1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * Copyright (c) 2007-2014 Nicira, Inc.
4 */
5
6#include <linux/etherdevice.h>
7#include <linux/if.h>
8#include <linux/if_vlan.h>
9#include <linux/jhash.h>
10#include <linux/kernel.h>
11#include <linux/list.h>
12#include <linux/mutex.h>
13#include <linux/percpu.h>
14#include <linux/rcupdate.h>
15#include <linux/rtnetlink.h>
16#include <linux/compat.h>
17#include <net/net_namespace.h>
18#include <linux/module.h>
19
20#include "datapath.h"
21#include "vport.h"
22#include "vport-internal_dev.h"
23
24static LIST_HEAD(vport_ops_list);
25
26/* Protected by RCU read lock for reading, ovs_mutex for writing. */
27static struct hlist_head *dev_table;
28#define VPORT_HASH_BUCKETS 1024
29
30/**
31 *	ovs_vport_init - initialize vport subsystem
32 *
33 * Called at module load time to initialize the vport subsystem.
34 */
35int ovs_vport_init(void)
36{
37	dev_table = kcalloc(VPORT_HASH_BUCKETS, sizeof(struct hlist_head),
38			    GFP_KERNEL);
39	if (!dev_table)
40		return -ENOMEM;
41
42	return 0;
43}
44
45/**
46 *	ovs_vport_exit - shutdown vport subsystem
47 *
48 * Called at module exit time to shutdown the vport subsystem.
49 */
50void ovs_vport_exit(void)
51{
52	kfree(dev_table);
53}
54
55static struct hlist_head *hash_bucket(const struct net *net, const char *name)
56{
57	unsigned int hash = jhash(name, strlen(name), (unsigned long) net);
58	return &dev_table[hash & (VPORT_HASH_BUCKETS - 1)];
59}
60
61int __ovs_vport_ops_register(struct vport_ops *ops)
62{
63	int err = -EEXIST;
64	struct vport_ops *o;
65
66	ovs_lock();
67	list_for_each_entry(o, &vport_ops_list, list)
68		if (ops->type == o->type)
69			goto errout;
70
71	list_add_tail(&ops->list, &vport_ops_list);
72	err = 0;
73errout:
74	ovs_unlock();
75	return err;
76}
77EXPORT_SYMBOL_GPL(__ovs_vport_ops_register);
78
79void ovs_vport_ops_unregister(struct vport_ops *ops)
80{
81	ovs_lock();
82	list_del(&ops->list);
83	ovs_unlock();
84}
85EXPORT_SYMBOL_GPL(ovs_vport_ops_unregister);
86
87/**
88 *	ovs_vport_locate - find a port that has already been created
89 *
90 * @net: network namespace
91 * @name: name of port to find
92 *
93 * Must be called with ovs or RCU read lock.
94 */
95struct vport *ovs_vport_locate(const struct net *net, const char *name)
96{
97	struct hlist_head *bucket = hash_bucket(net, name);
98	struct vport *vport;
99
100	hlist_for_each_entry_rcu(vport, bucket, hash_node,
101				 lockdep_ovsl_is_held())
102		if (!strcmp(name, ovs_vport_name(vport)) &&
103		    net_eq(ovs_dp_get_net(vport->dp), net))
104			return vport;
105
106	return NULL;
107}
108
109/**
110 *	ovs_vport_alloc - allocate and initialize new vport
111 *
112 * @priv_size: Size of private data area to allocate.
113 * @ops: vport device ops
114 *
115 * Allocate and initialize a new vport defined by @ops.  The vport will contain
116 * a private data area of size @priv_size that can be accessed using
117 * vport_priv().  vports that are no longer needed should be released with
118 * vport_free().
119 */
120struct vport *ovs_vport_alloc(int priv_size, const struct vport_ops *ops,
121			      const struct vport_parms *parms)
122{
123	struct vport *vport;
124	size_t alloc_size;
125
126	alloc_size = sizeof(struct vport);
127	if (priv_size) {
128		alloc_size = ALIGN(alloc_size, VPORT_ALIGN);
129		alloc_size += priv_size;
130	}
131
132	vport = kzalloc(alloc_size, GFP_KERNEL);
133	if (!vport)
134		return ERR_PTR(-ENOMEM);
135
136	vport->dp = parms->dp;
137	vport->port_no = parms->port_no;
138	vport->ops = ops;
139	INIT_HLIST_NODE(&vport->dp_hash_node);
140
141	if (ovs_vport_set_upcall_portids(vport, parms->upcall_portids)) {
142		kfree(vport);
143		return ERR_PTR(-EINVAL);
144	}
145
146	return vport;
147}
148EXPORT_SYMBOL_GPL(ovs_vport_alloc);
149
150/**
151 *	ovs_vport_free - uninitialize and free vport
152 *
153 * @vport: vport to free
154 *
155 * Frees a vport allocated with vport_alloc() when it is no longer needed.
156 *
157 * The caller must ensure that an RCU grace period has passed since the last
158 * time @vport was in a datapath.
159 */
160void ovs_vport_free(struct vport *vport)
161{
162	/* vport is freed from RCU callback or error path, Therefore
163	 * it is safe to use raw dereference.
164	 */
165	kfree(rcu_dereference_raw(vport->upcall_portids));
166	kfree(vport);
167}
168EXPORT_SYMBOL_GPL(ovs_vport_free);
169
170static struct vport_ops *ovs_vport_lookup(const struct vport_parms *parms)
171{
172	struct vport_ops *ops;
173
174	list_for_each_entry(ops, &vport_ops_list, list)
175		if (ops->type == parms->type)
176			return ops;
177
178	return NULL;
179}
180
181/**
182 *	ovs_vport_add - add vport device (for kernel callers)
183 *
184 * @parms: Information about new vport.
185 *
186 * Creates a new vport with the specified configuration (which is dependent on
187 * device type).  ovs_mutex must be held.
188 */
189struct vport *ovs_vport_add(const struct vport_parms *parms)
190{
191	struct vport_ops *ops;
192	struct vport *vport;
193
194	ops = ovs_vport_lookup(parms);
195	if (ops) {
196		struct hlist_head *bucket;
197
198		if (!try_module_get(ops->owner))
199			return ERR_PTR(-EAFNOSUPPORT);
200
201		vport = ops->create(parms);
202		if (IS_ERR(vport)) {
203			module_put(ops->owner);
204			return vport;
205		}
206
207		bucket = hash_bucket(ovs_dp_get_net(vport->dp),
208				     ovs_vport_name(vport));
209		hlist_add_head_rcu(&vport->hash_node, bucket);
210		return vport;
211	}
212
213	/* Unlock to attempt module load and return -EAGAIN if load
214	 * was successful as we need to restart the port addition
215	 * workflow.
216	 */
217	ovs_unlock();
218	request_module("vport-type-%d", parms->type);
219	ovs_lock();
220
221	if (!ovs_vport_lookup(parms))
222		return ERR_PTR(-EAFNOSUPPORT);
223	else
224		return ERR_PTR(-EAGAIN);
225}
226
227/**
228 *	ovs_vport_set_options - modify existing vport device (for kernel callers)
229 *
230 * @vport: vport to modify.
231 * @options: New configuration.
232 *
233 * Modifies an existing device with the specified configuration (which is
234 * dependent on device type).  ovs_mutex must be held.
235 */
236int ovs_vport_set_options(struct vport *vport, struct nlattr *options)
237{
238	if (!vport->ops->set_options)
239		return -EOPNOTSUPP;
240	return vport->ops->set_options(vport, options);
241}
242
243/**
244 *	ovs_vport_del - delete existing vport device
245 *
246 * @vport: vport to delete.
247 *
248 * Detaches @vport from its datapath and destroys it.  ovs_mutex must
249 * be held.
250 */
251void ovs_vport_del(struct vport *vport)
252{
253	hlist_del_rcu(&vport->hash_node);
254	module_put(vport->ops->owner);
255	vport->ops->destroy(vport);
256}
257
258/**
259 *	ovs_vport_get_stats - retrieve device stats
260 *
261 * @vport: vport from which to retrieve the stats
262 * @stats: location to store stats
263 *
264 * Retrieves transmit, receive, and error stats for the given device.
265 *
266 * Must be called with ovs_mutex or rcu_read_lock.
267 */
268void ovs_vport_get_stats(struct vport *vport, struct ovs_vport_stats *stats)
269{
270	const struct rtnl_link_stats64 *dev_stats;
271	struct rtnl_link_stats64 temp;
272
273	dev_stats = dev_get_stats(vport->dev, &temp);
274	stats->rx_errors  = dev_stats->rx_errors;
275	stats->tx_errors  = dev_stats->tx_errors;
276	stats->tx_dropped = dev_stats->tx_dropped;
277	stats->rx_dropped = dev_stats->rx_dropped;
278
279	stats->rx_bytes	  = dev_stats->rx_bytes;
280	stats->rx_packets = dev_stats->rx_packets;
281	stats->tx_bytes	  = dev_stats->tx_bytes;
282	stats->tx_packets = dev_stats->tx_packets;
283}
284
285/**
286 *	ovs_vport_get_options - retrieve device options
287 *
288 * @vport: vport from which to retrieve the options.
289 * @skb: sk_buff where options should be appended.
290 *
291 * Retrieves the configuration of the given device, appending an
292 * %OVS_VPORT_ATTR_OPTIONS attribute that in turn contains nested
293 * vport-specific attributes to @skb.
294 *
295 * Returns 0 if successful, -EMSGSIZE if @skb has insufficient room, or another
296 * negative error code if a real error occurred.  If an error occurs, @skb is
297 * left unmodified.
298 *
299 * Must be called with ovs_mutex or rcu_read_lock.
300 */
301int ovs_vport_get_options(const struct vport *vport, struct sk_buff *skb)
302{
303	struct nlattr *nla;
304	int err;
305
306	if (!vport->ops->get_options)
307		return 0;
308
309	nla = nla_nest_start_noflag(skb, OVS_VPORT_ATTR_OPTIONS);
310	if (!nla)
311		return -EMSGSIZE;
312
313	err = vport->ops->get_options(vport, skb);
314	if (err) {
315		nla_nest_cancel(skb, nla);
316		return err;
317	}
318
319	nla_nest_end(skb, nla);
320	return 0;
321}
322
323/**
324 *	ovs_vport_set_upcall_portids - set upcall portids of @vport.
325 *
326 * @vport: vport to modify.
327 * @ids: new configuration, an array of port ids.
328 *
329 * Sets the vport's upcall_portids to @ids.
330 *
331 * Returns 0 if successful, -EINVAL if @ids is zero length or cannot be parsed
332 * as an array of U32.
333 *
334 * Must be called with ovs_mutex.
335 */
336int ovs_vport_set_upcall_portids(struct vport *vport, const struct nlattr *ids)
337{
338	struct vport_portids *old, *vport_portids;
339
340	if (!nla_len(ids) || nla_len(ids) % sizeof(u32))
341		return -EINVAL;
342
343	old = ovsl_dereference(vport->upcall_portids);
344
345	vport_portids = kmalloc(sizeof(*vport_portids) + nla_len(ids),
346				GFP_KERNEL);
347	if (!vport_portids)
348		return -ENOMEM;
349
350	vport_portids->n_ids = nla_len(ids) / sizeof(u32);
351	vport_portids->rn_ids = reciprocal_value(vport_portids->n_ids);
352	nla_memcpy(vport_portids->ids, ids, nla_len(ids));
353
354	rcu_assign_pointer(vport->upcall_portids, vport_portids);
355
356	if (old)
357		kfree_rcu(old, rcu);
358	return 0;
359}
360
361/**
362 *	ovs_vport_get_upcall_portids - get the upcall_portids of @vport.
363 *
364 * @vport: vport from which to retrieve the portids.
365 * @skb: sk_buff where portids should be appended.
366 *
367 * Retrieves the configuration of the given vport, appending the
368 * %OVS_VPORT_ATTR_UPCALL_PID attribute which is the array of upcall
369 * portids to @skb.
370 *
371 * Returns 0 if successful, -EMSGSIZE if @skb has insufficient room.
372 * If an error occurs, @skb is left unmodified.  Must be called with
373 * ovs_mutex or rcu_read_lock.
374 */
375int ovs_vport_get_upcall_portids(const struct vport *vport,
376				 struct sk_buff *skb)
377{
378	struct vport_portids *ids;
379
380	ids = rcu_dereference_ovsl(vport->upcall_portids);
381
382	if (vport->dp->user_features & OVS_DP_F_VPORT_PIDS)
383		return nla_put(skb, OVS_VPORT_ATTR_UPCALL_PID,
384			       ids->n_ids * sizeof(u32), (void *)ids->ids);
385	else
386		return nla_put_u32(skb, OVS_VPORT_ATTR_UPCALL_PID, ids->ids[0]);
387}
388
389/**
390 *	ovs_vport_find_upcall_portid - find the upcall portid to send upcall.
391 *
392 * @vport: vport from which the missed packet is received.
393 * @skb: skb that the missed packet was received.
394 *
395 * Uses the skb_get_hash() to select the upcall portid to send the
396 * upcall.
397 *
398 * Returns the portid of the target socket.  Must be called with rcu_read_lock.
399 */
400u32 ovs_vport_find_upcall_portid(const struct vport *vport,
401				 struct sk_buff *skb)
402{
403	struct vport_portids *ids;
404	u32 ids_index;
405	u32 hash;
406
407	ids = rcu_dereference(vport->upcall_portids);
408
409	/* If there is only one portid, select it in the fast-path. */
410	if (ids->n_ids == 1)
411		return ids->ids[0];
412
413	hash = skb_get_hash(skb);
414	ids_index = hash - ids->n_ids * reciprocal_divide(hash, ids->rn_ids);
415	return ids->ids[ids_index];
416}
417
418/**
419 *	ovs_vport_receive - pass up received packet to the datapath for processing
420 *
421 * @vport: vport that received the packet
422 * @skb: skb that was received
423 * @tun_info: tunnel (if any) that carried packet
424 *
425 * Must be called with rcu_read_lock.  The packet cannot be shared and
426 * skb->data should point to the Ethernet header.
427 */
428int ovs_vport_receive(struct vport *vport, struct sk_buff *skb,
429		      const struct ip_tunnel_info *tun_info)
430{
431	struct sw_flow_key key;
432	int error;
433
434	OVS_CB(skb)->input_vport = vport;
435	OVS_CB(skb)->mru = 0;
436	OVS_CB(skb)->cutlen = 0;
437	if (unlikely(dev_net(skb->dev) != ovs_dp_get_net(vport->dp))) {
438		u32 mark;
439
440		mark = skb->mark;
441		skb_scrub_packet(skb, true);
442		skb->mark = mark;
443		tun_info = NULL;
444	}
445
446	/* Extract flow from 'skb' into 'key'. */
447	error = ovs_flow_key_extract(tun_info, skb, &key);
448	if (unlikely(error)) {
449		kfree_skb(skb);
450		return error;
451	}
452	ovs_dp_process_packet(skb, &key);
453	return 0;
454}
455
456static int packet_length(const struct sk_buff *skb,
457			 struct net_device *dev)
458{
459	int length = skb->len - dev->hard_header_len;
460
461	if (!skb_vlan_tag_present(skb) &&
462	    eth_type_vlan(skb->protocol))
463		length -= VLAN_HLEN;
464
465	/* Don't subtract for multiple VLAN tags. Most (all?) drivers allow
466	 * (ETH_LEN + VLAN_HLEN) in addition to the mtu value, but almost none
467	 * account for 802.1ad. e.g. is_skb_forwardable().
468	 */
469
470	return length > 0 ? length : 0;
471}
472
473void ovs_vport_send(struct vport *vport, struct sk_buff *skb, u8 mac_proto)
474{
475	int mtu = vport->dev->mtu;
476
477	switch (vport->dev->type) {
478	case ARPHRD_NONE:
479		if (mac_proto == MAC_PROTO_ETHERNET) {
480			skb_reset_network_header(skb);
481			skb_reset_mac_len(skb);
482			skb->protocol = htons(ETH_P_TEB);
483		} else if (mac_proto != MAC_PROTO_NONE) {
484			WARN_ON_ONCE(1);
485			goto drop;
486		}
487		break;
488	case ARPHRD_ETHER:
489		if (mac_proto != MAC_PROTO_ETHERNET)
490			goto drop;
491		break;
492	default:
493		goto drop;
494	}
495
496	if (unlikely(packet_length(skb, vport->dev) > mtu &&
497		     !skb_is_gso(skb))) {
498		net_warn_ratelimited("%s: dropped over-mtu packet: %d > %d\n",
499				     vport->dev->name,
500				     packet_length(skb, vport->dev), mtu);
501		vport->dev->stats.tx_errors++;
502		goto drop;
503	}
504
505	skb->dev = vport->dev;
506	skb->tstamp = 0;
507	vport->ops->send(skb);
508	return;
509
510drop:
511	kfree_skb(skb);
512}
513