
156 * it's used by the neigh notification call.
157 */
158 refcount_t refcnt;
159
160 /* Save the last reported time offloaded traffic pass over one of the
161 * neigh hash entry flows. Use it to periodically update the neigh
162 * 'used' value and avoid neigh deleting by the kernel.
163 */
164 unsigned long reported_lastuse;
165
166 struct rcu_head rcu;
167 };
168
169 enum {
170 /* set when the encap entry is successfully offloaded into HW */
171 MLX5_ENCAP_ENTRY_VALID = BIT(0),
172 MLX5_REFORMAT_DECAP = BIT(1),
173 MLX5_ENCAP_ENTRY_NO_ROUTE = BIT(2),
174 };
175
176 struct mlx5e_decap_key {
177 struct ethhdr key;
178 };
179
180 struct mlx5e_decap_entry {
181 struct mlx5e_decap_key key;
182 struct list_head flows;
183 struct hlist_node hlist;
184 refcount_t refcnt;
185 struct completion res_ready;
186 int compl_result;
187 struct mlx5_pkt_reformat *pkt_reformat;
188 struct rcu_head rcu;
189 };
190
191 struct mlx5e_mpls_info {
192 u32 label;
193 u8 tc;
194 u8 bos;
195 u8 ttl;
196 };
197
198 struct mlx5e_encap_entry {
199 /* attached neigh hash entry */
200 struct mlx5e_neigh_hash_entry *nhe;
201 /* neigh hash entry list of encaps sharing the same neigh */
202 struct list_head encap_list;
203 /* a node of the eswitch encap hash table which keeping all the encap
204 * entries
205 */
206 struct hlist_node encap_hlist;
207 struct list_head flows;
208 struct list_head route_list;
209 struct mlx5_pkt_reformat *pkt_reformat;
210 const struct ip_tunnel_info *tun_info;
211 struct mlx5e_mpls_info mpls_info;
212 unsigned char h_dest[ETH_ALEN]; /* destination eth addr */
213
214 struct net_device *out_dev;
215 int route_dev_ifindex;
216 struct mlx5e_tc_tunnel *tunnel;
217 int reformat_type;
218 u8 flags;
219 char *encap_header;
220 int encap_size;
221 refcount_t refcnt;
222 struct completion res_ready;
223 int compl_result;
224 struct rcu_head rcu;
225 };
226
227 struct mlx5e_rep_sq {
228 struct mlx5_flow_handle *send_to_vport_rule;
229 struct mlx5_flow_handle *send_to_vport_rule_peer;
230 u32 sqn;
231 struct list_head list;
232 };
233
234 int mlx5e_rep_init(void);
235 void mlx5e_rep_cleanup(void);
236 int mlx5e_rep_bond_init(struct mlx5e_rep_priv *rpriv);
237 void mlx5e_rep_bond_cleanup(struct mlx5e_rep_priv *rpriv);
238 int mlx5e_rep_bond_enslave(struct mlx5_eswitch *esw, struct net_device *netdev,
239 struct net_device *lag_dev);
240 void mlx5e_rep_bond_unslave(struct mlx5_eswitch *esw,
241 const struct net_device *netdev,
242 const struct net_device *lag_dev);
243 int mlx5e_rep_bond_update(struct mlx5e_priv *priv, bool cleanup);
244
245 bool mlx5e_rep_has_offload_stats(const struct net_device *dev, int attr_id);
246 int mlx5e_rep_get_offload_stats(int attr_id, const struct net_device *dev,
247 void *sp);
248
249 bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv);
250 void mlx5e_rep_activate_channels(struct mlx5e_priv *priv);
251 void mlx5e_rep_deactivate_channels(struct mlx5e_priv *priv);
252
253 void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv);
254
255 bool mlx5e_eswitch_vf_rep(const struct net_device *netdev);
256 bool mlx5e_eswitch_uplink_rep(const struct net_device *netdev);
mlx5e_eswitch_rep(const struct net_device * netdev)257 static inline bool mlx5e_eswitch_rep(const struct net_device *netdev)
258 {
259 return mlx5e_eswitch_vf_rep(netdev) ||
260 mlx5e_eswitch_uplink_rep(netdev);
261 }
262
263 #else /* CONFIG_MLX5_ESWITCH */
mlx5e_is_uplink_rep(struct mlx5e_priv * priv)264 static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; }
mlx5e_rep_activate_channels(struct mlx5e_priv * priv)265 static inline void mlx5e_rep_activate_channels(struct mlx5e_priv *priv) {}
mlx5e_rep_deactivate_channels(struct mlx5e_priv * priv)266 static inline void mlx5e_rep_deactivate_channels(struct mlx5e_priv *priv) {}
mlx5e_rep_init(void)267 static inline int mlx5e_rep_init(void) { return 0; };
mlx5e_rep_cleanup(void)268 static inline void mlx5e_rep_cleanup(void) {};
mlx5e_rep_has_offload_stats(const struct net_device * dev,int attr_id)269 static inline bool mlx5e_rep_has_offload_stats(const struct net_device *dev,
270 int attr_id) { return false; }
mlx5e_rep_get_offload_stats(int attr_id,const struct net_device * dev,void * sp)271 static inline int mlx5e_rep_get_offload_stats(int attr_id,
272 const struct net_device *dev,
273 void *sp) { return -EOPNOTSUPP; }
274 #endif
275
mlx5e_is_vport_rep(struct mlx5e_priv * priv)276 static inline bool mlx5e_is_vport_rep(struct mlx5e_priv *priv)
277 {
278 return (MLX5_ESWITCH_MANAGER(priv->mdev) && priv->ppriv);
279 }
280 #endif /* __MLX5E_REP_H__ */
281