1 /*- 2 * Copyright (c) 2008 Isilon Inc http://www.isilon.com/ 3 * Authors: Doug Rabson <dfr@rabson.org> 4 * Developed with Red Inc: Alfred Perlstein <alfred@freebsd.org> 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice, this list of conditions and the following disclaimer. 11 * 2. Redistributions in binary form must reproduce the above copyright 12 * notice, this list of conditions and the following disclaimer in the 13 * documentation and/or other materials provided with the distribution. 14 * 15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 18 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 19 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 25 * SUCH DAMAGE. 26 */ 27 28 #include <sys/cdefs.h> 29 __FBSDID("$FreeBSD$"); 30 31 #include <sys/param.h> 32 #include <sys/hash.h> 33 #include <sys/kernel.h> 34 #include <sys/lock.h> 35 #include <sys/malloc.h> 36 #include <sys/mbuf.h> 37 #include <sys/mutex.h> 38 #include <sys/queue.h> 39 40 #include <rpc/rpc.h> 41 #include <rpc/replay.h> 42 43 struct replay_cache_entry { 44 int rce_hash; 45 struct rpc_msg rce_msg; 46 struct sockaddr_storage rce_addr; 47 struct rpc_msg rce_repmsg; 48 struct mbuf *rce_repbody; 49 50 TAILQ_ENTRY(replay_cache_entry) rce_link; 51 TAILQ_ENTRY(replay_cache_entry) rce_alllink; 52 }; 53 TAILQ_HEAD(replay_cache_list, replay_cache_entry); 54 55 static struct replay_cache_entry * 56 replay_alloc(struct replay_cache *rc, struct rpc_msg *msg, 57 struct sockaddr *addr, int h); 58 static void replay_free(struct replay_cache *rc, 59 struct replay_cache_entry *rce); 60 static void replay_prune(struct replay_cache *rc); 61 62 #define REPLAY_HASH_SIZE 256 63 #define REPLAY_MAX 1024 64 65 struct replay_cache { 66 struct replay_cache_list rc_cache[REPLAY_HASH_SIZE]; 67 struct replay_cache_list rc_all; 68 struct mtx rc_lock; 69 int rc_count; 70 size_t rc_size; 71 size_t rc_maxsize; 72 }; 73 74 struct replay_cache * 75 replay_newcache(size_t maxsize) 76 { 77 struct replay_cache *rc; 78 int i; 79 80 rc = malloc(sizeof(*rc), M_RPC, M_WAITOK|M_ZERO); 81 for (i = 0; i < REPLAY_HASH_SIZE; i++) 82 TAILQ_INIT(&rc->rc_cache[i]); 83 TAILQ_INIT(&rc->rc_all); 84 mtx_init(&rc->rc_lock, "rc_lock", NULL, MTX_DEF); 85 rc->rc_maxsize = maxsize; 86 87 return (rc); 88 } 89 90 void 91 replay_setsize(struct replay_cache *rc, size_t newmaxsize) 92 { 93 94 mtx_lock(&rc->rc_lock); 95 rc->rc_maxsize = newmaxsize; 96 replay_prune(rc); 97 mtx_unlock(&rc->rc_lock); 98 } 99 100 void 101 replay_freecache(struct replay_cache *rc) 102 { 103 104 mtx_lock(&rc->rc_lock); 105 while (TAILQ_FIRST(&rc->rc_all)) 106 replay_free(rc, TAILQ_FIRST(&rc->rc_all)); 107 mtx_destroy(&rc->rc_lock); 108 free(rc, M_RPC); 109 } 110 111 static struct replay_cache_entry * 112 replay_alloc(struct replay_cache *rc, 113 struct rpc_msg *msg, struct sockaddr *addr, int h) 114 { 115 struct replay_cache_entry *rce; 116 117 mtx_assert(&rc->rc_lock, MA_OWNED); 118 119 rc->rc_count++; 120 rce = malloc(sizeof(*rce), M_RPC, M_NOWAIT|M_ZERO); 121 if (!rce) 122 return (NULL); 123 rce->rce_hash = h; 124 rce->rce_msg = *msg; 125 bcopy(addr, &rce->rce_addr, addr->sa_len); 126 127 TAILQ_INSERT_HEAD(&rc->rc_cache[h], rce, rce_link); 128 TAILQ_INSERT_HEAD(&rc->rc_all, rce, rce_alllink); 129 130 return (rce); 131 } 132 133 static void 134 replay_free(struct replay_cache *rc, struct replay_cache_entry *rce) 135 { 136 137 mtx_assert(&rc->rc_lock, MA_OWNED); 138 139 rc->rc_count--; 140 TAILQ_REMOVE(&rc->rc_cache[rce->rce_hash], rce, rce_link); 141 TAILQ_REMOVE(&rc->rc_all, rce, rce_alllink); 142 if (rce->rce_repbody) { 143 rc->rc_size -= m_length(rce->rce_repbody, NULL); 144 m_freem(rce->rce_repbody); 145 } 146 free(rce, M_RPC); 147 } 148 149 static void 150 replay_prune(struct replay_cache *rc) 151 { 152 struct replay_cache_entry *rce; 153 154 mtx_assert(&rc->rc_lock, MA_OWNED); 155 156 if (rc->rc_count < REPLAY_MAX && rc->rc_size <= rc->rc_maxsize) 157 return; 158 159 do { 160 /* 161 * Try to free an entry. Don't free in-progress entries. 162 */ 163 TAILQ_FOREACH_REVERSE(rce, &rc->rc_all, replay_cache_list, 164 rce_alllink) { 165 if (rce->rce_repmsg.rm_xid) 166 break; 167 } 168 if (rce) 169 replay_free(rc, rce); 170 } while (rce && (rc->rc_count >= REPLAY_MAX 171 || rc->rc_size > rc->rc_maxsize)); 172 } 173 174 enum replay_state 175 replay_find(struct replay_cache *rc, struct rpc_msg *msg, 176 struct sockaddr *addr, struct rpc_msg *repmsg, struct mbuf **mp) 177 { 178 int h = HASHSTEP(HASHINIT, msg->rm_xid) % REPLAY_HASH_SIZE; 179 struct replay_cache_entry *rce; 180 181 mtx_lock(&rc->rc_lock); 182 TAILQ_FOREACH(rce, &rc->rc_cache[h], rce_link) { 183 if (rce->rce_msg.rm_xid == msg->rm_xid 184 && rce->rce_msg.rm_call.cb_prog == msg->rm_call.cb_prog 185 && rce->rce_msg.rm_call.cb_vers == msg->rm_call.cb_vers 186 && rce->rce_msg.rm_call.cb_proc == msg->rm_call.cb_proc 187 && rce->rce_addr.ss_len == addr->sa_len 188 && bcmp(&rce->rce_addr, addr, addr->sa_len) == 0) { 189 if (rce->rce_repmsg.rm_xid) { 190 /* 191 * We have a reply for this 192 * message. Copy it and return. Keep 193 * replay_all LRU sorted 194 */ 195 TAILQ_REMOVE(&rc->rc_all, rce, rce_alllink); 196 TAILQ_INSERT_HEAD(&rc->rc_all, rce, 197 rce_alllink); 198 *repmsg = rce->rce_repmsg; 199 if (rce->rce_repbody) { 200 *mp = m_copym(rce->rce_repbody, 201 0, M_COPYALL, M_NOWAIT); 202 mtx_unlock(&rc->rc_lock); 203 if (!*mp) 204 return (RS_ERROR); 205 } else { 206 mtx_unlock(&rc->rc_lock); 207 } 208 return (RS_DONE); 209 } else { 210 mtx_unlock(&rc->rc_lock); 211 return (RS_INPROGRESS); 212 } 213 } 214 } 215 216 replay_prune(rc); 217 218 rce = replay_alloc(rc, msg, addr, h); 219 220 mtx_unlock(&rc->rc_lock); 221 222 if (!rce) 223 return (RS_ERROR); 224 else 225 return (RS_NEW); 226 } 227 228 void 229 replay_setreply(struct replay_cache *rc, 230 struct rpc_msg *repmsg, struct sockaddr *addr, struct mbuf *m) 231 { 232 int h = HASHSTEP(HASHINIT, repmsg->rm_xid) % REPLAY_HASH_SIZE; 233 struct replay_cache_entry *rce; 234 235 /* 236 * Copy the reply before the lock so we can sleep. 237 */ 238 if (m) 239 m = m_copym(m, 0, M_COPYALL, M_WAITOK); 240 241 mtx_lock(&rc->rc_lock); 242 TAILQ_FOREACH(rce, &rc->rc_cache[h], rce_link) { 243 if (rce->rce_msg.rm_xid == repmsg->rm_xid 244 && rce->rce_addr.ss_len == addr->sa_len 245 && bcmp(&rce->rce_addr, addr, addr->sa_len) == 0) { 246 break; 247 } 248 } 249 if (rce) { 250 rce->rce_repmsg = *repmsg; 251 rce->rce_repbody = m; 252 if (m) 253 rc->rc_size += m_length(m, NULL); 254 } 255 mtx_unlock(&rc->rc_lock); 256 } 257