1/*
2 * Copyright (c) 2006 Oracle.  All rights reserved.
3 *
4 * This software is available to you under a choice of one of two
5 * licenses.  You may choose to be licensed under the terms of the GNU
6 * General Public License (GPL) Version 2, available from the file
7 * COPYING in the main directory of this source tree, or the
8 * OpenIB.org BSD license below:
9 *
10 *     Redistribution and use in source and binary forms, with or
11 *     without modification, are permitted provided that the following
12 *     conditions are met:
13 *
14 *      - Redistributions of source code must retain the above
15 *        copyright notice, this list of conditions and the following
16 *        disclaimer.
17 *
18 *      - Redistributions in binary form must reproduce the above
19 *        copyright notice, this list of conditions and the following
20 *        disclaimer in the documentation and/or other materials
21 *        provided with the distribution.
22 *
23 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
24 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
25 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
26 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
27 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
28 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
29 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
30 * SOFTWARE.
31 *
32 */
33#include <linux/kernel.h>
34#include <net/sock.h>
35#include <linux/in.h>
36#include <linux/if_arp.h>
37#include <linux/jhash.h>
38#include <linux/ratelimit.h>
39#include "rds.h"
40
41static struct rhashtable bind_hash_table;
42
43static struct rhashtable_params ht_parms = {
44	.nelem_hint = 768,
45	.key_len = sizeof(u64),
46	.key_offset = offsetof(struct rds_sock, rs_bound_key),
47	.head_offset = offsetof(struct rds_sock, rs_bound_node),
48	.max_size = 16384,
49	.min_size = 1024,
50};
51
52/*
53 * Return the rds_sock bound at the given local address.
54 *
55 * The rx path can race with rds_release.  We notice if rds_release() has
56 * marked this socket and don't return a rs ref to the rx path.
57 */
58struct rds_sock *rds_find_bound(__be32 addr, __be16 port)
59{
60	u64 key = ((u64)addr << 32) | port;
61	struct rds_sock *rs;
62
63	rs = rhashtable_lookup_fast(&bind_hash_table, &key, ht_parms);
64	if (rs && !sock_flag(rds_rs_to_sk(rs), SOCK_DEAD))
65		rds_sock_addref(rs);
66	else
67		rs = NULL;
68
69	rdsdebug("returning rs %p for %pI4:%u\n", rs, &addr,
70		ntohs(port));
71
72	return rs;
73}
74
75/* returns -ve errno or +ve port */
76static int rds_add_bound(struct rds_sock *rs, __be32 addr, __be16 *port)
77{
78	int ret = -EADDRINUSE;
79	u16 rover, last;
80	u64 key;
81
82	if (*port != 0) {
83		rover = be16_to_cpu(*port);
84		last = rover;
85	} else {
86		rover = max_t(u16, prandom_u32(), 2);
87		last = rover - 1;
88	}
89
90	do {
91		if (rover == 0)
92			rover++;
93
94		key = ((u64)addr << 32) | cpu_to_be16(rover);
95		if (rhashtable_lookup_fast(&bind_hash_table, &key, ht_parms))
96			continue;
97
98		rs->rs_bound_key = key;
99		rs->rs_bound_addr = addr;
100		rs->rs_bound_port = cpu_to_be16(rover);
101		rs->rs_bound_node.next = NULL;
102		rds_sock_addref(rs);
103		if (!rhashtable_insert_fast(&bind_hash_table,
104					    &rs->rs_bound_node, ht_parms)) {
105			*port = rs->rs_bound_port;
106			ret = 0;
107			rdsdebug("rs %p binding to %pI4:%d\n",
108			  rs, &addr, (int)ntohs(*port));
109			break;
110		} else {
111			rds_sock_put(rs);
112			ret = -ENOMEM;
113			break;
114		}
115	} while (rover++ != last);
116
117	return ret;
118}
119
120void rds_remove_bound(struct rds_sock *rs)
121{
122
123	if (!rs->rs_bound_addr)
124		return;
125
126	rdsdebug("rs %p unbinding from %pI4:%d\n",
127		 rs, &rs->rs_bound_addr,
128		 ntohs(rs->rs_bound_port));
129
130	rhashtable_remove_fast(&bind_hash_table, &rs->rs_bound_node, ht_parms);
131	rds_sock_put(rs);
132	rs->rs_bound_addr = 0;
133}
134
135int rds_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
136{
137	struct sock *sk = sock->sk;
138	struct sockaddr_in *sin = (struct sockaddr_in *)uaddr;
139	struct rds_sock *rs = rds_sk_to_rs(sk);
140	struct rds_transport *trans;
141	int ret = 0;
142
143	lock_sock(sk);
144
145	if (addr_len != sizeof(struct sockaddr_in) ||
146	    sin->sin_family != AF_INET ||
147	    rs->rs_bound_addr ||
148	    sin->sin_addr.s_addr == htonl(INADDR_ANY)) {
149		ret = -EINVAL;
150		goto out;
151	}
152
153	ret = rds_add_bound(rs, sin->sin_addr.s_addr, &sin->sin_port);
154	if (ret)
155		goto out;
156
157	if (rs->rs_transport) { /* previously bound */
158		trans = rs->rs_transport;
159		if (trans->laddr_check(sock_net(sock->sk),
160				       sin->sin_addr.s_addr) != 0) {
161			ret = -ENOPROTOOPT;
162			rds_remove_bound(rs);
163		} else {
164			ret = 0;
165		}
166		goto out;
167	}
168	trans = rds_trans_get_preferred(sock_net(sock->sk),
169					sin->sin_addr.s_addr);
170	if (!trans) {
171		ret = -EADDRNOTAVAIL;
172		rds_remove_bound(rs);
173		printk_ratelimited(KERN_INFO "RDS: rds_bind() could not find a transport, "
174				"load rds_tcp or rds_rdma?\n");
175		goto out;
176	}
177
178	rs->rs_transport = trans;
179	ret = 0;
180
181out:
182	release_sock(sk);
183	return ret;
184}
185
186void rds_bind_lock_destroy(void)
187{
188	rhashtable_destroy(&bind_hash_table);
189}
190
191int rds_bind_lock_init(void)
192{
193	return rhashtable_init(&bind_hash_table, &ht_parms);
194}
195