peer_object.c 9.8 KB
Newer Older
1
/* RxRPC remote transport endpoint record management
2
 *
3
 * Copyright (C) 2007, 2016 Red Hat, Inc. All Rights Reserved.
4 5 6 7 8 9 10 11
 * Written by David Howells (dhowells@redhat.com)
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

12 13
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

14 15 16 17 18
#include <linux/module.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/udp.h>
#include <linux/in.h>
D
David Howells 已提交
19
#include <linux/in6.h>
20
#include <linux/slab.h>
21
#include <linux/hashtable.h>
22 23 24
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include <net/ip.h>
25
#include <net/route.h>
D
David Howells 已提交
26
#include <net/ip6_route.h>
27 28
#include "ar-internal.h"

29 30
static DEFINE_HASHTABLE(rxrpc_peer_hash, 10);
static DEFINE_SPINLOCK(rxrpc_peer_hash_lock);
31

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
/*
 * Hash a peer key.
 */
static unsigned long rxrpc_peer_hash_key(struct rxrpc_local *local,
					 const struct sockaddr_rxrpc *srx)
{
	const u16 *p;
	unsigned int i, size;
	unsigned long hash_key;

	_enter("");

	hash_key = (unsigned long)local / __alignof__(*local);
	hash_key += srx->transport_type;
	hash_key += srx->transport_len;
	hash_key += srx->transport.family;

	switch (srx->transport.family) {
	case AF_INET:
		hash_key += (u16 __force)srx->transport.sin.sin_port;
		size = sizeof(srx->transport.sin.sin_addr);
		p = (u16 *)&srx->transport.sin.sin_addr;
		break;
55
#ifdef CONFIG_AF_RXRPC_IPV6
D
David Howells 已提交
56 57 58 59 60
	case AF_INET6:
		hash_key += (u16 __force)srx->transport.sin.sin_port;
		size = sizeof(srx->transport.sin6.sin6_addr);
		p = (u16 *)&srx->transport.sin6.sin6_addr;
		break;
61
#endif
62 63 64
	default:
		WARN(1, "AF_RXRPC: Unsupported transport address family\n");
		return 0;
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
	}

	/* Step through the peer address in 16-bit portions for speed */
	for (i = 0; i < size; i += sizeof(*p), p++)
		hash_key += *p;

	_leave(" 0x%lx", hash_key);
	return hash_key;
}

/*
 * Compare a peer to a key.  Return -ve, 0 or +ve to indicate less than, same
 * or greater than.
 *
 * Unfortunately, the primitives in linux/hashtable.h don't allow for sorted
 * buckets and mid-bucket insertion, so we don't make full use of this
 * information at this point.
 */
static long rxrpc_peer_cmp_key(const struct rxrpc_peer *peer,
			       struct rxrpc_local *local,
			       const struct sockaddr_rxrpc *srx,
			       unsigned long hash_key)
{
	long diff;

	diff = ((peer->hash_key - hash_key) ?:
		((unsigned long)peer->local - (unsigned long)local) ?:
		(peer->srx.transport_type - srx->transport_type) ?:
		(peer->srx.transport_len - srx->transport_len) ?:
		(peer->srx.transport.family - srx->transport.family));
	if (diff != 0)
		return diff;

	switch (srx->transport.family) {
	case AF_INET:
		return ((u16 __force)peer->srx.transport.sin.sin_port -
			(u16 __force)srx->transport.sin.sin_port) ?:
			memcmp(&peer->srx.transport.sin.sin_addr,
			       &srx->transport.sin.sin_addr,
			       sizeof(struct in_addr));
105
#ifdef CONFIG_AF_RXRPC_IPV6
D
David Howells 已提交
106 107 108 109 110 111
	case AF_INET6:
		return ((u16 __force)peer->srx.transport.sin6.sin6_port -
			(u16 __force)srx->transport.sin6.sin6_port) ?:
			memcmp(&peer->srx.transport.sin6.sin6_addr,
			       &srx->transport.sin6.sin6_addr,
			       sizeof(struct in6_addr));
112
#endif
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
	default:
		BUG();
	}
}

/*
 * Look up a remote transport endpoint for the specified address using RCU.
 */
static struct rxrpc_peer *__rxrpc_lookup_peer_rcu(
	struct rxrpc_local *local,
	const struct sockaddr_rxrpc *srx,
	unsigned long hash_key)
{
	struct rxrpc_peer *peer;

	hash_for_each_possible_rcu(rxrpc_peer_hash, peer, hash_link, hash_key) {
		if (rxrpc_peer_cmp_key(peer, local, srx, hash_key) == 0) {
			if (atomic_read(&peer->usage) == 0)
				return NULL;
			return peer;
		}
	}

	return NULL;
}

/*
 * Look up a remote transport endpoint for the specified address using RCU.
 */
struct rxrpc_peer *rxrpc_lookup_peer_rcu(struct rxrpc_local *local,
					 const struct sockaddr_rxrpc *srx)
{
	struct rxrpc_peer *peer;
	unsigned long hash_key = rxrpc_peer_hash_key(local, srx);

	peer = __rxrpc_lookup_peer_rcu(local, srx, hash_key);
	if (peer) {
D
David Howells 已提交
150
		_net("PEER %d {%pISp}", peer->debug_id, &peer->srx.transport);
151 152 153 154
		_leave(" = %p {u=%d}", peer, atomic_read(&peer->usage));
	}
	return peer;
}
155

156 157 158 159 160 161
/*
 * assess the MTU size for the network interface through which this peer is
 * reached
 */
static void rxrpc_assess_MTU_size(struct rxrpc_peer *peer)
{
D
David Howells 已提交
162
	struct dst_entry *dst;
163
	struct rtable *rt;
D
David Howells 已提交
164 165
	struct flowi fl;
	struct flowi4 *fl4 = &fl.u.ip4;
166
#ifdef CONFIG_AF_RXRPC_IPV6
D
David Howells 已提交
167
	struct flowi6 *fl6 = &fl.u.ip6;
168
#endif
169 170 171

	peer->if_mtu = 1500;

D
David Howells 已提交
172 173 174 175 176 177 178 179 180 181 182 183 184 185
	memset(&fl, 0, sizeof(fl));
	switch (peer->srx.transport.family) {
	case AF_INET:
		rt = ip_route_output_ports(
			&init_net, fl4, NULL,
			peer->srx.transport.sin.sin_addr.s_addr, 0,
			htons(7000), htons(7001), IPPROTO_UDP, 0, 0);
		if (IS_ERR(rt)) {
			_leave(" [route err %ld]", PTR_ERR(rt));
			return;
		}
		dst = &rt->dst;
		break;

186
#ifdef CONFIG_AF_RXRPC_IPV6
D
David Howells 已提交
187 188 189 190 191 192 193 194 195
	case AF_INET6:
		fl6->flowi6_iif = LOOPBACK_IFINDEX;
		fl6->flowi6_scope = RT_SCOPE_UNIVERSE;
		fl6->flowi6_proto = IPPROTO_UDP;
		memcpy(&fl6->daddr, &peer->srx.transport.sin6.sin6_addr,
		       sizeof(struct in6_addr));
		fl6->fl6_dport = htons(7001);
		fl6->fl6_sport = htons(7000);
		dst = ip6_route_output(&init_net, NULL, fl6);
196 197
		if (dst->error) {
			_leave(" [route err %d]", dst->error);
D
David Howells 已提交
198 199 200
			return;
		}
		break;
201
#endif
D
David Howells 已提交
202 203 204

	default:
		BUG();
205 206
	}

D
David Howells 已提交
207 208
	peer->if_mtu = dst_mtu(dst);
	dst_release(dst);
209

D
David Howells 已提交
210
	_leave(" [if_mtu %u]", peer->if_mtu);
211 212
}

213
/*
214
 * Allocate a peer.
215
 */
216
struct rxrpc_peer *rxrpc_alloc_peer(struct rxrpc_local *local, gfp_t gfp)
217 218 219 220 221 222 223
{
	struct rxrpc_peer *peer;

	_enter("");

	peer = kzalloc(sizeof(struct rxrpc_peer), gfp);
	if (peer) {
224 225
		atomic_set(&peer->usage, 1);
		peer->local = local;
226 227 228
		INIT_HLIST_HEAD(&peer->error_targets);
		INIT_WORK(&peer->error_distributor,
			  &rxrpc_peer_error_distributor);
229
		peer->service_conns = RB_ROOT;
230
		seqlock_init(&peer->service_conn_lock);
231 232
		spin_lock_init(&peer->lock);
		peer->debug_id = atomic_inc_return(&rxrpc_debug_id);
233 234 235 236 237 238
	}

	_leave(" = %p", peer);
	return peer;
}

239 240 241 242 243
/*
 * Initialise peer record.
 */
static void rxrpc_init_peer(struct rxrpc_peer *peer, unsigned long hash_key)
{
244
	peer->hash_key = hash_key;
245 246
	rxrpc_assess_MTU_size(peer);
	peer->mtu = peer->if_mtu;
247
	peer->rtt_last_req = ktime_get_real();
248

D
David Howells 已提交
249 250
	switch (peer->srx.transport.family) {
	case AF_INET:
251
		peer->hdrsize = sizeof(struct iphdr);
D
David Howells 已提交
252
		break;
253
#ifdef CONFIG_AF_RXRPC_IPV6
D
David Howells 已提交
254 255 256
	case AF_INET6:
		peer->hdrsize = sizeof(struct ipv6hdr);
		break;
257
#endif
D
David Howells 已提交
258 259 260 261 262 263 264 265 266
	default:
		BUG();
	}

	switch (peer->srx.transport_type) {
	case SOCK_DGRAM:
		peer->hdrsize += sizeof(struct udphdr);
		break;
	default:
267 268 269 270 271 272 273
		BUG();
	}

	peer->hdrsize += sizeof(struct rxrpc_wire_header);
	peer->maxdata = peer->mtu - peer->hdrsize;
}

274 275 276 277 278 279 280 281 282 283 284 285 286 287
/*
 * Set up a new peer.
 */
static struct rxrpc_peer *rxrpc_create_peer(struct rxrpc_local *local,
					    struct sockaddr_rxrpc *srx,
					    unsigned long hash_key,
					    gfp_t gfp)
{
	struct rxrpc_peer *peer;

	_enter("");

	peer = rxrpc_alloc_peer(local, gfp);
	if (peer) {
288
		memcpy(&peer->srx, srx, sizeof(*srx));
289 290
		rxrpc_init_peer(peer, hash_key);
	}
291

292 293 294
	_leave(" = %p", peer);
	return peer;
}
295

296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
/*
 * Set up a new incoming peer.  The address is prestored in the preallocated
 * peer.
 */
struct rxrpc_peer *rxrpc_lookup_incoming_peer(struct rxrpc_local *local,
					      struct rxrpc_peer *prealloc)
{
	struct rxrpc_peer *peer;
	unsigned long hash_key;

	hash_key = rxrpc_peer_hash_key(local, &prealloc->srx);
	prealloc->local = local;
	rxrpc_init_peer(prealloc, hash_key);

	spin_lock(&rxrpc_peer_hash_lock);

	/* Need to check that we aren't racing with someone else */
	peer = __rxrpc_lookup_peer_rcu(local, &prealloc->srx, hash_key);
	if (peer && !rxrpc_get_peer_maybe(peer))
		peer = NULL;
	if (!peer) {
		peer = prealloc;
		hash_add_rcu(rxrpc_peer_hash, &peer->hash_link, hash_key);
319 320
	}

321
	spin_unlock(&rxrpc_peer_hash_lock);
322 323 324 325 326 327
	return peer;
}

/*
 * obtain a remote transport endpoint for the specified address
 */
328 329
struct rxrpc_peer *rxrpc_lookup_peer(struct rxrpc_local *local,
				     struct sockaddr_rxrpc *srx, gfp_t gfp)
330 331
{
	struct rxrpc_peer *peer, *candidate;
332
	unsigned long hash_key = rxrpc_peer_hash_key(local, srx);
333

D
David Howells 已提交
334
	_enter("{%pISp}", &srx->transport);
335 336

	/* search the peer list first */
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
	rcu_read_lock();
	peer = __rxrpc_lookup_peer_rcu(local, srx, hash_key);
	if (peer && !rxrpc_get_peer_maybe(peer))
		peer = NULL;
	rcu_read_unlock();

	if (!peer) {
		/* The peer is not yet present in hash - create a candidate
		 * for a new record and then redo the search.
		 */
		candidate = rxrpc_create_peer(local, srx, hash_key, gfp);
		if (!candidate) {
			_leave(" = NULL [nomem]");
			return NULL;
		}
352

353
		spin_lock_bh(&rxrpc_peer_hash_lock);
354

355 356 357 358 359 360 361
		/* Need to check that we aren't racing with someone else */
		peer = __rxrpc_lookup_peer_rcu(local, srx, hash_key);
		if (peer && !rxrpc_get_peer_maybe(peer))
			peer = NULL;
		if (!peer)
			hash_add_rcu(rxrpc_peer_hash,
				     &candidate->hash_link, hash_key);
362

363
		spin_unlock_bh(&rxrpc_peer_hash_lock);
364

365 366 367 368 369
		if (peer)
			kfree(candidate);
		else
			peer = candidate;
	}
370

D
David Howells 已提交
371
	_net("PEER %d {%pISp}", peer->debug_id, &peer->srx.transport);
372

373
	_leave(" = %p {u=%d}", peer, atomic_read(&peer->usage));
374 375 376 377
	return peer;
}

/*
378
 * Discard a ref on a remote peer record.
379
 */
380
void __rxrpc_put_peer(struct rxrpc_peer *peer)
381
{
382
	ASSERT(hlist_empty(&peer->error_targets));
383

384
	spin_lock_bh(&rxrpc_peer_hash_lock);
385
	hash_del_rcu(&peer->hash_link);
386
	spin_unlock_bh(&rxrpc_peer_hash_lock);
387

388
	kfree_rcu(peer, rcu);
389
}
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404

/**
 * rxrpc_kernel_get_peer - Get the peer address of a call
 * @sock: The socket on which the call is in progress.
 * @call: The call to query
 * @_srx: Where to place the result
 *
 * Get the address of the remote peer in a call.
 */
void rxrpc_kernel_get_peer(struct socket *sock, struct rxrpc_call *call,
			   struct sockaddr_rxrpc *_srx)
{
	*_srx = call->peer->srx;
}
EXPORT_SYMBOL(rxrpc_kernel_get_peer);