xfrm6_policy.c 10.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10
/*
 * xfrm6_policy.c: based on xfrm4_policy.c
 *
 * Authors:
 *	Mitsuru KANDA @USAGI
 * 	Kazunori MIYAZAWA @USAGI
 * 	Kunihiro Ishiguro <kunihiro@ipinfusion.com>
 * 		IPv6 support
 * 	YOSHIFUJI Hideaki
 * 		Split up af-specific portion
11
 *
L
Linus Torvalds 已提交
12 13
 */

H
Herbert Xu 已提交
14 15 16
#include <linux/compiler.h>
#include <linux/netdevice.h>
#include <net/addrconf.h>
L
Linus Torvalds 已提交
17 18 19 20
#include <net/xfrm.h>
#include <net/ip.h>
#include <net/ipv6.h>
#include <net/ip6_route.h>
21 22 23
#ifdef CONFIG_IPV6_MIP6
#include <net/mip6.h>
#endif
L
Linus Torvalds 已提交
24 25 26 27

static struct dst_ops xfrm6_dst_ops;
static struct xfrm_policy_afinfo xfrm6_policy_afinfo;

28
static int xfrm6_dst_lookup(struct xfrm_dst **xdst, struct flowi *fl)
L
Linus Torvalds 已提交
29
{
30 31 32 33 34 35
	struct dst_entry *dst = ip6_route_output(NULL, fl);
	int err = dst->error;
	if (!err)
		*xdst = (struct xfrm_dst *) dst;
	else
		dst_release(dst);
L
Linus Torvalds 已提交
36 37 38
	return err;
}

39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
static int xfrm6_get_saddr(xfrm_address_t *saddr, xfrm_address_t *daddr)
{
	struct rt6_info *rt;
	struct flowi fl_tunnel = {
		.nl_u = {
			.ip6_u = {
				.daddr = *(struct in6_addr *)&daddr->a6,
			},
		},
	};

	if (!xfrm6_dst_lookup((struct xfrm_dst **)&rt, &fl_tunnel)) {
		ipv6_get_saddr(&rt->u.dst, (struct in6_addr *)&daddr->a6,
			       (struct in6_addr *)&saddr->a6);
		dst_release(&rt->u.dst);
		return 0;
	}
	return -EHOSTUNREACH;
}

L
Linus Torvalds 已提交
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
static struct dst_entry *
__xfrm6_find_bundle(struct flowi *fl, struct xfrm_policy *policy)
{
	struct dst_entry *dst;

	/* Still not clear if we should set fl->fl6_{src,dst}... */
	read_lock_bh(&policy->lock);
	for (dst = policy->bundles; dst; dst = dst->next) {
		struct xfrm_dst *xdst = (struct xfrm_dst*)dst;
		struct in6_addr fl_dst_prefix, fl_src_prefix;

		ipv6_addr_prefix(&fl_dst_prefix,
				 &fl->fl6_dst,
				 xdst->u.rt6.rt6i_dst.plen);
		ipv6_addr_prefix(&fl_src_prefix,
				 &fl->fl6_src,
				 xdst->u.rt6.rt6i_src.plen);
		if (ipv6_addr_equal(&xdst->u.rt6.rt6i_dst.addr, &fl_dst_prefix) &&
		    ipv6_addr_equal(&xdst->u.rt6.rt6i_src.addr, &fl_src_prefix) &&
78
		    xfrm_bundle_ok(policy, xdst, fl, AF_INET6,
79 80
				   (xdst->u.rt6.rt6i_dst.plen != 128 ||
				    xdst->u.rt6.rt6i_src.plen != 128))) {
L
Linus Torvalds 已提交
81 82 83 84 85 86 87 88
			dst_clone(dst);
			break;
		}
	}
	read_unlock_bh(&policy->lock);
	return dst;
}

89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
static inline struct in6_addr*
__xfrm6_bundle_addr_remote(struct xfrm_state *x, struct in6_addr *addr)
{
	return (x->type->remote_addr) ?
		(struct in6_addr*)x->type->remote_addr(x, (xfrm_address_t *)addr) :
		(struct in6_addr*)&x->id.daddr;
}

static inline struct in6_addr*
__xfrm6_bundle_addr_local(struct xfrm_state *x, struct in6_addr *addr)
{
	return (x->type->local_addr) ?
		(struct in6_addr*)x->type->local_addr(x, (xfrm_address_t *)addr) :
		(struct in6_addr*)&x->props.saddr;
}

105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122
static inline void
__xfrm6_bundle_len_inc(int *len, int *nflen, struct xfrm_state *x)
{
	if (x->type->flags & XFRM_TYPE_NON_FRAGMENT)
		*nflen += x->props.header_len;
	else
		*len += x->props.header_len;
}

static inline void
__xfrm6_bundle_len_dec(int *len, int *nflen, struct xfrm_state *x)
{
	if (x->type->flags & XFRM_TYPE_NON_FRAGMENT)
		*nflen -= x->props.header_len;
	else
		*len -= x->props.header_len;
}

L
Linus Torvalds 已提交
123 124 125 126 127 128 129 130 131 132 133 134 135 136
/* Allocate chain of dst_entry's, attach known xfrm's, calculate
 * all the metrics... Shortly, bundle a bundle.
 */

static int
__xfrm6_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int nx,
		      struct flowi *fl, struct dst_entry **dst_p)
{
	struct dst_entry *dst, *dst_prev;
	struct rt6_info *rt0 = (struct rt6_info*)(*dst_p);
	struct rt6_info *rt  = rt0;
	struct flowi fl_tunnel = {
		.nl_u = {
			.ip6_u = {
137 138
				.saddr = fl->fl6_src,
				.daddr = fl->fl6_dst,
L
Linus Torvalds 已提交
139 140 141 142 143 144
			}
		}
	};
	int i;
	int err = 0;
	int header_len = 0;
145
	int nfheader_len = 0;
L
Linus Torvalds 已提交
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
	int trailer_len = 0;

	dst = dst_prev = NULL;
	dst_hold(&rt->u.dst);

	for (i = 0; i < nx; i++) {
		struct dst_entry *dst1 = dst_alloc(&xfrm6_dst_ops);
		struct xfrm_dst *xdst;

		if (unlikely(dst1 == NULL)) {
			err = -ENOBUFS;
			dst_release(&rt->u.dst);
			goto error;
		}

		if (!dst)
			dst = dst1;
		else {
			dst_prev->child = dst1;
			dst1->flags |= DST_NOHASH;
			dst_clone(dst1);
		}

		xdst = (struct xfrm_dst *)dst1;
		xdst->route = &rt->u.dst;
171
		xdst->genid = xfrm[i]->genid;
172 173
		if (rt->rt6i_node)
			xdst->route_cookie = rt->rt6i_node->fn_sernum;
L
Linus Torvalds 已提交
174 175 176

		dst1->next = dst_prev;
		dst_prev = dst1;
177

178
		__xfrm6_bundle_len_inc(&header_len, &nfheader_len, xfrm[i]);
L
Linus Torvalds 已提交
179 180
		trailer_len += xfrm[i]->props.trailer_len;

181 182
		if (xfrm[i]->props.mode == XFRM_MODE_TUNNEL ||
		    xfrm[i]->props.mode == XFRM_MODE_ROUTEOPTIMIZATION) {
183 184 185 186 187 188 189
			unsigned short encap_family = xfrm[i]->props.family;
			switch(encap_family) {
			case AF_INET:
				fl_tunnel.fl4_dst = xfrm[i]->id.daddr.a4;
				fl_tunnel.fl4_src = xfrm[i]->props.saddr.a4;
				break;
			case AF_INET6:
190 191
				ipv6_addr_copy(&fl_tunnel.fl6_dst, __xfrm6_bundle_addr_remote(xfrm[i], &fl->fl6_dst));

192
				ipv6_addr_copy(&fl_tunnel.fl6_src, __xfrm6_bundle_addr_local(xfrm[i], &fl->fl6_src));
193 194 195 196 197
				break;
			default:
				BUG_ON(1);
			}

L
Linus Torvalds 已提交
198
			err = xfrm_dst_lookup((struct xfrm_dst **) &rt,
199
					      &fl_tunnel, encap_family);
L
Linus Torvalds 已提交
200 201 202 203 204 205 206 207
			if (err)
				goto error;
		} else
			dst_hold(&rt->u.dst);
	}

	dst_prev->child = &rt->u.dst;
	dst->path = &rt->u.dst;
208 209
	if (rt->rt6i_node)
		((struct xfrm_dst *)dst)->path_cookie = rt->rt6i_node->fn_sernum;
L
Linus Torvalds 已提交
210 211 212 213 214 215 216 217

	*dst_p = dst;
	dst = dst_prev;

	dst_prev = *dst_p;
	i = 0;
	for (; dst_prev != &rt->u.dst; dst_prev = dst_prev->child) {
		struct xfrm_dst *x = (struct xfrm_dst*)dst_prev;
218
		struct xfrm_state_afinfo *afinfo;
L
Linus Torvalds 已提交
219 220 221 222 223 224 225 226 227

		dst_prev->xfrm = xfrm[i++];
		dst_prev->dev = rt->u.dst.dev;
		if (rt->u.dst.dev)
			dev_hold(rt->u.dst.dev);
		dst_prev->obsolete	= -1;
		dst_prev->flags	       |= DST_HOST;
		dst_prev->lastuse	= jiffies;
		dst_prev->header_len	= header_len;
228
		dst_prev->nfheader_len	= nfheader_len;
L
Linus Torvalds 已提交
229 230 231 232 233 234
		dst_prev->trailer_len	= trailer_len;
		memcpy(&dst_prev->metrics, &x->route->metrics, sizeof(dst_prev->metrics));

		/* Copy neighbour for reachability confirmation */
		dst_prev->neighbour	= neigh_clone(rt->u.dst.neighbour);
		dst_prev->input		= rt->u.dst.input;
235 236 237 238 239 240 241 242 243 244 245
		/* XXX: When IPv4 is implemented as module and can be unloaded,
		 * we should manage reference to xfrm4_output in afinfo->output.
		 * Miyazawa
		 */
		afinfo = xfrm_state_get_afinfo(dst_prev->xfrm->props.family);
		if (!afinfo) {
			dst = *dst_p;
			goto error;
		};
		dst_prev->output = afinfo->output;
		xfrm_state_put_afinfo(afinfo);
L
Linus Torvalds 已提交
246 247 248 249 250 251
		/* Sheit... I remember I did this right. Apparently,
		 * it was magically lost, so this code needs audit */
		x->u.rt6.rt6i_flags    = rt0->rt6i_flags&(RTCF_BROADCAST|RTCF_MULTICAST|RTCF_LOCAL);
		x->u.rt6.rt6i_metric   = rt0->rt6i_metric;
		x->u.rt6.rt6i_node     = rt0->rt6i_node;
		x->u.rt6.rt6i_gateway  = rt0->rt6i_gateway;
252
		memcpy(&x->u.rt6.rt6i_gateway, &rt0->rt6i_gateway, sizeof(x->u.rt6.rt6i_gateway));
L
Linus Torvalds 已提交
253
		x->u.rt6.rt6i_dst      = rt0->rt6i_dst;
254
		x->u.rt6.rt6i_src      = rt0->rt6i_src;
H
Herbert Xu 已提交
255 256
		x->u.rt6.rt6i_idev     = rt0->rt6i_idev;
		in6_dev_hold(rt0->rt6i_idev);
257
		__xfrm6_bundle_len_dec(&header_len, &nfheader_len, x->u.dst.xfrm);
L
Linus Torvalds 已提交
258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
		trailer_len -= x->u.dst.xfrm->props.trailer_len;
	}

	xfrm_init_pmtu(dst);
	return 0;

error:
	if (dst)
		dst_free(dst);
	return err;
}

static inline void
_decode_session6(struct sk_buff *skb, struct flowi *fl)
{
273
	u16 offset = skb->h.raw - skb->nh.raw;
274
	struct ipv6hdr *hdr = ipv6_hdr(skb);
275
	struct ipv6_opt_hdr *exthdr;
276 277
	const unsigned char *nh = skb_network_header(skb);
	u8 nexthdr = nh[IP6CB(skb)->nhoff];
L
Linus Torvalds 已提交
278 279 280 281 282

	memset(fl, 0, sizeof(struct flowi));
	ipv6_addr_copy(&fl->fl6_dst, &hdr->daddr);
	ipv6_addr_copy(&fl->fl6_src, &hdr->saddr);

283 284 285
	while (pskb_may_pull(skb, nh + offset + 1 - skb->data)) {
		nh = skb_network_header(skb);
		exthdr = (struct ipv6_opt_hdr *)(nh + offset);
286

L
Linus Torvalds 已提交
287 288 289 290 291 292
		switch (nexthdr) {
		case NEXTHDR_ROUTING:
		case NEXTHDR_HOP:
		case NEXTHDR_DEST:
			offset += ipv6_optlen(exthdr);
			nexthdr = exthdr->nexthdr;
293
			exthdr = (struct ipv6_opt_hdr *)(nh + offset);
L
Linus Torvalds 已提交
294 295 296
			break;

		case IPPROTO_UDP:
297
		case IPPROTO_UDPLITE:
L
Linus Torvalds 已提交
298 299
		case IPPROTO_TCP:
		case IPPROTO_SCTP:
300
		case IPPROTO_DCCP:
301
			if (pskb_may_pull(skb, nh + offset + 4 - skb->data)) {
A
Al Viro 已提交
302
				__be16 *ports = (__be16 *)exthdr;
L
Linus Torvalds 已提交
303 304 305 306 307 308 309 310

				fl->fl_ip_sport = ports[0];
				fl->fl_ip_dport = ports[1];
			}
			fl->proto = nexthdr;
			return;

		case IPPROTO_ICMPV6:
311
			if (pskb_may_pull(skb, nh + offset + 2 - skb->data)) {
L
Linus Torvalds 已提交
312 313 314 315 316 317 318 319
				u8 *icmp = (u8 *)exthdr;

				fl->fl_icmp_type = icmp[0];
				fl->fl_icmp_code = icmp[1];
			}
			fl->proto = nexthdr;
			return;

320 321
#ifdef CONFIG_IPV6_MIP6
		case IPPROTO_MH:
322
			if (pskb_may_pull(skb, nh + offset + 3 - skb->data)) {
323 324 325 326 327 328 329 330 331
				struct ip6_mh *mh;
				mh = (struct ip6_mh *)exthdr;

				fl->fl_mh_type = mh->ip6mh_type;
			}
			fl->proto = nexthdr;
			return;
#endif

L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
		/* XXX Why are there these headers? */
		case IPPROTO_AH:
		case IPPROTO_ESP:
		case IPPROTO_COMP:
		default:
			fl->fl_ipsec_spi = 0;
			fl->proto = nexthdr;
			return;
		};
	}
}

static inline int xfrm6_garbage_collect(void)
{
	xfrm6_policy_afinfo.garbage_collect();
	return (atomic_read(&xfrm6_dst_ops.entries) > xfrm6_dst_ops.gc_thresh*2);
}

static void xfrm6_update_pmtu(struct dst_entry *dst, u32 mtu)
{
	struct xfrm_dst *xdst = (struct xfrm_dst *)dst;
	struct dst_entry *path = xdst->route;

	path->ops->update_pmtu(path, mtu);
}

H
Herbert Xu 已提交
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
static void xfrm6_dst_destroy(struct dst_entry *dst)
{
	struct xfrm_dst *xdst = (struct xfrm_dst *)dst;

	if (likely(xdst->u.rt6.rt6i_idev))
		in6_dev_put(xdst->u.rt6.rt6i_idev);
	xfrm_dst_destroy(xdst);
}

static void xfrm6_dst_ifdown(struct dst_entry *dst, struct net_device *dev,
			     int unregister)
{
	struct xfrm_dst *xdst;

	if (!unregister)
		return;

	xdst = (struct xfrm_dst *)dst;
	if (xdst->u.rt6.rt6i_idev->dev == dev) {
		struct inet6_dev *loopback_idev = in6_dev_get(&loopback_dev);
		BUG_ON(!loopback_idev);

		do {
			in6_dev_put(xdst->u.rt6.rt6i_idev);
			xdst->u.rt6.rt6i_idev = loopback_idev;
			in6_dev_hold(loopback_idev);
			xdst = (struct xfrm_dst *)xdst->u.dst.child;
		} while (xdst->u.dst.xfrm);

		__in6_dev_put(loopback_idev);
	}

	xfrm_dst_ifdown(dst, dev);
}

L
Linus Torvalds 已提交
393 394 395 396 397
static struct dst_ops xfrm6_dst_ops = {
	.family =		AF_INET6,
	.protocol =		__constant_htons(ETH_P_IPV6),
	.gc =			xfrm6_garbage_collect,
	.update_pmtu =		xfrm6_update_pmtu,
H
Herbert Xu 已提交
398 399
	.destroy =		xfrm6_dst_destroy,
	.ifdown =		xfrm6_dst_ifdown,
L
Linus Torvalds 已提交
400 401 402 403 404 405 406 407
	.gc_thresh =		1024,
	.entry_size =		sizeof(struct xfrm_dst),
};

static struct xfrm_policy_afinfo xfrm6_policy_afinfo = {
	.family =		AF_INET6,
	.dst_ops =		&xfrm6_dst_ops,
	.dst_lookup =		xfrm6_dst_lookup,
408
	.get_saddr = 		xfrm6_get_saddr,
L
Linus Torvalds 已提交
409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
	.find_bundle =		__xfrm6_find_bundle,
	.bundle_create =	__xfrm6_bundle_create,
	.decode_session =	_decode_session6,
};

static void __init xfrm6_policy_init(void)
{
	xfrm_policy_register_afinfo(&xfrm6_policy_afinfo);
}

static void xfrm6_policy_fini(void)
{
	xfrm_policy_unregister_afinfo(&xfrm6_policy_afinfo);
}

void __init xfrm6_init(void)
{
	xfrm6_policy_init();
	xfrm6_state_init();
}

void xfrm6_fini(void)
{
	//xfrm6_input_fini();
	xfrm6_policy_fini();
	xfrm6_state_fini();
}