Loading...
1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * Copyright (c) 2007-2012 Nicira, Inc.
4 */
5
6#include <linux/if_vlan.h>
7#include <linux/kernel.h>
8#include <linux/netdevice.h>
9#include <linux/etherdevice.h>
10#include <linux/ethtool.h>
11#include <linux/skbuff.h>
12
13#include <net/dst.h>
14#include <net/xfrm.h>
15#include <net/rtnetlink.h>
16
17#include "datapath.h"
18#include "vport-internal_dev.h"
19#include "vport-netdev.h"
20
21struct internal_dev {
22 struct vport *vport;
23};
24
25static struct vport_ops ovs_internal_vport_ops;
26
27static struct internal_dev *internal_dev_priv(struct net_device *netdev)
28{
29 return netdev_priv(netdev);
30}
31
32/* Called with rcu_read_lock_bh. */
33static netdev_tx_t
34internal_dev_xmit(struct sk_buff *skb, struct net_device *netdev)
35{
36 int len, err;
37
38 len = skb->len;
39 rcu_read_lock();
40 err = ovs_vport_receive(internal_dev_priv(netdev)->vport, skb, NULL);
41 rcu_read_unlock();
42
43 if (likely(!err)) {
44 struct pcpu_sw_netstats *tstats = this_cpu_ptr(netdev->tstats);
45
46 u64_stats_update_begin(&tstats->syncp);
47 tstats->tx_bytes += len;
48 tstats->tx_packets++;
49 u64_stats_update_end(&tstats->syncp);
50 } else {
51 netdev->stats.tx_errors++;
52 }
53 return NETDEV_TX_OK;
54}
55
56static int internal_dev_open(struct net_device *netdev)
57{
58 netif_start_queue(netdev);
59 return 0;
60}
61
62static int internal_dev_stop(struct net_device *netdev)
63{
64 netif_stop_queue(netdev);
65 return 0;
66}
67
68static void internal_dev_getinfo(struct net_device *netdev,
69 struct ethtool_drvinfo *info)
70{
71 strlcpy(info->driver, "openvswitch", sizeof(info->driver));
72}
73
74static const struct ethtool_ops internal_dev_ethtool_ops = {
75 .get_drvinfo = internal_dev_getinfo,
76 .get_link = ethtool_op_get_link,
77};
78
79static void internal_dev_destructor(struct net_device *dev)
80{
81 struct vport *vport = ovs_internal_dev_get_vport(dev);
82
83 ovs_vport_free(vport);
84}
85
86static void
87internal_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats)
88{
89 int i;
90
91 memset(stats, 0, sizeof(*stats));
92 stats->rx_errors = dev->stats.rx_errors;
93 stats->tx_errors = dev->stats.tx_errors;
94 stats->tx_dropped = dev->stats.tx_dropped;
95 stats->rx_dropped = dev->stats.rx_dropped;
96
97 for_each_possible_cpu(i) {
98 const struct pcpu_sw_netstats *percpu_stats;
99 struct pcpu_sw_netstats local_stats;
100 unsigned int start;
101
102 percpu_stats = per_cpu_ptr(dev->tstats, i);
103
104 do {
105 start = u64_stats_fetch_begin_irq(&percpu_stats->syncp);
106 local_stats = *percpu_stats;
107 } while (u64_stats_fetch_retry_irq(&percpu_stats->syncp, start));
108
109 stats->rx_bytes += local_stats.rx_bytes;
110 stats->rx_packets += local_stats.rx_packets;
111 stats->tx_bytes += local_stats.tx_bytes;
112 stats->tx_packets += local_stats.tx_packets;
113 }
114}
115
116static const struct net_device_ops internal_dev_netdev_ops = {
117 .ndo_open = internal_dev_open,
118 .ndo_stop = internal_dev_stop,
119 .ndo_start_xmit = internal_dev_xmit,
120 .ndo_set_mac_address = eth_mac_addr,
121 .ndo_get_stats64 = internal_get_stats,
122};
123
124static struct rtnl_link_ops internal_dev_link_ops __read_mostly = {
125 .kind = "openvswitch",
126};
127
128static void do_setup(struct net_device *netdev)
129{
130 ether_setup(netdev);
131
132 netdev->max_mtu = ETH_MAX_MTU;
133
134 netdev->netdev_ops = &internal_dev_netdev_ops;
135
136 netdev->priv_flags &= ~IFF_TX_SKB_SHARING;
137 netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_OPENVSWITCH |
138 IFF_NO_QUEUE;
139 netdev->needs_free_netdev = true;
140 netdev->priv_destructor = NULL;
141 netdev->ethtool_ops = &internal_dev_ethtool_ops;
142 netdev->rtnl_link_ops = &internal_dev_link_ops;
143
144 netdev->features = NETIF_F_LLTX | NETIF_F_SG | NETIF_F_FRAGLIST |
145 NETIF_F_HIGHDMA | NETIF_F_HW_CSUM |
146 NETIF_F_GSO_SOFTWARE | NETIF_F_GSO_ENCAP_ALL;
147
148 netdev->vlan_features = netdev->features;
149 netdev->hw_enc_features = netdev->features;
150 netdev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX;
151 netdev->hw_features = netdev->features & ~NETIF_F_LLTX;
152
153 eth_hw_addr_random(netdev);
154}
155
156static struct vport *internal_dev_create(const struct vport_parms *parms)
157{
158 struct vport *vport;
159 struct internal_dev *internal_dev;
160 struct net_device *dev;
161 int err;
162
163 vport = ovs_vport_alloc(0, &ovs_internal_vport_ops, parms);
164 if (IS_ERR(vport)) {
165 err = PTR_ERR(vport);
166 goto error;
167 }
168
169 dev = alloc_netdev(sizeof(struct internal_dev),
170 parms->name, NET_NAME_USER, do_setup);
171 vport->dev = dev;
172 if (!vport->dev) {
173 err = -ENOMEM;
174 goto error_free_vport;
175 }
176 vport->dev->tstats = netdev_alloc_pcpu_stats(struct pcpu_sw_netstats);
177 if (!vport->dev->tstats) {
178 err = -ENOMEM;
179 goto error_free_netdev;
180 }
181
182 dev_net_set(vport->dev, ovs_dp_get_net(vport->dp));
183 internal_dev = internal_dev_priv(vport->dev);
184 internal_dev->vport = vport;
185
186 /* Restrict bridge port to current netns. */
187 if (vport->port_no == OVSP_LOCAL)
188 vport->dev->features |= NETIF_F_NETNS_LOCAL;
189
190 rtnl_lock();
191 err = register_netdevice(vport->dev);
192 if (err)
193 goto error_unlock;
194 vport->dev->priv_destructor = internal_dev_destructor;
195
196 dev_set_promiscuity(vport->dev, 1);
197 rtnl_unlock();
198 netif_start_queue(vport->dev);
199
200 return vport;
201
202error_unlock:
203 rtnl_unlock();
204 free_percpu(dev->tstats);
205error_free_netdev:
206 free_netdev(dev);
207error_free_vport:
208 ovs_vport_free(vport);
209error:
210 return ERR_PTR(err);
211}
212
213static void internal_dev_destroy(struct vport *vport)
214{
215 netif_stop_queue(vport->dev);
216 rtnl_lock();
217 dev_set_promiscuity(vport->dev, -1);
218
219 /* unregister_netdevice() waits for an RCU grace period. */
220 unregister_netdevice(vport->dev);
221 free_percpu(vport->dev->tstats);
222 rtnl_unlock();
223}
224
225static netdev_tx_t internal_dev_recv(struct sk_buff *skb)
226{
227 struct net_device *netdev = skb->dev;
228 struct pcpu_sw_netstats *stats;
229
230 if (unlikely(!(netdev->flags & IFF_UP))) {
231 kfree_skb(skb);
232 netdev->stats.rx_dropped++;
233 return NETDEV_TX_OK;
234 }
235
236 skb_dst_drop(skb);
237 nf_reset_ct(skb);
238 secpath_reset(skb);
239
240 skb->pkt_type = PACKET_HOST;
241 skb->protocol = eth_type_trans(skb, netdev);
242 skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
243
244 stats = this_cpu_ptr(netdev->tstats);
245 u64_stats_update_begin(&stats->syncp);
246 stats->rx_packets++;
247 stats->rx_bytes += skb->len;
248 u64_stats_update_end(&stats->syncp);
249
250 netif_rx(skb);
251 return NETDEV_TX_OK;
252}
253
254static struct vport_ops ovs_internal_vport_ops = {
255 .type = OVS_VPORT_TYPE_INTERNAL,
256 .create = internal_dev_create,
257 .destroy = internal_dev_destroy,
258 .send = internal_dev_recv,
259};
260
261int ovs_is_internal_dev(const struct net_device *netdev)
262{
263 return netdev->netdev_ops == &internal_dev_netdev_ops;
264}
265
266struct vport *ovs_internal_dev_get_vport(struct net_device *netdev)
267{
268 if (!ovs_is_internal_dev(netdev))
269 return NULL;
270
271 return internal_dev_priv(netdev)->vport;
272}
273
274int ovs_internal_dev_rtnl_link_register(void)
275{
276 int err;
277
278 err = rtnl_link_register(&internal_dev_link_ops);
279 if (err < 0)
280 return err;
281
282 err = ovs_vport_ops_register(&ovs_internal_vport_ops);
283 if (err < 0)
284 rtnl_link_unregister(&internal_dev_link_ops);
285
286 return err;
287}
288
289void ovs_internal_dev_rtnl_link_unregister(void)
290{
291 ovs_vport_ops_unregister(&ovs_internal_vport_ops);
292 rtnl_link_unregister(&internal_dev_link_ops);
293}
1/*
2 * Copyright (c) 2007-2012 Nicira, Inc.
3 *
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of version 2 of the GNU General Public
6 * License as published by the Free Software Foundation.
7 *
8 * This program is distributed in the hope that it will be useful, but
9 * WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * General Public License for more details.
12 *
13 * You should have received a copy of the GNU General Public License
14 * along with this program; if not, write to the Free Software
15 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
16 * 02110-1301, USA
17 */
18
19#include <linux/hardirq.h>
20#include <linux/if_vlan.h>
21#include <linux/kernel.h>
22#include <linux/netdevice.h>
23#include <linux/etherdevice.h>
24#include <linux/ethtool.h>
25#include <linux/skbuff.h>
26
27#include <net/dst.h>
28#include <net/xfrm.h>
29#include <net/rtnetlink.h>
30
31#include "datapath.h"
32#include "vport-internal_dev.h"
33#include "vport-netdev.h"
34
35struct internal_dev {
36 struct vport *vport;
37};
38
39static struct vport_ops ovs_internal_vport_ops;
40
41static struct internal_dev *internal_dev_priv(struct net_device *netdev)
42{
43 return netdev_priv(netdev);
44}
45
46/* Called with rcu_read_lock_bh. */
47static int internal_dev_xmit(struct sk_buff *skb, struct net_device *netdev)
48{
49 int len, err;
50
51 len = skb->len;
52 rcu_read_lock();
53 err = ovs_vport_receive(internal_dev_priv(netdev)->vport, skb, NULL);
54 rcu_read_unlock();
55
56 if (likely(!err)) {
57 struct pcpu_sw_netstats *tstats = this_cpu_ptr(netdev->tstats);
58
59 u64_stats_update_begin(&tstats->syncp);
60 tstats->tx_bytes += len;
61 tstats->tx_packets++;
62 u64_stats_update_end(&tstats->syncp);
63 } else {
64 netdev->stats.tx_errors++;
65 }
66 return 0;
67}
68
69static int internal_dev_open(struct net_device *netdev)
70{
71 netif_start_queue(netdev);
72 return 0;
73}
74
75static int internal_dev_stop(struct net_device *netdev)
76{
77 netif_stop_queue(netdev);
78 return 0;
79}
80
81static void internal_dev_getinfo(struct net_device *netdev,
82 struct ethtool_drvinfo *info)
83{
84 strlcpy(info->driver, "openvswitch", sizeof(info->driver));
85}
86
87static const struct ethtool_ops internal_dev_ethtool_ops = {
88 .get_drvinfo = internal_dev_getinfo,
89 .get_link = ethtool_op_get_link,
90};
91
92static int internal_dev_change_mtu(struct net_device *netdev, int new_mtu)
93{
94 if (new_mtu < 68)
95 return -EINVAL;
96
97 netdev->mtu = new_mtu;
98 return 0;
99}
100
101static void internal_dev_destructor(struct net_device *dev)
102{
103 struct vport *vport = ovs_internal_dev_get_vport(dev);
104
105 ovs_vport_free(vport);
106 free_netdev(dev);
107}
108
109static struct rtnl_link_stats64 *
110internal_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats)
111{
112 int i;
113
114 memset(stats, 0, sizeof(*stats));
115 stats->rx_errors = dev->stats.rx_errors;
116 stats->tx_errors = dev->stats.tx_errors;
117 stats->tx_dropped = dev->stats.tx_dropped;
118 stats->rx_dropped = dev->stats.rx_dropped;
119
120 for_each_possible_cpu(i) {
121 const struct pcpu_sw_netstats *percpu_stats;
122 struct pcpu_sw_netstats local_stats;
123 unsigned int start;
124
125 percpu_stats = per_cpu_ptr(dev->tstats, i);
126
127 do {
128 start = u64_stats_fetch_begin_irq(&percpu_stats->syncp);
129 local_stats = *percpu_stats;
130 } while (u64_stats_fetch_retry_irq(&percpu_stats->syncp, start));
131
132 stats->rx_bytes += local_stats.rx_bytes;
133 stats->rx_packets += local_stats.rx_packets;
134 stats->tx_bytes += local_stats.tx_bytes;
135 stats->tx_packets += local_stats.tx_packets;
136 }
137
138 return stats;
139}
140
141static void internal_set_rx_headroom(struct net_device *dev, int new_hr)
142{
143 dev->needed_headroom = new_hr;
144}
145
146static const struct net_device_ops internal_dev_netdev_ops = {
147 .ndo_open = internal_dev_open,
148 .ndo_stop = internal_dev_stop,
149 .ndo_start_xmit = internal_dev_xmit,
150 .ndo_set_mac_address = eth_mac_addr,
151 .ndo_change_mtu = internal_dev_change_mtu,
152 .ndo_get_stats64 = internal_get_stats,
153 .ndo_set_rx_headroom = internal_set_rx_headroom,
154};
155
156static struct rtnl_link_ops internal_dev_link_ops __read_mostly = {
157 .kind = "openvswitch",
158};
159
160static void do_setup(struct net_device *netdev)
161{
162 ether_setup(netdev);
163
164 netdev->netdev_ops = &internal_dev_netdev_ops;
165
166 netdev->priv_flags &= ~IFF_TX_SKB_SHARING;
167 netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_OPENVSWITCH |
168 IFF_PHONY_HEADROOM;
169 netdev->destructor = internal_dev_destructor;
170 netdev->ethtool_ops = &internal_dev_ethtool_ops;
171 netdev->rtnl_link_ops = &internal_dev_link_ops;
172 netdev->tx_queue_len = 0;
173
174 netdev->features = NETIF_F_LLTX | NETIF_F_SG | NETIF_F_FRAGLIST |
175 NETIF_F_HIGHDMA | NETIF_F_HW_CSUM |
176 NETIF_F_GSO_SOFTWARE | NETIF_F_GSO_ENCAP_ALL;
177
178 netdev->vlan_features = netdev->features;
179 netdev->hw_enc_features = netdev->features;
180 netdev->features |= NETIF_F_HW_VLAN_CTAG_TX;
181 netdev->hw_features = netdev->features & ~NETIF_F_LLTX;
182
183 eth_hw_addr_random(netdev);
184}
185
186static struct vport *internal_dev_create(const struct vport_parms *parms)
187{
188 struct vport *vport;
189 struct internal_dev *internal_dev;
190 int err;
191
192 vport = ovs_vport_alloc(0, &ovs_internal_vport_ops, parms);
193 if (IS_ERR(vport)) {
194 err = PTR_ERR(vport);
195 goto error;
196 }
197
198 vport->dev = alloc_netdev(sizeof(struct internal_dev),
199 parms->name, NET_NAME_UNKNOWN, do_setup);
200 if (!vport->dev) {
201 err = -ENOMEM;
202 goto error_free_vport;
203 }
204 vport->dev->tstats = netdev_alloc_pcpu_stats(struct pcpu_sw_netstats);
205 if (!vport->dev->tstats) {
206 err = -ENOMEM;
207 goto error_free_netdev;
208 }
209 vport->dev->needed_headroom = vport->dp->max_headroom;
210
211 dev_net_set(vport->dev, ovs_dp_get_net(vport->dp));
212 internal_dev = internal_dev_priv(vport->dev);
213 internal_dev->vport = vport;
214
215 /* Restrict bridge port to current netns. */
216 if (vport->port_no == OVSP_LOCAL)
217 vport->dev->features |= NETIF_F_NETNS_LOCAL;
218
219 rtnl_lock();
220 err = register_netdevice(vport->dev);
221 if (err)
222 goto error_unlock;
223
224 dev_set_promiscuity(vport->dev, 1);
225 rtnl_unlock();
226 netif_start_queue(vport->dev);
227
228 return vport;
229
230error_unlock:
231 rtnl_unlock();
232 free_percpu(vport->dev->tstats);
233error_free_netdev:
234 free_netdev(vport->dev);
235error_free_vport:
236 ovs_vport_free(vport);
237error:
238 return ERR_PTR(err);
239}
240
241static void internal_dev_destroy(struct vport *vport)
242{
243 netif_stop_queue(vport->dev);
244 rtnl_lock();
245 dev_set_promiscuity(vport->dev, -1);
246
247 /* unregister_netdevice() waits for an RCU grace period. */
248 unregister_netdevice(vport->dev);
249 free_percpu(vport->dev->tstats);
250 rtnl_unlock();
251}
252
253static netdev_tx_t internal_dev_recv(struct sk_buff *skb)
254{
255 struct net_device *netdev = skb->dev;
256 struct pcpu_sw_netstats *stats;
257
258 if (unlikely(!(netdev->flags & IFF_UP))) {
259 kfree_skb(skb);
260 netdev->stats.rx_dropped++;
261 return NETDEV_TX_OK;
262 }
263
264 skb_dst_drop(skb);
265 nf_reset(skb);
266 secpath_reset(skb);
267
268 skb->pkt_type = PACKET_HOST;
269 skb->protocol = eth_type_trans(skb, netdev);
270 skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
271
272 stats = this_cpu_ptr(netdev->tstats);
273 u64_stats_update_begin(&stats->syncp);
274 stats->rx_packets++;
275 stats->rx_bytes += skb->len;
276 u64_stats_update_end(&stats->syncp);
277
278 netif_rx(skb);
279 return NETDEV_TX_OK;
280}
281
282static struct vport_ops ovs_internal_vport_ops = {
283 .type = OVS_VPORT_TYPE_INTERNAL,
284 .create = internal_dev_create,
285 .destroy = internal_dev_destroy,
286 .send = internal_dev_recv,
287};
288
289int ovs_is_internal_dev(const struct net_device *netdev)
290{
291 return netdev->netdev_ops == &internal_dev_netdev_ops;
292}
293
294struct vport *ovs_internal_dev_get_vport(struct net_device *netdev)
295{
296 if (!ovs_is_internal_dev(netdev))
297 return NULL;
298
299 return internal_dev_priv(netdev)->vport;
300}
301
302int ovs_internal_dev_rtnl_link_register(void)
303{
304 int err;
305
306 err = rtnl_link_register(&internal_dev_link_ops);
307 if (err < 0)
308 return err;
309
310 err = ovs_vport_ops_register(&ovs_internal_vport_ops);
311 if (err < 0)
312 rtnl_link_unregister(&internal_dev_link_ops);
313
314 return err;
315}
316
317void ovs_internal_dev_rtnl_link_unregister(void)
318{
319 ovs_vport_ops_unregister(&ovs_internal_vport_ops);
320 rtnl_link_unregister(&internal_dev_link_ops);
321}