Loading...
Note: File does not exist in v3.5.6.
1#include <linux/types.h>
2#include <linux/skbuff.h>
3#include <linux/socket.h>
4#include <linux/sysctl.h>
5#include <linux/net.h>
6#include <linux/module.h>
7#include <linux/if_arp.h>
8#include <linux/ipv6.h>
9#include <linux/mpls.h>
10#include <linux/vmalloc.h>
11#include <net/ip.h>
12#include <net/dst.h>
13#include <net/sock.h>
14#include <net/arp.h>
15#include <net/ip_fib.h>
16#include <net/netevent.h>
17#include <net/netns/generic.h>
18#if IS_ENABLED(CONFIG_IPV6)
19#include <net/ipv6.h>
20#include <net/addrconf.h>
21#endif
22#include <net/nexthop.h>
23#include "internal.h"
24
25/* Maximum number of labels to look ahead at when selecting a path of
26 * a multipath route
27 */
28#define MAX_MP_SELECT_LABELS 4
29
30#define MPLS_NEIGH_TABLE_UNSPEC (NEIGH_LINK_TABLE + 1)
31
32static int zero = 0;
33static int label_limit = (1 << 20) - 1;
34
35static void rtmsg_lfib(int event, u32 label, struct mpls_route *rt,
36 struct nlmsghdr *nlh, struct net *net, u32 portid,
37 unsigned int nlm_flags);
38
39static struct mpls_route *mpls_route_input_rcu(struct net *net, unsigned index)
40{
41 struct mpls_route *rt = NULL;
42
43 if (index < net->mpls.platform_labels) {
44 struct mpls_route __rcu **platform_label =
45 rcu_dereference(net->mpls.platform_label);
46 rt = rcu_dereference(platform_label[index]);
47 }
48 return rt;
49}
50
51static inline struct mpls_dev *mpls_dev_get(const struct net_device *dev)
52{
53 return rcu_dereference_rtnl(dev->mpls_ptr);
54}
55
56bool mpls_output_possible(const struct net_device *dev)
57{
58 return dev && (dev->flags & IFF_UP) && netif_carrier_ok(dev);
59}
60EXPORT_SYMBOL_GPL(mpls_output_possible);
61
62static u8 *__mpls_nh_via(struct mpls_route *rt, struct mpls_nh *nh)
63{
64 u8 *nh0_via = PTR_ALIGN((u8 *)&rt->rt_nh[rt->rt_nhn], VIA_ALEN_ALIGN);
65 int nh_index = nh - rt->rt_nh;
66
67 return nh0_via + rt->rt_max_alen * nh_index;
68}
69
70static const u8 *mpls_nh_via(const struct mpls_route *rt,
71 const struct mpls_nh *nh)
72{
73 return __mpls_nh_via((struct mpls_route *)rt, (struct mpls_nh *)nh);
74}
75
76static unsigned int mpls_nh_header_size(const struct mpls_nh *nh)
77{
78 /* The size of the layer 2.5 labels to be added for this route */
79 return nh->nh_labels * sizeof(struct mpls_shim_hdr);
80}
81
82unsigned int mpls_dev_mtu(const struct net_device *dev)
83{
84 /* The amount of data the layer 2 frame can hold */
85 return dev->mtu;
86}
87EXPORT_SYMBOL_GPL(mpls_dev_mtu);
88
89bool mpls_pkt_too_big(const struct sk_buff *skb, unsigned int mtu)
90{
91 if (skb->len <= mtu)
92 return false;
93
94 if (skb_is_gso(skb) && skb_gso_network_seglen(skb) <= mtu)
95 return false;
96
97 return true;
98}
99EXPORT_SYMBOL_GPL(mpls_pkt_too_big);
100
101static u32 mpls_multipath_hash(struct mpls_route *rt,
102 struct sk_buff *skb, bool bos)
103{
104 struct mpls_entry_decoded dec;
105 struct mpls_shim_hdr *hdr;
106 bool eli_seen = false;
107 int label_index;
108 u32 hash = 0;
109
110 for (label_index = 0; label_index < MAX_MP_SELECT_LABELS && !bos;
111 label_index++) {
112 if (!pskb_may_pull(skb, sizeof(*hdr) * label_index))
113 break;
114
115 /* Read and decode the current label */
116 hdr = mpls_hdr(skb) + label_index;
117 dec = mpls_entry_decode(hdr);
118
119 /* RFC6790 - reserved labels MUST NOT be used as keys
120 * for the load-balancing function
121 */
122 if (likely(dec.label >= MPLS_LABEL_FIRST_UNRESERVED)) {
123 hash = jhash_1word(dec.label, hash);
124
125 /* The entropy label follows the entropy label
126 * indicator, so this means that the entropy
127 * label was just added to the hash - no need to
128 * go any deeper either in the label stack or in the
129 * payload
130 */
131 if (eli_seen)
132 break;
133 } else if (dec.label == MPLS_LABEL_ENTROPY) {
134 eli_seen = true;
135 }
136
137 bos = dec.bos;
138 if (bos && pskb_may_pull(skb, sizeof(*hdr) * label_index +
139 sizeof(struct iphdr))) {
140 const struct iphdr *v4hdr;
141
142 v4hdr = (const struct iphdr *)(mpls_hdr(skb) +
143 label_index);
144 if (v4hdr->version == 4) {
145 hash = jhash_3words(ntohl(v4hdr->saddr),
146 ntohl(v4hdr->daddr),
147 v4hdr->protocol, hash);
148 } else if (v4hdr->version == 6 &&
149 pskb_may_pull(skb, sizeof(*hdr) * label_index +
150 sizeof(struct ipv6hdr))) {
151 const struct ipv6hdr *v6hdr;
152
153 v6hdr = (const struct ipv6hdr *)(mpls_hdr(skb) +
154 label_index);
155
156 hash = __ipv6_addr_jhash(&v6hdr->saddr, hash);
157 hash = __ipv6_addr_jhash(&v6hdr->daddr, hash);
158 hash = jhash_1word(v6hdr->nexthdr, hash);
159 }
160 }
161 }
162
163 return hash;
164}
165
166static struct mpls_nh *mpls_select_multipath(struct mpls_route *rt,
167 struct sk_buff *skb, bool bos)
168{
169 int alive = ACCESS_ONCE(rt->rt_nhn_alive);
170 u32 hash = 0;
171 int nh_index = 0;
172 int n = 0;
173
174 /* No need to look further into packet if there's only
175 * one path
176 */
177 if (rt->rt_nhn == 1)
178 goto out;
179
180 if (alive <= 0)
181 return NULL;
182
183 hash = mpls_multipath_hash(rt, skb, bos);
184 nh_index = hash % alive;
185 if (alive == rt->rt_nhn)
186 goto out;
187 for_nexthops(rt) {
188 if (nh->nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
189 continue;
190 if (n == nh_index)
191 return nh;
192 n++;
193 } endfor_nexthops(rt);
194
195out:
196 return &rt->rt_nh[nh_index];
197}
198
199static bool mpls_egress(struct mpls_route *rt, struct sk_buff *skb,
200 struct mpls_entry_decoded dec)
201{
202 enum mpls_payload_type payload_type;
203 bool success = false;
204
205 /* The IPv4 code below accesses through the IPv4 header
206 * checksum, which is 12 bytes into the packet.
207 * The IPv6 code below accesses through the IPv6 hop limit
208 * which is 8 bytes into the packet.
209 *
210 * For all supported cases there should always be at least 12
211 * bytes of packet data present. The IPv4 header is 20 bytes
212 * without options and the IPv6 header is always 40 bytes
213 * long.
214 */
215 if (!pskb_may_pull(skb, 12))
216 return false;
217
218 payload_type = rt->rt_payload_type;
219 if (payload_type == MPT_UNSPEC)
220 payload_type = ip_hdr(skb)->version;
221
222 switch (payload_type) {
223 case MPT_IPV4: {
224 struct iphdr *hdr4 = ip_hdr(skb);
225 skb->protocol = htons(ETH_P_IP);
226 csum_replace2(&hdr4->check,
227 htons(hdr4->ttl << 8),
228 htons(dec.ttl << 8));
229 hdr4->ttl = dec.ttl;
230 success = true;
231 break;
232 }
233 case MPT_IPV6: {
234 struct ipv6hdr *hdr6 = ipv6_hdr(skb);
235 skb->protocol = htons(ETH_P_IPV6);
236 hdr6->hop_limit = dec.ttl;
237 success = true;
238 break;
239 }
240 case MPT_UNSPEC:
241 break;
242 }
243
244 return success;
245}
246
247static int mpls_forward(struct sk_buff *skb, struct net_device *dev,
248 struct packet_type *pt, struct net_device *orig_dev)
249{
250 struct net *net = dev_net(dev);
251 struct mpls_shim_hdr *hdr;
252 struct mpls_route *rt;
253 struct mpls_nh *nh;
254 struct mpls_entry_decoded dec;
255 struct net_device *out_dev;
256 struct mpls_dev *mdev;
257 unsigned int hh_len;
258 unsigned int new_header_size;
259 unsigned int mtu;
260 int err;
261
262 /* Careful this entire function runs inside of an rcu critical section */
263
264 mdev = mpls_dev_get(dev);
265 if (!mdev || !mdev->input_enabled)
266 goto drop;
267
268 if (skb->pkt_type != PACKET_HOST)
269 goto drop;
270
271 if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
272 goto drop;
273
274 if (!pskb_may_pull(skb, sizeof(*hdr)))
275 goto drop;
276
277 /* Read and decode the label */
278 hdr = mpls_hdr(skb);
279 dec = mpls_entry_decode(hdr);
280
281 /* Pop the label */
282 skb_pull(skb, sizeof(*hdr));
283 skb_reset_network_header(skb);
284
285 skb_orphan(skb);
286
287 rt = mpls_route_input_rcu(net, dec.label);
288 if (!rt)
289 goto drop;
290
291 nh = mpls_select_multipath(rt, skb, dec.bos);
292 if (!nh)
293 goto drop;
294
295 /* Find the output device */
296 out_dev = rcu_dereference(nh->nh_dev);
297 if (!mpls_output_possible(out_dev))
298 goto drop;
299
300 if (skb_warn_if_lro(skb))
301 goto drop;
302
303 skb_forward_csum(skb);
304
305 /* Verify ttl is valid */
306 if (dec.ttl <= 1)
307 goto drop;
308 dec.ttl -= 1;
309
310 /* Verify the destination can hold the packet */
311 new_header_size = mpls_nh_header_size(nh);
312 mtu = mpls_dev_mtu(out_dev);
313 if (mpls_pkt_too_big(skb, mtu - new_header_size))
314 goto drop;
315
316 hh_len = LL_RESERVED_SPACE(out_dev);
317 if (!out_dev->header_ops)
318 hh_len = 0;
319
320 /* Ensure there is enough space for the headers in the skb */
321 if (skb_cow(skb, hh_len + new_header_size))
322 goto drop;
323
324 skb->dev = out_dev;
325 skb->protocol = htons(ETH_P_MPLS_UC);
326
327 if (unlikely(!new_header_size && dec.bos)) {
328 /* Penultimate hop popping */
329 if (!mpls_egress(rt, skb, dec))
330 goto drop;
331 } else {
332 bool bos;
333 int i;
334 skb_push(skb, new_header_size);
335 skb_reset_network_header(skb);
336 /* Push the new labels */
337 hdr = mpls_hdr(skb);
338 bos = dec.bos;
339 for (i = nh->nh_labels - 1; i >= 0; i--) {
340 hdr[i] = mpls_entry_encode(nh->nh_label[i],
341 dec.ttl, 0, bos);
342 bos = false;
343 }
344 }
345
346 /* If via wasn't specified then send out using device address */
347 if (nh->nh_via_table == MPLS_NEIGH_TABLE_UNSPEC)
348 err = neigh_xmit(NEIGH_LINK_TABLE, out_dev,
349 out_dev->dev_addr, skb);
350 else
351 err = neigh_xmit(nh->nh_via_table, out_dev,
352 mpls_nh_via(rt, nh), skb);
353 if (err)
354 net_dbg_ratelimited("%s: packet transmission failed: %d\n",
355 __func__, err);
356 return 0;
357
358drop:
359 kfree_skb(skb);
360 return NET_RX_DROP;
361}
362
363static struct packet_type mpls_packet_type __read_mostly = {
364 .type = cpu_to_be16(ETH_P_MPLS_UC),
365 .func = mpls_forward,
366};
367
368static const struct nla_policy rtm_mpls_policy[RTA_MAX+1] = {
369 [RTA_DST] = { .type = NLA_U32 },
370 [RTA_OIF] = { .type = NLA_U32 },
371};
372
373struct mpls_route_config {
374 u32 rc_protocol;
375 u32 rc_ifindex;
376 u8 rc_via_table;
377 u8 rc_via_alen;
378 u8 rc_via[MAX_VIA_ALEN];
379 u32 rc_label;
380 u8 rc_output_labels;
381 u32 rc_output_label[MAX_NEW_LABELS];
382 u32 rc_nlflags;
383 enum mpls_payload_type rc_payload_type;
384 struct nl_info rc_nlinfo;
385 struct rtnexthop *rc_mp;
386 int rc_mp_len;
387};
388
389static struct mpls_route *mpls_rt_alloc(int num_nh, u8 max_alen)
390{
391 u8 max_alen_aligned = ALIGN(max_alen, VIA_ALEN_ALIGN);
392 struct mpls_route *rt;
393
394 rt = kzalloc(ALIGN(sizeof(*rt) + num_nh * sizeof(*rt->rt_nh),
395 VIA_ALEN_ALIGN) +
396 num_nh * max_alen_aligned,
397 GFP_KERNEL);
398 if (rt) {
399 rt->rt_nhn = num_nh;
400 rt->rt_nhn_alive = num_nh;
401 rt->rt_max_alen = max_alen_aligned;
402 }
403
404 return rt;
405}
406
407static void mpls_rt_free(struct mpls_route *rt)
408{
409 if (rt)
410 kfree_rcu(rt, rt_rcu);
411}
412
413static void mpls_notify_route(struct net *net, unsigned index,
414 struct mpls_route *old, struct mpls_route *new,
415 const struct nl_info *info)
416{
417 struct nlmsghdr *nlh = info ? info->nlh : NULL;
418 unsigned portid = info ? info->portid : 0;
419 int event = new ? RTM_NEWROUTE : RTM_DELROUTE;
420 struct mpls_route *rt = new ? new : old;
421 unsigned nlm_flags = (old && new) ? NLM_F_REPLACE : 0;
422 /* Ignore reserved labels for now */
423 if (rt && (index >= MPLS_LABEL_FIRST_UNRESERVED))
424 rtmsg_lfib(event, index, rt, nlh, net, portid, nlm_flags);
425}
426
427static void mpls_route_update(struct net *net, unsigned index,
428 struct mpls_route *new,
429 const struct nl_info *info)
430{
431 struct mpls_route __rcu **platform_label;
432 struct mpls_route *rt;
433
434 ASSERT_RTNL();
435
436 platform_label = rtnl_dereference(net->mpls.platform_label);
437 rt = rtnl_dereference(platform_label[index]);
438 rcu_assign_pointer(platform_label[index], new);
439
440 mpls_notify_route(net, index, rt, new, info);
441
442 /* If we removed a route free it now */
443 mpls_rt_free(rt);
444}
445
446static unsigned find_free_label(struct net *net)
447{
448 struct mpls_route __rcu **platform_label;
449 size_t platform_labels;
450 unsigned index;
451
452 platform_label = rtnl_dereference(net->mpls.platform_label);
453 platform_labels = net->mpls.platform_labels;
454 for (index = MPLS_LABEL_FIRST_UNRESERVED; index < platform_labels;
455 index++) {
456 if (!rtnl_dereference(platform_label[index]))
457 return index;
458 }
459 return LABEL_NOT_SPECIFIED;
460}
461
462#if IS_ENABLED(CONFIG_INET)
463static struct net_device *inet_fib_lookup_dev(struct net *net,
464 const void *addr)
465{
466 struct net_device *dev;
467 struct rtable *rt;
468 struct in_addr daddr;
469
470 memcpy(&daddr, addr, sizeof(struct in_addr));
471 rt = ip_route_output(net, daddr.s_addr, 0, 0, 0);
472 if (IS_ERR(rt))
473 return ERR_CAST(rt);
474
475 dev = rt->dst.dev;
476 dev_hold(dev);
477
478 ip_rt_put(rt);
479
480 return dev;
481}
482#else
483static struct net_device *inet_fib_lookup_dev(struct net *net,
484 const void *addr)
485{
486 return ERR_PTR(-EAFNOSUPPORT);
487}
488#endif
489
490#if IS_ENABLED(CONFIG_IPV6)
491static struct net_device *inet6_fib_lookup_dev(struct net *net,
492 const void *addr)
493{
494 struct net_device *dev;
495 struct dst_entry *dst;
496 struct flowi6 fl6;
497 int err;
498
499 if (!ipv6_stub)
500 return ERR_PTR(-EAFNOSUPPORT);
501
502 memset(&fl6, 0, sizeof(fl6));
503 memcpy(&fl6.daddr, addr, sizeof(struct in6_addr));
504 err = ipv6_stub->ipv6_dst_lookup(net, NULL, &dst, &fl6);
505 if (err)
506 return ERR_PTR(err);
507
508 dev = dst->dev;
509 dev_hold(dev);
510 dst_release(dst);
511
512 return dev;
513}
514#else
515static struct net_device *inet6_fib_lookup_dev(struct net *net,
516 const void *addr)
517{
518 return ERR_PTR(-EAFNOSUPPORT);
519}
520#endif
521
522static struct net_device *find_outdev(struct net *net,
523 struct mpls_route *rt,
524 struct mpls_nh *nh, int oif)
525{
526 struct net_device *dev = NULL;
527
528 if (!oif) {
529 switch (nh->nh_via_table) {
530 case NEIGH_ARP_TABLE:
531 dev = inet_fib_lookup_dev(net, mpls_nh_via(rt, nh));
532 break;
533 case NEIGH_ND_TABLE:
534 dev = inet6_fib_lookup_dev(net, mpls_nh_via(rt, nh));
535 break;
536 case NEIGH_LINK_TABLE:
537 break;
538 }
539 } else {
540 dev = dev_get_by_index(net, oif);
541 }
542
543 if (!dev)
544 return ERR_PTR(-ENODEV);
545
546 if (IS_ERR(dev))
547 return dev;
548
549 /* The caller is holding rtnl anyways, so release the dev reference */
550 dev_put(dev);
551
552 return dev;
553}
554
555static int mpls_nh_assign_dev(struct net *net, struct mpls_route *rt,
556 struct mpls_nh *nh, int oif)
557{
558 struct net_device *dev = NULL;
559 int err = -ENODEV;
560
561 dev = find_outdev(net, rt, nh, oif);
562 if (IS_ERR(dev)) {
563 err = PTR_ERR(dev);
564 dev = NULL;
565 goto errout;
566 }
567
568 /* Ensure this is a supported device */
569 err = -EINVAL;
570 if (!mpls_dev_get(dev))
571 goto errout;
572
573 if ((nh->nh_via_table == NEIGH_LINK_TABLE) &&
574 (dev->addr_len != nh->nh_via_alen))
575 goto errout;
576
577 RCU_INIT_POINTER(nh->nh_dev, dev);
578
579 if (!(dev->flags & IFF_UP)) {
580 nh->nh_flags |= RTNH_F_DEAD;
581 } else {
582 unsigned int flags;
583
584 flags = dev_get_flags(dev);
585 if (!(flags & (IFF_RUNNING | IFF_LOWER_UP)))
586 nh->nh_flags |= RTNH_F_LINKDOWN;
587 }
588
589 return 0;
590
591errout:
592 return err;
593}
594
595static int mpls_nh_build_from_cfg(struct mpls_route_config *cfg,
596 struct mpls_route *rt)
597{
598 struct net *net = cfg->rc_nlinfo.nl_net;
599 struct mpls_nh *nh = rt->rt_nh;
600 int err;
601 int i;
602
603 if (!nh)
604 return -ENOMEM;
605
606 err = -EINVAL;
607 /* Ensure only a supported number of labels are present */
608 if (cfg->rc_output_labels > MAX_NEW_LABELS)
609 goto errout;
610
611 nh->nh_labels = cfg->rc_output_labels;
612 for (i = 0; i < nh->nh_labels; i++)
613 nh->nh_label[i] = cfg->rc_output_label[i];
614
615 nh->nh_via_table = cfg->rc_via_table;
616 memcpy(__mpls_nh_via(rt, nh), cfg->rc_via, cfg->rc_via_alen);
617 nh->nh_via_alen = cfg->rc_via_alen;
618
619 err = mpls_nh_assign_dev(net, rt, nh, cfg->rc_ifindex);
620 if (err)
621 goto errout;
622
623 if (nh->nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
624 rt->rt_nhn_alive--;
625
626 return 0;
627
628errout:
629 return err;
630}
631
632static int mpls_nh_build(struct net *net, struct mpls_route *rt,
633 struct mpls_nh *nh, int oif, struct nlattr *via,
634 struct nlattr *newdst)
635{
636 int err = -ENOMEM;
637
638 if (!nh)
639 goto errout;
640
641 if (newdst) {
642 err = nla_get_labels(newdst, MAX_NEW_LABELS,
643 &nh->nh_labels, nh->nh_label);
644 if (err)
645 goto errout;
646 }
647
648 if (via) {
649 err = nla_get_via(via, &nh->nh_via_alen, &nh->nh_via_table,
650 __mpls_nh_via(rt, nh));
651 if (err)
652 goto errout;
653 } else {
654 nh->nh_via_table = MPLS_NEIGH_TABLE_UNSPEC;
655 }
656
657 err = mpls_nh_assign_dev(net, rt, nh, oif);
658 if (err)
659 goto errout;
660
661 return 0;
662
663errout:
664 return err;
665}
666
667static int mpls_count_nexthops(struct rtnexthop *rtnh, int len,
668 u8 cfg_via_alen, u8 *max_via_alen)
669{
670 int nhs = 0;
671 int remaining = len;
672
673 if (!rtnh) {
674 *max_via_alen = cfg_via_alen;
675 return 1;
676 }
677
678 *max_via_alen = 0;
679
680 while (rtnh_ok(rtnh, remaining)) {
681 struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
682 int attrlen;
683
684 attrlen = rtnh_attrlen(rtnh);
685 nla = nla_find(attrs, attrlen, RTA_VIA);
686 if (nla && nla_len(nla) >=
687 offsetof(struct rtvia, rtvia_addr)) {
688 int via_alen = nla_len(nla) -
689 offsetof(struct rtvia, rtvia_addr);
690
691 if (via_alen <= MAX_VIA_ALEN)
692 *max_via_alen = max_t(u16, *max_via_alen,
693 via_alen);
694 }
695
696 nhs++;
697 rtnh = rtnh_next(rtnh, &remaining);
698 }
699
700 /* leftover implies invalid nexthop configuration, discard it */
701 return remaining > 0 ? 0 : nhs;
702}
703
704static int mpls_nh_build_multi(struct mpls_route_config *cfg,
705 struct mpls_route *rt)
706{
707 struct rtnexthop *rtnh = cfg->rc_mp;
708 struct nlattr *nla_via, *nla_newdst;
709 int remaining = cfg->rc_mp_len;
710 int nhs = 0;
711 int err = 0;
712
713 change_nexthops(rt) {
714 int attrlen;
715
716 nla_via = NULL;
717 nla_newdst = NULL;
718
719 err = -EINVAL;
720 if (!rtnh_ok(rtnh, remaining))
721 goto errout;
722
723 /* neither weighted multipath nor any flags
724 * are supported
725 */
726 if (rtnh->rtnh_hops || rtnh->rtnh_flags)
727 goto errout;
728
729 attrlen = rtnh_attrlen(rtnh);
730 if (attrlen > 0) {
731 struct nlattr *attrs = rtnh_attrs(rtnh);
732
733 nla_via = nla_find(attrs, attrlen, RTA_VIA);
734 nla_newdst = nla_find(attrs, attrlen, RTA_NEWDST);
735 }
736
737 err = mpls_nh_build(cfg->rc_nlinfo.nl_net, rt, nh,
738 rtnh->rtnh_ifindex, nla_via, nla_newdst);
739 if (err)
740 goto errout;
741
742 if (nh->nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
743 rt->rt_nhn_alive--;
744
745 rtnh = rtnh_next(rtnh, &remaining);
746 nhs++;
747 } endfor_nexthops(rt);
748
749 rt->rt_nhn = nhs;
750
751 return 0;
752
753errout:
754 return err;
755}
756
757static int mpls_route_add(struct mpls_route_config *cfg)
758{
759 struct mpls_route __rcu **platform_label;
760 struct net *net = cfg->rc_nlinfo.nl_net;
761 struct mpls_route *rt, *old;
762 int err = -EINVAL;
763 u8 max_via_alen;
764 unsigned index;
765 int nhs;
766
767 index = cfg->rc_label;
768
769 /* If a label was not specified during insert pick one */
770 if ((index == LABEL_NOT_SPECIFIED) &&
771 (cfg->rc_nlflags & NLM_F_CREATE)) {
772 index = find_free_label(net);
773 }
774
775 /* Reserved labels may not be set */
776 if (index < MPLS_LABEL_FIRST_UNRESERVED)
777 goto errout;
778
779 /* The full 20 bit range may not be supported. */
780 if (index >= net->mpls.platform_labels)
781 goto errout;
782
783 /* Append makes no sense with mpls */
784 err = -EOPNOTSUPP;
785 if (cfg->rc_nlflags & NLM_F_APPEND)
786 goto errout;
787
788 err = -EEXIST;
789 platform_label = rtnl_dereference(net->mpls.platform_label);
790 old = rtnl_dereference(platform_label[index]);
791 if ((cfg->rc_nlflags & NLM_F_EXCL) && old)
792 goto errout;
793
794 err = -EEXIST;
795 if (!(cfg->rc_nlflags & NLM_F_REPLACE) && old)
796 goto errout;
797
798 err = -ENOENT;
799 if (!(cfg->rc_nlflags & NLM_F_CREATE) && !old)
800 goto errout;
801
802 err = -EINVAL;
803 nhs = mpls_count_nexthops(cfg->rc_mp, cfg->rc_mp_len,
804 cfg->rc_via_alen, &max_via_alen);
805 if (nhs == 0)
806 goto errout;
807
808 err = -ENOMEM;
809 rt = mpls_rt_alloc(nhs, max_via_alen);
810 if (!rt)
811 goto errout;
812
813 rt->rt_protocol = cfg->rc_protocol;
814 rt->rt_payload_type = cfg->rc_payload_type;
815
816 if (cfg->rc_mp)
817 err = mpls_nh_build_multi(cfg, rt);
818 else
819 err = mpls_nh_build_from_cfg(cfg, rt);
820 if (err)
821 goto freert;
822
823 mpls_route_update(net, index, rt, &cfg->rc_nlinfo);
824
825 return 0;
826
827freert:
828 mpls_rt_free(rt);
829errout:
830 return err;
831}
832
833static int mpls_route_del(struct mpls_route_config *cfg)
834{
835 struct net *net = cfg->rc_nlinfo.nl_net;
836 unsigned index;
837 int err = -EINVAL;
838
839 index = cfg->rc_label;
840
841 /* Reserved labels may not be removed */
842 if (index < MPLS_LABEL_FIRST_UNRESERVED)
843 goto errout;
844
845 /* The full 20 bit range may not be supported */
846 if (index >= net->mpls.platform_labels)
847 goto errout;
848
849 mpls_route_update(net, index, NULL, &cfg->rc_nlinfo);
850
851 err = 0;
852errout:
853 return err;
854}
855
856#define MPLS_PERDEV_SYSCTL_OFFSET(field) \
857 (&((struct mpls_dev *)0)->field)
858
859static const struct ctl_table mpls_dev_table[] = {
860 {
861 .procname = "input",
862 .maxlen = sizeof(int),
863 .mode = 0644,
864 .proc_handler = proc_dointvec,
865 .data = MPLS_PERDEV_SYSCTL_OFFSET(input_enabled),
866 },
867 { }
868};
869
870static int mpls_dev_sysctl_register(struct net_device *dev,
871 struct mpls_dev *mdev)
872{
873 char path[sizeof("net/mpls/conf/") + IFNAMSIZ];
874 struct ctl_table *table;
875 int i;
876
877 table = kmemdup(&mpls_dev_table, sizeof(mpls_dev_table), GFP_KERNEL);
878 if (!table)
879 goto out;
880
881 /* Table data contains only offsets relative to the base of
882 * the mdev at this point, so make them absolute.
883 */
884 for (i = 0; i < ARRAY_SIZE(mpls_dev_table); i++)
885 table[i].data = (char *)mdev + (uintptr_t)table[i].data;
886
887 snprintf(path, sizeof(path), "net/mpls/conf/%s", dev->name);
888
889 mdev->sysctl = register_net_sysctl(dev_net(dev), path, table);
890 if (!mdev->sysctl)
891 goto free;
892
893 return 0;
894
895free:
896 kfree(table);
897out:
898 return -ENOBUFS;
899}
900
901static void mpls_dev_sysctl_unregister(struct mpls_dev *mdev)
902{
903 struct ctl_table *table;
904
905 table = mdev->sysctl->ctl_table_arg;
906 unregister_net_sysctl_table(mdev->sysctl);
907 kfree(table);
908}
909
910static struct mpls_dev *mpls_add_dev(struct net_device *dev)
911{
912 struct mpls_dev *mdev;
913 int err = -ENOMEM;
914
915 ASSERT_RTNL();
916
917 mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
918 if (!mdev)
919 return ERR_PTR(err);
920
921 err = mpls_dev_sysctl_register(dev, mdev);
922 if (err)
923 goto free;
924
925 rcu_assign_pointer(dev->mpls_ptr, mdev);
926
927 return mdev;
928
929free:
930 kfree(mdev);
931 return ERR_PTR(err);
932}
933
934static void mpls_ifdown(struct net_device *dev, int event)
935{
936 struct mpls_route __rcu **platform_label;
937 struct net *net = dev_net(dev);
938 unsigned index;
939
940 platform_label = rtnl_dereference(net->mpls.platform_label);
941 for (index = 0; index < net->mpls.platform_labels; index++) {
942 struct mpls_route *rt = rtnl_dereference(platform_label[index]);
943
944 if (!rt)
945 continue;
946
947 change_nexthops(rt) {
948 if (rtnl_dereference(nh->nh_dev) != dev)
949 continue;
950 switch (event) {
951 case NETDEV_DOWN:
952 case NETDEV_UNREGISTER:
953 nh->nh_flags |= RTNH_F_DEAD;
954 /* fall through */
955 case NETDEV_CHANGE:
956 nh->nh_flags |= RTNH_F_LINKDOWN;
957 ACCESS_ONCE(rt->rt_nhn_alive) = rt->rt_nhn_alive - 1;
958 break;
959 }
960 if (event == NETDEV_UNREGISTER)
961 RCU_INIT_POINTER(nh->nh_dev, NULL);
962 } endfor_nexthops(rt);
963 }
964
965
966 return;
967}
968
969static void mpls_ifup(struct net_device *dev, unsigned int nh_flags)
970{
971 struct mpls_route __rcu **platform_label;
972 struct net *net = dev_net(dev);
973 unsigned index;
974 int alive;
975
976 platform_label = rtnl_dereference(net->mpls.platform_label);
977 for (index = 0; index < net->mpls.platform_labels; index++) {
978 struct mpls_route *rt = rtnl_dereference(platform_label[index]);
979
980 if (!rt)
981 continue;
982
983 alive = 0;
984 change_nexthops(rt) {
985 struct net_device *nh_dev =
986 rtnl_dereference(nh->nh_dev);
987
988 if (!(nh->nh_flags & nh_flags)) {
989 alive++;
990 continue;
991 }
992 if (nh_dev != dev)
993 continue;
994 alive++;
995 nh->nh_flags &= ~nh_flags;
996 } endfor_nexthops(rt);
997
998 ACCESS_ONCE(rt->rt_nhn_alive) = alive;
999 }
1000
1001 return;
1002}
1003
1004static int mpls_dev_notify(struct notifier_block *this, unsigned long event,
1005 void *ptr)
1006{
1007 struct net_device *dev = netdev_notifier_info_to_dev(ptr);
1008 struct mpls_dev *mdev;
1009 unsigned int flags;
1010
1011 if (event == NETDEV_REGISTER) {
1012 /* For now just support ethernet devices */
1013 if ((dev->type == ARPHRD_ETHER) ||
1014 (dev->type == ARPHRD_LOOPBACK)) {
1015 mdev = mpls_add_dev(dev);
1016 if (IS_ERR(mdev))
1017 return notifier_from_errno(PTR_ERR(mdev));
1018 }
1019 return NOTIFY_OK;
1020 }
1021
1022 mdev = mpls_dev_get(dev);
1023 if (!mdev)
1024 return NOTIFY_OK;
1025
1026 switch (event) {
1027 case NETDEV_DOWN:
1028 mpls_ifdown(dev, event);
1029 break;
1030 case NETDEV_UP:
1031 flags = dev_get_flags(dev);
1032 if (flags & (IFF_RUNNING | IFF_LOWER_UP))
1033 mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
1034 else
1035 mpls_ifup(dev, RTNH_F_DEAD);
1036 break;
1037 case NETDEV_CHANGE:
1038 flags = dev_get_flags(dev);
1039 if (flags & (IFF_RUNNING | IFF_LOWER_UP))
1040 mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
1041 else
1042 mpls_ifdown(dev, event);
1043 break;
1044 case NETDEV_UNREGISTER:
1045 mpls_ifdown(dev, event);
1046 mdev = mpls_dev_get(dev);
1047 if (mdev) {
1048 mpls_dev_sysctl_unregister(mdev);
1049 RCU_INIT_POINTER(dev->mpls_ptr, NULL);
1050 kfree_rcu(mdev, rcu);
1051 }
1052 break;
1053 case NETDEV_CHANGENAME:
1054 mdev = mpls_dev_get(dev);
1055 if (mdev) {
1056 int err;
1057
1058 mpls_dev_sysctl_unregister(mdev);
1059 err = mpls_dev_sysctl_register(dev, mdev);
1060 if (err)
1061 return notifier_from_errno(err);
1062 }
1063 break;
1064 }
1065 return NOTIFY_OK;
1066}
1067
1068static struct notifier_block mpls_dev_notifier = {
1069 .notifier_call = mpls_dev_notify,
1070};
1071
1072static int nla_put_via(struct sk_buff *skb,
1073 u8 table, const void *addr, int alen)
1074{
1075 static const int table_to_family[NEIGH_NR_TABLES + 1] = {
1076 AF_INET, AF_INET6, AF_DECnet, AF_PACKET,
1077 };
1078 struct nlattr *nla;
1079 struct rtvia *via;
1080 int family = AF_UNSPEC;
1081
1082 nla = nla_reserve(skb, RTA_VIA, alen + 2);
1083 if (!nla)
1084 return -EMSGSIZE;
1085
1086 if (table <= NEIGH_NR_TABLES)
1087 family = table_to_family[table];
1088
1089 via = nla_data(nla);
1090 via->rtvia_family = family;
1091 memcpy(via->rtvia_addr, addr, alen);
1092 return 0;
1093}
1094
1095int nla_put_labels(struct sk_buff *skb, int attrtype,
1096 u8 labels, const u32 label[])
1097{
1098 struct nlattr *nla;
1099 struct mpls_shim_hdr *nla_label;
1100 bool bos;
1101 int i;
1102 nla = nla_reserve(skb, attrtype, labels*4);
1103 if (!nla)
1104 return -EMSGSIZE;
1105
1106 nla_label = nla_data(nla);
1107 bos = true;
1108 for (i = labels - 1; i >= 0; i--) {
1109 nla_label[i] = mpls_entry_encode(label[i], 0, 0, bos);
1110 bos = false;
1111 }
1112
1113 return 0;
1114}
1115EXPORT_SYMBOL_GPL(nla_put_labels);
1116
1117int nla_get_labels(const struct nlattr *nla,
1118 u32 max_labels, u8 *labels, u32 label[])
1119{
1120 unsigned len = nla_len(nla);
1121 unsigned nla_labels;
1122 struct mpls_shim_hdr *nla_label;
1123 bool bos;
1124 int i;
1125
1126 /* len needs to be an even multiple of 4 (the label size) */
1127 if (len & 3)
1128 return -EINVAL;
1129
1130 /* Limit the number of new labels allowed */
1131 nla_labels = len/4;
1132 if (nla_labels > max_labels)
1133 return -EINVAL;
1134
1135 nla_label = nla_data(nla);
1136 bos = true;
1137 for (i = nla_labels - 1; i >= 0; i--, bos = false) {
1138 struct mpls_entry_decoded dec;
1139 dec = mpls_entry_decode(nla_label + i);
1140
1141 /* Ensure the bottom of stack flag is properly set
1142 * and ttl and tc are both clear.
1143 */
1144 if ((dec.bos != bos) || dec.ttl || dec.tc)
1145 return -EINVAL;
1146
1147 switch (dec.label) {
1148 case MPLS_LABEL_IMPLNULL:
1149 /* RFC3032: This is a label that an LSR may
1150 * assign and distribute, but which never
1151 * actually appears in the encapsulation.
1152 */
1153 return -EINVAL;
1154 }
1155
1156 label[i] = dec.label;
1157 }
1158 *labels = nla_labels;
1159 return 0;
1160}
1161EXPORT_SYMBOL_GPL(nla_get_labels);
1162
1163int nla_get_via(const struct nlattr *nla, u8 *via_alen,
1164 u8 *via_table, u8 via_addr[])
1165{
1166 struct rtvia *via = nla_data(nla);
1167 int err = -EINVAL;
1168 int alen;
1169
1170 if (nla_len(nla) < offsetof(struct rtvia, rtvia_addr))
1171 goto errout;
1172 alen = nla_len(nla) -
1173 offsetof(struct rtvia, rtvia_addr);
1174 if (alen > MAX_VIA_ALEN)
1175 goto errout;
1176
1177 /* Validate the address family */
1178 switch (via->rtvia_family) {
1179 case AF_PACKET:
1180 *via_table = NEIGH_LINK_TABLE;
1181 break;
1182 case AF_INET:
1183 *via_table = NEIGH_ARP_TABLE;
1184 if (alen != 4)
1185 goto errout;
1186 break;
1187 case AF_INET6:
1188 *via_table = NEIGH_ND_TABLE;
1189 if (alen != 16)
1190 goto errout;
1191 break;
1192 default:
1193 /* Unsupported address family */
1194 goto errout;
1195 }
1196
1197 memcpy(via_addr, via->rtvia_addr, alen);
1198 *via_alen = alen;
1199 err = 0;
1200
1201errout:
1202 return err;
1203}
1204
1205static int rtm_to_route_config(struct sk_buff *skb, struct nlmsghdr *nlh,
1206 struct mpls_route_config *cfg)
1207{
1208 struct rtmsg *rtm;
1209 struct nlattr *tb[RTA_MAX+1];
1210 int index;
1211 int err;
1212
1213 err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_mpls_policy);
1214 if (err < 0)
1215 goto errout;
1216
1217 err = -EINVAL;
1218 rtm = nlmsg_data(nlh);
1219 memset(cfg, 0, sizeof(*cfg));
1220
1221 if (rtm->rtm_family != AF_MPLS)
1222 goto errout;
1223 if (rtm->rtm_dst_len != 20)
1224 goto errout;
1225 if (rtm->rtm_src_len != 0)
1226 goto errout;
1227 if (rtm->rtm_tos != 0)
1228 goto errout;
1229 if (rtm->rtm_table != RT_TABLE_MAIN)
1230 goto errout;
1231 /* Any value is acceptable for rtm_protocol */
1232
1233 /* As mpls uses destination specific addresses
1234 * (or source specific address in the case of multicast)
1235 * all addresses have universal scope.
1236 */
1237 if (rtm->rtm_scope != RT_SCOPE_UNIVERSE)
1238 goto errout;
1239 if (rtm->rtm_type != RTN_UNICAST)
1240 goto errout;
1241 if (rtm->rtm_flags != 0)
1242 goto errout;
1243
1244 cfg->rc_label = LABEL_NOT_SPECIFIED;
1245 cfg->rc_protocol = rtm->rtm_protocol;
1246 cfg->rc_via_table = MPLS_NEIGH_TABLE_UNSPEC;
1247 cfg->rc_nlflags = nlh->nlmsg_flags;
1248 cfg->rc_nlinfo.portid = NETLINK_CB(skb).portid;
1249 cfg->rc_nlinfo.nlh = nlh;
1250 cfg->rc_nlinfo.nl_net = sock_net(skb->sk);
1251
1252 for (index = 0; index <= RTA_MAX; index++) {
1253 struct nlattr *nla = tb[index];
1254 if (!nla)
1255 continue;
1256
1257 switch(index) {
1258 case RTA_OIF:
1259 cfg->rc_ifindex = nla_get_u32(nla);
1260 break;
1261 case RTA_NEWDST:
1262 if (nla_get_labels(nla, MAX_NEW_LABELS,
1263 &cfg->rc_output_labels,
1264 cfg->rc_output_label))
1265 goto errout;
1266 break;
1267 case RTA_DST:
1268 {
1269 u8 label_count;
1270 if (nla_get_labels(nla, 1, &label_count,
1271 &cfg->rc_label))
1272 goto errout;
1273
1274 /* Reserved labels may not be set */
1275 if (cfg->rc_label < MPLS_LABEL_FIRST_UNRESERVED)
1276 goto errout;
1277
1278 break;
1279 }
1280 case RTA_VIA:
1281 {
1282 if (nla_get_via(nla, &cfg->rc_via_alen,
1283 &cfg->rc_via_table, cfg->rc_via))
1284 goto errout;
1285 break;
1286 }
1287 case RTA_MULTIPATH:
1288 {
1289 cfg->rc_mp = nla_data(nla);
1290 cfg->rc_mp_len = nla_len(nla);
1291 break;
1292 }
1293 default:
1294 /* Unsupported attribute */
1295 goto errout;
1296 }
1297 }
1298
1299 err = 0;
1300errout:
1301 return err;
1302}
1303
1304static int mpls_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh)
1305{
1306 struct mpls_route_config cfg;
1307 int err;
1308
1309 err = rtm_to_route_config(skb, nlh, &cfg);
1310 if (err < 0)
1311 return err;
1312
1313 return mpls_route_del(&cfg);
1314}
1315
1316
1317static int mpls_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh)
1318{
1319 struct mpls_route_config cfg;
1320 int err;
1321
1322 err = rtm_to_route_config(skb, nlh, &cfg);
1323 if (err < 0)
1324 return err;
1325
1326 return mpls_route_add(&cfg);
1327}
1328
1329static int mpls_dump_route(struct sk_buff *skb, u32 portid, u32 seq, int event,
1330 u32 label, struct mpls_route *rt, int flags)
1331{
1332 struct net_device *dev;
1333 struct nlmsghdr *nlh;
1334 struct rtmsg *rtm;
1335
1336 nlh = nlmsg_put(skb, portid, seq, event, sizeof(*rtm), flags);
1337 if (nlh == NULL)
1338 return -EMSGSIZE;
1339
1340 rtm = nlmsg_data(nlh);
1341 rtm->rtm_family = AF_MPLS;
1342 rtm->rtm_dst_len = 20;
1343 rtm->rtm_src_len = 0;
1344 rtm->rtm_tos = 0;
1345 rtm->rtm_table = RT_TABLE_MAIN;
1346 rtm->rtm_protocol = rt->rt_protocol;
1347 rtm->rtm_scope = RT_SCOPE_UNIVERSE;
1348 rtm->rtm_type = RTN_UNICAST;
1349 rtm->rtm_flags = 0;
1350
1351 if (nla_put_labels(skb, RTA_DST, 1, &label))
1352 goto nla_put_failure;
1353 if (rt->rt_nhn == 1) {
1354 const struct mpls_nh *nh = rt->rt_nh;
1355
1356 if (nh->nh_labels &&
1357 nla_put_labels(skb, RTA_NEWDST, nh->nh_labels,
1358 nh->nh_label))
1359 goto nla_put_failure;
1360 if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC &&
1361 nla_put_via(skb, nh->nh_via_table, mpls_nh_via(rt, nh),
1362 nh->nh_via_alen))
1363 goto nla_put_failure;
1364 dev = rtnl_dereference(nh->nh_dev);
1365 if (dev && nla_put_u32(skb, RTA_OIF, dev->ifindex))
1366 goto nla_put_failure;
1367 if (nh->nh_flags & RTNH_F_LINKDOWN)
1368 rtm->rtm_flags |= RTNH_F_LINKDOWN;
1369 if (nh->nh_flags & RTNH_F_DEAD)
1370 rtm->rtm_flags |= RTNH_F_DEAD;
1371 } else {
1372 struct rtnexthop *rtnh;
1373 struct nlattr *mp;
1374 int dead = 0;
1375 int linkdown = 0;
1376
1377 mp = nla_nest_start(skb, RTA_MULTIPATH);
1378 if (!mp)
1379 goto nla_put_failure;
1380
1381 for_nexthops(rt) {
1382 rtnh = nla_reserve_nohdr(skb, sizeof(*rtnh));
1383 if (!rtnh)
1384 goto nla_put_failure;
1385
1386 dev = rtnl_dereference(nh->nh_dev);
1387 if (dev)
1388 rtnh->rtnh_ifindex = dev->ifindex;
1389 if (nh->nh_flags & RTNH_F_LINKDOWN) {
1390 rtnh->rtnh_flags |= RTNH_F_LINKDOWN;
1391 linkdown++;
1392 }
1393 if (nh->nh_flags & RTNH_F_DEAD) {
1394 rtnh->rtnh_flags |= RTNH_F_DEAD;
1395 dead++;
1396 }
1397
1398 if (nh->nh_labels && nla_put_labels(skb, RTA_NEWDST,
1399 nh->nh_labels,
1400 nh->nh_label))
1401 goto nla_put_failure;
1402 if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC &&
1403 nla_put_via(skb, nh->nh_via_table,
1404 mpls_nh_via(rt, nh),
1405 nh->nh_via_alen))
1406 goto nla_put_failure;
1407
1408 /* length of rtnetlink header + attributes */
1409 rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *)rtnh;
1410 } endfor_nexthops(rt);
1411
1412 if (linkdown == rt->rt_nhn)
1413 rtm->rtm_flags |= RTNH_F_LINKDOWN;
1414 if (dead == rt->rt_nhn)
1415 rtm->rtm_flags |= RTNH_F_DEAD;
1416
1417 nla_nest_end(skb, mp);
1418 }
1419
1420 nlmsg_end(skb, nlh);
1421 return 0;
1422
1423nla_put_failure:
1424 nlmsg_cancel(skb, nlh);
1425 return -EMSGSIZE;
1426}
1427
1428static int mpls_dump_routes(struct sk_buff *skb, struct netlink_callback *cb)
1429{
1430 struct net *net = sock_net(skb->sk);
1431 struct mpls_route __rcu **platform_label;
1432 size_t platform_labels;
1433 unsigned int index;
1434
1435 ASSERT_RTNL();
1436
1437 index = cb->args[0];
1438 if (index < MPLS_LABEL_FIRST_UNRESERVED)
1439 index = MPLS_LABEL_FIRST_UNRESERVED;
1440
1441 platform_label = rtnl_dereference(net->mpls.platform_label);
1442 platform_labels = net->mpls.platform_labels;
1443 for (; index < platform_labels; index++) {
1444 struct mpls_route *rt;
1445 rt = rtnl_dereference(platform_label[index]);
1446 if (!rt)
1447 continue;
1448
1449 if (mpls_dump_route(skb, NETLINK_CB(cb->skb).portid,
1450 cb->nlh->nlmsg_seq, RTM_NEWROUTE,
1451 index, rt, NLM_F_MULTI) < 0)
1452 break;
1453 }
1454 cb->args[0] = index;
1455
1456 return skb->len;
1457}
1458
1459static inline size_t lfib_nlmsg_size(struct mpls_route *rt)
1460{
1461 size_t payload =
1462 NLMSG_ALIGN(sizeof(struct rtmsg))
1463 + nla_total_size(4); /* RTA_DST */
1464
1465 if (rt->rt_nhn == 1) {
1466 struct mpls_nh *nh = rt->rt_nh;
1467
1468 if (nh->nh_dev)
1469 payload += nla_total_size(4); /* RTA_OIF */
1470 if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC) /* RTA_VIA */
1471 payload += nla_total_size(2 + nh->nh_via_alen);
1472 if (nh->nh_labels) /* RTA_NEWDST */
1473 payload += nla_total_size(nh->nh_labels * 4);
1474 } else {
1475 /* each nexthop is packed in an attribute */
1476 size_t nhsize = 0;
1477
1478 for_nexthops(rt) {
1479 nhsize += nla_total_size(sizeof(struct rtnexthop));
1480 /* RTA_VIA */
1481 if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC)
1482 nhsize += nla_total_size(2 + nh->nh_via_alen);
1483 if (nh->nh_labels)
1484 nhsize += nla_total_size(nh->nh_labels * 4);
1485 } endfor_nexthops(rt);
1486 /* nested attribute */
1487 payload += nla_total_size(nhsize);
1488 }
1489
1490 return payload;
1491}
1492
1493static void rtmsg_lfib(int event, u32 label, struct mpls_route *rt,
1494 struct nlmsghdr *nlh, struct net *net, u32 portid,
1495 unsigned int nlm_flags)
1496{
1497 struct sk_buff *skb;
1498 u32 seq = nlh ? nlh->nlmsg_seq : 0;
1499 int err = -ENOBUFS;
1500
1501 skb = nlmsg_new(lfib_nlmsg_size(rt), GFP_KERNEL);
1502 if (skb == NULL)
1503 goto errout;
1504
1505 err = mpls_dump_route(skb, portid, seq, event, label, rt, nlm_flags);
1506 if (err < 0) {
1507 /* -EMSGSIZE implies BUG in lfib_nlmsg_size */
1508 WARN_ON(err == -EMSGSIZE);
1509 kfree_skb(skb);
1510 goto errout;
1511 }
1512 rtnl_notify(skb, net, portid, RTNLGRP_MPLS_ROUTE, nlh, GFP_KERNEL);
1513
1514 return;
1515errout:
1516 if (err < 0)
1517 rtnl_set_sk_err(net, RTNLGRP_MPLS_ROUTE, err);
1518}
1519
1520static int resize_platform_label_table(struct net *net, size_t limit)
1521{
1522 size_t size = sizeof(struct mpls_route *) * limit;
1523 size_t old_limit;
1524 size_t cp_size;
1525 struct mpls_route __rcu **labels = NULL, **old;
1526 struct mpls_route *rt0 = NULL, *rt2 = NULL;
1527 unsigned index;
1528
1529 if (size) {
1530 labels = kzalloc(size, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY);
1531 if (!labels)
1532 labels = vzalloc(size);
1533
1534 if (!labels)
1535 goto nolabels;
1536 }
1537
1538 /* In case the predefined labels need to be populated */
1539 if (limit > MPLS_LABEL_IPV4NULL) {
1540 struct net_device *lo = net->loopback_dev;
1541 rt0 = mpls_rt_alloc(1, lo->addr_len);
1542 if (!rt0)
1543 goto nort0;
1544 RCU_INIT_POINTER(rt0->rt_nh->nh_dev, lo);
1545 rt0->rt_protocol = RTPROT_KERNEL;
1546 rt0->rt_payload_type = MPT_IPV4;
1547 rt0->rt_nh->nh_via_table = NEIGH_LINK_TABLE;
1548 rt0->rt_nh->nh_via_alen = lo->addr_len;
1549 memcpy(__mpls_nh_via(rt0, rt0->rt_nh), lo->dev_addr,
1550 lo->addr_len);
1551 }
1552 if (limit > MPLS_LABEL_IPV6NULL) {
1553 struct net_device *lo = net->loopback_dev;
1554 rt2 = mpls_rt_alloc(1, lo->addr_len);
1555 if (!rt2)
1556 goto nort2;
1557 RCU_INIT_POINTER(rt2->rt_nh->nh_dev, lo);
1558 rt2->rt_protocol = RTPROT_KERNEL;
1559 rt2->rt_payload_type = MPT_IPV6;
1560 rt2->rt_nh->nh_via_table = NEIGH_LINK_TABLE;
1561 rt2->rt_nh->nh_via_alen = lo->addr_len;
1562 memcpy(__mpls_nh_via(rt2, rt2->rt_nh), lo->dev_addr,
1563 lo->addr_len);
1564 }
1565
1566 rtnl_lock();
1567 /* Remember the original table */
1568 old = rtnl_dereference(net->mpls.platform_label);
1569 old_limit = net->mpls.platform_labels;
1570
1571 /* Free any labels beyond the new table */
1572 for (index = limit; index < old_limit; index++)
1573 mpls_route_update(net, index, NULL, NULL);
1574
1575 /* Copy over the old labels */
1576 cp_size = size;
1577 if (old_limit < limit)
1578 cp_size = old_limit * sizeof(struct mpls_route *);
1579
1580 memcpy(labels, old, cp_size);
1581
1582 /* If needed set the predefined labels */
1583 if ((old_limit <= MPLS_LABEL_IPV6NULL) &&
1584 (limit > MPLS_LABEL_IPV6NULL)) {
1585 RCU_INIT_POINTER(labels[MPLS_LABEL_IPV6NULL], rt2);
1586 rt2 = NULL;
1587 }
1588
1589 if ((old_limit <= MPLS_LABEL_IPV4NULL) &&
1590 (limit > MPLS_LABEL_IPV4NULL)) {
1591 RCU_INIT_POINTER(labels[MPLS_LABEL_IPV4NULL], rt0);
1592 rt0 = NULL;
1593 }
1594
1595 /* Update the global pointers */
1596 net->mpls.platform_labels = limit;
1597 rcu_assign_pointer(net->mpls.platform_label, labels);
1598
1599 rtnl_unlock();
1600
1601 mpls_rt_free(rt2);
1602 mpls_rt_free(rt0);
1603
1604 if (old) {
1605 synchronize_rcu();
1606 kvfree(old);
1607 }
1608 return 0;
1609
1610nort2:
1611 mpls_rt_free(rt0);
1612nort0:
1613 kvfree(labels);
1614nolabels:
1615 return -ENOMEM;
1616}
1617
1618static int mpls_platform_labels(struct ctl_table *table, int write,
1619 void __user *buffer, size_t *lenp, loff_t *ppos)
1620{
1621 struct net *net = table->data;
1622 int platform_labels = net->mpls.platform_labels;
1623 int ret;
1624 struct ctl_table tmp = {
1625 .procname = table->procname,
1626 .data = &platform_labels,
1627 .maxlen = sizeof(int),
1628 .mode = table->mode,
1629 .extra1 = &zero,
1630 .extra2 = &label_limit,
1631 };
1632
1633 ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
1634
1635 if (write && ret == 0)
1636 ret = resize_platform_label_table(net, platform_labels);
1637
1638 return ret;
1639}
1640
1641static const struct ctl_table mpls_table[] = {
1642 {
1643 .procname = "platform_labels",
1644 .data = NULL,
1645 .maxlen = sizeof(int),
1646 .mode = 0644,
1647 .proc_handler = mpls_platform_labels,
1648 },
1649 { }
1650};
1651
1652static int mpls_net_init(struct net *net)
1653{
1654 struct ctl_table *table;
1655
1656 net->mpls.platform_labels = 0;
1657 net->mpls.platform_label = NULL;
1658
1659 table = kmemdup(mpls_table, sizeof(mpls_table), GFP_KERNEL);
1660 if (table == NULL)
1661 return -ENOMEM;
1662
1663 table[0].data = net;
1664 net->mpls.ctl = register_net_sysctl(net, "net/mpls", table);
1665 if (net->mpls.ctl == NULL) {
1666 kfree(table);
1667 return -ENOMEM;
1668 }
1669
1670 return 0;
1671}
1672
1673static void mpls_net_exit(struct net *net)
1674{
1675 struct mpls_route __rcu **platform_label;
1676 size_t platform_labels;
1677 struct ctl_table *table;
1678 unsigned int index;
1679
1680 table = net->mpls.ctl->ctl_table_arg;
1681 unregister_net_sysctl_table(net->mpls.ctl);
1682 kfree(table);
1683
1684 /* An rcu grace period has passed since there was a device in
1685 * the network namespace (and thus the last in flight packet)
1686 * left this network namespace. This is because
1687 * unregister_netdevice_many and netdev_run_todo has completed
1688 * for each network device that was in this network namespace.
1689 *
1690 * As such no additional rcu synchronization is necessary when
1691 * freeing the platform_label table.
1692 */
1693 rtnl_lock();
1694 platform_label = rtnl_dereference(net->mpls.platform_label);
1695 platform_labels = net->mpls.platform_labels;
1696 for (index = 0; index < platform_labels; index++) {
1697 struct mpls_route *rt = rtnl_dereference(platform_label[index]);
1698 RCU_INIT_POINTER(platform_label[index], NULL);
1699 mpls_rt_free(rt);
1700 }
1701 rtnl_unlock();
1702
1703 kvfree(platform_label);
1704}
1705
1706static struct pernet_operations mpls_net_ops = {
1707 .init = mpls_net_init,
1708 .exit = mpls_net_exit,
1709};
1710
1711static int __init mpls_init(void)
1712{
1713 int err;
1714
1715 BUILD_BUG_ON(sizeof(struct mpls_shim_hdr) != 4);
1716
1717 err = register_pernet_subsys(&mpls_net_ops);
1718 if (err)
1719 goto out;
1720
1721 err = register_netdevice_notifier(&mpls_dev_notifier);
1722 if (err)
1723 goto out_unregister_pernet;
1724
1725 dev_add_pack(&mpls_packet_type);
1726
1727 rtnl_register(PF_MPLS, RTM_NEWROUTE, mpls_rtm_newroute, NULL, NULL);
1728 rtnl_register(PF_MPLS, RTM_DELROUTE, mpls_rtm_delroute, NULL, NULL);
1729 rtnl_register(PF_MPLS, RTM_GETROUTE, NULL, mpls_dump_routes, NULL);
1730 err = 0;
1731out:
1732 return err;
1733
1734out_unregister_pernet:
1735 unregister_pernet_subsys(&mpls_net_ops);
1736 goto out;
1737}
1738module_init(mpls_init);
1739
1740static void __exit mpls_exit(void)
1741{
1742 rtnl_unregister_all(PF_MPLS);
1743 dev_remove_pack(&mpls_packet_type);
1744 unregister_netdevice_notifier(&mpls_dev_notifier);
1745 unregister_pernet_subsys(&mpls_net_ops);
1746}
1747module_exit(mpls_exit);
1748
1749MODULE_DESCRIPTION("MultiProtocol Label Switching");
1750MODULE_LICENSE("GPL v2");
1751MODULE_ALIAS_NETPROTO(PF_MPLS);