1 /* 2 * Copyright (c) 2006, 2018 Oracle and/or its affiliates. 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/ipv6.h> 37 #include <linux/if_arp.h> 38 #include <linux/jhash.h> 39 #include <linux/ratelimit.h> 40 #include "rds.h" 41 42 static struct rhashtable bind_hash_table; 43 44 static const struct rhashtable_params ht_parms = { 45 .nelem_hint = 768, 46 .key_len = RDS_BOUND_KEY_LEN, 47 .key_offset = offsetof(struct rds_sock, rs_bound_key), 48 .head_offset = offsetof(struct rds_sock, rs_bound_node), 49 .max_size = 16384, 50 .min_size = 1024, 51 }; 52 53 /* Create a key for the bind hash table manipulation. Port is in network byte 54 * order. 55 */ 56 static inline void __rds_create_bind_key(u8 *key, const struct in6_addr *addr, 57 __be16 port, __u32 scope_id) 58 { 59 memcpy(key, addr, sizeof(*addr)); 60 key += sizeof(*addr); 61 memcpy(key, &port, sizeof(port)); 62 key += sizeof(port); 63 memcpy(key, &scope_id, sizeof(scope_id)); 64 } 65 66 /* 67 * Return the rds_sock bound at the given local address. 68 * 69 * The rx path can race with rds_release. We notice if rds_release() has 70 * marked this socket and don't return a rs ref to the rx path. 71 */ 72 struct rds_sock *rds_find_bound(const struct in6_addr *addr, __be16 port, 73 __u32 scope_id) 74 { 75 u8 key[RDS_BOUND_KEY_LEN]; 76 struct rds_sock *rs; 77 78 __rds_create_bind_key(key, addr, port, scope_id); 79 rs = rhashtable_lookup_fast(&bind_hash_table, key, ht_parms); 80 if (rs && !sock_flag(rds_rs_to_sk(rs), SOCK_DEAD)) 81 rds_sock_addref(rs); 82 else 83 rs = NULL; 84 85 rdsdebug("returning rs %p for %pI6c:%u\n", rs, addr, 86 ntohs(port)); 87 88 return rs; 89 } 90 91 /* returns -ve errno or +ve port */ 92 static int rds_add_bound(struct rds_sock *rs, const struct in6_addr *addr, 93 __be16 *port, __u32 scope_id) 94 { 95 int ret = -EADDRINUSE; 96 u16 rover, last; 97 u8 key[RDS_BOUND_KEY_LEN]; 98 99 if (*port != 0) { 100 rover = be16_to_cpu(*port); 101 if (rover == RDS_FLAG_PROBE_PORT) 102 return -EINVAL; 103 last = rover; 104 } else { 105 rover = max_t(u16, prandom_u32(), 2); 106 last = rover - 1; 107 } 108 109 do { 110 if (rover == 0) 111 rover++; 112 113 if (rover == RDS_FLAG_PROBE_PORT) 114 continue; 115 __rds_create_bind_key(key, addr, cpu_to_be16(rover), 116 scope_id); 117 if (rhashtable_lookup_fast(&bind_hash_table, key, ht_parms)) 118 continue; 119 120 memcpy(rs->rs_bound_key, key, sizeof(rs->rs_bound_key)); 121 rs->rs_bound_addr = *addr; 122 net_get_random_once(&rs->rs_hash_initval, 123 sizeof(rs->rs_hash_initval)); 124 rs->rs_bound_port = cpu_to_be16(rover); 125 rs->rs_bound_node.next = NULL; 126 rds_sock_addref(rs); 127 if (!rhashtable_insert_fast(&bind_hash_table, 128 &rs->rs_bound_node, ht_parms)) { 129 *port = rs->rs_bound_port; 130 rs->rs_bound_scope_id = scope_id; 131 ret = 0; 132 rdsdebug("rs %p binding to %pI6c:%d\n", 133 rs, addr, (int)ntohs(*port)); 134 break; 135 } else { 136 rs->rs_bound_addr = in6addr_any; 137 rds_sock_put(rs); 138 ret = -ENOMEM; 139 break; 140 } 141 } while (rover++ != last); 142 143 return ret; 144 } 145 146 void rds_remove_bound(struct rds_sock *rs) 147 { 148 149 if (ipv6_addr_any(&rs->rs_bound_addr)) 150 return; 151 152 rdsdebug("rs %p unbinding from %pI6c:%d\n", 153 rs, &rs->rs_bound_addr, 154 ntohs(rs->rs_bound_port)); 155 156 rhashtable_remove_fast(&bind_hash_table, &rs->rs_bound_node, ht_parms); 157 rds_sock_put(rs); 158 rs->rs_bound_addr = in6addr_any; 159 } 160 161 int rds_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) 162 { 163 struct sock *sk = sock->sk; 164 struct rds_sock *rs = rds_sk_to_rs(sk); 165 struct in6_addr v6addr, *binding_addr; 166 struct rds_transport *trans; 167 __u32 scope_id = 0; 168 int ret = 0; 169 __be16 port; 170 171 /* We allow an RDS socket to be bound to either IPv4 or IPv6 172 * address. 173 */ 174 if (uaddr->sa_family == AF_INET) { 175 struct sockaddr_in *sin = (struct sockaddr_in *)uaddr; 176 177 if (addr_len < sizeof(struct sockaddr_in) || 178 sin->sin_addr.s_addr == htonl(INADDR_ANY) || 179 sin->sin_addr.s_addr == htonl(INADDR_BROADCAST) || 180 IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) 181 return -EINVAL; 182 ipv6_addr_set_v4mapped(sin->sin_addr.s_addr, &v6addr); 183 binding_addr = &v6addr; 184 port = sin->sin_port; 185 #if IS_ENABLED(CONFIG_IPV6) 186 } else if (uaddr->sa_family == AF_INET6) { 187 struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)uaddr; 188 int addr_type; 189 190 if (addr_len < sizeof(struct sockaddr_in6)) 191 return -EINVAL; 192 addr_type = ipv6_addr_type(&sin6->sin6_addr); 193 if (!(addr_type & IPV6_ADDR_UNICAST)) { 194 __be32 addr4; 195 196 if (!(addr_type & IPV6_ADDR_MAPPED)) 197 return -EINVAL; 198 199 /* It is a mapped address. Need to do some sanity 200 * checks. 201 */ 202 addr4 = sin6->sin6_addr.s6_addr32[3]; 203 if (addr4 == htonl(INADDR_ANY) || 204 addr4 == htonl(INADDR_BROADCAST) || 205 IN_MULTICAST(ntohl(addr4))) 206 return -EINVAL; 207 } 208 /* The scope ID must be specified for link local address. */ 209 if (addr_type & IPV6_ADDR_LINKLOCAL) { 210 if (sin6->sin6_scope_id == 0) 211 return -EINVAL; 212 scope_id = sin6->sin6_scope_id; 213 } 214 binding_addr = &sin6->sin6_addr; 215 port = sin6->sin6_port; 216 #endif 217 } else { 218 return -EINVAL; 219 } 220 lock_sock(sk); 221 222 /* RDS socket does not allow re-binding. */ 223 if (!ipv6_addr_any(&rs->rs_bound_addr)) { 224 ret = -EINVAL; 225 goto out; 226 } 227 /* Socket is connected. The binding address should have the same 228 * scope ID as the connected address, except the case when one is 229 * non-link local address (scope_id is 0). 230 */ 231 if (!ipv6_addr_any(&rs->rs_conn_addr) && scope_id && 232 rs->rs_bound_scope_id && 233 scope_id != rs->rs_bound_scope_id) { 234 ret = -EINVAL; 235 goto out; 236 } 237 238 ret = rds_add_bound(rs, binding_addr, &port, scope_id); 239 if (ret) 240 goto out; 241 242 if (rs->rs_transport) { /* previously bound */ 243 trans = rs->rs_transport; 244 if (trans->laddr_check(sock_net(sock->sk), 245 binding_addr, scope_id) != 0) { 246 ret = -ENOPROTOOPT; 247 rds_remove_bound(rs); 248 } else { 249 ret = 0; 250 } 251 goto out; 252 } 253 trans = rds_trans_get_preferred(sock_net(sock->sk), binding_addr, 254 scope_id); 255 if (!trans) { 256 ret = -EADDRNOTAVAIL; 257 rds_remove_bound(rs); 258 pr_info_ratelimited("RDS: %s could not find a transport for %pI6c, load rds_tcp or rds_rdma?\n", 259 __func__, binding_addr); 260 goto out; 261 } 262 263 rs->rs_transport = trans; 264 ret = 0; 265 266 out: 267 release_sock(sk); 268 return ret; 269 } 270 271 void rds_bind_lock_destroy(void) 272 { 273 rhashtable_destroy(&bind_hash_table); 274 } 275 276 int rds_bind_lock_init(void) 277 { 278 return rhashtable_init(&bind_hash_table, &ht_parms); 279 } 280