Blame view
drivers/net/vrf.c
33.9 KB
2874c5fd2
|
1 |
// SPDX-License-Identifier: GPL-2.0-or-later |
193125dbd
|
2 3 4 5 6 7 8 9 |
/* * vrf.c: device driver to encapsulate a VRF space * * Copyright (c) 2015 Cumulus Networks. All rights reserved. * Copyright (c) 2015 Shrijeet Mukherjee <shm@cumulusnetworks.com> * Copyright (c) 2015 David Ahern <dsa@cumulusnetworks.com> * * Based on dummy, team and ipvlan drivers |
193125dbd
|
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
*/ #include <linux/module.h> #include <linux/kernel.h> #include <linux/netdevice.h> #include <linux/etherdevice.h> #include <linux/ip.h> #include <linux/init.h> #include <linux/moduleparam.h> #include <linux/netfilter.h> #include <linux/rtnetlink.h> #include <net/rtnetlink.h> #include <linux/u64_stats_sync.h> #include <linux/hashtable.h> #include <linux/inetdevice.h> |
8f58336d3
|
26 |
#include <net/arp.h> |
193125dbd
|
27 28 |
#include <net/ip.h> #include <net/ip_fib.h> |
35402e313
|
29 |
#include <net/ip6_fib.h> |
193125dbd
|
30 |
#include <net/ip6_route.h> |
193125dbd
|
31 32 |
#include <net/route.h> #include <net/addrconf.h> |
ee15ee5d9
|
33 |
#include <net/l3mdev.h> |
1aa6c4f6b
|
34 |
#include <net/fib_rules.h> |
097d3c950
|
35 |
#include <net/netns/generic.h> |
193125dbd
|
36 37 38 |
#define DRV_NAME "vrf" #define DRV_VERSION "1.0" |
1aa6c4f6b
|
39 |
#define FIB_RULE_PREF 1000 /* default preference for FIB rules */ |
097d3c950
|
40 41 |
static unsigned int vrf_net_id; |
1aa6c4f6b
|
42 |
|
ec539514e
|
43 |
struct net_vrf { |
b0e95ccdd
|
44 45 |
struct rtable __rcu *rth; struct rt6_info __rcu *rt6; |
43b059a31
|
46 47 48 |
#if IS_ENABLED(CONFIG_IPV6) struct fib6_table *fib6_table; #endif |
ec539514e
|
49 50 |
u32 tb_id; }; |
193125dbd
|
51 52 53 54 55 56 |
struct pcpu_dstats { u64 tx_pkts; u64 tx_bytes; u64 tx_drps; u64 rx_pkts; u64 rx_bytes; |
afe80a499
|
57 |
u64 rx_drps; |
193125dbd
|
58 59 |
struct u64_stats_sync syncp; }; |
afe80a499
|
60 61 62 63 64 65 66 67 68 |
static void vrf_rx_stats(struct net_device *dev, int len) { struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats); u64_stats_update_begin(&dstats->syncp); dstats->rx_pkts++; dstats->rx_bytes += len; u64_stats_update_end(&dstats->syncp); } |
57b8efa1a
|
69 70 71 72 73 |
static void vrf_tx_error(struct net_device *vrf_dev, struct sk_buff *skb) { vrf_dev->stats.tx_errors++; kfree_skb(skb); } |
bc1f44709
|
74 75 |
static void vrf_get_stats64(struct net_device *dev, struct rtnl_link_stats64 *stats) |
193125dbd
|
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
{ int i; for_each_possible_cpu(i) { const struct pcpu_dstats *dstats; u64 tbytes, tpkts, tdrops, rbytes, rpkts; unsigned int start; dstats = per_cpu_ptr(dev->dstats, i); do { start = u64_stats_fetch_begin_irq(&dstats->syncp); tbytes = dstats->tx_bytes; tpkts = dstats->tx_pkts; tdrops = dstats->tx_drps; rbytes = dstats->rx_bytes; rpkts = dstats->rx_pkts; } while (u64_stats_fetch_retry_irq(&dstats->syncp, start)); stats->tx_bytes += tbytes; stats->tx_packets += tpkts; stats->tx_dropped += tdrops; stats->rx_bytes += rbytes; stats->rx_packets += rpkts; } |
193125dbd
|
99 |
} |
dcdd43c41
|
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 |
/* by default VRF devices do not have a qdisc and are expected * to be created with only a single queue. */ static bool qdisc_tx_is_default(const struct net_device *dev) { struct netdev_queue *txq; struct Qdisc *qdisc; if (dev->num_tx_queues > 1) return false; txq = netdev_get_tx_queue(dev, 0); qdisc = rcu_access_pointer(txq->qdisc); return !qdisc->enqueue; } |
afe80a499
|
116 117 118 119 120 121 122 123 124 125 126 |
/* Local traffic destined to local address. Reinsert the packet to rx * path, similar to loopback handling. */ static int vrf_local_xmit(struct sk_buff *skb, struct net_device *dev, struct dst_entry *dst) { int len = skb->len; skb_orphan(skb); skb_dst_set(skb, dst); |
afe80a499
|
127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 |
/* set pkt_type to avoid skb hitting packet taps twice - * once on Tx and again in Rx processing */ skb->pkt_type = PACKET_LOOPBACK; skb->protocol = eth_type_trans(skb, dev); if (likely(netif_rx(skb) == NET_RX_SUCCESS)) vrf_rx_stats(dev, len); else this_cpu_inc(dev->dstats->rx_drps); return NETDEV_TX_OK; } |
35402e313
|
142 |
#if IS_ENABLED(CONFIG_IPV6) |
4c1feac58
|
143 144 145 146 147 148 149 150 151 152 153 154 155 |
static int vrf_ip6_local_out(struct net *net, struct sock *sk, struct sk_buff *skb) { int err; err = nf_hook(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net, sk, skb, NULL, skb_dst(skb)->dev, dst_output); if (likely(err == 1)) err = dst_output(net, sk, skb); return err; } |
35402e313
|
156 157 158 |
static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb, struct net_device *dev) { |
107e47cc8
|
159 |
const struct ipv6hdr *iph; |
35402e313
|
160 |
struct net *net = dev_net(skb->dev); |
107e47cc8
|
161 |
struct flowi6 fl6; |
35402e313
|
162 163 164 |
int ret = NET_XMIT_DROP; struct dst_entry *dst; struct dst_entry *dst_null = &net->ipv6.ip6_null_entry->dst; |
107e47cc8
|
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 |
if (!pskb_may_pull(skb, ETH_HLEN + sizeof(struct ipv6hdr))) goto err; iph = ipv6_hdr(skb); memset(&fl6, 0, sizeof(fl6)); /* needed to match OIF rule */ fl6.flowi6_oif = dev->ifindex; fl6.flowi6_iif = LOOPBACK_IFINDEX; fl6.daddr = iph->daddr; fl6.saddr = iph->saddr; fl6.flowlabel = ip6_flowinfo(iph); fl6.flowi6_mark = skb->mark; fl6.flowi6_proto = iph->nexthdr; fl6.flowi6_flags = FLOWI_FLAG_SKIP_NH_OIF; |
35402e313
|
180 181 182 183 184 |
dst = ip6_route_output(net, NULL, &fl6); if (dst == dst_null) goto err; skb_dst_drop(skb); |
b4869aa2f
|
185 186 187 |
/* if dst.dev is loopback or the VRF device again this is locally * originated traffic destined to a local address. Short circuit |
4f04256c9
|
188 |
* to Rx path |
b4869aa2f
|
189 |
*/ |
4f04256c9
|
190 191 |
if (dst->dev == dev) return vrf_local_xmit(skb, dev, dst); |
b4869aa2f
|
192 |
|
35402e313
|
193 |
skb_dst_set(skb, dst); |
911a66fbc
|
194 195 |
/* strip the ethernet header added for pass through VRF device */ __skb_pull(skb, skb_network_offset(skb)); |
4c1feac58
|
196 |
ret = vrf_ip6_local_out(net, skb->sk, skb); |
35402e313
|
197 198 199 200 201 202 203 204 205 206 207 |
if (unlikely(net_xmit_eval(ret))) dev->stats.tx_errors++; else ret = NET_XMIT_SUCCESS; return ret; err: vrf_tx_error(dev, skb); return NET_XMIT_DROP; } #else |
193125dbd
|
208 209 210 |
static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb, struct net_device *dev) { |
57b8efa1a
|
211 212 |
vrf_tx_error(dev, skb); return NET_XMIT_DROP; |
193125dbd
|
213 |
} |
35402e313
|
214 |
#endif |
193125dbd
|
215 |
|
ebfc102c5
|
216 217 218 219 220 221 222 223 224 225 226 227 228 |
/* based on ip_local_out; can't use it b/c the dst is switched pointing to us */ static int vrf_ip_local_out(struct net *net, struct sock *sk, struct sk_buff *skb) { int err; err = nf_hook(NFPROTO_IPV4, NF_INET_LOCAL_OUT, net, sk, skb, NULL, skb_dst(skb)->dev, dst_output); if (likely(err == 1)) err = dst_output(net, sk, skb); return err; } |
193125dbd
|
229 230 231 |
static netdev_tx_t vrf_process_v4_outbound(struct sk_buff *skb, struct net_device *vrf_dev) { |
107e47cc8
|
232 |
struct iphdr *ip4h; |
193125dbd
|
233 |
int ret = NET_XMIT_DROP; |
107e47cc8
|
234 |
struct flowi4 fl4; |
911a66fbc
|
235 236 |
struct net *net = dev_net(vrf_dev); struct rtable *rt; |
107e47cc8
|
237 238 239 240 241 242 243 244 245 246 247 248 249 250 |
if (!pskb_may_pull(skb, ETH_HLEN + sizeof(struct iphdr))) goto err; ip4h = ip_hdr(skb); memset(&fl4, 0, sizeof(fl4)); /* needed to match OIF rule */ fl4.flowi4_oif = vrf_dev->ifindex; fl4.flowi4_iif = LOOPBACK_IFINDEX; fl4.flowi4_tos = RT_TOS(ip4h->tos); fl4.flowi4_flags = FLOWI_FLAG_ANYSRC | FLOWI_FLAG_SKIP_NH_OIF; fl4.flowi4_proto = ip4h->protocol; fl4.daddr = ip4h->daddr; fl4.saddr = ip4h->saddr; |
911a66fbc
|
251 252 253 |
rt = ip_route_output_flow(net, &fl4, NULL); if (IS_ERR(rt)) goto err; |
193125dbd
|
254 |
|
911a66fbc
|
255 |
skb_dst_drop(skb); |
afe80a499
|
256 257 258 |
/* if dst.dev is loopback or the VRF device again this is locally * originated traffic destined to a local address. Short circuit |
4f04256c9
|
259 |
* to Rx path |
afe80a499
|
260 |
*/ |
4f04256c9
|
261 262 |
if (rt->dst.dev == vrf_dev) return vrf_local_xmit(skb, vrf_dev, &rt->dst); |
afe80a499
|
263 |
|
911a66fbc
|
264 265 266 267 |
skb_dst_set(skb, &rt->dst); /* strip the ethernet header added for pass through VRF device */ __skb_pull(skb, skb_network_offset(skb)); |
193125dbd
|
268 269 270 271 272 |
if (!ip4h->saddr) { ip4h->saddr = inet_select_addr(skb_dst(skb)->dev, 0, RT_SCOPE_LINK); } |
ebfc102c5
|
273 |
ret = vrf_ip_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb); |
193125dbd
|
274 275 276 277 278 279 280 281 |
if (unlikely(net_xmit_eval(ret))) vrf_dev->stats.tx_errors++; else ret = NET_XMIT_SUCCESS; out: return ret; err: |
57b8efa1a
|
282 |
vrf_tx_error(vrf_dev, skb); |
193125dbd
|
283 284 285 286 287 288 289 290 291 292 293 |
goto out; } static netdev_tx_t is_ip_tx_frame(struct sk_buff *skb, struct net_device *dev) { switch (skb->protocol) { case htons(ETH_P_IP): return vrf_process_v4_outbound(skb, dev); case htons(ETH_P_IPV6): return vrf_process_v6_outbound(skb, dev); default: |
57b8efa1a
|
294 |
vrf_tx_error(dev, skb); |
193125dbd
|
295 296 297 298 299 300 |
return NET_XMIT_DROP; } } static netdev_tx_t vrf_xmit(struct sk_buff *skb, struct net_device *dev) { |
f7887d40e
|
301 |
int len = skb->len; |
193125dbd
|
302 303 304 305 306 307 308 |
netdev_tx_t ret = is_ip_tx_frame(skb, dev); if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN)) { struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats); u64_stats_update_begin(&dstats->syncp); dstats->tx_pkts++; |
f7887d40e
|
309 |
dstats->tx_bytes += len; |
193125dbd
|
310 311 312 313 314 315 316 |
u64_stats_update_end(&dstats->syncp); } else { this_cpu_inc(dev->dstats->tx_drps); } return ret; } |
dcdd43c41
|
317 318 319 320 321 322 323 |
static int vrf_finish_direct(struct net *net, struct sock *sk, struct sk_buff *skb) { struct net_device *vrf_dev = skb->dev; if (!list_empty(&vrf_dev->ptype_all) && likely(skb_headroom(skb) >= ETH_HLEN)) { |
d58ff3512
|
324 |
struct ethhdr *eth = skb_push(skb, ETH_HLEN); |
dcdd43c41
|
325 326 327 328 329 330 331 332 333 334 335 336 337 338 |
ether_addr_copy(eth->h_source, vrf_dev->dev_addr); eth_zero_addr(eth->h_dest); eth->h_proto = skb->protocol; rcu_read_lock_bh(); dev_queue_xmit_nit(skb, vrf_dev); rcu_read_unlock_bh(); skb_pull(skb, ETH_HLEN); } return 1; } |
35402e313
|
339 |
#if IS_ENABLED(CONFIG_IPV6) |
35402e313
|
340 341 342 343 344 345 |
/* modelled after ip6_finish_output2 */ static int vrf_finish_output6(struct net *net, struct sock *sk, struct sk_buff *skb) { struct dst_entry *dst = skb_dst(skb); struct net_device *dev = dst->dev; |
9b1c1ef13
|
346 |
const struct in6_addr *nexthop; |
35402e313
|
347 |
struct neighbour *neigh; |
35402e313
|
348 |
int ret; |
895b5c9f2
|
349 |
nf_reset_ct(skb); |
eb63ecc17
|
350 |
|
35402e313
|
351 352 353 354 355 356 357 358 359 |
skb->protocol = htons(ETH_P_IPV6); skb->dev = dev; rcu_read_lock_bh(); nexthop = rt6_nexthop((struct rt6_info *)dst, &ipv6_hdr(skb)->daddr); neigh = __ipv6_neigh_lookup_noref(dst->dev, nexthop); if (unlikely(!neigh)) neigh = __neigh_create(&nd_tbl, nexthop, dst->dev, false); if (!IS_ERR(neigh)) { |
4ff062035
|
360 |
sock_confirm_neigh(skb, neigh); |
0353f2823
|
361 |
ret = neigh_output(neigh, skb, false); |
35402e313
|
362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 |
rcu_read_unlock_bh(); return ret; } rcu_read_unlock_bh(); IP6_INC_STATS(dev_net(dst->dev), ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES); kfree_skb(skb); return -EINVAL; } /* modelled after ip6_output */ static int vrf_output6(struct net *net, struct sock *sk, struct sk_buff *skb) { return NF_HOOK_COND(NFPROTO_IPV6, NF_INET_POST_ROUTING, net, sk, skb, NULL, skb_dst(skb)->dev, vrf_finish_output6, !(IP6CB(skb)->flags & IP6SKB_REROUTED)); } |
4c1feac58
|
381 382 383 384 |
/* set dst on skb to send packet to us via dev_xmit path. Allows * packet to go through device based features such as qdisc, netfilter * hooks and packet sockets with skb->dev set to vrf device. */ |
a9ec54d1b
|
385 386 |
static struct sk_buff *vrf_ip6_out_redirect(struct net_device *vrf_dev, struct sk_buff *skb) |
4c1feac58
|
387 388 389 390 |
{ struct net_vrf *vrf = netdev_priv(vrf_dev); struct dst_entry *dst = NULL; struct rt6_info *rt6; |
4c1feac58
|
391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 |
rcu_read_lock(); rt6 = rcu_dereference(vrf->rt6); if (likely(rt6)) { dst = &rt6->dst; dst_hold(dst); } rcu_read_unlock(); if (unlikely(!dst)) { vrf_tx_error(vrf_dev, skb); return NULL; } skb_dst_drop(skb); skb_dst_set(skb, dst); return skb; } |
a9ec54d1b
|
411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 |
static int vrf_output6_direct(struct net *net, struct sock *sk, struct sk_buff *skb) { skb->protocol = htons(ETH_P_IPV6); return NF_HOOK_COND(NFPROTO_IPV6, NF_INET_POST_ROUTING, net, sk, skb, NULL, skb->dev, vrf_finish_direct, !(IPCB(skb)->flags & IPSKB_REROUTED)); } static struct sk_buff *vrf_ip6_out_direct(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb) { struct net *net = dev_net(vrf_dev); int err; skb->dev = vrf_dev; err = nf_hook(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net, sk, skb, NULL, vrf_dev, vrf_output6_direct); if (likely(err == 1)) err = vrf_output6_direct(net, sk, skb); /* reset skb device */ if (likely(err == 1)) |
895b5c9f2
|
439 |
nf_reset_ct(skb); |
a9ec54d1b
|
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 |
else skb = NULL; return skb; } static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb) { /* don't divert link scope packets */ if (rt6_need_strict(&ipv6_hdr(skb)->daddr)) return skb; if (qdisc_tx_is_default(vrf_dev)) return vrf_ip6_out_direct(vrf_dev, sk, skb); return vrf_ip6_out_redirect(vrf_dev, skb); } |
b0e95ccdd
|
459 |
/* holding rtnl */ |
810e530bf
|
460 |
static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf) |
35402e313
|
461 |
{ |
b0e95ccdd
|
462 |
struct rt6_info *rt6 = rtnl_dereference(vrf->rt6); |
810e530bf
|
463 464 |
struct net *net = dev_net(dev); struct dst_entry *dst; |
b0e95ccdd
|
465 |
|
b4869aa2f
|
466 |
RCU_INIT_POINTER(vrf->rt6, NULL); |
b4869aa2f
|
467 |
synchronize_rcu(); |
b0e95ccdd
|
468 |
|
810e530bf
|
469 470 471 472 473 474 475 476 477 478 |
/* move dev in dst's to loopback so this VRF device can be deleted * - based on dst_ifdown */ if (rt6) { dst = &rt6->dst; dev_put(dst->dev); dst->dev = net->loopback_dev; dev_hold(dst->dev); dst_release(dst); } |
35402e313
|
479 480 481 482 |
} static int vrf_rt6_create(struct net_device *dev) { |
a4c2fd7f7
|
483 |
int flags = DST_HOST | DST_NOPOLICY | DST_NOXFRM; |
35402e313
|
484 |
struct net_vrf *vrf = netdev_priv(dev); |
9ab179d83
|
485 |
struct net *net = dev_net(dev); |
4f04256c9
|
486 |
struct rt6_info *rt6; |
35402e313
|
487 |
int rc = -ENOMEM; |
e43486371
|
488 489 490 |
/* IPv6 can be CONFIG enabled and then disabled runtime */ if (!ipv6_mod_enabled()) return 0; |
43b059a31
|
491 492 |
vrf->fib6_table = fib6_new_table(net, vrf->tb_id); if (!vrf->fib6_table) |
b3b4663c9
|
493 |
goto out; |
b4869aa2f
|
494 495 |
/* create a dst for routing packets out a VRF device */ rt6 = ip6_dst_alloc(net, dev, flags); |
35402e313
|
496 497 |
if (!rt6) goto out; |
b3b4663c9
|
498 |
rt6->dst.output = vrf_output6; |
b4869aa2f
|
499 |
|
b0e95ccdd
|
500 |
rcu_assign_pointer(vrf->rt6, rt6); |
35402e313
|
501 502 503 504 505 |
rc = 0; out: return rc; } #else |
4c1feac58
|
506 507 508 509 510 511 |
static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb) { return skb; } |
810e530bf
|
512 |
static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf) |
35402e313
|
513 514 515 516 517 518 519 520 |
{ } static int vrf_rt6_create(struct net_device *dev) { return 0; } #endif |
8f58336d3
|
521 |
/* modelled after ip_finish_output2 */ |
0c4b51f00
|
522 |
static int vrf_finish_output(struct net *net, struct sock *sk, struct sk_buff *skb) |
193125dbd
|
523 |
{ |
8f58336d3
|
524 525 526 527 528 |
struct dst_entry *dst = skb_dst(skb); struct rtable *rt = (struct rtable *)dst; struct net_device *dev = dst->dev; unsigned int hh_len = LL_RESERVED_SPACE(dev); struct neighbour *neigh; |
5c9f7c1df
|
529 |
bool is_v6gw = false; |
8f58336d3
|
530 |
int ret = -EINVAL; |
895b5c9f2
|
531 |
nf_reset_ct(skb); |
eb63ecc17
|
532 |
|
8f58336d3
|
533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 |
/* Be paranoid, rather than too clever. */ if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { struct sk_buff *skb2; skb2 = skb_realloc_headroom(skb, LL_RESERVED_SPACE(dev)); if (!skb2) { ret = -ENOMEM; goto err; } if (skb->sk) skb_set_owner_w(skb2, skb->sk); consume_skb(skb); skb = skb2; } rcu_read_lock_bh(); |
5c9f7c1df
|
550 |
neigh = ip_neigh_for_gw(rt, skb, &is_v6gw); |
4ff062035
|
551 552 |
if (!IS_ERR(neigh)) { sock_confirm_neigh(skb, neigh); |
5c9f7c1df
|
553 554 |
/* if crossing protocols, can not use the cached header */ ret = neigh_output(neigh, skb, is_v6gw); |
82dd0d2a9
|
555 556 |
rcu_read_unlock_bh(); return ret; |
4ff062035
|
557 |
} |
8f58336d3
|
558 559 560 |
rcu_read_unlock_bh(); err: |
82dd0d2a9
|
561 |
vrf_tx_error(skb->dev, skb); |
8f58336d3
|
562 |
return ret; |
193125dbd
|
563 |
} |
ede2059db
|
564 |
static int vrf_output(struct net *net, struct sock *sk, struct sk_buff *skb) |
193125dbd
|
565 566 |
{ struct net_device *dev = skb_dst(skb)->dev; |
29a26a568
|
567 |
IP_UPD_PO_STATS(net, IPSTATS_MIB_OUT, skb->len); |
193125dbd
|
568 569 570 |
skb->dev = dev; skb->protocol = htons(ETH_P_IP); |
29a26a568
|
571 572 |
return NF_HOOK_COND(NFPROTO_IPV4, NF_INET_POST_ROUTING, net, sk, skb, NULL, dev, |
8f58336d3
|
573 |
vrf_finish_output, |
193125dbd
|
574 575 |
!(IPCB(skb)->flags & IPSKB_REROUTED)); } |
ebfc102c5
|
576 577 578 579 |
/* set dst on skb to send packet to us via dev_xmit path. Allows * packet to go through device based features such as qdisc, netfilter * hooks and packet sockets with skb->dev set to vrf device. */ |
dcdd43c41
|
580 581 |
static struct sk_buff *vrf_ip_out_redirect(struct net_device *vrf_dev, struct sk_buff *skb) |
ebfc102c5
|
582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 |
{ struct net_vrf *vrf = netdev_priv(vrf_dev); struct dst_entry *dst = NULL; struct rtable *rth; rcu_read_lock(); rth = rcu_dereference(vrf->rth); if (likely(rth)) { dst = &rth->dst; dst_hold(dst); } rcu_read_unlock(); if (unlikely(!dst)) { vrf_tx_error(vrf_dev, skb); return NULL; } skb_dst_drop(skb); skb_dst_set(skb, dst); return skb; } |
dcdd43c41
|
607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 |
static int vrf_output_direct(struct net *net, struct sock *sk, struct sk_buff *skb) { skb->protocol = htons(ETH_P_IP); return NF_HOOK_COND(NFPROTO_IPV4, NF_INET_POST_ROUTING, net, sk, skb, NULL, skb->dev, vrf_finish_direct, !(IPCB(skb)->flags & IPSKB_REROUTED)); } static struct sk_buff *vrf_ip_out_direct(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb) { struct net *net = dev_net(vrf_dev); int err; skb->dev = vrf_dev; err = nf_hook(NFPROTO_IPV4, NF_INET_LOCAL_OUT, net, sk, skb, NULL, vrf_dev, vrf_output_direct); if (likely(err == 1)) err = vrf_output_direct(net, sk, skb); /* reset skb device */ if (likely(err == 1)) |
895b5c9f2
|
635 |
nf_reset_ct(skb); |
dcdd43c41
|
636 637 638 639 640 641 642 643 644 645 |
else skb = NULL; return skb; } static struct sk_buff *vrf_ip_out(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb) { |
1e19c4d68
|
646 647 648 |
/* don't divert multicast or local broadcast */ if (ipv4_is_multicast(ip_hdr(skb)->daddr) || ipv4_is_lbcast(ip_hdr(skb)->daddr)) |
dcdd43c41
|
649 650 651 652 653 654 655 |
return skb; if (qdisc_tx_is_default(vrf_dev)) return vrf_ip_out_direct(vrf_dev, sk, skb); return vrf_ip_out_redirect(vrf_dev, skb); } |
ebfc102c5
|
656 657 658 659 660 661 662 663 664 |
/* called with rcu lock held */ static struct sk_buff *vrf_l3_out(struct net_device *vrf_dev, struct sock *sk, struct sk_buff *skb, u16 proto) { switch (proto) { case AF_INET: return vrf_ip_out(vrf_dev, sk, skb); |
4c1feac58
|
665 666 |
case AF_INET6: return vrf_ip6_out(vrf_dev, sk, skb); |
ebfc102c5
|
667 668 669 670 |
} return skb; } |
b0e95ccdd
|
671 |
/* holding rtnl */ |
810e530bf
|
672 |
static void vrf_rtable_release(struct net_device *dev, struct net_vrf *vrf) |
193125dbd
|
673 |
{ |
b0e95ccdd
|
674 |
struct rtable *rth = rtnl_dereference(vrf->rth); |
810e530bf
|
675 676 |
struct net *net = dev_net(dev); struct dst_entry *dst; |
b0e95ccdd
|
677 |
|
afe80a499
|
678 |
RCU_INIT_POINTER(vrf->rth, NULL); |
afe80a499
|
679 |
synchronize_rcu(); |
193125dbd
|
680 |
|
810e530bf
|
681 682 683 684 685 686 687 688 689 690 |
/* move dev in dst's to loopback so this VRF device can be deleted * - based on dst_ifdown */ if (rth) { dst = &rth->dst; dev_put(dst->dev); dst->dev = net->loopback_dev; dev_hold(dst->dev); dst_release(dst); } |
193125dbd
|
691 |
} |
b0e95ccdd
|
692 |
static int vrf_rtable_create(struct net_device *dev) |
193125dbd
|
693 |
{ |
b7503e0cd
|
694 |
struct net_vrf *vrf = netdev_priv(dev); |
4f04256c9
|
695 |
struct rtable *rth; |
193125dbd
|
696 |
|
b3b4663c9
|
697 |
if (!fib_new_table(dev_net(dev), vrf->tb_id)) |
b0e95ccdd
|
698 |
return -ENOMEM; |
b3b4663c9
|
699 |
|
afe80a499
|
700 |
/* create a dst for routing packets out through a VRF device */ |
9ab179d83
|
701 |
rth = rt_dst_alloc(dev, 0, RTN_UNICAST, 1, 1, 0); |
b0e95ccdd
|
702 703 |
if (!rth) return -ENOMEM; |
193125dbd
|
704 |
|
b0e95ccdd
|
705 |
rth->dst.output = vrf_output; |
b0e95ccdd
|
706 707 708 709 |
rcu_assign_pointer(vrf->rth, rth); return 0; |
193125dbd
|
710 711 712 713 714 |
} /**************************** device handling ********************/ /* cycle interface to flush neighbor cache and move routes across tables */ |
dc1aea1e0
|
715 716 |
static void cycle_netdev(struct net_device *dev, struct netlink_ext_ack *extack) |
193125dbd
|
717 718 719 720 721 722 |
{ unsigned int flags = dev->flags; int ret; if (!netif_running(dev)) return; |
567c5e13b
|
723 |
ret = dev_change_flags(dev, flags & ~IFF_UP, extack); |
193125dbd
|
724 |
if (ret >= 0) |
567c5e13b
|
725 |
ret = dev_change_flags(dev, flags, extack); |
193125dbd
|
726 727 728 729 730 731 732 733 |
if (ret < 0) { netdev_err(dev, "Failed to cycle device %s; route tables might be wrong! ", dev->name); } } |
42ab19ee9
|
734 735 |
static int do_vrf_add_slave(struct net_device *dev, struct net_device *port_dev, struct netlink_ext_ack *extack) |
193125dbd
|
736 |
{ |
bad531623
|
737 |
int ret; |
193125dbd
|
738 |
|
26d31ac11
|
739 740 741 |
/* do not allow loopback device to be enslaved to a VRF. * The vrf device acts as the loopback for the vrf. */ |
de3baa3ed
|
742 743 744 |
if (port_dev == dev_net(dev)->loopback_dev) { NL_SET_ERR_MSG(extack, "Can not enslave loopback device to a VRF"); |
26d31ac11
|
745 |
return -EOPNOTSUPP; |
de3baa3ed
|
746 |
} |
26d31ac11
|
747 |
|
fdeea7be8
|
748 |
port_dev->priv_flags |= IFF_L3MDEV_SLAVE; |
42ab19ee9
|
749 |
ret = netdev_master_upper_dev_link(port_dev, dev, NULL, NULL, extack); |
193125dbd
|
750 |
if (ret < 0) |
fdeea7be8
|
751 |
goto err; |
193125dbd
|
752 |
|
dc1aea1e0
|
753 |
cycle_netdev(port_dev, extack); |
193125dbd
|
754 755 |
return 0; |
fdeea7be8
|
756 757 758 759 |
err: port_dev->priv_flags &= ~IFF_L3MDEV_SLAVE; return ret; |
193125dbd
|
760 |
} |
33eaf2a6e
|
761 762 |
static int vrf_add_slave(struct net_device *dev, struct net_device *port_dev, struct netlink_ext_ack *extack) |
193125dbd
|
763 |
{ |
de3baa3ed
|
764 765 766 767 768 769 770 |
if (netif_is_l3_master(port_dev)) { NL_SET_ERR_MSG(extack, "Can not enslave an L3 master device to a VRF"); return -EINVAL; } if (netif_is_l3_slave(port_dev)) |
193125dbd
|
771 |
return -EINVAL; |
42ab19ee9
|
772 |
return do_vrf_add_slave(dev, port_dev, extack); |
193125dbd
|
773 774 775 776 777 |
} /* inverse of do_vrf_add_slave */ static int do_vrf_del_slave(struct net_device *dev, struct net_device *port_dev) { |
193125dbd
|
778 |
netdev_upper_dev_unlink(port_dev, dev); |
fee6d4c77
|
779 |
port_dev->priv_flags &= ~IFF_L3MDEV_SLAVE; |
193125dbd
|
780 |
|
dc1aea1e0
|
781 |
cycle_netdev(port_dev, NULL); |
193125dbd
|
782 |
|
193125dbd
|
783 784 785 786 787 |
return 0; } static int vrf_del_slave(struct net_device *dev, struct net_device *port_dev) { |
193125dbd
|
788 789 790 791 792 793 |
return do_vrf_del_slave(dev, port_dev); } static void vrf_dev_uninit(struct net_device *dev) { struct net_vrf *vrf = netdev_priv(dev); |
193125dbd
|
794 |
|
810e530bf
|
795 796 |
vrf_rtable_release(dev, vrf); vrf_rt6_release(dev, vrf); |
193125dbd
|
797 |
|
3a4a27d3b
|
798 |
free_percpu(dev->dstats); |
193125dbd
|
799 800 801 802 803 804 |
dev->dstats = NULL; } static int vrf_dev_init(struct net_device *dev) { struct net_vrf *vrf = netdev_priv(dev); |
193125dbd
|
805 806 807 808 809 |
dev->dstats = netdev_alloc_pcpu_stats(struct pcpu_dstats); if (!dev->dstats) goto out_nomem; /* create the default dst which points back to us */ |
b0e95ccdd
|
810 |
if (vrf_rtable_create(dev) != 0) |
193125dbd
|
811 |
goto out_stats; |
35402e313
|
812 813 |
if (vrf_rt6_create(dev) != 0) goto out_rth; |
193125dbd
|
814 |
dev->flags = IFF_MASTER | IFF_NOARP; |
b87ab6b8e
|
815 816 817 818 819 |
/* MTU is irrelevant for VRF device; set to 64k similar to lo */ dev->mtu = 64 * 1024; /* similarly, oper state is irrelevant; set to up to avoid confusion */ dev->operstate = IF_OPER_UP; |
193125dbd
|
820 |
return 0; |
35402e313
|
821 |
out_rth: |
810e530bf
|
822 |
vrf_rtable_release(dev, vrf); |
193125dbd
|
823 824 825 826 827 828 829 830 831 832 833 |
out_stats: free_percpu(dev->dstats); dev->dstats = NULL; out_nomem: return -ENOMEM; } static const struct net_device_ops vrf_netdev_ops = { .ndo_init = vrf_dev_init, .ndo_uninit = vrf_dev_uninit, .ndo_start_xmit = vrf_xmit, |
6819e3f6d
|
834 |
.ndo_set_mac_address = eth_mac_addr, |
193125dbd
|
835 836 837 838 |
.ndo_get_stats64 = vrf_get_stats64, .ndo_add_slave = vrf_add_slave, .ndo_del_slave = vrf_del_slave, }; |
ee15ee5d9
|
839 840 841 842 843 844 |
static u32 vrf_fib_table(const struct net_device *dev) { struct net_vrf *vrf = netdev_priv(dev); return vrf->tb_id; } |
73e20b761
|
845 846 |
static int vrf_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb) { |
1a4a5bf52
|
847 |
kfree_skb(skb); |
73e20b761
|
848 849 850 851 852 853 854 855 |
return 0; } static struct sk_buff *vrf_rcv_nfhook(u8 pf, unsigned int hook, struct sk_buff *skb, struct net_device *dev) { struct net *net = dev_net(dev); |
1a4a5bf52
|
856 |
if (nf_hook(pf, hook, net, NULL, skb, dev, NULL, vrf_rcv_finish) != 1) |
73e20b761
|
857 858 859 860 |
skb = NULL; /* kfree_skb(skb) handled by nf code */ return skb; } |
35402e313
|
861 |
#if IS_ENABLED(CONFIG_IPV6) |
74b20582a
|
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 |
/* neighbor handling is done with actual device; do not want * to flip skb->dev for those ndisc packets. This really fails * for multiple next protocols (e.g., NEXTHDR_HOP). But it is * a start. */ static bool ipv6_ndisc_frame(const struct sk_buff *skb) { const struct ipv6hdr *iph = ipv6_hdr(skb); bool rc = false; if (iph->nexthdr == NEXTHDR_ICMP) { const struct icmp6hdr *icmph; struct icmp6hdr _icmph; icmph = skb_header_pointer(skb, sizeof(*iph), sizeof(_icmph), &_icmph); if (!icmph) goto out; switch (icmph->icmp6_type) { case NDISC_ROUTER_SOLICITATION: case NDISC_ROUTER_ADVERTISEMENT: case NDISC_NEIGHBOUR_SOLICITATION: case NDISC_NEIGHBOUR_ADVERTISEMENT: case NDISC_REDIRECT: rc = true; break; } } out: return rc; } |
9ff743846
|
895 896 897 898 |
static struct rt6_info *vrf_ip6_route_lookup(struct net *net, const struct net_device *dev, struct flowi6 *fl6, int ifindex, |
b75cc8f90
|
899 |
const struct sk_buff *skb, |
9ff743846
|
900 901 902 |
int flags) { struct net_vrf *vrf = netdev_priv(dev); |
9ff743846
|
903 |
|
43b059a31
|
904 |
return ip6_pol_route(net, vrf->fib6_table, ifindex, fl6, skb, flags); |
9ff743846
|
905 906 907 908 909 910 911 |
} static void vrf_ip6_input_dst(struct sk_buff *skb, struct net_device *vrf_dev, int ifindex) { const struct ipv6hdr *iph = ipv6_hdr(skb); struct flowi6 fl6 = { |
ecf091171
|
912 913 914 |
.flowi6_iif = ifindex, .flowi6_mark = skb->mark, .flowi6_proto = iph->nexthdr, |
9ff743846
|
915 916 917 |
.daddr = iph->daddr, .saddr = iph->saddr, .flowlabel = ip6_flowinfo(iph), |
9ff743846
|
918 919 920 |
}; struct net *net = dev_net(vrf_dev); struct rt6_info *rt6; |
b75cc8f90
|
921 |
rt6 = vrf_ip6_route_lookup(net, vrf_dev, &fl6, ifindex, skb, |
9ff743846
|
922 923 924 925 926 927 928 929 930 |
RT6_LOOKUP_F_HAS_SADDR | RT6_LOOKUP_F_IFACE); if (unlikely(!rt6)) return; if (unlikely(&rt6->dst == &net->ipv6.ip6_null_entry->dst)) return; skb_dst_set(skb, &rt6->dst); } |
74b20582a
|
931 932 933 |
static struct sk_buff *vrf_ip6_rcv(struct net_device *vrf_dev, struct sk_buff *skb) { |
9ff743846
|
934 |
int orig_iif = skb->skb_iif; |
6f12fa775
|
935 936 |
bool need_strict = rt6_need_strict(&ipv6_hdr(skb)->daddr); bool is_ndisc = ipv6_ndisc_frame(skb); |
9ff743846
|
937 |
|
6f12fa775
|
938 939 |
/* loopback, multicast & non-ND link-local traffic; do not push through * packet taps again. Reset pkt_type for upper layers to process skb |
b4869aa2f
|
940 |
*/ |
6f12fa775
|
941 |
if (skb->pkt_type == PACKET_LOOPBACK || (need_strict && !is_ndisc)) { |
b4869aa2f
|
942 943 |
skb->dev = vrf_dev; skb->skb_iif = vrf_dev->ifindex; |
a04a480d4
|
944 |
IP6CB(skb)->flags |= IP6SKB_L3SLAVE; |
6f12fa775
|
945 946 |
if (skb->pkt_type == PACKET_LOOPBACK) skb->pkt_type = PACKET_HOST; |
b4869aa2f
|
947 948 |
goto out; } |
6f12fa775
|
949 950 |
/* if packet is NDISC then keep the ingress interface */ if (!is_ndisc) { |
926d93a33
|
951 |
vrf_rx_stats(vrf_dev, skb->len); |
74b20582a
|
952 953 |
skb->dev = vrf_dev; skb->skb_iif = vrf_dev->ifindex; |
a9ec54d1b
|
954 955 956 957 958 |
if (!list_empty(&vrf_dev->ptype_all)) { skb_push(skb, skb->mac_len); dev_queue_xmit_nit(skb, vrf_dev); skb_pull(skb, skb->mac_len); } |
74b20582a
|
959 960 961 |
IP6CB(skb)->flags |= IP6SKB_L3SLAVE; } |
9ff743846
|
962 963 |
if (need_strict) vrf_ip6_input_dst(skb, vrf_dev, orig_iif); |
73e20b761
|
964 |
skb = vrf_rcv_nfhook(NFPROTO_IPV6, NF_INET_PRE_ROUTING, skb, vrf_dev); |
b4869aa2f
|
965 |
out: |
74b20582a
|
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 |
return skb; } #else static struct sk_buff *vrf_ip6_rcv(struct net_device *vrf_dev, struct sk_buff *skb) { return skb; } #endif static struct sk_buff *vrf_ip_rcv(struct net_device *vrf_dev, struct sk_buff *skb) { skb->dev = vrf_dev; skb->skb_iif = vrf_dev->ifindex; |
a04a480d4
|
982 |
IPCB(skb)->flags |= IPSKB_L3SLAVE; |
74b20582a
|
983 |
|
e58e41596
|
984 985 |
if (ipv4_is_multicast(ip_hdr(skb)->daddr)) goto out; |
afe80a499
|
986 987 988 989 990 991 992 |
/* loopback traffic; do not push through packet taps again. * Reset pkt_type for upper layers to process skb */ if (skb->pkt_type == PACKET_LOOPBACK) { skb->pkt_type = PACKET_HOST; goto out; } |
926d93a33
|
993 |
vrf_rx_stats(vrf_dev, skb->len); |
dcdd43c41
|
994 995 996 997 998 |
if (!list_empty(&vrf_dev->ptype_all)) { skb_push(skb, skb->mac_len); dev_queue_xmit_nit(skb, vrf_dev); skb_pull(skb, skb->mac_len); } |
74b20582a
|
999 |
|
73e20b761
|
1000 |
skb = vrf_rcv_nfhook(NFPROTO_IPV4, NF_INET_PRE_ROUTING, skb, vrf_dev); |
afe80a499
|
1001 |
out: |
74b20582a
|
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 |
return skb; } /* called with rcu lock held */ static struct sk_buff *vrf_l3_rcv(struct net_device *vrf_dev, struct sk_buff *skb, u16 proto) { switch (proto) { case AF_INET: return vrf_ip_rcv(vrf_dev, skb); case AF_INET6: return vrf_ip6_rcv(vrf_dev, skb); } return skb; } #if IS_ENABLED(CONFIG_IPV6) |
4c1feac58
|
1021 1022 |
/* send to link-local or multicast address via interface enslaved to * VRF device. Force lookup to VRF table without changing flow struct |
7d9e5f422
|
1023 1024 |
* Note: Caller to this function must hold rcu_read_lock() and no refcnt * is taken on the dst by this function. |
4c1feac58
|
1025 1026 1027 |
*/ static struct dst_entry *vrf_link_scope_lookup(const struct net_device *dev, struct flowi6 *fl6) |
35402e313
|
1028 |
{ |
9ff743846
|
1029 |
struct net *net = dev_net(dev); |
7d9e5f422
|
1030 |
int flags = RT6_LOOKUP_F_IFACE | RT6_LOOKUP_F_DST_NOREF; |
b0e95ccdd
|
1031 |
struct dst_entry *dst = NULL; |
9ff743846
|
1032 |
struct rt6_info *rt; |
35402e313
|
1033 |
|
4c1feac58
|
1034 1035 1036 1037 1038 1039 |
/* VRF device does not have a link-local address and * sending packets to link-local or mcast addresses over * a VRF device does not make sense */ if (fl6->flowi6_oif == dev->ifindex) { dst = &net->ipv6.ip6_null_entry->dst; |
4c1feac58
|
1040 |
return dst; |
35402e313
|
1041 |
} |
4c1feac58
|
1042 1043 |
if (!ipv6_addr_any(&fl6->saddr)) flags |= RT6_LOOKUP_F_HAS_SADDR; |
b75cc8f90
|
1044 |
rt = vrf_ip6_route_lookup(net, dev, fl6, fl6->flowi6_oif, NULL, flags); |
4c1feac58
|
1045 1046 |
if (rt) dst = &rt->dst; |
9ff743846
|
1047 |
|
b0e95ccdd
|
1048 |
return dst; |
35402e313
|
1049 1050 |
} #endif |
ee15ee5d9
|
1051 1052 |
static const struct l3mdev_ops vrf_l3mdev_ops = { .l3mdev_fib_table = vrf_fib_table, |
74b20582a
|
1053 |
.l3mdev_l3_rcv = vrf_l3_rcv, |
ebfc102c5
|
1054 |
.l3mdev_l3_out = vrf_l3_out, |
35402e313
|
1055 |
#if IS_ENABLED(CONFIG_IPV6) |
4c1feac58
|
1056 |
.l3mdev_link_scope_lookup = vrf_link_scope_lookup, |
35402e313
|
1057 |
#endif |
ee15ee5d9
|
1058 |
}; |
193125dbd
|
1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 |
static void vrf_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { strlcpy(info->driver, DRV_NAME, sizeof(info->driver)); strlcpy(info->version, DRV_VERSION, sizeof(info->version)); } static const struct ethtool_ops vrf_ethtool_ops = { .get_drvinfo = vrf_get_drvinfo, }; |
1aa6c4f6b
|
1069 1070 1071 1072 1073 1074 1075 |
static inline size_t vrf_fib_rule_nl_size(void) { size_t sz; sz = NLMSG_ALIGN(sizeof(struct fib_rule_hdr)); sz += nla_total_size(sizeof(u8)); /* FRA_L3MDEV */ sz += nla_total_size(sizeof(u32)); /* FRA_PRIORITY */ |
1b71af605
|
1076 |
sz += nla_total_size(sizeof(u8)); /* FRA_PROTOCOL */ |
1aa6c4f6b
|
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 |
return sz; } static int vrf_fib_rule(const struct net_device *dev, __u8 family, bool add_it) { struct fib_rule_hdr *frh; struct nlmsghdr *nlh; struct sk_buff *skb; int err; |
dac91170f
|
1087 1088 |
if ((family == AF_INET6 || family == RTNL_FAMILY_IP6MR) && !ipv6_mod_enabled()) |
e43486371
|
1089 |
return 0; |
1aa6c4f6b
|
1090 1091 1092 1093 1094 1095 1096 1097 1098 |
skb = nlmsg_new(vrf_fib_rule_nl_size(), GFP_KERNEL); if (!skb) return -ENOMEM; nlh = nlmsg_put(skb, 0, 0, 0, sizeof(*frh), 0); if (!nlh) goto nla_put_failure; /* rule only needs to appear once */ |
426c87caa
|
1099 |
nlh->nlmsg_flags |= NLM_F_EXCL; |
1aa6c4f6b
|
1100 1101 1102 1103 1104 |
frh = nlmsg_data(nlh); memset(frh, 0, sizeof(*frh)); frh->family = family; frh->action = FR_ACT_TO_TBL; |
1b71af605
|
1105 1106 1107 |
if (nla_put_u8(skb, FRA_PROTOCOL, RTPROT_KERNEL)) goto nla_put_failure; |
1aa6c4f6b
|
1108 |
|
18129a249
|
1109 |
if (nla_put_u8(skb, FRA_L3MDEV, 1)) |
1aa6c4f6b
|
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 |
goto nla_put_failure; if (nla_put_u32(skb, FRA_PRIORITY, FIB_RULE_PREF)) goto nla_put_failure; nlmsg_end(skb, nlh); /* fib_nl_{new,del}rule handling looks for net from skb->sk */ skb->sk = dev_net(dev)->rtnl; if (add_it) { |
c21ef3e34
|
1120 |
err = fib_nl_newrule(skb, nlh, NULL); |
1aa6c4f6b
|
1121 1122 1123 |
if (err == -EEXIST) err = 0; } else { |
c21ef3e34
|
1124 |
err = fib_nl_delrule(skb, nlh, NULL); |
1aa6c4f6b
|
1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 |
if (err == -ENOENT) err = 0; } nlmsg_free(skb); return err; nla_put_failure: nlmsg_free(skb); return -EMSGSIZE; } static int vrf_add_fib_rules(const struct net_device *dev) { int err; err = vrf_fib_rule(dev, AF_INET, true); if (err < 0) goto out_err; err = vrf_fib_rule(dev, AF_INET6, true); if (err < 0) goto ipv6_err; |
e58e41596
|
1149 1150 1151 1152 1153 |
#if IS_ENABLED(CONFIG_IP_MROUTE_MULTIPLE_TABLES) err = vrf_fib_rule(dev, RTNL_FAMILY_IPMR, true); if (err < 0) goto ipmr_err; #endif |
e4a38c0c4
|
1154 1155 1156 1157 1158 |
#if IS_ENABLED(CONFIG_IPV6_MROUTE_MULTIPLE_TABLES) err = vrf_fib_rule(dev, RTNL_FAMILY_IP6MR, true); if (err < 0) goto ip6mr_err; #endif |
1aa6c4f6b
|
1159 |
return 0; |
e4a38c0c4
|
1160 1161 1162 1163 |
#if IS_ENABLED(CONFIG_IPV6_MROUTE_MULTIPLE_TABLES) ip6mr_err: vrf_fib_rule(dev, RTNL_FAMILY_IPMR, false); #endif |
e58e41596
|
1164 1165 1166 1167 |
#if IS_ENABLED(CONFIG_IP_MROUTE_MULTIPLE_TABLES) ipmr_err: vrf_fib_rule(dev, AF_INET6, false); #endif |
1aa6c4f6b
|
1168 1169 1170 1171 1172 1173 1174 1175 |
ipv6_err: vrf_fib_rule(dev, AF_INET, false); out_err: netdev_err(dev, "Failed to add FIB rules. "); return err; } |
193125dbd
|
1176 1177 1178 1179 1180 1181 |
static void vrf_setup(struct net_device *dev) { ether_setup(dev); /* Initialize the device structure. */ dev->netdev_ops = &vrf_netdev_ops; |
ee15ee5d9
|
1182 |
dev->l3mdev_ops = &vrf_l3mdev_ops; |
193125dbd
|
1183 |
dev->ethtool_ops = &vrf_ethtool_ops; |
cf124db56
|
1184 |
dev->needs_free_netdev = true; |
193125dbd
|
1185 1186 1187 1188 1189 1190 1191 1192 1193 |
/* Fill in device structure with ethernet-generic values. */ eth_hw_addr_random(dev); /* don't acquire vrf device's netif_tx_lock when transmitting */ dev->features |= NETIF_F_LLTX; /* don't allow vrf devices to change network namespaces. */ dev->features |= NETIF_F_NETNS_LOCAL; |
7889681f4
|
1194 1195 1196 1197 1198 1199 |
/* does not make sense for a VLAN to be added to a vrf device */ dev->features |= NETIF_F_VLAN_CHALLENGED; /* enable offload features */ dev->features |= NETIF_F_GSO_SOFTWARE; |
cb1603948
|
1200 |
dev->features |= NETIF_F_RXCSUM | NETIF_F_HW_CSUM | NETIF_F_SCTP_CRC; |
7889681f4
|
1201 1202 1203 1204 1205 1206 1207 |
dev->features |= NETIF_F_SG | NETIF_F_FRAGLIST | NETIF_F_HIGHDMA; dev->hw_features = dev->features; dev->hw_enc_features = dev->features; /* default to no qdisc; user can add if desired */ dev->priv_flags |= IFF_NO_QUEUE; |
1017e0987
|
1208 |
dev->priv_flags |= IFF_NO_RX_HANDLER; |
6819e3f6d
|
1209 |
dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; |
ad49bc636
|
1210 |
|
5055376a3
|
1211 1212 1213 1214 1215 1216 |
/* VRF devices do not care about MTU, but if the MTU is set * too low then the ipv4 and ipv6 protocols are disabled * which breaks networking. */ dev->min_mtu = IPV6_MIN_MTU; dev->max_mtu = ETH_MAX_MTU; |
193125dbd
|
1217 |
} |
a8b8a889e
|
1218 1219 |
static int vrf_validate(struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) |
193125dbd
|
1220 1221 |
{ if (tb[IFLA_ADDRESS]) { |
53b948356
|
1222 1223 |
if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN) { NL_SET_ERR_MSG(extack, "Invalid hardware address"); |
193125dbd
|
1224 |
return -EINVAL; |
53b948356
|
1225 1226 1227 |
} if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS]))) { NL_SET_ERR_MSG(extack, "Invalid hardware address"); |
193125dbd
|
1228 |
return -EADDRNOTAVAIL; |
53b948356
|
1229 |
} |
193125dbd
|
1230 1231 1232 1233 1234 1235 |
} return 0; } static void vrf_dellink(struct net_device *dev, struct list_head *head) { |
f630c38ef
|
1236 1237 1238 1239 1240 |
struct net_device *port_dev; struct list_head *iter; netdev_for_each_lower_dev(dev, port_dev, iter) vrf_del_slave(dev, port_dev); |
193125dbd
|
1241 1242 1243 1244 |
unregister_netdevice_queue(dev, head); } static int vrf_newlink(struct net *src_net, struct net_device *dev, |
7a3f4a185
|
1245 1246 |
struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) |
193125dbd
|
1247 1248 |
{ struct net_vrf *vrf = netdev_priv(dev); |
097d3c950
|
1249 1250 |
bool *add_fib_rules; struct net *net; |
1aa6c4f6b
|
1251 |
int err; |
193125dbd
|
1252 |
|
53b948356
|
1253 1254 |
if (!data || !data[IFLA_VRF_TABLE]) { NL_SET_ERR_MSG(extack, "VRF table id is missing"); |
193125dbd
|
1255 |
return -EINVAL; |
53b948356
|
1256 |
} |
193125dbd
|
1257 1258 |
vrf->tb_id = nla_get_u32(data[IFLA_VRF_TABLE]); |
53b948356
|
1259 1260 1261 |
if (vrf->tb_id == RT_TABLE_UNSPEC) { NL_SET_ERR_MSG_ATTR(extack, data[IFLA_VRF_TABLE], "Invalid VRF table id"); |
24c63bbc1
|
1262 |
return -EINVAL; |
53b948356
|
1263 |
} |
193125dbd
|
1264 |
|
007979eaf
|
1265 |
dev->priv_flags |= IFF_L3MDEV_MASTER; |
193125dbd
|
1266 |
|
1aa6c4f6b
|
1267 1268 1269 |
err = register_netdevice(dev); if (err) goto out; |
097d3c950
|
1270 1271 1272 |
net = dev_net(dev); add_fib_rules = net_generic(net, vrf_net_id); if (*add_fib_rules) { |
1aa6c4f6b
|
1273 1274 1275 1276 1277 |
err = vrf_add_fib_rules(dev); if (err) { unregister_netdevice(dev); goto out; } |
097d3c950
|
1278 |
*add_fib_rules = false; |
1aa6c4f6b
|
1279 1280 1281 1282 |
} out: return err; |
193125dbd
|
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 |
} static size_t vrf_nl_getsize(const struct net_device *dev) { return nla_total_size(sizeof(u32)); /* IFLA_VRF_TABLE */ } static int vrf_fillinfo(struct sk_buff *skb, const struct net_device *dev) { struct net_vrf *vrf = netdev_priv(dev); return nla_put_u32(skb, IFLA_VRF_TABLE, vrf->tb_id); } |
67eb03318
|
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 |
static size_t vrf_get_slave_size(const struct net_device *bond_dev, const struct net_device *slave_dev) { return nla_total_size(sizeof(u32)); /* IFLA_VRF_PORT_TABLE */ } static int vrf_fill_slave_info(struct sk_buff *skb, const struct net_device *vrf_dev, const struct net_device *slave_dev) { struct net_vrf *vrf = netdev_priv(vrf_dev); if (nla_put_u32(skb, IFLA_VRF_PORT_TABLE, vrf->tb_id)) return -EMSGSIZE; return 0; } |
193125dbd
|
1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 |
static const struct nla_policy vrf_nl_policy[IFLA_VRF_MAX + 1] = { [IFLA_VRF_TABLE] = { .type = NLA_U32 }, }; static struct rtnl_link_ops vrf_link_ops __read_mostly = { .kind = DRV_NAME, .priv_size = sizeof(struct net_vrf), .get_size = vrf_nl_getsize, .policy = vrf_nl_policy, .validate = vrf_validate, .fill_info = vrf_fillinfo, |
67eb03318
|
1326 1327 |
.get_slave_size = vrf_get_slave_size, .fill_slave_info = vrf_fill_slave_info, |
193125dbd
|
1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 |
.newlink = vrf_newlink, .dellink = vrf_dellink, .setup = vrf_setup, .maxtype = IFLA_VRF_MAX, }; static int vrf_device_event(struct notifier_block *unused, unsigned long event, void *ptr) { struct net_device *dev = netdev_notifier_info_to_dev(ptr); /* only care about unregister events to drop slave references */ if (event == NETDEV_UNREGISTER) { |
193125dbd
|
1341 |
struct net_device *vrf_dev; |
fee6d4c77
|
1342 |
if (!netif_is_l3_slave(dev)) |
193125dbd
|
1343 |
goto out; |
58aa90875
|
1344 1345 |
vrf_dev = netdev_master_upper_dev_get(dev); vrf_del_slave(vrf_dev, dev); |
193125dbd
|
1346 1347 1348 1349 1350 1351 1352 1353 |
} out: return NOTIFY_DONE; } static struct notifier_block vrf_notifier_block __read_mostly = { .notifier_call = vrf_device_event, }; |
097d3c950
|
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 |
/* Initialize per network namespace state */ static int __net_init vrf_netns_init(struct net *net) { bool *add_fib_rules = net_generic(net, vrf_net_id); *add_fib_rules = true; return 0; } static struct pernet_operations vrf_net_ops __net_initdata = { .init = vrf_netns_init, .id = &vrf_net_id, .size = sizeof(bool), }; |
193125dbd
|
1369 1370 1371 |
static int __init vrf_init_module(void) { int rc; |
193125dbd
|
1372 |
register_netdevice_notifier(&vrf_notifier_block); |
097d3c950
|
1373 |
rc = register_pernet_subsys(&vrf_net_ops); |
193125dbd
|
1374 1375 |
if (rc < 0) goto error; |
097d3c950
|
1376 1377 1378 1379 1380 |
rc = rtnl_link_register(&vrf_link_ops); if (rc < 0) { unregister_pernet_subsys(&vrf_net_ops); goto error; } |
193125dbd
|
1381 1382 1383 1384 |
return 0; error: unregister_netdevice_notifier(&vrf_notifier_block); |
193125dbd
|
1385 1386 |
return rc; } |
193125dbd
|
1387 |
module_init(vrf_init_module); |
193125dbd
|
1388 1389 1390 1391 1392 |
MODULE_AUTHOR("Shrijeet Mukherjee, David Ahern"); MODULE_DESCRIPTION("Device driver to instantiate VRF domains"); MODULE_LICENSE("GPL"); MODULE_ALIAS_RTNL_LINK(DRV_NAME); MODULE_VERSION(DRV_VERSION); |