af_mpls.c 60.7 KB
Newer Older
E
Eric W. Biederman 已提交
1 2 3
#include <linux/types.h>
#include <linux/skbuff.h>
#include <linux/socket.h>
4
#include <linux/sysctl.h>
E
Eric W. Biederman 已提交
5 6 7 8 9
#include <linux/net.h>
#include <linux/module.h>
#include <linux/if_arp.h>
#include <linux/ipv6.h>
#include <linux/mpls.h>
D
David Ahern 已提交
10
#include <linux/netconf.h>
11
#include <linux/nospec.h>
12
#include <linux/vmalloc.h>
R
Robert Shearman 已提交
13
#include <linux/percpu.h>
E
Eric W. Biederman 已提交
14 15 16 17 18 19
#include <net/ip.h>
#include <net/dst.h>
#include <net/sock.h>
#include <net/arp.h>
#include <net/ip_fib.h>
#include <net/netevent.h>
20
#include <net/ip_tunnels.h>
E
Eric W. Biederman 已提交
21
#include <net/netns/generic.h>
22 23 24
#if IS_ENABLED(CONFIG_IPV6)
#include <net/ipv6.h>
#endif
R
Robert Shearman 已提交
25
#include <net/addrconf.h>
R
Roopa Prabhu 已提交
26
#include <net/nexthop.h>
E
Eric W. Biederman 已提交
27 28
#include "internal.h"

29 30 31
/* max memory we will use for mpls_route */
#define MAX_MPLS_ROUTE_MEM	4096

32 33 34 35 36
/* Maximum number of labels to look ahead at when selecting a path of
 * a multipath route
 */
#define MAX_MP_SELECT_LABELS 4

37 38
#define MPLS_NEIGH_TABLE_UNSPEC (NEIGH_LINK_TABLE + 1)

39
static int zero = 0;
40
static int one = 1;
41
static int label_limit = (1 << 20) - 1;
42
static int ttl_max = 255;
43

44
#if IS_ENABLED(CONFIG_NET_IP_TUNNEL)
45
static size_t ipgre_mpls_encap_hlen(struct ip_tunnel_encap *e)
46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
{
	return sizeof(struct mpls_shim_hdr);
}

static const struct ip_tunnel_encap_ops mpls_iptun_ops = {
	.encap_hlen	= ipgre_mpls_encap_hlen,
};

static int ipgre_tunnel_encap_add_mpls_ops(void)
{
	return ip_tunnel_encap_add_ops(&mpls_iptun_ops, TUNNEL_ENCAP_MPLS);
}

static void ipgre_tunnel_encap_del_mpls_ops(void)
{
	ip_tunnel_encap_del_ops(&mpls_iptun_ops, TUNNEL_ENCAP_MPLS);
}
#else
static int ipgre_tunnel_encap_add_mpls_ops(void)
{
	return 0;
}

static void ipgre_tunnel_encap_del_mpls_ops(void)
{
}
#endif

74 75 76 77
static void rtmsg_lfib(int event, u32 label, struct mpls_route *rt,
		       struct nlmsghdr *nlh, struct net *net, u32 portid,
		       unsigned int nlm_flags);

E
Eric W. Biederman 已提交
78 79 80 81 82 83 84 85 86 87 88 89
static struct mpls_route *mpls_route_input_rcu(struct net *net, unsigned index)
{
	struct mpls_route *rt = NULL;

	if (index < net->mpls.platform_labels) {
		struct mpls_route __rcu **platform_label =
			rcu_dereference(net->mpls.platform_label);
		rt = rcu_dereference(platform_label[index]);
	}
	return rt;
}

90
bool mpls_output_possible(const struct net_device *dev)
E
Eric W. Biederman 已提交
91 92 93
{
	return dev && (dev->flags & IFF_UP) && netif_carrier_ok(dev);
}
94
EXPORT_SYMBOL_GPL(mpls_output_possible);
E
Eric W. Biederman 已提交
95

96 97
static u8 *__mpls_nh_via(struct mpls_route *rt, struct mpls_nh *nh)
{
98
	return (u8 *)nh + rt->rt_via_offset;
99 100 101 102 103 104 105 106
}

static const u8 *mpls_nh_via(const struct mpls_route *rt,
			     const struct mpls_nh *nh)
{
	return __mpls_nh_via((struct mpls_route *)rt, (struct mpls_nh *)nh);
}

R
Roopa Prabhu 已提交
107
static unsigned int mpls_nh_header_size(const struct mpls_nh *nh)
E
Eric W. Biederman 已提交
108 109
{
	/* The size of the layer 2.5 labels to be added for this route */
R
Roopa Prabhu 已提交
110
	return nh->nh_labels * sizeof(struct mpls_shim_hdr);
E
Eric W. Biederman 已提交
111 112
}

113
unsigned int mpls_dev_mtu(const struct net_device *dev)
E
Eric W. Biederman 已提交
114 115 116 117
{
	/* The amount of data the layer 2 frame can hold */
	return dev->mtu;
}
118
EXPORT_SYMBOL_GPL(mpls_dev_mtu);
E
Eric W. Biederman 已提交
119

120
bool mpls_pkt_too_big(const struct sk_buff *skb, unsigned int mtu)
E
Eric W. Biederman 已提交
121 122 123 124
{
	if (skb->len <= mtu)
		return false;

125
	if (skb_is_gso(skb) && skb_gso_validate_network_len(skb, mtu))
E
Eric W. Biederman 已提交
126 127 128 129
		return false;

	return true;
}
130
EXPORT_SYMBOL_GPL(mpls_pkt_too_big);
E
Eric W. Biederman 已提交
131

R
Robert Shearman 已提交
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
void mpls_stats_inc_outucastpkts(struct net_device *dev,
				 const struct sk_buff *skb)
{
	struct mpls_dev *mdev;

	if (skb->protocol == htons(ETH_P_MPLS_UC)) {
		mdev = mpls_dev_get(dev);
		if (mdev)
			MPLS_INC_STATS_LEN(mdev, skb->len,
					   tx_packets,
					   tx_bytes);
	} else if (skb->protocol == htons(ETH_P_IP)) {
		IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUT, skb->len);
#if IS_ENABLED(CONFIG_IPV6)
	} else if (skb->protocol == htons(ETH_P_IPV6)) {
		struct inet6_dev *in6dev = __in6_dev_get(dev);

		if (in6dev)
			IP6_UPD_PO_STATS(dev_net(dev), in6dev,
					 IPSTATS_MIB_OUT, skb->len);
#endif
	}
}
EXPORT_SYMBOL_GPL(mpls_stats_inc_outucastpkts);

157
static u32 mpls_multipath_hash(struct mpls_route *rt, struct sk_buff *skb)
R
Roopa Prabhu 已提交
158
{
159
	struct mpls_entry_decoded dec;
160
	unsigned int mpls_hdr_len = 0;
161 162 163 164 165
	struct mpls_shim_hdr *hdr;
	bool eli_seen = false;
	int label_index;
	u32 hash = 0;

166
	for (label_index = 0; label_index < MAX_MP_SELECT_LABELS;
167
	     label_index++) {
168 169
		mpls_hdr_len += sizeof(*hdr);
		if (!pskb_may_pull(skb, mpls_hdr_len))
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
			break;

		/* Read and decode the current label */
		hdr = mpls_hdr(skb) + label_index;
		dec = mpls_entry_decode(hdr);

		/* RFC6790 - reserved labels MUST NOT be used as keys
		 * for the load-balancing function
		 */
		if (likely(dec.label >= MPLS_LABEL_FIRST_UNRESERVED)) {
			hash = jhash_1word(dec.label, hash);

			/* The entropy label follows the entropy label
			 * indicator, so this means that the entropy
			 * label was just added to the hash - no need to
			 * go any deeper either in the label stack or in the
			 * payload
			 */
			if (eli_seen)
				break;
		} else if (dec.label == MPLS_LABEL_ENTROPY) {
			eli_seen = true;
		}

194 195 196 197 198
		if (!dec.bos)
			continue;

		/* found bottom label; does skb have room for a header? */
		if (pskb_may_pull(skb, mpls_hdr_len + sizeof(struct iphdr))) {
199 200
			const struct iphdr *v4hdr;

201
			v4hdr = (const struct iphdr *)(hdr + 1);
202 203 204 205 206
			if (v4hdr->version == 4) {
				hash = jhash_3words(ntohl(v4hdr->saddr),
						    ntohl(v4hdr->daddr),
						    v4hdr->protocol, hash);
			} else if (v4hdr->version == 6 &&
207 208
				   pskb_may_pull(skb, mpls_hdr_len +
						 sizeof(struct ipv6hdr))) {
209 210
				const struct ipv6hdr *v6hdr;

211
				v6hdr = (const struct ipv6hdr *)(hdr + 1);
212 213 214 215 216
				hash = __ipv6_addr_jhash(&v6hdr->saddr, hash);
				hash = __ipv6_addr_jhash(&v6hdr->daddr, hash);
				hash = jhash_1word(v6hdr->nexthdr, hash);
			}
		}
217 218

		break;
219 220
	}

R
Roopa Prabhu 已提交
221 222 223
	return hash;
}

224 225 226 227 228
static struct mpls_nh *mpls_get_nexthop(struct mpls_route *rt, u8 index)
{
	return (struct mpls_nh *)((u8 *)rt->rt_nh + index * rt->rt_nh_size);
}

229 230 231 232 233
/* number of alive nexthops (rt->rt_nhn_alive) and the flags for
 * a next hop (nh->nh_flags) are modified by netdev event handlers.
 * Since those fields can change at any moment, use READ_ONCE to
 * access both.
 */
R
Roopa Prabhu 已提交
234
static struct mpls_nh *mpls_select_multipath(struct mpls_route *rt,
235
					     struct sk_buff *skb)
R
Roopa Prabhu 已提交
236 237 238 239
{
	u32 hash = 0;
	int nh_index = 0;
	int n = 0;
240
	u8 alive;
R
Roopa Prabhu 已提交
241 242 243 244 245

	/* No need to look further into packet if there's only
	 * one path
	 */
	if (rt->rt_nhn == 1)
246
		return rt->rt_nh;
R
Roopa Prabhu 已提交
247

248 249
	alive = READ_ONCE(rt->rt_nhn_alive);
	if (alive == 0)
R
Roopa Prabhu 已提交
250 251
		return NULL;

252
	hash = mpls_multipath_hash(rt, skb);
R
Roopa Prabhu 已提交
253 254 255 256
	nh_index = hash % alive;
	if (alive == rt->rt_nhn)
		goto out;
	for_nexthops(rt) {
257 258 259
		unsigned int nh_flags = READ_ONCE(nh->nh_flags);

		if (nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
R
Roopa Prabhu 已提交
260 261 262 263 264 265
			continue;
		if (n == nh_index)
			return nh;
		n++;
	} endfor_nexthops(rt);

266
out:
267
	return mpls_get_nexthop(rt, nh_index);
R
Roopa Prabhu 已提交
268 269
}

270 271
static bool mpls_egress(struct net *net, struct mpls_route *rt,
			struct sk_buff *skb, struct mpls_entry_decoded dec)
E
Eric W. Biederman 已提交
272
{
273 274
	enum mpls_payload_type payload_type;
	bool success = false;
E
Eric W. Biederman 已提交
275

276 277 278 279 280 281 282 283 284 285 286 287 288
	/* The IPv4 code below accesses through the IPv4 header
	 * checksum, which is 12 bytes into the packet.
	 * The IPv6 code below accesses through the IPv6 hop limit
	 * which is 8 bytes into the packet.
	 *
	 * For all supported cases there should always be at least 12
	 * bytes of packet data present.  The IPv4 header is 20 bytes
	 * without options and the IPv6 header is always 40 bytes
	 * long.
	 */
	if (!pskb_may_pull(skb, 12))
		return false;

289 290 291 292 293 294 295
	payload_type = rt->rt_payload_type;
	if (payload_type == MPT_UNSPEC)
		payload_type = ip_hdr(skb)->version;

	switch (payload_type) {
	case MPT_IPV4: {
		struct iphdr *hdr4 = ip_hdr(skb);
296
		u8 new_ttl;
E
Eric W. Biederman 已提交
297
		skb->protocol = htons(ETH_P_IP);
298 299 300 301 302 303 304 305 306 307 308 309

		/* If propagating TTL, take the decremented TTL from
		 * the incoming MPLS header, otherwise decrement the
		 * TTL, but only if not 0 to avoid underflow.
		 */
		if (rt->rt_ttl_propagate == MPLS_TTL_PROP_ENABLED ||
		    (rt->rt_ttl_propagate == MPLS_TTL_PROP_DEFAULT &&
		     net->mpls.ip_ttl_propagate))
			new_ttl = dec.ttl;
		else
			new_ttl = hdr4->ttl ? hdr4->ttl - 1 : 0;

E
Eric W. Biederman 已提交
310 311
		csum_replace2(&hdr4->check,
			      htons(hdr4->ttl << 8),
312 313
			      htons(new_ttl << 8));
		hdr4->ttl = new_ttl;
314 315
		success = true;
		break;
E
Eric W. Biederman 已提交
316
	}
317
	case MPT_IPV6: {
E
Eric W. Biederman 已提交
318 319
		struct ipv6hdr *hdr6 = ipv6_hdr(skb);
		skb->protocol = htons(ETH_P_IPV6);
320 321 322 323 324 325 326 327 328 329 330

		/* If propagating TTL, take the decremented TTL from
		 * the incoming MPLS header, otherwise decrement the
		 * hop limit, but only if not 0 to avoid underflow.
		 */
		if (rt->rt_ttl_propagate == MPLS_TTL_PROP_ENABLED ||
		    (rt->rt_ttl_propagate == MPLS_TTL_PROP_DEFAULT &&
		     net->mpls.ip_ttl_propagate))
			hdr6->hop_limit = dec.ttl;
		else if (hdr6->hop_limit)
			hdr6->hop_limit = hdr6->hop_limit - 1;
331 332
		success = true;
		break;
E
Eric W. Biederman 已提交
333
	}
334
	case MPT_UNSPEC:
335
		/* Should have decided which protocol it is by now */
336 337 338
		break;
	}

E
Eric W. Biederman 已提交
339 340 341 342 343 344 345 346 347
	return success;
}

static int mpls_forward(struct sk_buff *skb, struct net_device *dev,
			struct packet_type *pt, struct net_device *orig_dev)
{
	struct net *net = dev_net(dev);
	struct mpls_shim_hdr *hdr;
	struct mpls_route *rt;
R
Roopa Prabhu 已提交
348
	struct mpls_nh *nh;
E
Eric W. Biederman 已提交
349 350
	struct mpls_entry_decoded dec;
	struct net_device *out_dev;
R
Robert Shearman 已提交
351
	struct mpls_dev *out_mdev;
R
Robert Shearman 已提交
352
	struct mpls_dev *mdev;
E
Eric W. Biederman 已提交
353 354 355 356 357 358 359
	unsigned int hh_len;
	unsigned int new_header_size;
	unsigned int mtu;
	int err;

	/* Careful this entire function runs inside of an rcu critical section */

R
Robert Shearman 已提交
360
	mdev = mpls_dev_get(dev);
R
Robert Shearman 已提交
361
	if (!mdev)
R
Robert Shearman 已提交
362 363
		goto drop;

R
Robert Shearman 已提交
364 365 366 367 368
	MPLS_INC_STATS_LEN(mdev, skb->len, rx_packets,
			   rx_bytes);

	if (!mdev->input_enabled) {
		MPLS_INC_STATS(mdev, rx_dropped);
E
Eric W. Biederman 已提交
369
		goto drop;
R
Robert Shearman 已提交
370 371 372 373
	}

	if (skb->pkt_type != PACKET_HOST)
		goto err;
E
Eric W. Biederman 已提交
374 375

	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
R
Robert Shearman 已提交
376
		goto err;
E
Eric W. Biederman 已提交
377 378

	if (!pskb_may_pull(skb, sizeof(*hdr)))
R
Robert Shearman 已提交
379
		goto err;
E
Eric W. Biederman 已提交
380 381 382 383 384 385

	/* Read and decode the label */
	hdr = mpls_hdr(skb);
	dec = mpls_entry_decode(hdr);

	rt = mpls_route_input_rcu(net, dec.label);
R
Robert Shearman 已提交
386 387
	if (!rt) {
		MPLS_INC_STATS(mdev, rx_noroute);
E
Eric W. Biederman 已提交
388
		goto drop;
R
Robert Shearman 已提交
389
	}
E
Eric W. Biederman 已提交
390

391
	nh = mpls_select_multipath(rt, skb);
R
Roopa Prabhu 已提交
392
	if (!nh)
R
Robert Shearman 已提交
393
		goto err;
E
Eric W. Biederman 已提交
394

395 396 397 398 399 400
	/* Pop the label */
	skb_pull(skb, sizeof(*hdr));
	skb_reset_network_header(skb);

	skb_orphan(skb);

E
Eric W. Biederman 已提交
401
	if (skb_warn_if_lro(skb))
R
Robert Shearman 已提交
402
		goto err;
E
Eric W. Biederman 已提交
403 404 405 406

	skb_forward_csum(skb);

	/* Verify ttl is valid */
407
	if (dec.ttl <= 1)
R
Robert Shearman 已提交
408
		goto err;
E
Eric W. Biederman 已提交
409 410
	dec.ttl -= 1;

R
Robert Shearman 已提交
411 412 413 414 415
	/* Find the output device */
	out_dev = rcu_dereference(nh->nh_dev);
	if (!mpls_output_possible(out_dev))
		goto tx_err;

E
Eric W. Biederman 已提交
416
	/* Verify the destination can hold the packet */
R
Roopa Prabhu 已提交
417
	new_header_size = mpls_nh_header_size(nh);
E
Eric W. Biederman 已提交
418 419
	mtu = mpls_dev_mtu(out_dev);
	if (mpls_pkt_too_big(skb, mtu - new_header_size))
R
Robert Shearman 已提交
420
		goto tx_err;
E
Eric W. Biederman 已提交
421 422 423 424 425 426 427

	hh_len = LL_RESERVED_SPACE(out_dev);
	if (!out_dev->header_ops)
		hh_len = 0;

	/* Ensure there is enough space for the headers in the skb */
	if (skb_cow(skb, hh_len + new_header_size))
R
Robert Shearman 已提交
428
		goto tx_err;
E
Eric W. Biederman 已提交
429 430 431 432 433 434

	skb->dev = out_dev;
	skb->protocol = htons(ETH_P_MPLS_UC);

	if (unlikely(!new_header_size && dec.bos)) {
		/* Penultimate hop popping */
435
		if (!mpls_egress(dev_net(out_dev), rt, skb, dec))
R
Robert Shearman 已提交
436
			goto err;
E
Eric W. Biederman 已提交
437 438 439 440 441 442 443 444
	} else {
		bool bos;
		int i;
		skb_push(skb, new_header_size);
		skb_reset_network_header(skb);
		/* Push the new labels */
		hdr = mpls_hdr(skb);
		bos = dec.bos;
R
Roopa Prabhu 已提交
445 446 447
		for (i = nh->nh_labels - 1; i >= 0; i--) {
			hdr[i] = mpls_entry_encode(nh->nh_label[i],
						   dec.ttl, 0, bos);
E
Eric W. Biederman 已提交
448 449 450 451
			bos = false;
		}
	}

R
Robert Shearman 已提交
452 453
	mpls_stats_inc_outucastpkts(out_dev, skb);

454 455 456 457 458 459 460
	/* If via wasn't specified then send out using device address */
	if (nh->nh_via_table == MPLS_NEIGH_TABLE_UNSPEC)
		err = neigh_xmit(NEIGH_LINK_TABLE, out_dev,
				 out_dev->dev_addr, skb);
	else
		err = neigh_xmit(nh->nh_via_table, out_dev,
				 mpls_nh_via(rt, nh), skb);
E
Eric W. Biederman 已提交
461 462 463 464 465
	if (err)
		net_dbg_ratelimited("%s: packet transmission failed: %d\n",
				    __func__, err);
	return 0;

R
Robert Shearman 已提交
466 467 468 469 470 471 472
tx_err:
	out_mdev = out_dev ? mpls_dev_get(out_dev) : NULL;
	if (out_mdev)
		MPLS_INC_STATS(out_mdev, tx_errors);
	goto drop;
err:
	MPLS_INC_STATS(mdev, rx_errors);
E
Eric W. Biederman 已提交
473 474 475 476 477 478 479 480 481 482
drop:
	kfree_skb(skb);
	return NET_RX_DROP;
}

static struct packet_type mpls_packet_type __read_mostly = {
	.type = cpu_to_be16(ETH_P_MPLS_UC),
	.func = mpls_forward,
};

483
static const struct nla_policy rtm_mpls_policy[RTA_MAX+1] = {
484 485
	[RTA_DST]		= { .type = NLA_U32 },
	[RTA_OIF]		= { .type = NLA_U32 },
486
	[RTA_TTL_PROPAGATE]	= { .type = NLA_U8 },
487 488
};

489
struct mpls_route_config {
490 491
	u32			rc_protocol;
	u32			rc_ifindex;
R
Roopa Prabhu 已提交
492 493
	u8			rc_via_table;
	u8			rc_via_alen;
494 495
	u8			rc_via[MAX_VIA_ALEN];
	u32			rc_label;
496
	u8			rc_ttl_propagate;
R
Roopa Prabhu 已提交
497
	u8			rc_output_labels;
498 499 500 501
	u32			rc_output_label[MAX_NEW_LABELS];
	u32			rc_nlflags;
	enum mpls_payload_type	rc_payload_type;
	struct nl_info		rc_nlinfo;
R
Roopa Prabhu 已提交
502 503
	struct rtnexthop	*rc_mp;
	int			rc_mp_len;
504 505
};

506 507 508 509
/* all nexthops within a route have the same size based on max
 * number of labels and max via length for a hop
 */
static struct mpls_route *mpls_rt_alloc(u8 num_nh, u8 max_alen, u8 max_labels)
E
Eric W. Biederman 已提交
510
{
511
	u8 nh_size = MPLS_NH_SIZE(max_labels, max_alen);
E
Eric W. Biederman 已提交
512
	struct mpls_route *rt;
513
	size_t size;
E
Eric W. Biederman 已提交
514

515 516 517 518 519 520 521 522 523 524 525 526
	size = sizeof(*rt) + num_nh * nh_size;
	if (size > MAX_MPLS_ROUTE_MEM)
		return ERR_PTR(-EINVAL);

	rt = kzalloc(size, GFP_KERNEL);
	if (!rt)
		return ERR_PTR(-ENOMEM);

	rt->rt_nhn = num_nh;
	rt->rt_nhn_alive = num_nh;
	rt->rt_nh_size = nh_size;
	rt->rt_via_offset = MPLS_NH_VIA_OFF(max_labels);
R
Roopa Prabhu 已提交
527

E
Eric W. Biederman 已提交
528 529 530 531 532 533 534 535 536
	return rt;
}

static void mpls_rt_free(struct mpls_route *rt)
{
	if (rt)
		kfree_rcu(rt, rt_rcu);
}

537 538 539 540 541 542 543 544 545 546
static void mpls_notify_route(struct net *net, unsigned index,
			      struct mpls_route *old, struct mpls_route *new,
			      const struct nl_info *info)
{
	struct nlmsghdr *nlh = info ? info->nlh : NULL;
	unsigned portid = info ? info->portid : 0;
	int event = new ? RTM_NEWROUTE : RTM_DELROUTE;
	struct mpls_route *rt = new ? new : old;
	unsigned nlm_flags = (old && new) ? NLM_F_REPLACE : 0;
	/* Ignore reserved labels for now */
547
	if (rt && (index >= MPLS_LABEL_FIRST_UNRESERVED))
548 549 550
		rtmsg_lfib(event, index, rt, nlh, net, portid, nlm_flags);
}

E
Eric W. Biederman 已提交
551
static void mpls_route_update(struct net *net, unsigned index,
R
Roopa Prabhu 已提交
552
			      struct mpls_route *new,
E
Eric W. Biederman 已提交
553 554
			      const struct nl_info *info)
{
555
	struct mpls_route __rcu **platform_label;
R
Roopa Prabhu 已提交
556
	struct mpls_route *rt;
E
Eric W. Biederman 已提交
557 558 559

	ASSERT_RTNL();

560 561
	platform_label = rtnl_dereference(net->mpls.platform_label);
	rt = rtnl_dereference(platform_label[index]);
R
Roopa Prabhu 已提交
562
	rcu_assign_pointer(platform_label[index], new);
E
Eric W. Biederman 已提交
563

R
Roopa Prabhu 已提交
564
	mpls_notify_route(net, index, rt, new, info);
565

E
Eric W. Biederman 已提交
566
	/* If we removed a route free it now */
R
Roopa Prabhu 已提交
567
	mpls_rt_free(rt);
E
Eric W. Biederman 已提交
568 569
}

570 571
static unsigned find_free_label(struct net *net)
{
572 573
	struct mpls_route __rcu **platform_label;
	size_t platform_labels;
574
	unsigned index;
575 576 577

	platform_label = rtnl_dereference(net->mpls.platform_label);
	platform_labels = net->mpls.platform_labels;
578 579
	for (index = MPLS_LABEL_FIRST_UNRESERVED; index < platform_labels;
	     index++) {
580
		if (!rtnl_dereference(platform_label[index]))
581 582 583 584 585
			return index;
	}
	return LABEL_NOT_SPECIFIED;
}

586
#if IS_ENABLED(CONFIG_INET)
587 588
static struct net_device *inet_fib_lookup_dev(struct net *net,
					      const void *addr)
R
Roopa Prabhu 已提交
589
{
590
	struct net_device *dev;
R
Roopa Prabhu 已提交
591 592 593 594 595 596
	struct rtable *rt;
	struct in_addr daddr;

	memcpy(&daddr, addr, sizeof(struct in_addr));
	rt = ip_route_output(net, daddr.s_addr, 0, 0, 0);
	if (IS_ERR(rt))
597
		return ERR_CAST(rt);
R
Roopa Prabhu 已提交
598 599 600 601 602 603 604

	dev = rt->dst.dev;
	dev_hold(dev);

	ip_rt_put(rt);

	return dev;
605 606
}
#else
607 608
static struct net_device *inet_fib_lookup_dev(struct net *net,
					      const void *addr)
609 610
{
	return ERR_PTR(-EAFNOSUPPORT);
R
Roopa Prabhu 已提交
611
}
612
#endif
R
Roopa Prabhu 已提交
613

614
#if IS_ENABLED(CONFIG_IPV6)
615 616
static struct net_device *inet6_fib_lookup_dev(struct net *net,
					       const void *addr)
R
Roopa Prabhu 已提交
617
{
618
	struct net_device *dev;
R
Roopa Prabhu 已提交
619 620
	struct dst_entry *dst;
	struct flowi6 fl6;
621 622 623 624
	int err;

	if (!ipv6_stub)
		return ERR_PTR(-EAFNOSUPPORT);
R
Roopa Prabhu 已提交
625 626 627

	memset(&fl6, 0, sizeof(fl6));
	memcpy(&fl6.daddr, addr, sizeof(struct in6_addr));
628 629
	err = ipv6_stub->ipv6_dst_lookup(net, NULL, &dst, &fl6);
	if (err)
630
		return ERR_PTR(err);
R
Roopa Prabhu 已提交
631 632 633 634 635 636 637

	dev = dst->dev;
	dev_hold(dev);
	dst_release(dst);

	return dev;
}
638
#else
639 640
static struct net_device *inet6_fib_lookup_dev(struct net *net,
					       const void *addr)
641 642 643 644
{
	return ERR_PTR(-EAFNOSUPPORT);
}
#endif
R
Roopa Prabhu 已提交
645 646

static struct net_device *find_outdev(struct net *net,
647
				      struct mpls_route *rt,
R
Roopa Prabhu 已提交
648
				      struct mpls_nh *nh, int oif)
R
Roopa Prabhu 已提交
649 650 651
{
	struct net_device *dev = NULL;

R
Roopa Prabhu 已提交
652 653
	if (!oif) {
		switch (nh->nh_via_table) {
R
Roopa Prabhu 已提交
654
		case NEIGH_ARP_TABLE:
655
			dev = inet_fib_lookup_dev(net, mpls_nh_via(rt, nh));
R
Roopa Prabhu 已提交
656 657
			break;
		case NEIGH_ND_TABLE:
658
			dev = inet6_fib_lookup_dev(net, mpls_nh_via(rt, nh));
R
Roopa Prabhu 已提交
659 660 661 662 663
			break;
		case NEIGH_LINK_TABLE:
			break;
		}
	} else {
R
Roopa Prabhu 已提交
664
		dev = dev_get_by_index(net, oif);
R
Roopa Prabhu 已提交
665 666
	}

667 668 669
	if (!dev)
		return ERR_PTR(-ENODEV);

670 671 672
	if (IS_ERR(dev))
		return dev;

R
Roopa Prabhu 已提交
673 674 675
	/* The caller is holding rtnl anyways, so release the dev reference */
	dev_put(dev);

R
Roopa Prabhu 已提交
676 677 678
	return dev;
}

679 680
static int mpls_nh_assign_dev(struct net *net, struct mpls_route *rt,
			      struct mpls_nh *nh, int oif)
R
Roopa Prabhu 已提交
681 682 683 684
{
	struct net_device *dev = NULL;
	int err = -ENODEV;

685
	dev = find_outdev(net, rt, nh, oif);
R
Roopa Prabhu 已提交
686 687 688 689 690 691 692 693 694 695 696
	if (IS_ERR(dev)) {
		err = PTR_ERR(dev);
		dev = NULL;
		goto errout;
	}

	/* Ensure this is a supported device */
	err = -EINVAL;
	if (!mpls_dev_get(dev))
		goto errout;

697 698 699 700
	if ((nh->nh_via_table == NEIGH_LINK_TABLE) &&
	    (dev->addr_len != nh->nh_via_alen))
		goto errout;

R
Roopa Prabhu 已提交
701 702
	RCU_INIT_POINTER(nh->nh_dev, dev);

R
Roopa Prabhu 已提交
703 704 705 706 707 708 709 710 711 712
	if (!(dev->flags & IFF_UP)) {
		nh->nh_flags |= RTNH_F_DEAD;
	} else {
		unsigned int flags;

		flags = dev_get_flags(dev);
		if (!(flags & (IFF_RUNNING | IFF_LOWER_UP)))
			nh->nh_flags |= RTNH_F_LINKDOWN;
	}

R
Roopa Prabhu 已提交
713 714 715 716 717 718
	return 0;

errout:
	return err;
}

719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766
static int nla_get_via(const struct nlattr *nla, u8 *via_alen, u8 *via_table,
		       u8 via_addr[], struct netlink_ext_ack *extack)
{
	struct rtvia *via = nla_data(nla);
	int err = -EINVAL;
	int alen;

	if (nla_len(nla) < offsetof(struct rtvia, rtvia_addr)) {
		NL_SET_ERR_MSG_ATTR(extack, nla,
				    "Invalid attribute length for RTA_VIA");
		goto errout;
	}
	alen = nla_len(nla) -
			offsetof(struct rtvia, rtvia_addr);
	if (alen > MAX_VIA_ALEN) {
		NL_SET_ERR_MSG_ATTR(extack, nla,
				    "Invalid address length for RTA_VIA");
		goto errout;
	}

	/* Validate the address family */
	switch (via->rtvia_family) {
	case AF_PACKET:
		*via_table = NEIGH_LINK_TABLE;
		break;
	case AF_INET:
		*via_table = NEIGH_ARP_TABLE;
		if (alen != 4)
			goto errout;
		break;
	case AF_INET6:
		*via_table = NEIGH_ND_TABLE;
		if (alen != 16)
			goto errout;
		break;
	default:
		/* Unsupported address family */
		goto errout;
	}

	memcpy(via_addr, via->rtvia_addr, alen);
	*via_alen = alen;
	err = 0;

errout:
	return err;
}

R
Roopa Prabhu 已提交
767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782
static int mpls_nh_build_from_cfg(struct mpls_route_config *cfg,
				  struct mpls_route *rt)
{
	struct net *net = cfg->rc_nlinfo.nl_net;
	struct mpls_nh *nh = rt->rt_nh;
	int err;
	int i;

	if (!nh)
		return -ENOMEM;

	nh->nh_labels = cfg->rc_output_labels;
	for (i = 0; i < nh->nh_labels; i++)
		nh->nh_label[i] = cfg->rc_output_label[i];

	nh->nh_via_table = cfg->rc_via_table;
783
	memcpy(__mpls_nh_via(rt, nh), cfg->rc_via, cfg->rc_via_alen);
R
Roopa Prabhu 已提交
784 785
	nh->nh_via_alen = cfg->rc_via_alen;

786
	err = mpls_nh_assign_dev(net, rt, nh, cfg->rc_ifindex);
R
Roopa Prabhu 已提交
787 788 789
	if (err)
		goto errout;

R
Roopa Prabhu 已提交
790 791 792
	if (nh->nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
		rt->rt_nhn_alive--;

R
Roopa Prabhu 已提交
793 794 795 796 797 798
	return 0;

errout:
	return err;
}

799
static int mpls_nh_build(struct net *net, struct mpls_route *rt,
R
Roopa Prabhu 已提交
800
			 struct mpls_nh *nh, int oif, struct nlattr *via,
801 802
			 struct nlattr *newdst, u8 max_labels,
			 struct netlink_ext_ack *extack)
R
Roopa Prabhu 已提交
803 804 805 806 807 808 809
{
	int err = -ENOMEM;

	if (!nh)
		goto errout;

	if (newdst) {
810
		err = nla_get_labels(newdst, max_labels, &nh->nh_labels,
811
				     nh->nh_label, extack);
R
Roopa Prabhu 已提交
812 813 814 815
		if (err)
			goto errout;
	}

816 817
	if (via) {
		err = nla_get_via(via, &nh->nh_via_alen, &nh->nh_via_table,
818
				  __mpls_nh_via(rt, nh), extack);
819 820 821 822 823
		if (err)
			goto errout;
	} else {
		nh->nh_via_table = MPLS_NEIGH_TABLE_UNSPEC;
	}
R
Roopa Prabhu 已提交
824

825
	err = mpls_nh_assign_dev(net, rt, nh, oif);
R
Roopa Prabhu 已提交
826 827 828 829 830 831 832 833 834
	if (err)
		goto errout;

	return 0;

errout:
	return err;
}

835
static u8 mpls_count_nexthops(struct rtnexthop *rtnh, int len,
836 837
			      u8 cfg_via_alen, u8 *max_via_alen,
			      u8 *max_labels)
R
Roopa Prabhu 已提交
838 839
{
	int remaining = len;
840
	u8 nhs = 0;
R
Roopa Prabhu 已提交
841

842
	*max_via_alen = 0;
843
	*max_labels = 0;
844

R
Roopa Prabhu 已提交
845
	while (rtnh_ok(rtnh, remaining)) {
846 847
		struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
		int attrlen;
848
		u8 n_labels = 0;
849 850 851 852 853 854 855 856 857 858 859 860 861

		attrlen = rtnh_attrlen(rtnh);
		nla = nla_find(attrs, attrlen, RTA_VIA);
		if (nla && nla_len(nla) >=
		    offsetof(struct rtvia, rtvia_addr)) {
			int via_alen = nla_len(nla) -
				offsetof(struct rtvia, rtvia_addr);

			if (via_alen <= MAX_VIA_ALEN)
				*max_via_alen = max_t(u16, *max_via_alen,
						      via_alen);
		}

862 863
		nla = nla_find(attrs, attrlen, RTA_NEWDST);
		if (nla &&
864 865
		    nla_get_labels(nla, MAX_NEW_LABELS, &n_labels,
				   NULL, NULL) != 0)
866 867 868 869
			return 0;

		*max_labels = max_t(u8, *max_labels, n_labels);

870 871 872 873 874
		/* number of nexthops is tracked by a u8.
		 * Check for overflow.
		 */
		if (nhs == 255)
			return 0;
R
Roopa Prabhu 已提交
875
		nhs++;
876

R
Roopa Prabhu 已提交
877 878 879 880 881 882 883 884
		rtnh = rtnh_next(rtnh, &remaining);
	}

	/* leftover implies invalid nexthop configuration, discard it */
	return remaining > 0 ? 0 : nhs;
}

static int mpls_nh_build_multi(struct mpls_route_config *cfg,
885 886
			       struct mpls_route *rt, u8 max_labels,
			       struct netlink_ext_ack *extack)
R
Roopa Prabhu 已提交
887 888 889 890 891
{
	struct rtnexthop *rtnh = cfg->rc_mp;
	struct nlattr *nla_via, *nla_newdst;
	int remaining = cfg->rc_mp_len;
	int err = 0;
892
	u8 nhs = 0;
R
Roopa Prabhu 已提交
893 894 895 896 897 898 899 900 901 902 903

	change_nexthops(rt) {
		int attrlen;

		nla_via = NULL;
		nla_newdst = NULL;

		err = -EINVAL;
		if (!rtnh_ok(rtnh, remaining))
			goto errout;

904 905 906 907 908 909
		/* neither weighted multipath nor any flags
		 * are supported
		 */
		if (rtnh->rtnh_hops || rtnh->rtnh_flags)
			goto errout;

R
Roopa Prabhu 已提交
910 911 912 913 914 915 916 917
		attrlen = rtnh_attrlen(rtnh);
		if (attrlen > 0) {
			struct nlattr *attrs = rtnh_attrs(rtnh);

			nla_via = nla_find(attrs, attrlen, RTA_VIA);
			nla_newdst = nla_find(attrs, attrlen, RTA_NEWDST);
		}

918
		err = mpls_nh_build(cfg->rc_nlinfo.nl_net, rt, nh,
919
				    rtnh->rtnh_ifindex, nla_via, nla_newdst,
920
				    max_labels, extack);
R
Roopa Prabhu 已提交
921 922 923
		if (err)
			goto errout;

R
Roopa Prabhu 已提交
924 925 926
		if (nh->nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN))
			rt->rt_nhn_alive--;

R
Roopa Prabhu 已提交
927 928 929 930 931 932 933 934 935 936 937 938
		rtnh = rtnh_next(rtnh, &remaining);
		nhs++;
	} endfor_nexthops(rt);

	rt->rt_nhn = nhs;

	return 0;

errout:
	return err;
}

939
static bool mpls_label_ok(struct net *net, unsigned int *index,
940
			  struct netlink_ext_ack *extack)
941
{
942 943
	bool is_ok = true;

944
	/* Reserved labels may not be set */
945
	if (*index < MPLS_LABEL_FIRST_UNRESERVED) {
946 947
		NL_SET_ERR_MSG(extack,
			       "Invalid label - must be MPLS_LABEL_FIRST_UNRESERVED or higher");
948
		is_ok = false;
949
	}
950 951

	/* The full 20 bit range may not be supported. */
952
	if (is_ok && *index >= net->mpls.platform_labels) {
953 954
		NL_SET_ERR_MSG(extack,
			       "Label >= configured maximum in platform_labels");
955
		is_ok = false;
956
	}
957

958 959
	*index = array_index_nospec(*index, net->mpls.platform_labels);
	return is_ok;
960 961
}

962 963
static int mpls_route_add(struct mpls_route_config *cfg,
			  struct netlink_ext_ack *extack)
964
{
965
	struct mpls_route __rcu **platform_label;
966 967 968
	struct net *net = cfg->rc_nlinfo.nl_net;
	struct mpls_route *rt, *old;
	int err = -EINVAL;
969
	u8 max_via_alen;
R
Roopa Prabhu 已提交
970
	unsigned index;
971
	u8 max_labels;
972
	u8 nhs;
973 974 975 976 977 978 979 980 981

	index = cfg->rc_label;

	/* If a label was not specified during insert pick one */
	if ((index == LABEL_NOT_SPECIFIED) &&
	    (cfg->rc_nlflags & NLM_F_CREATE)) {
		index = find_free_label(net);
	}

982
	if (!mpls_label_ok(net, &index, extack))
983 984 985
		goto errout;

	/* Append makes no sense with mpls */
986
	err = -EOPNOTSUPP;
987 988
	if (cfg->rc_nlflags & NLM_F_APPEND) {
		NL_SET_ERR_MSG(extack, "MPLS does not support route append");
989
		goto errout;
990
	}
991 992

	err = -EEXIST;
993 994
	platform_label = rtnl_dereference(net->mpls.platform_label);
	old = rtnl_dereference(platform_label[index]);
995 996 997 998 999 1000 1001 1002 1003 1004 1005
	if ((cfg->rc_nlflags & NLM_F_EXCL) && old)
		goto errout;

	err = -EEXIST;
	if (!(cfg->rc_nlflags & NLM_F_REPLACE) && old)
		goto errout;

	err = -ENOENT;
	if (!(cfg->rc_nlflags & NLM_F_CREATE) && !old)
		goto errout;

1006
	err = -EINVAL;
1007 1008 1009 1010 1011 1012 1013 1014 1015 1016
	if (cfg->rc_mp) {
		nhs = mpls_count_nexthops(cfg->rc_mp, cfg->rc_mp_len,
					  cfg->rc_via_alen, &max_via_alen,
					  &max_labels);
	} else {
		max_via_alen = cfg->rc_via_alen;
		max_labels = cfg->rc_output_labels;
		nhs = 1;
	}

1017 1018
	if (nhs == 0) {
		NL_SET_ERR_MSG(extack, "Route does not contain a nexthop");
1019
		goto errout;
1020
	}
R
Roopa Prabhu 已提交
1021

1022
	err = -ENOMEM;
1023
	rt = mpls_rt_alloc(nhs, max_via_alen, max_labels);
1024 1025
	if (IS_ERR(rt)) {
		err = PTR_ERR(rt);
1026
		goto errout;
1027
	}
1028 1029

	rt->rt_protocol = cfg->rc_protocol;
1030
	rt->rt_payload_type = cfg->rc_payload_type;
1031
	rt->rt_ttl_propagate = cfg->rc_ttl_propagate;
1032

R
Roopa Prabhu 已提交
1033
	if (cfg->rc_mp)
1034
		err = mpls_nh_build_multi(cfg, rt, max_labels, extack);
R
Roopa Prabhu 已提交
1035 1036 1037 1038 1039 1040
	else
		err = mpls_nh_build_from_cfg(cfg, rt);
	if (err)
		goto freert;

	mpls_route_update(net, index, rt, &cfg->rc_nlinfo);
1041 1042 1043

	return 0;

R
Roopa Prabhu 已提交
1044 1045
freert:
	mpls_rt_free(rt);
1046 1047 1048 1049
errout:
	return err;
}

1050 1051
static int mpls_route_del(struct mpls_route_config *cfg,
			  struct netlink_ext_ack *extack)
1052 1053 1054 1055 1056 1057 1058
{
	struct net *net = cfg->rc_nlinfo.nl_net;
	unsigned index;
	int err = -EINVAL;

	index = cfg->rc_label;

1059
	if (!mpls_label_ok(net, &index, extack))
1060 1061
		goto errout;

R
Roopa Prabhu 已提交
1062
	mpls_route_update(net, index, NULL, &cfg->rc_nlinfo);
1063 1064 1065 1066 1067 1068

	err = 0;
errout:
	return err;
}

R
Robert Shearman 已提交
1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
static void mpls_get_stats(struct mpls_dev *mdev,
			   struct mpls_link_stats *stats)
{
	struct mpls_pcpu_stats *p;
	int i;

	memset(stats, 0, sizeof(*stats));

	for_each_possible_cpu(i) {
		struct mpls_link_stats local;
		unsigned int start;

		p = per_cpu_ptr(mdev->stats, i);
		do {
			start = u64_stats_fetch_begin(&p->syncp);
			local = p->stats;
		} while (u64_stats_fetch_retry(&p->syncp, start));

		stats->rx_packets	+= local.rx_packets;
		stats->rx_bytes		+= local.rx_bytes;
		stats->tx_packets	+= local.tx_packets;
		stats->tx_bytes		+= local.tx_bytes;
		stats->rx_errors	+= local.rx_errors;
		stats->tx_errors	+= local.tx_errors;
		stats->rx_dropped	+= local.rx_dropped;
		stats->tx_dropped	+= local.tx_dropped;
		stats->rx_noroute	+= local.rx_noroute;
	}
}

static int mpls_fill_stats_af(struct sk_buff *skb,
			      const struct net_device *dev)
{
	struct mpls_link_stats *stats;
	struct mpls_dev *mdev;
	struct nlattr *nla;

	mdev = mpls_dev_get(dev);
	if (!mdev)
		return -ENODATA;

	nla = nla_reserve_64bit(skb, MPLS_STATS_LINK,
				sizeof(struct mpls_link_stats),
				MPLS_STATS_UNSPEC);
	if (!nla)
		return -EMSGSIZE;

	stats = nla_data(nla);
	mpls_get_stats(mdev, stats);

	return 0;
}

static size_t mpls_get_stats_af_size(const struct net_device *dev)
{
	struct mpls_dev *mdev;

	mdev = mpls_dev_get(dev);
	if (!mdev)
		return 0;

	return nla_total_size_64bit(sizeof(struct mpls_link_stats));
}

D
David Ahern 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
static int mpls_netconf_fill_devconf(struct sk_buff *skb, struct mpls_dev *mdev,
				     u32 portid, u32 seq, int event,
				     unsigned int flags, int type)
{
	struct nlmsghdr  *nlh;
	struct netconfmsg *ncm;
	bool all = false;

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(struct netconfmsg),
			flags);
	if (!nlh)
		return -EMSGSIZE;

	if (type == NETCONFA_ALL)
		all = true;

	ncm = nlmsg_data(nlh);
	ncm->ncm_family = AF_MPLS;

	if (nla_put_s32(skb, NETCONFA_IFINDEX, mdev->dev->ifindex) < 0)
		goto nla_put_failure;

	if ((all || type == NETCONFA_INPUT) &&
	    nla_put_s32(skb, NETCONFA_INPUT,
			mdev->input_enabled) < 0)
		goto nla_put_failure;

	nlmsg_end(skb, nlh);
	return 0;

nla_put_failure:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
}

static int mpls_netconf_msgsize_devconf(int type)
{
	int size = NLMSG_ALIGN(sizeof(struct netconfmsg))
			+ nla_total_size(4); /* NETCONFA_IFINDEX */
	bool all = false;

	if (type == NETCONFA_ALL)
		all = true;

	if (all || type == NETCONFA_INPUT)
		size += nla_total_size(4);

	return size;
}

1183 1184
static void mpls_netconf_notify_devconf(struct net *net, int event,
					int type, struct mpls_dev *mdev)
D
David Ahern 已提交
1185 1186 1187 1188 1189 1190 1191 1192
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

	skb = nlmsg_new(mpls_netconf_msgsize_devconf(type), GFP_KERNEL);
	if (!skb)
		goto errout;

1193
	err = mpls_netconf_fill_devconf(skb, mdev, 0, 0, event, 0, type);
D
David Ahern 已提交
1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
	if (err < 0) {
		/* -EMSGSIZE implies BUG in mpls_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}

	rtnl_notify(skb, net, 0, RTNLGRP_MPLS_NETCONF, NULL, GFP_KERNEL);
	return;
errout:
	if (err < 0)
		rtnl_set_sk_err(net, RTNLGRP_MPLS_NETCONF, err);
}

static const struct nla_policy devconf_mpls_policy[NETCONFA_MAX + 1] = {
	[NETCONFA_IFINDEX]	= { .len = sizeof(int) },
};

static int mpls_netconf_get_devconf(struct sk_buff *in_skb,
1213 1214
				    struct nlmsghdr *nlh,
				    struct netlink_ext_ack *extack)
D
David Ahern 已提交
1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225
{
	struct net *net = sock_net(in_skb->sk);
	struct nlattr *tb[NETCONFA_MAX + 1];
	struct netconfmsg *ncm;
	struct net_device *dev;
	struct mpls_dev *mdev;
	struct sk_buff *skb;
	int ifindex;
	int err;

	err = nlmsg_parse(nlh, sizeof(*ncm), tb, NETCONFA_MAX,
D
David Ahern 已提交
1226
			  devconf_mpls_policy, extack);
D
David Ahern 已提交
1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
	if (err < 0)
		goto errout;

	err = -EINVAL;
	if (!tb[NETCONFA_IFINDEX])
		goto errout;

	ifindex = nla_get_s32(tb[NETCONFA_IFINDEX]);
	dev = __dev_get_by_index(net, ifindex);
	if (!dev)
		goto errout;

	mdev = mpls_dev_get(dev);
	if (!mdev)
		goto errout;

	err = -ENOBUFS;
	skb = nlmsg_new(mpls_netconf_msgsize_devconf(NETCONFA_ALL), GFP_KERNEL);
	if (!skb)
		goto errout;

	err = mpls_netconf_fill_devconf(skb, mdev,
					NETLINK_CB(in_skb).portid,
					nlh->nlmsg_seq, RTM_NEWNETCONF, 0,
					NETCONFA_ALL);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in mpls_netconf_msgsize_devconf() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	err = rtnl_unicast(skb, net, NETLINK_CB(in_skb).portid);
errout:
	return err;
}

static int mpls_netconf_dump_devconf(struct sk_buff *skb,
				     struct netlink_callback *cb)
{
1266
	const struct nlmsghdr *nlh = cb->nlh;
D
David Ahern 已提交
1267 1268 1269 1270 1271 1272 1273
	struct net *net = sock_net(skb->sk);
	struct hlist_head *head;
	struct net_device *dev;
	struct mpls_dev *mdev;
	int idx, s_idx;
	int h, s_h;

1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288
	if (cb->strict_check) {
		struct netlink_ext_ack *extack = cb->extack;
		struct netconfmsg *ncm;

		if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ncm))) {
			NL_SET_ERR_MSG_MOD(extack, "Invalid header for netconf dump request");
			return -EINVAL;
		}

		if (nlmsg_attrlen(nlh, sizeof(*ncm))) {
			NL_SET_ERR_MSG_MOD(extack, "Invalid data after header in netconf dump request");
			return -EINVAL;
		}
	}

D
David Ahern 已提交
1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304
	s_h = cb->args[0];
	s_idx = idx = cb->args[1];

	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
		idx = 0;
		head = &net->dev_index_head[h];
		rcu_read_lock();
		cb->seq = net->dev_base_seq;
		hlist_for_each_entry_rcu(dev, head, index_hlist) {
			if (idx < s_idx)
				goto cont;
			mdev = mpls_dev_get(dev);
			if (!mdev)
				goto cont;
			if (mpls_netconf_fill_devconf(skb, mdev,
						      NETLINK_CB(cb->skb).portid,
1305
						      nlh->nlmsg_seq,
D
David Ahern 已提交
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
						      RTM_NEWNETCONF,
						      NLM_F_MULTI,
						      NETCONFA_ALL) < 0) {
				rcu_read_unlock();
				goto done;
			}
			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
cont:
			idx++;
		}
		rcu_read_unlock();
	}
done:
	cb->args[0] = h;
	cb->args[1] = idx;

	return skb->len;
}

1325 1326 1327
#define MPLS_PERDEV_SYSCTL_OFFSET(field)	\
	(&((struct mpls_dev *)0)->field)

D
David Ahern 已提交
1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342
static int mpls_conf_proc(struct ctl_table *ctl, int write,
			  void __user *buffer,
			  size_t *lenp, loff_t *ppos)
{
	int oval = *(int *)ctl->data;
	int ret = proc_dointvec(ctl, write, buffer, lenp, ppos);

	if (write) {
		struct mpls_dev *mdev = ctl->extra1;
		int i = (int *)ctl->data - (int *)mdev;
		struct net *net = ctl->extra2;
		int val = *(int *)ctl->data;

		if (i == offsetof(struct mpls_dev, input_enabled) &&
		    val != oval) {
1343 1344
			mpls_netconf_notify_devconf(net, RTM_NEWNETCONF,
						    NETCONFA_INPUT, mdev);
D
David Ahern 已提交
1345 1346 1347 1348 1349 1350
		}
	}

	return ret;
}

1351 1352 1353 1354 1355
static const struct ctl_table mpls_dev_table[] = {
	{
		.procname	= "input",
		.maxlen		= sizeof(int),
		.mode		= 0644,
D
David Ahern 已提交
1356
		.proc_handler	= mpls_conf_proc,
1357 1358 1359 1360 1361 1362 1363 1364 1365
		.data		= MPLS_PERDEV_SYSCTL_OFFSET(input_enabled),
	},
	{ }
};

static int mpls_dev_sysctl_register(struct net_device *dev,
				    struct mpls_dev *mdev)
{
	char path[sizeof("net/mpls/conf/") + IFNAMSIZ];
D
David Ahern 已提交
1366
	struct net *net = dev_net(dev);
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
	struct ctl_table *table;
	int i;

	table = kmemdup(&mpls_dev_table, sizeof(mpls_dev_table), GFP_KERNEL);
	if (!table)
		goto out;

	/* Table data contains only offsets relative to the base of
	 * the mdev at this point, so make them absolute.
	 */
D
David Ahern 已提交
1377
	for (i = 0; i < ARRAY_SIZE(mpls_dev_table); i++) {
1378
		table[i].data = (char *)mdev + (uintptr_t)table[i].data;
D
David Ahern 已提交
1379 1380 1381
		table[i].extra1 = mdev;
		table[i].extra2 = net;
	}
1382 1383 1384

	snprintf(path, sizeof(path), "net/mpls/conf/%s", dev->name);

1385
	mdev->sysctl = register_net_sysctl(net, path, table);
1386 1387 1388
	if (!mdev->sysctl)
		goto free;

1389
	mpls_netconf_notify_devconf(net, RTM_NEWNETCONF, NETCONFA_ALL, mdev);
1390 1391 1392 1393 1394 1395 1396 1397
	return 0;

free:
	kfree(table);
out:
	return -ENOBUFS;
}

1398 1399
static void mpls_dev_sysctl_unregister(struct net_device *dev,
				       struct mpls_dev *mdev)
1400
{
1401
	struct net *net = dev_net(dev);
1402 1403 1404 1405 1406
	struct ctl_table *table;

	table = mdev->sysctl->ctl_table_arg;
	unregister_net_sysctl_table(mdev->sysctl);
	kfree(table);
1407 1408

	mpls_netconf_notify_devconf(net, RTM_DELNETCONF, 0, mdev);
1409 1410
}

R
Robert Shearman 已提交
1411 1412 1413 1414
static struct mpls_dev *mpls_add_dev(struct net_device *dev)
{
	struct mpls_dev *mdev;
	int err = -ENOMEM;
R
Robert Shearman 已提交
1415
	int i;
R
Robert Shearman 已提交
1416 1417 1418 1419 1420 1421 1422

	ASSERT_RTNL();

	mdev = kzalloc(sizeof(*mdev), GFP_KERNEL);
	if (!mdev)
		return ERR_PTR(err);

R
Robert Shearman 已提交
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433
	mdev->stats = alloc_percpu(struct mpls_pcpu_stats);
	if (!mdev->stats)
		goto free;

	for_each_possible_cpu(i) {
		struct mpls_pcpu_stats *mpls_stats;

		mpls_stats = per_cpu_ptr(mdev->stats, i);
		u64_stats_init(&mpls_stats->syncp);
	}

1434 1435
	mdev->dev = dev;

1436 1437 1438 1439
	err = mpls_dev_sysctl_register(dev, mdev);
	if (err)
		goto free;

R
Robert Shearman 已提交
1440 1441 1442
	rcu_assign_pointer(dev->mpls_ptr, mdev);

	return mdev;
1443 1444

free:
R
Robert Shearman 已提交
1445
	free_percpu(mdev->stats);
1446 1447
	kfree(mdev);
	return ERR_PTR(err);
R
Robert Shearman 已提交
1448 1449
}

R
Robert Shearman 已提交
1450 1451 1452 1453 1454 1455 1456 1457
static void mpls_dev_destroy_rcu(struct rcu_head *head)
{
	struct mpls_dev *mdev = container_of(head, struct mpls_dev, rcu);

	free_percpu(mdev->stats);
	kfree(mdev);
}

R
Roopa Prabhu 已提交
1458
static void mpls_ifdown(struct net_device *dev, int event)
E
Eric W. Biederman 已提交
1459
{
1460
	struct mpls_route __rcu **platform_label;
E
Eric W. Biederman 已提交
1461
	struct net *net = dev_net(dev);
1462
	u8 alive, deleted;
E
Eric W. Biederman 已提交
1463 1464
	unsigned index;

1465
	platform_label = rtnl_dereference(net->mpls.platform_label);
E
Eric W. Biederman 已提交
1466
	for (index = 0; index < net->mpls.platform_labels; index++) {
1467
		struct mpls_route *rt = rtnl_dereference(platform_label[index]);
R
Roopa Prabhu 已提交
1468

E
Eric W. Biederman 已提交
1469 1470
		if (!rt)
			continue;
R
Roopa Prabhu 已提交
1471

1472
		alive = 0;
1473
		deleted = 0;
R
Roopa Prabhu 已提交
1474
		change_nexthops(rt) {
1475 1476
			unsigned int nh_flags = nh->nh_flags;

R
Roopa Prabhu 已提交
1477
			if (rtnl_dereference(nh->nh_dev) != dev)
1478 1479
				goto next;

R
Roopa Prabhu 已提交
1480 1481 1482
			switch (event) {
			case NETDEV_DOWN:
			case NETDEV_UNREGISTER:
1483
				nh_flags |= RTNH_F_DEAD;
R
Roopa Prabhu 已提交
1484 1485
				/* fall through */
			case NETDEV_CHANGE:
1486
				nh_flags |= RTNH_F_LINKDOWN;
R
Roopa Prabhu 已提交
1487 1488 1489 1490
				break;
			}
			if (event == NETDEV_UNREGISTER)
				RCU_INIT_POINTER(nh->nh_dev, NULL);
1491 1492 1493

			if (nh->nh_flags != nh_flags)
				WRITE_ONCE(nh->nh_flags, nh_flags);
1494
next:
1495
			if (!(nh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN)))
1496
				alive++;
1497 1498
			if (!rtnl_dereference(nh->nh_dev))
				deleted++;
R
Roopa Prabhu 已提交
1499
		} endfor_nexthops(rt);
1500 1501

		WRITE_ONCE(rt->rt_nhn_alive, alive);
1502 1503 1504 1505

		/* if there are no more nexthops, delete the route */
		if (event == NETDEV_UNREGISTER && deleted == rt->rt_nhn)
			mpls_route_update(net, index, NULL, NULL);
E
Eric W. Biederman 已提交
1506
	}
R
Roopa Prabhu 已提交
1507 1508
}

1509
static void mpls_ifup(struct net_device *dev, unsigned int flags)
R
Roopa Prabhu 已提交
1510 1511 1512 1513
{
	struct mpls_route __rcu **platform_label;
	struct net *net = dev_net(dev);
	unsigned index;
1514
	u8 alive;
R
Roopa Prabhu 已提交
1515 1516 1517 1518 1519 1520 1521

	platform_label = rtnl_dereference(net->mpls.platform_label);
	for (index = 0; index < net->mpls.platform_labels; index++) {
		struct mpls_route *rt = rtnl_dereference(platform_label[index]);

		if (!rt)
			continue;
1522

R
Roopa Prabhu 已提交
1523 1524
		alive = 0;
		change_nexthops(rt) {
1525
			unsigned int nh_flags = nh->nh_flags;
R
Roopa Prabhu 已提交
1526 1527
			struct net_device *nh_dev =
				rtnl_dereference(nh->nh_dev);
R
Robert Shearman 已提交
1528

1529
			if (!(nh_flags & flags)) {
R
Roopa Prabhu 已提交
1530 1531 1532 1533 1534 1535
				alive++;
				continue;
			}
			if (nh_dev != dev)
				continue;
			alive++;
1536
			nh_flags &= ~flags;
1537
			WRITE_ONCE(nh->nh_flags, nh_flags);
R
Roopa Prabhu 已提交
1538 1539
		} endfor_nexthops(rt);

1540
		WRITE_ONCE(rt->rt_nhn_alive, alive);
R
Roopa Prabhu 已提交
1541
	}
E
Eric W. Biederman 已提交
1542 1543 1544 1545 1546 1547
}

static int mpls_dev_notify(struct notifier_block *this, unsigned long event,
			   void *ptr)
{
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
R
Robert Shearman 已提交
1548
	struct mpls_dev *mdev;
R
Roopa Prabhu 已提交
1549
	unsigned int flags;
E
Eric W. Biederman 已提交
1550

R
Roopa Prabhu 已提交
1551
	if (event == NETDEV_REGISTER) {
1552 1553 1554 1555

		/* For now just support Ethernet, IPGRE, IP6GRE, SIT and
		 * IPIP devices
		 */
1556 1557
		if (dev->type == ARPHRD_ETHER ||
		    dev->type == ARPHRD_LOOPBACK ||
1558
		    dev->type == ARPHRD_IPGRE ||
1559
		    dev->type == ARPHRD_IP6GRE ||
1560 1561
		    dev->type == ARPHRD_SIT ||
		    dev->type == ARPHRD_TUNNEL) {
R
Robert Shearman 已提交
1562 1563 1564 1565
			mdev = mpls_add_dev(dev);
			if (IS_ERR(mdev))
				return notifier_from_errno(PTR_ERR(mdev));
		}
R
Roopa Prabhu 已提交
1566 1567
		return NOTIFY_OK;
	}
R
Robert Shearman 已提交
1568

R
Roopa Prabhu 已提交
1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
	mdev = mpls_dev_get(dev);
	if (!mdev)
		return NOTIFY_OK;

	switch (event) {
	case NETDEV_DOWN:
		mpls_ifdown(dev, event);
		break;
	case NETDEV_UP:
		flags = dev_get_flags(dev);
		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
			mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
		else
			mpls_ifup(dev, RTNH_F_DEAD);
		break;
	case NETDEV_CHANGE:
		flags = dev_get_flags(dev);
		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
			mpls_ifup(dev, RTNH_F_DEAD | RTNH_F_LINKDOWN);
		else
			mpls_ifdown(dev, event);
		break;
E
Eric W. Biederman 已提交
1591
	case NETDEV_UNREGISTER:
R
Roopa Prabhu 已提交
1592 1593 1594
		mpls_ifdown(dev, event);
		mdev = mpls_dev_get(dev);
		if (mdev) {
1595
			mpls_dev_sysctl_unregister(dev, mdev);
R
Roopa Prabhu 已提交
1596
			RCU_INIT_POINTER(dev->mpls_ptr, NULL);
R
Robert Shearman 已提交
1597
			call_rcu(&mdev->rcu, mpls_dev_destroy_rcu);
R
Roopa Prabhu 已提交
1598
		}
E
Eric W. Biederman 已提交
1599
		break;
1600 1601 1602 1603 1604
	case NETDEV_CHANGENAME:
		mdev = mpls_dev_get(dev);
		if (mdev) {
			int err;

1605
			mpls_dev_sysctl_unregister(dev, mdev);
1606 1607 1608 1609 1610
			err = mpls_dev_sysctl_register(dev, mdev);
			if (err)
				return notifier_from_errno(err);
		}
		break;
E
Eric W. Biederman 已提交
1611 1612 1613 1614 1615 1616 1617 1618
	}
	return NOTIFY_OK;
}

static struct notifier_block mpls_dev_notifier = {
	.notifier_call = mpls_dev_notify,
};

1619
static int nla_put_via(struct sk_buff *skb,
1620
		       u8 table, const void *addr, int alen)
1621
{
1622 1623 1624
	static const int table_to_family[NEIGH_NR_TABLES + 1] = {
		AF_INET, AF_INET6, AF_DECnet, AF_PACKET,
	};
1625 1626
	struct nlattr *nla;
	struct rtvia *via;
1627
	int family = AF_UNSPEC;
1628 1629 1630 1631 1632

	nla = nla_reserve(skb, RTA_VIA, alen + 2);
	if (!nla)
		return -EMSGSIZE;

1633 1634 1635
	if (table <= NEIGH_NR_TABLES)
		family = table_to_family[table];

1636 1637 1638 1639 1640 1641
	via = nla_data(nla);
	via->rtvia_family = family;
	memcpy(via->rtvia_addr, addr, alen);
	return 0;
}

1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661
int nla_put_labels(struct sk_buff *skb, int attrtype,
		   u8 labels, const u32 label[])
{
	struct nlattr *nla;
	struct mpls_shim_hdr *nla_label;
	bool bos;
	int i;
	nla = nla_reserve(skb, attrtype, labels*4);
	if (!nla)
		return -EMSGSIZE;

	nla_label = nla_data(nla);
	bos = true;
	for (i = labels - 1; i >= 0; i--) {
		nla_label[i] = mpls_entry_encode(label[i], 0, 0, bos);
		bos = false;
	}

	return 0;
}
1662
EXPORT_SYMBOL_GPL(nla_put_labels);
1663

1664 1665
int nla_get_labels(const struct nlattr *nla, u8 max_labels, u8 *labels,
		   u32 label[], struct netlink_ext_ack *extack)
1666 1667 1668
{
	unsigned len = nla_len(nla);
	struct mpls_shim_hdr *nla_label;
1669
	u8 nla_labels;
1670 1671 1672
	bool bos;
	int i;

1673 1674 1675
	/* len needs to be an even multiple of 4 (the label size). Number
	 * of labels is a u8 so check for overflow.
	 */
1676 1677 1678
	if (len & 3 || len / 4 > 255) {
		NL_SET_ERR_MSG_ATTR(extack, nla,
				    "Invalid length for labels attribute");
1679
		return -EINVAL;
1680
	}
1681 1682 1683

	/* Limit the number of new labels allowed */
	nla_labels = len/4;
1684 1685
	if (nla_labels > max_labels) {
		NL_SET_ERR_MSG(extack, "Too many labels");
1686
		return -EINVAL;
1687
	}
1688

1689 1690 1691 1692
	/* when label == NULL, caller wants number of labels */
	if (!label)
		goto out;

1693 1694 1695 1696 1697 1698 1699 1700 1701
	nla_label = nla_data(nla);
	bos = true;
	for (i = nla_labels - 1; i >= 0; i--, bos = false) {
		struct mpls_entry_decoded dec;
		dec = mpls_entry_decode(nla_label + i);

		/* Ensure the bottom of stack flag is properly set
		 * and ttl and tc are both clear.
		 */
1702 1703 1704 1705 1706 1707 1708 1709 1710
		if (dec.ttl) {
			NL_SET_ERR_MSG_ATTR(extack, nla,
					    "TTL in label must be 0");
			return -EINVAL;
		}

		if (dec.tc) {
			NL_SET_ERR_MSG_ATTR(extack, nla,
					    "Traffic class in label must be 0");
1711
			return -EINVAL;
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724
		}

		if (dec.bos != bos) {
			NL_SET_BAD_ATTR(extack, nla);
			if (bos) {
				NL_SET_ERR_MSG(extack,
					       "BOS bit must be set in first label");
			} else {
				NL_SET_ERR_MSG(extack,
					       "BOS bit can only be set in first label");
			}
			return -EINVAL;
		}
1725

1726
		switch (dec.label) {
1727
		case MPLS_LABEL_IMPLNULL:
1728 1729 1730 1731
			/* RFC3032: This is a label that an LSR may
			 * assign and distribute, but which never
			 * actually appears in the encapsulation.
			 */
1732 1733
			NL_SET_ERR_MSG_ATTR(extack, nla,
					    "Implicit NULL Label (3) can not be used in encapsulation");
1734 1735 1736
			return -EINVAL;
		}

1737 1738
		label[i] = dec.label;
	}
1739
out:
1740 1741 1742
	*labels = nla_labels;
	return 0;
}
1743
EXPORT_SYMBOL_GPL(nla_get_labels);
1744

1745 1746 1747 1748
static int rtm_to_route_config(struct sk_buff *skb,
			       struct nlmsghdr *nlh,
			       struct mpls_route_config *cfg,
			       struct netlink_ext_ack *extack)
1749 1750 1751 1752 1753 1754
{
	struct rtmsg *rtm;
	struct nlattr *tb[RTA_MAX+1];
	int index;
	int err;

1755
	err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_mpls_policy,
1756
			  extack);
1757 1758 1759 1760 1761 1762
	if (err < 0)
		goto errout;

	err = -EINVAL;
	rtm = nlmsg_data(nlh);

1763 1764
	if (rtm->rtm_family != AF_MPLS) {
		NL_SET_ERR_MSG(extack, "Invalid address family in rtmsg");
1765
		goto errout;
1766 1767 1768
	}
	if (rtm->rtm_dst_len != 20) {
		NL_SET_ERR_MSG(extack, "rtm_dst_len must be 20 for MPLS");
1769
		goto errout;
1770 1771 1772
	}
	if (rtm->rtm_src_len != 0) {
		NL_SET_ERR_MSG(extack, "rtm_src_len must be 0 for MPLS");
1773
		goto errout;
1774 1775 1776
	}
	if (rtm->rtm_tos != 0) {
		NL_SET_ERR_MSG(extack, "rtm_tos must be 0 for MPLS");
1777
		goto errout;
1778 1779 1780 1781
	}
	if (rtm->rtm_table != RT_TABLE_MAIN) {
		NL_SET_ERR_MSG(extack,
			       "MPLS only supports the main route table");
1782
		goto errout;
1783
	}
1784 1785 1786 1787 1788 1789
	/* Any value is acceptable for rtm_protocol */

	/* As mpls uses destination specific addresses
	 * (or source specific address in the case of multicast)
	 * all addresses have universal scope.
	 */
1790 1791 1792
	if (rtm->rtm_scope != RT_SCOPE_UNIVERSE) {
		NL_SET_ERR_MSG(extack,
			       "Invalid route scope  - MPLS only supports UNIVERSE");
1793
		goto errout;
1794 1795 1796 1797
	}
	if (rtm->rtm_type != RTN_UNICAST) {
		NL_SET_ERR_MSG(extack,
			       "Invalid route type - MPLS only supports UNICAST");
1798
		goto errout;
1799 1800 1801
	}
	if (rtm->rtm_flags != 0) {
		NL_SET_ERR_MSG(extack, "rtm_flags must be 0 for MPLS");
1802
		goto errout;
1803
	}
1804 1805 1806

	cfg->rc_label		= LABEL_NOT_SPECIFIED;
	cfg->rc_protocol	= rtm->rtm_protocol;
1807
	cfg->rc_via_table	= MPLS_NEIGH_TABLE_UNSPEC;
1808
	cfg->rc_ttl_propagate	= MPLS_TTL_PROP_DEFAULT;
1809 1810 1811 1812 1813 1814 1815 1816 1817 1818
	cfg->rc_nlflags		= nlh->nlmsg_flags;
	cfg->rc_nlinfo.portid	= NETLINK_CB(skb).portid;
	cfg->rc_nlinfo.nlh	= nlh;
	cfg->rc_nlinfo.nl_net	= sock_net(skb->sk);

	for (index = 0; index <= RTA_MAX; index++) {
		struct nlattr *nla = tb[index];
		if (!nla)
			continue;

1819
		switch (index) {
1820 1821 1822 1823 1824 1825
		case RTA_OIF:
			cfg->rc_ifindex = nla_get_u32(nla);
			break;
		case RTA_NEWDST:
			if (nla_get_labels(nla, MAX_NEW_LABELS,
					   &cfg->rc_output_labels,
1826
					   cfg->rc_output_label, extack))
1827 1828 1829 1830
				goto errout;
			break;
		case RTA_DST:
		{
R
Roopa Prabhu 已提交
1831
			u8 label_count;
1832
			if (nla_get_labels(nla, 1, &label_count,
1833
					   &cfg->rc_label, extack))
1834 1835
				goto errout;

1836
			if (!mpls_label_ok(cfg->rc_nlinfo.nl_net,
1837
					   &cfg->rc_label, extack))
1838 1839 1840 1841 1842
				goto errout;
			break;
		}
		case RTA_VIA:
		{
R
Roopa Prabhu 已提交
1843
			if (nla_get_via(nla, &cfg->rc_via_alen,
1844 1845
					&cfg->rc_via_table, cfg->rc_via,
					extack))
1846
				goto errout;
R
Roopa Prabhu 已提交
1847 1848 1849 1850 1851 1852
			break;
		}
		case RTA_MULTIPATH:
		{
			cfg->rc_mp = nla_data(nla);
			cfg->rc_mp_len = nla_len(nla);
1853 1854
			break;
		}
1855 1856 1857 1858
		case RTA_TTL_PROPAGATE:
		{
			u8 ttl_propagate = nla_get_u8(nla);

1859 1860 1861
			if (ttl_propagate > 1) {
				NL_SET_ERR_MSG_ATTR(extack, nla,
						    "RTA_TTL_PROPAGATE can only be 0 or 1");
1862
				goto errout;
1863
			}
1864 1865 1866 1867 1868
			cfg->rc_ttl_propagate = ttl_propagate ?
				MPLS_TTL_PROP_ENABLED :
				MPLS_TTL_PROP_DISABLED;
			break;
		}
1869
		default:
1870
			NL_SET_ERR_MSG_ATTR(extack, nla, "Unknown attribute");
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880
			/* Unsupported attribute */
			goto errout;
		}
	}

	err = 0;
errout:
	return err;
}

1881 1882
static int mpls_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh,
			     struct netlink_ext_ack *extack)
1883
{
1884
	struct mpls_route_config *cfg;
1885 1886
	int err;

1887 1888 1889 1890
	cfg = kzalloc(sizeof(*cfg), GFP_KERNEL);
	if (!cfg)
		return -ENOMEM;

1891
	err = rtm_to_route_config(skb, nlh, cfg, extack);
1892
	if (err < 0)
1893
		goto out;
1894

1895
	err = mpls_route_del(cfg, extack);
1896 1897 1898 1899
out:
	kfree(cfg);

	return err;
1900 1901 1902
}


1903 1904
static int mpls_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh,
			     struct netlink_ext_ack *extack)
1905
{
1906
	struct mpls_route_config *cfg;
1907 1908
	int err;

1909 1910 1911 1912
	cfg = kzalloc(sizeof(*cfg), GFP_KERNEL);
	if (!cfg)
		return -ENOMEM;

1913
	err = rtm_to_route_config(skb, nlh, cfg, extack);
1914
	if (err < 0)
1915
		goto out;
1916

1917
	err = mpls_route_add(cfg, extack);
1918 1919 1920 1921
out:
	kfree(cfg);

	return err;
1922 1923 1924 1925 1926
}

static int mpls_dump_route(struct sk_buff *skb, u32 portid, u32 seq, int event,
			   u32 label, struct mpls_route *rt, int flags)
{
1927
	struct net_device *dev;
1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947
	struct nlmsghdr *nlh;
	struct rtmsg *rtm;

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*rtm), flags);
	if (nlh == NULL)
		return -EMSGSIZE;

	rtm = nlmsg_data(nlh);
	rtm->rtm_family = AF_MPLS;
	rtm->rtm_dst_len = 20;
	rtm->rtm_src_len = 0;
	rtm->rtm_tos = 0;
	rtm->rtm_table = RT_TABLE_MAIN;
	rtm->rtm_protocol = rt->rt_protocol;
	rtm->rtm_scope = RT_SCOPE_UNIVERSE;
	rtm->rtm_type = RTN_UNICAST;
	rtm->rtm_flags = 0;

	if (nla_put_labels(skb, RTA_DST, 1, &label))
		goto nla_put_failure;
1948 1949 1950 1951 1952 1953 1954 1955 1956

	if (rt->rt_ttl_propagate != MPLS_TTL_PROP_DEFAULT) {
		bool ttl_propagate =
			rt->rt_ttl_propagate == MPLS_TTL_PROP_ENABLED;

		if (nla_put_u8(skb, RTA_TTL_PROPAGATE,
			       ttl_propagate))
			goto nla_put_failure;
	}
R
Roopa Prabhu 已提交
1957
	if (rt->rt_nhn == 1) {
1958
		const struct mpls_nh *nh = rt->rt_nh;
R
Roopa Prabhu 已提交
1959 1960 1961 1962 1963

		if (nh->nh_labels &&
		    nla_put_labels(skb, RTA_NEWDST, nh->nh_labels,
				   nh->nh_label))
			goto nla_put_failure;
1964
		if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC &&
1965
		    nla_put_via(skb, nh->nh_via_table, mpls_nh_via(rt, nh),
R
Roopa Prabhu 已提交
1966 1967 1968 1969 1970
				nh->nh_via_alen))
			goto nla_put_failure;
		dev = rtnl_dereference(nh->nh_dev);
		if (dev && nla_put_u32(skb, RTA_OIF, dev->ifindex))
			goto nla_put_failure;
R
Roopa Prabhu 已提交
1971 1972 1973 1974
		if (nh->nh_flags & RTNH_F_LINKDOWN)
			rtm->rtm_flags |= RTNH_F_LINKDOWN;
		if (nh->nh_flags & RTNH_F_DEAD)
			rtm->rtm_flags |= RTNH_F_DEAD;
R
Roopa Prabhu 已提交
1975 1976 1977
	} else {
		struct rtnexthop *rtnh;
		struct nlattr *mp;
1978 1979
		u8 linkdown = 0;
		u8 dead = 0;
R
Roopa Prabhu 已提交
1980 1981 1982 1983 1984 1985

		mp = nla_nest_start(skb, RTA_MULTIPATH);
		if (!mp)
			goto nla_put_failure;

		for_nexthops(rt) {
1986 1987 1988 1989
			dev = rtnl_dereference(nh->nh_dev);
			if (!dev)
				continue;

R
Roopa Prabhu 已提交
1990 1991 1992 1993
			rtnh = nla_reserve_nohdr(skb, sizeof(*rtnh));
			if (!rtnh)
				goto nla_put_failure;

1994
			rtnh->rtnh_ifindex = dev->ifindex;
R
Roopa Prabhu 已提交
1995 1996 1997 1998 1999 2000 2001 2002 2003
			if (nh->nh_flags & RTNH_F_LINKDOWN) {
				rtnh->rtnh_flags |= RTNH_F_LINKDOWN;
				linkdown++;
			}
			if (nh->nh_flags & RTNH_F_DEAD) {
				rtnh->rtnh_flags |= RTNH_F_DEAD;
				dead++;
			}

R
Roopa Prabhu 已提交
2004 2005 2006 2007
			if (nh->nh_labels && nla_put_labels(skb, RTA_NEWDST,
							    nh->nh_labels,
							    nh->nh_label))
				goto nla_put_failure;
2008 2009
			if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC &&
			    nla_put_via(skb, nh->nh_via_table,
2010
					mpls_nh_via(rt, nh),
R
Roopa Prabhu 已提交
2011 2012 2013 2014 2015 2016 2017
					nh->nh_via_alen))
				goto nla_put_failure;

			/* length of rtnetlink header + attributes */
			rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *)rtnh;
		} endfor_nexthops(rt);

R
Roopa Prabhu 已提交
2018 2019 2020 2021 2022
		if (linkdown == rt->rt_nhn)
			rtm->rtm_flags |= RTNH_F_LINKDOWN;
		if (dead == rt->rt_nhn)
			rtm->rtm_flags |= RTNH_F_DEAD;

R
Roopa Prabhu 已提交
2023 2024
		nla_nest_end(skb, mp);
	}
2025 2026 2027 2028 2029 2030 2031 2032 2033

	nlmsg_end(skb, nlh);
	return 0;

nla_put_failure:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
}

2034
#if IS_ENABLED(CONFIG_INET)
D
David Ahern 已提交
2035 2036
static int mpls_valid_fib_dump_req(struct net *net, const struct nlmsghdr *nlh,
				   struct fib_dump_filter *filter,
2037
				   struct netlink_callback *cb)
2038
{
2039
	return ip_valid_fib_dump_req(net, nlh, filter, cb);
2040 2041
}
#else
D
David Ahern 已提交
2042 2043
static int mpls_valid_fib_dump_req(struct net *net, const struct nlmsghdr *nlh,
				   struct fib_dump_filter *filter,
2044
				   struct netlink_callback *cb)
2045
{
2046
	struct netlink_ext_ack *extack = cb->extack;
2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070
	struct rtmsg *rtm;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*rtm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid header for FIB dump request");
		return -EINVAL;
	}

	rtm = nlmsg_data(nlh);
	if (rtm->rtm_dst_len || rtm->rtm_src_len  || rtm->rtm_tos   ||
	    rtm->rtm_table   || rtm->rtm_protocol || rtm->rtm_scope ||
	    rtm->rtm_type    || rtm->rtm_flags) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid values in header for FIB dump request");
		return -EINVAL;
	}

	if (nlmsg_attrlen(nlh, sizeof(*rtm))) {
		NL_SET_ERR_MSG_MOD(extack, "Invalid data after header in FIB dump request");
		return -EINVAL;
	}

	return 0;
}
#endif

2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092
static bool mpls_rt_uses_dev(struct mpls_route *rt,
			     const struct net_device *dev)
{
	struct net_device *nh_dev;

	if (rt->rt_nhn == 1) {
		struct mpls_nh *nh = rt->rt_nh;

		nh_dev = rtnl_dereference(nh->nh_dev);
		if (dev == nh_dev)
			return true;
	} else {
		for_nexthops(rt) {
			nh_dev = rtnl_dereference(nh->nh_dev);
			if (nh_dev == dev)
				return true;
		} endfor_nexthops(rt);
	}

	return false;
}

2093 2094
static int mpls_dump_routes(struct sk_buff *skb, struct netlink_callback *cb)
{
2095
	const struct nlmsghdr *nlh = cb->nlh;
2096
	struct net *net = sock_net(skb->sk);
2097
	struct mpls_route __rcu **platform_label;
D
David Ahern 已提交
2098
	struct fib_dump_filter filter = {};
2099
	unsigned int flags = NLM_F_MULTI;
2100
	size_t platform_labels;
2101 2102 2103 2104
	unsigned int index;

	ASSERT_RTNL();

2105
	if (cb->strict_check) {
D
David Ahern 已提交
2106
		int err;
2107

2108
		err = mpls_valid_fib_dump_req(net, nlh, &filter, cb);
2109 2110
		if (err < 0)
			return err;
2111 2112 2113 2114 2115 2116 2117 2118

		/* for MPLS, there is only 1 table with fixed type and flags.
		 * If either are set in the filter then return nothing.
		 */
		if ((filter.table_id && filter.table_id != RT_TABLE_MAIN) ||
		    (filter.rt_type && filter.rt_type != RTN_UNICAST) ||
		     filter.flags)
			return skb->len;
2119 2120
	}

2121
	index = cb->args[0];
2122 2123
	if (index < MPLS_LABEL_FIRST_UNRESERVED)
		index = MPLS_LABEL_FIRST_UNRESERVED;
2124

2125 2126
	platform_label = rtnl_dereference(net->mpls.platform_label);
	platform_labels = net->mpls.platform_labels;
2127 2128 2129 2130

	if (filter.filter_set)
		flags |= NLM_F_DUMP_FILTERED;

2131
	for (; index < platform_labels; index++) {
2132
		struct mpls_route *rt;
2133

2134
		rt = rtnl_dereference(platform_label[index]);
2135 2136 2137
		if (!rt)
			continue;

2138 2139 2140 2141
		if ((filter.dev && !mpls_rt_uses_dev(rt, filter.dev)) ||
		    (filter.protocol && rt->rt_protocol != filter.protocol))
			continue;

2142 2143
		if (mpls_dump_route(skb, NETLINK_CB(cb->skb).portid,
				    cb->nlh->nlmsg_seq, RTM_NEWROUTE,
2144
				    index, rt, flags) < 0)
2145 2146 2147 2148 2149 2150 2151
			break;
	}
	cb->args[0] = index;

	return skb->len;
}

2152 2153 2154 2155
static inline size_t lfib_nlmsg_size(struct mpls_route *rt)
{
	size_t payload =
		NLMSG_ALIGN(sizeof(struct rtmsg))
2156 2157
		+ nla_total_size(4)			/* RTA_DST */
		+ nla_total_size(1);			/* RTA_TTL_PROPAGATE */
R
Roopa Prabhu 已提交
2158 2159 2160 2161 2162 2163

	if (rt->rt_nhn == 1) {
		struct mpls_nh *nh = rt->rt_nh;

		if (nh->nh_dev)
			payload += nla_total_size(4); /* RTA_OIF */
2164
		if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC) /* RTA_VIA */
2165
			payload += nla_total_size(2 + nh->nh_via_alen);
R
Roopa Prabhu 已提交
2166 2167 2168 2169 2170 2171 2172
		if (nh->nh_labels) /* RTA_NEWDST */
			payload += nla_total_size(nh->nh_labels * 4);
	} else {
		/* each nexthop is packed in an attribute */
		size_t nhsize = 0;

		for_nexthops(rt) {
2173 2174
			if (!rtnl_dereference(nh->nh_dev))
				continue;
R
Roopa Prabhu 已提交
2175
			nhsize += nla_total_size(sizeof(struct rtnexthop));
2176 2177 2178
			/* RTA_VIA */
			if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC)
				nhsize += nla_total_size(2 + nh->nh_via_alen);
R
Roopa Prabhu 已提交
2179 2180 2181 2182 2183 2184 2185
			if (nh->nh_labels)
				nhsize += nla_total_size(nh->nh_labels * 4);
		} endfor_nexthops(rt);
		/* nested attribute */
		payload += nla_total_size(nhsize);
	}

2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215
	return payload;
}

static void rtmsg_lfib(int event, u32 label, struct mpls_route *rt,
		       struct nlmsghdr *nlh, struct net *net, u32 portid,
		       unsigned int nlm_flags)
{
	struct sk_buff *skb;
	u32 seq = nlh ? nlh->nlmsg_seq : 0;
	int err = -ENOBUFS;

	skb = nlmsg_new(lfib_nlmsg_size(rt), GFP_KERNEL);
	if (skb == NULL)
		goto errout;

	err = mpls_dump_route(skb, portid, seq, event, label, rt, nlm_flags);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in lfib_nlmsg_size */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	rtnl_notify(skb, net, portid, RTNLGRP_MPLS_ROUTE, nlh, GFP_KERNEL);

	return;
errout:
	if (err < 0)
		rtnl_set_sk_err(net, RTNLGRP_MPLS_ROUTE, err);
}

R
Roopa Prabhu 已提交
2216 2217 2218 2219 2220
static int mpls_getroute(struct sk_buff *in_skb, struct nlmsghdr *in_nlh,
			 struct netlink_ext_ack *extack)
{
	struct net *net = sock_net(in_skb->sk);
	u32 portid = NETLINK_CB(in_skb).portid;
2221
	u32 in_label = LABEL_NOT_SPECIFIED;
R
Roopa Prabhu 已提交
2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232
	struct nlattr *tb[RTA_MAX + 1];
	u32 labels[MAX_NEW_LABELS];
	struct mpls_shim_hdr *hdr;
	unsigned int hdr_size = 0;
	struct net_device *dev;
	struct mpls_route *rt;
	struct rtmsg *rtm, *r;
	struct nlmsghdr *nlh;
	struct sk_buff *skb;
	struct mpls_nh *nh;
	u8 n_labels;
2233
	int err;
R
Roopa Prabhu 已提交
2234 2235

	err = nlmsg_parse(in_nlh, sizeof(*rtm), tb, RTA_MAX,
2236
			  rtm_mpls_policy, extack);
R
Roopa Prabhu 已提交
2237 2238 2239 2240 2241 2242 2243 2244 2245
	if (err < 0)
		goto errout;

	rtm = nlmsg_data(in_nlh);

	if (tb[RTA_DST]) {
		u8 label_count;

		if (nla_get_labels(tb[RTA_DST], 1, &label_count,
2246 2247
				   &in_label, extack)) {
			err = -EINVAL;
R
Roopa Prabhu 已提交
2248
			goto errout;
2249
		}
R
Roopa Prabhu 已提交
2250

2251
		if (!mpls_label_ok(net, &in_label, extack)) {
2252
			err = -EINVAL;
R
Roopa Prabhu 已提交
2253
			goto errout;
2254
		}
R
Roopa Prabhu 已提交
2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379
	}

	rt = mpls_route_input_rcu(net, in_label);
	if (!rt) {
		err = -ENETUNREACH;
		goto errout;
	}

	if (rtm->rtm_flags & RTM_F_FIB_MATCH) {
		skb = nlmsg_new(lfib_nlmsg_size(rt), GFP_KERNEL);
		if (!skb) {
			err = -ENOBUFS;
			goto errout;
		}

		err = mpls_dump_route(skb, portid, in_nlh->nlmsg_seq,
				      RTM_NEWROUTE, in_label, rt, 0);
		if (err < 0) {
			/* -EMSGSIZE implies BUG in lfib_nlmsg_size */
			WARN_ON(err == -EMSGSIZE);
			goto errout_free;
		}

		return rtnl_unicast(skb, net, portid);
	}

	if (tb[RTA_NEWDST]) {
		if (nla_get_labels(tb[RTA_NEWDST], MAX_NEW_LABELS, &n_labels,
				   labels, extack) != 0) {
			err = -EINVAL;
			goto errout;
		}

		hdr_size = n_labels * sizeof(struct mpls_shim_hdr);
	}

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb) {
		err = -ENOBUFS;
		goto errout;
	}

	skb->protocol = htons(ETH_P_MPLS_UC);

	if (hdr_size) {
		bool bos;
		int i;

		if (skb_cow(skb, hdr_size)) {
			err = -ENOBUFS;
			goto errout_free;
		}

		skb_reserve(skb, hdr_size);
		skb_push(skb, hdr_size);
		skb_reset_network_header(skb);

		/* Push new labels */
		hdr = mpls_hdr(skb);
		bos = true;
		for (i = n_labels - 1; i >= 0; i--) {
			hdr[i] = mpls_entry_encode(labels[i],
						   1, 0, bos);
			bos = false;
		}
	}

	nh = mpls_select_multipath(rt, skb);
	if (!nh) {
		err = -ENETUNREACH;
		goto errout_free;
	}

	if (hdr_size) {
		skb_pull(skb, hdr_size);
		skb_reset_network_header(skb);
	}

	nlh = nlmsg_put(skb, portid, in_nlh->nlmsg_seq,
			RTM_NEWROUTE, sizeof(*r), 0);
	if (!nlh) {
		err = -EMSGSIZE;
		goto errout_free;
	}

	r = nlmsg_data(nlh);
	r->rtm_family	 = AF_MPLS;
	r->rtm_dst_len	= 20;
	r->rtm_src_len	= 0;
	r->rtm_table	= RT_TABLE_MAIN;
	r->rtm_type	= RTN_UNICAST;
	r->rtm_scope	= RT_SCOPE_UNIVERSE;
	r->rtm_protocol = rt->rt_protocol;
	r->rtm_flags	= 0;

	if (nla_put_labels(skb, RTA_DST, 1, &in_label))
		goto nla_put_failure;

	if (nh->nh_labels &&
	    nla_put_labels(skb, RTA_NEWDST, nh->nh_labels,
			   nh->nh_label))
		goto nla_put_failure;

	if (nh->nh_via_table != MPLS_NEIGH_TABLE_UNSPEC &&
	    nla_put_via(skb, nh->nh_via_table, mpls_nh_via(rt, nh),
			nh->nh_via_alen))
		goto nla_put_failure;
	dev = rtnl_dereference(nh->nh_dev);
	if (dev && nla_put_u32(skb, RTA_OIF, dev->ifindex))
		goto nla_put_failure;

	nlmsg_end(skb, nlh);

	err = rtnl_unicast(skb, net, portid);
errout:
	return err;

nla_put_failure:
	nlmsg_cancel(skb, nlh);
	err = -EMSGSIZE;
errout_free:
	kfree_skb(skb);
	return err;
}

2380 2381 2382 2383 2384 2385 2386 2387 2388 2389
static int resize_platform_label_table(struct net *net, size_t limit)
{
	size_t size = sizeof(struct mpls_route *) * limit;
	size_t old_limit;
	size_t cp_size;
	struct mpls_route __rcu **labels = NULL, **old;
	struct mpls_route *rt0 = NULL, *rt2 = NULL;
	unsigned index;

	if (size) {
2390
		labels = kvzalloc(size, GFP_KERNEL);
2391 2392 2393 2394 2395
		if (!labels)
			goto nolabels;
	}

	/* In case the predefined labels need to be populated */
2396
	if (limit > MPLS_LABEL_IPV4NULL) {
2397
		struct net_device *lo = net->loopback_dev;
2398
		rt0 = mpls_rt_alloc(1, lo->addr_len, 0);
2399
		if (IS_ERR(rt0))
2400
			goto nort0;
R
Roopa Prabhu 已提交
2401
		RCU_INIT_POINTER(rt0->rt_nh->nh_dev, lo);
2402
		rt0->rt_protocol = RTPROT_KERNEL;
2403
		rt0->rt_payload_type = MPT_IPV4;
2404
		rt0->rt_ttl_propagate = MPLS_TTL_PROP_DEFAULT;
R
Roopa Prabhu 已提交
2405
		rt0->rt_nh->nh_via_table = NEIGH_LINK_TABLE;
2406
		rt0->rt_nh->nh_via_alen = lo->addr_len;
2407 2408
		memcpy(__mpls_nh_via(rt0, rt0->rt_nh), lo->dev_addr,
		       lo->addr_len);
2409
	}
2410
	if (limit > MPLS_LABEL_IPV6NULL) {
2411
		struct net_device *lo = net->loopback_dev;
2412
		rt2 = mpls_rt_alloc(1, lo->addr_len, 0);
2413
		if (IS_ERR(rt2))
2414
			goto nort2;
R
Roopa Prabhu 已提交
2415
		RCU_INIT_POINTER(rt2->rt_nh->nh_dev, lo);
2416
		rt2->rt_protocol = RTPROT_KERNEL;
2417
		rt2->rt_payload_type = MPT_IPV6;
2418
		rt2->rt_ttl_propagate = MPLS_TTL_PROP_DEFAULT;
R
Roopa Prabhu 已提交
2419
		rt2->rt_nh->nh_via_table = NEIGH_LINK_TABLE;
2420
		rt2->rt_nh->nh_via_alen = lo->addr_len;
2421 2422
		memcpy(__mpls_nh_via(rt2, rt2->rt_nh), lo->dev_addr,
		       lo->addr_len);
2423 2424 2425 2426
	}

	rtnl_lock();
	/* Remember the original table */
2427
	old = rtnl_dereference(net->mpls.platform_label);
2428 2429 2430 2431
	old_limit = net->mpls.platform_labels;

	/* Free any labels beyond the new table */
	for (index = limit; index < old_limit; index++)
R
Roopa Prabhu 已提交
2432
		mpls_route_update(net, index, NULL, NULL);
2433 2434 2435 2436 2437 2438 2439 2440 2441

	/* Copy over the old labels */
	cp_size = size;
	if (old_limit < limit)
		cp_size = old_limit * sizeof(struct mpls_route *);

	memcpy(labels, old, cp_size);

	/* If needed set the predefined labels */
2442 2443 2444
	if ((old_limit <= MPLS_LABEL_IPV6NULL) &&
	    (limit > MPLS_LABEL_IPV6NULL)) {
		RCU_INIT_POINTER(labels[MPLS_LABEL_IPV6NULL], rt2);
2445 2446 2447
		rt2 = NULL;
	}

2448 2449 2450
	if ((old_limit <= MPLS_LABEL_IPV4NULL) &&
	    (limit > MPLS_LABEL_IPV4NULL)) {
		RCU_INIT_POINTER(labels[MPLS_LABEL_IPV4NULL], rt0);
2451 2452 2453 2454 2455
		rt0 = NULL;
	}

	/* Update the global pointers */
	net->mpls.platform_labels = limit;
2456
	rcu_assign_pointer(net->mpls.platform_label, labels);
2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499

	rtnl_unlock();

	mpls_rt_free(rt2);
	mpls_rt_free(rt0);

	if (old) {
		synchronize_rcu();
		kvfree(old);
	}
	return 0;

nort2:
	mpls_rt_free(rt0);
nort0:
	kvfree(labels);
nolabels:
	return -ENOMEM;
}

static int mpls_platform_labels(struct ctl_table *table, int write,
				void __user *buffer, size_t *lenp, loff_t *ppos)
{
	struct net *net = table->data;
	int platform_labels = net->mpls.platform_labels;
	int ret;
	struct ctl_table tmp = {
		.procname	= table->procname,
		.data		= &platform_labels,
		.maxlen		= sizeof(int),
		.mode		= table->mode,
		.extra1		= &zero,
		.extra2		= &label_limit,
	};

	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);

	if (write && ret == 0)
		ret = resize_platform_label_table(net, platform_labels);

	return ret;
}

2500 2501 2502
#define MPLS_NS_SYSCTL_OFFSET(field)		\
	(&((struct net *)0)->field)

2503
static const struct ctl_table mpls_table[] = {
2504 2505 2506 2507 2508 2509 2510
	{
		.procname	= "platform_labels",
		.data		= NULL,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= mpls_platform_labels,
	},
2511 2512 2513 2514 2515 2516 2517 2518 2519
	{
		.procname	= "ip_ttl_propagate",
		.data		= MPLS_NS_SYSCTL_OFFSET(mpls.ip_ttl_propagate),
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one,
	},
2520 2521 2522 2523 2524 2525 2526 2527 2528
	{
		.procname	= "default_ttl",
		.data		= MPLS_NS_SYSCTL_OFFSET(mpls.default_ttl),
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &one,
		.extra2		= &ttl_max,
	},
2529 2530 2531
	{ }
};

E
Eric W. Biederman 已提交
2532 2533
static int mpls_net_init(struct net *net)
{
2534
	struct ctl_table *table;
2535
	int i;
2536

E
Eric W. Biederman 已提交
2537 2538
	net->mpls.platform_labels = 0;
	net->mpls.platform_label = NULL;
2539
	net->mpls.ip_ttl_propagate = 1;
2540
	net->mpls.default_ttl = 255;
E
Eric W. Biederman 已提交
2541

2542 2543 2544 2545
	table = kmemdup(mpls_table, sizeof(mpls_table), GFP_KERNEL);
	if (table == NULL)
		return -ENOMEM;

2546 2547 2548 2549 2550 2551
	/* Table data contains only offsets relative to the base of
	 * the mdev at this point, so make them absolute.
	 */
	for (i = 0; i < ARRAY_SIZE(mpls_table) - 1; i++)
		table[i].data = (char *)net + (uintptr_t)table[i].data;

2552
	net->mpls.ctl = register_net_sysctl(net, "net/mpls", table);
2553 2554
	if (net->mpls.ctl == NULL) {
		kfree(table);
2555
		return -ENOMEM;
2556
	}
2557

E
Eric W. Biederman 已提交
2558 2559 2560 2561 2562
	return 0;
}

static void mpls_net_exit(struct net *net)
{
2563 2564
	struct mpls_route __rcu **platform_label;
	size_t platform_labels;
2565
	struct ctl_table *table;
E
Eric W. Biederman 已提交
2566 2567
	unsigned int index;

2568 2569 2570 2571
	table = net->mpls.ctl->ctl_table_arg;
	unregister_net_sysctl_table(net->mpls.ctl);
	kfree(table);

2572 2573
	/* An rcu grace period has passed since there was a device in
	 * the network namespace (and thus the last in flight packet)
E
Eric W. Biederman 已提交
2574 2575 2576 2577 2578 2579 2580 2581
	 * left this network namespace.  This is because
	 * unregister_netdevice_many and netdev_run_todo has completed
	 * for each network device that was in this network namespace.
	 *
	 * As such no additional rcu synchronization is necessary when
	 * freeing the platform_label table.
	 */
	rtnl_lock();
2582 2583 2584 2585 2586
	platform_label = rtnl_dereference(net->mpls.platform_label);
	platform_labels = net->mpls.platform_labels;
	for (index = 0; index < platform_labels; index++) {
		struct mpls_route *rt = rtnl_dereference(platform_label[index]);
		RCU_INIT_POINTER(platform_label[index], NULL);
2587
		mpls_notify_route(net, index, rt, NULL, NULL);
E
Eric W. Biederman 已提交
2588 2589 2590 2591
		mpls_rt_free(rt);
	}
	rtnl_unlock();

2592
	kvfree(platform_label);
E
Eric W. Biederman 已提交
2593 2594 2595 2596 2597 2598 2599
}

static struct pernet_operations mpls_net_ops = {
	.init = mpls_net_init,
	.exit = mpls_net_exit,
};

R
Robert Shearman 已提交
2600 2601 2602 2603 2604 2605
static struct rtnl_af_ops mpls_af_ops __read_mostly = {
	.family		   = AF_MPLS,
	.fill_stats_af	   = mpls_fill_stats_af,
	.get_stats_af_size = mpls_get_stats_af_size,
};

E
Eric W. Biederman 已提交
2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621
static int __init mpls_init(void)
{
	int err;

	BUILD_BUG_ON(sizeof(struct mpls_shim_hdr) != 4);

	err = register_pernet_subsys(&mpls_net_ops);
	if (err)
		goto out;

	err = register_netdevice_notifier(&mpls_dev_notifier);
	if (err)
		goto out_unregister_pernet;

	dev_add_pack(&mpls_packet_type);

R
Robert Shearman 已提交
2622 2623
	rtnl_af_register(&mpls_af_ops);

2624 2625 2626 2627 2628 2629 2630 2631 2632
	rtnl_register_module(THIS_MODULE, PF_MPLS, RTM_NEWROUTE,
			     mpls_rtm_newroute, NULL, 0);
	rtnl_register_module(THIS_MODULE, PF_MPLS, RTM_DELROUTE,
			     mpls_rtm_delroute, NULL, 0);
	rtnl_register_module(THIS_MODULE, PF_MPLS, RTM_GETROUTE,
			     mpls_getroute, mpls_dump_routes, 0);
	rtnl_register_module(THIS_MODULE, PF_MPLS, RTM_GETNETCONF,
			     mpls_netconf_get_devconf,
			     mpls_netconf_dump_devconf, 0);
2633 2634 2635 2636
	err = ipgre_tunnel_encap_add_mpls_ops();
	if (err)
		pr_err("Can't add mpls over gre tunnel ops\n");

E
Eric W. Biederman 已提交
2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648
	err = 0;
out:
	return err;

out_unregister_pernet:
	unregister_pernet_subsys(&mpls_net_ops);
	goto out;
}
module_init(mpls_init);

static void __exit mpls_exit(void)
{
2649
	rtnl_unregister_all(PF_MPLS);
R
Robert Shearman 已提交
2650
	rtnl_af_unregister(&mpls_af_ops);
E
Eric W. Biederman 已提交
2651 2652 2653
	dev_remove_pack(&mpls_packet_type);
	unregister_netdevice_notifier(&mpls_dev_notifier);
	unregister_pernet_subsys(&mpls_net_ops);
2654
	ipgre_tunnel_encap_del_mpls_ops();
E
Eric W. Biederman 已提交
2655 2656 2657 2658 2659 2660
}
module_exit(mpls_exit);

MODULE_DESCRIPTION("MultiProtocol Label Switching");
MODULE_LICENSE("GPL v2");
MODULE_ALIAS_NETPROTO(PF_MPLS);