1 /* 2 * CDDL HEADER START 3 * 4 * The contents of this file are subject to the terms of the 5 * Common Development and Distribution License (the "License"). 6 * You may not use this file except in compliance with the License. 7 * 8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE 9 * or http://www.opensolaris.org/os/licensing. 10 * See the License for the specific language governing permissions 11 * and limitations under the License. 12 * 13 * When distributing Covered Code, include this CDDL HEADER in each 14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE. 15 * If applicable, add the following below this CDDL HEADER, with the 16 * fields enclosed by brackets "[]" replaced with your own identifying 17 * information: Portions Copyright [yyyy] [name of copyright owner] 18 * 19 * CDDL HEADER END 20 */ 21 22 /* 23 * Copyright 2009 Sun Microsystems, Inc. All rights reserved. 24 * Use is subject to license terms. 25 */ 26 27 /* 28 * FMD Transport Subsystem 29 * 30 * A transport module uses some underlying mechanism to transport events. 31 * This mechanism may use any underlying link-layer protocol and may support 32 * additional link-layer packets unrelated to FMA. Some appropriate link- 33 * layer mechanism to create the underlying connection is expected to be 34 * called prior to calling fmd_xprt_open() itself. Alternatively, a transport 35 * may be created in the suspended state by specifying the FMD_XPRT_SUSPENDED 36 * flag as part of the call to fmd_xprt_open(), and then may be resumed later. 37 * The underlying transport mechanism is *required* to provide ordering: that 38 * is, the sequences of bytes written across the transport must be read by 39 * the remote peer in the order that they are written, even across separate 40 * calls to fmdo_send(). As an example, the Internet TCP protocol would be 41 * a valid transport as it guarantees ordering, whereas the Internet UDP 42 * protocol would not because UDP datagrams may be delivered in any order 43 * as a result of delays introduced when datagrams pass through routers. 44 * 45 * Similar to sending events, a transport module receives events that are from 46 * its peer remote endpoint using some transport-specific mechanism that is 47 * unknown to FMD. As each event is received, the transport module is 48 * responsible for constructing a valid nvlist_t object from the data and then 49 * calling fmd_xprt_post() to post the event to the containing FMD's dispatch 50 * queue, making it available to all local modules that are not transport 51 * modules that have subscribed to the event. 52 * 53 * The following state machine is used for each transport. The initial state 54 * is either SYN, ACK, or RUN, depending on the flags specified to xprt_create. 55 * 56 * FMD_XPRT_ACCEPT !FMD_XPRT_ACCEPT 57 * | | 58 * waiting +--v--+ +--v--+ waiting 59 * for syn | SYN |--+ --+| ACK | for ack 60 * event +-----+ \ / +-----+ event 61 * | \ / | 62 * drop all +--v--+ X +--v--+ send subscriptions, 63 * events | ERR |<---+ +--->| SUB | recv subscriptions, 64 * +-----+ +-----+ wait for run event 65 * ^ | 66 * | +-----+ | 67 * +-----| RUN |<----+ 68 * +--^--+ 69 * | 70 * FMD_XPRT_RDONLY 71 * 72 * When fmd_xprt_open() is called without FMD_XPRT_ACCEPT, the Common Transport 73 * Layer enqueues a "syn" event for the module in its event queue and sets the 74 * state to ACK. In state ACK, we are waiting for the transport to get an 75 * "ack" event and call fmd_xprt_post() on this event. Other events will be 76 * discarded. If an "ack" is received, we transition to state SUB. If a 77 * configurable timeout occurs or if the "ack" is invalid (e.g. invalid version 78 * exchange), we transition to state ERR. Once in state ERR, no further 79 * operations are valid except fmd_xprt_close() and fmd_xprt_error() will 80 * return a non-zero value to the caller indicating the transport has failed. 81 * 82 * When fmd_xprt_open() is called with FMD_XPRT_ACCEPT, the Common Transport 83 * Layer assumes this transport is being used to accept a virtual connection 84 * from a remote peer that is sending a "syn", and sets the initial state to 85 * SYN. In this state, the transport waits for a "syn" event, validates it, 86 * and then transitions to state SUB if it is valid or state ERR if it is not. 87 * 88 * Once in state SUB, the transport module is expected to receive a sequence of 89 * zero or more "subscribe" events from the remote peer, followed by a "run" 90 * event. Once in state RUN, the transport is active and any events can be 91 * sent or received. The transport module is free to call fmd_xprt_close() 92 * from any state. The fmd_xprt_error() function will return zero if the 93 * transport is not in the ERR state, or non-zero if it is in the ERR state. 94 * 95 * Once the state machine reaches RUN, other FMA protocol events can be sent 96 * and received across the transport in addition to the various control events. 97 * 98 * Table of Common Transport Layer Control Events 99 * ============================================== 100 * 101 * FMA Class Payload 102 * --------- ------- 103 * resource.fm.xprt.uuclose string (uuid of case) 104 * resource.fm.xprt.uuresolved string (uuid of case) 105 * resource.fm.xprt.updated string (uuid of case) 106 * resource.fm.xprt.subscribe string (class pattern) 107 * resource.fm.xprt.unsubscribe string (class pattern) 108 * resource.fm.xprt.unsuback string (class pattern) 109 * resource.fm.xprt.syn version information 110 * resource.fm.xprt.ack version information 111 * resource.fm.xprt.run version information 112 * 113 * Control events are used to add and delete proxy subscriptions on the remote 114 * transport peer module, and to set up connections. When a "syn" event is 115 * sent, FMD will include in the payload the highest version of the FMA event 116 * protocol that is supported by the sender. When a "syn" event is received, 117 * the receiving FMD will use the minimum of this version and its version of 118 * the protocol, and reply with this new minimum version in the "ack" event. 119 * The receiver will then use this new minimum for subsequent event semantics. 120 */ 121 122 #include <sys/fm/protocol.h> 123 #include <strings.h> 124 #include <limits.h> 125 126 #include <fmd_alloc.h> 127 #include <fmd_error.h> 128 #include <fmd_conf.h> 129 #include <fmd_subr.h> 130 #include <fmd_string.h> 131 #include <fmd_protocol.h> 132 #include <fmd_thread.h> 133 #include <fmd_eventq.h> 134 #include <fmd_dispq.h> 135 #include <fmd_ctl.h> 136 #include <fmd_log.h> 137 #include <fmd_ustat.h> 138 #include <fmd_case.h> 139 #include <fmd_api.h> 140 #include <fmd_fmri.h> 141 #include <fmd_asru.h> 142 #include <fmd_xprt.h> 143 144 #include <fmd.h> 145 146 /* 147 * The states shown above in the transport state machine diagram are encoded 148 * using arrays of class patterns and a corresponding action function. These 149 * arrays are then passed to fmd_xprt_transition() to change transport states. 150 */ 151 152 const fmd_xprt_rule_t _fmd_xprt_state_syn[] = { 153 { "resource.fm.xprt.syn", fmd_xprt_event_syn }, 154 { "*", fmd_xprt_event_error }, 155 { NULL, NULL } 156 }; 157 158 const fmd_xprt_rule_t _fmd_xprt_state_ack[] = { 159 { "resource.fm.xprt.ack", fmd_xprt_event_ack }, 160 { "*", fmd_xprt_event_error }, 161 }; 162 163 const fmd_xprt_rule_t _fmd_xprt_state_err[] = { 164 { "*", fmd_xprt_event_drop }, 165 { NULL, NULL } 166 }; 167 168 const fmd_xprt_rule_t _fmd_xprt_state_sub[] = { 169 { "resource.fm.xprt.subscribe", fmd_xprt_event_sub }, 170 { "resource.fm.xprt.run", fmd_xprt_event_run }, 171 { "resource.fm.xprt.*", fmd_xprt_event_error }, 172 { "*", fmd_xprt_event_drop }, 173 { NULL, NULL } 174 }; 175 176 const fmd_xprt_rule_t _fmd_xprt_state_run[] = { 177 { "resource.fm.xprt.subscribe", fmd_xprt_event_sub }, 178 { "resource.fm.xprt.unsubscribe", fmd_xprt_event_unsub }, 179 { "resource.fm.xprt.unsuback", fmd_xprt_event_unsuback }, 180 { "resource.fm.xprt.uuclose", fmd_xprt_event_uuclose }, 181 { "resource.fm.xprt.uuresolved", fmd_xprt_event_uuresolved }, 182 { "resource.fm.xprt.updated", fmd_xprt_event_updated }, 183 { "resource.fm.xprt.*", fmd_xprt_event_error }, 184 { NULL, NULL } 185 }; 186 187 /* 188 * Template for per-transport statistics installed by fmd on behalf of each 189 * transport. These are used to initialize the per-transport xi_stats. For 190 * each statistic, the name is prepended with "fmd.xprt.%u", where %u is the 191 * transport ID (xi_id) and then are inserted into the per-module stats hash. 192 * The values in this array must match fmd_xprt_stat_t from <fmd_xprt.h>. 193 */ 194 static const fmd_xprt_stat_t _fmd_xprt_stat_tmpl = { 195 { 196 { "dispatched", FMD_TYPE_UINT64, "total events dispatched to transport" }, 197 { "dequeued", FMD_TYPE_UINT64, "total events dequeued by transport" }, 198 { "prdequeued", FMD_TYPE_UINT64, "protocol events dequeued by transport" }, 199 { "dropped", FMD_TYPE_UINT64, "total events dropped on queue overflow" }, 200 { "wcnt", FMD_TYPE_UINT32, "count of events waiting on queue" }, 201 { "wtime", FMD_TYPE_TIME, "total wait time on queue" }, 202 { "wlentime", FMD_TYPE_TIME, "total wait length * time product" }, 203 { "wlastupdate", FMD_TYPE_TIME, "hrtime of last wait queue update" }, 204 { "dtime", FMD_TYPE_TIME, "total processing time after dequeue" }, 205 { "dlastupdate", FMD_TYPE_TIME, "hrtime of last event dequeue completion" }, 206 }, 207 { "module", FMD_TYPE_STRING, "module that owns this transport" }, 208 { "authority", FMD_TYPE_STRING, "authority associated with this transport" }, 209 { "state", FMD_TYPE_STRING, "current transport state" }, 210 { "received", FMD_TYPE_UINT64, "events received by transport" }, 211 { "discarded", FMD_TYPE_UINT64, "bad events discarded by transport" }, 212 { "retried", FMD_TYPE_UINT64, "retries requested of transport" }, 213 { "replayed", FMD_TYPE_UINT64, "events replayed by transport" }, 214 { "lost", FMD_TYPE_UINT64, "events lost by transport" }, 215 { "timeouts", FMD_TYPE_UINT64, "events received by transport with ttl=0" }, 216 { "subscriptions", FMD_TYPE_UINT64, "subscriptions registered to transport" }, 217 }; 218 219 static void 220 fmd_xprt_class_hash_create(fmd_xprt_class_hash_t *xch, fmd_eventq_t *eq) 221 { 222 uint_t hashlen = fmd.d_str_buckets; 223 224 xch->xch_queue = eq; 225 xch->xch_hashlen = hashlen; 226 xch->xch_hash = fmd_zalloc(sizeof (void *) * hashlen, FMD_SLEEP); 227 } 228 229 static void 230 fmd_xprt_class_hash_destroy(fmd_xprt_class_hash_t *xch) 231 { 232 fmd_eventq_t *eq = xch->xch_queue; 233 fmd_xprt_class_t *xcp, *ncp; 234 uint_t i; 235 236 for (i = 0; i < xch->xch_hashlen; i++) { 237 for (xcp = xch->xch_hash[i]; xcp != NULL; xcp = ncp) { 238 ncp = xcp->xc_next; 239 240 if (eq != NULL) 241 fmd_dispq_delete(fmd.d_disp, eq, xcp->xc_class); 242 243 fmd_strfree(xcp->xc_class); 244 fmd_free(xcp, sizeof (fmd_xprt_class_t)); 245 } 246 } 247 248 fmd_free(xch->xch_hash, sizeof (void *) * xch->xch_hashlen); 249 } 250 251 /* 252 * Insert the specified class into the specified class hash, and return the 253 * reference count. A return value of one indicates this is the first insert. 254 * If an eventq is associated with the hash, insert a dispq subscription for it. 255 */ 256 static uint_t 257 fmd_xprt_class_hash_insert(fmd_xprt_impl_t *xip, 258 fmd_xprt_class_hash_t *xch, const char *class) 259 { 260 uint_t h = fmd_strhash(class) % xch->xch_hashlen; 261 fmd_xprt_class_t *xcp; 262 263 ASSERT(MUTEX_HELD(&xip->xi_lock)); 264 265 for (xcp = xch->xch_hash[h]; xcp != NULL; xcp = xcp->xc_next) { 266 if (strcmp(class, xcp->xc_class) == 0) 267 return (++xcp->xc_refs); 268 } 269 270 xcp = fmd_alloc(sizeof (fmd_xprt_class_t), FMD_SLEEP); 271 xcp->xc_class = fmd_strdup(class, FMD_SLEEP); 272 xcp->xc_next = xch->xch_hash[h]; 273 xcp->xc_refs = 1; 274 xch->xch_hash[h] = xcp; 275 276 if (xch->xch_queue != NULL) 277 fmd_dispq_insert(fmd.d_disp, xch->xch_queue, class); 278 279 return (xcp->xc_refs); 280 } 281 282 /* 283 * Delete the specified class from the specified class hash, and return the 284 * reference count. A return value of zero indicates the class was deleted. 285 * If an eventq is associated with the hash, delete the dispq subscription. 286 */ 287 static uint_t 288 fmd_xprt_class_hash_delete(fmd_xprt_impl_t *xip, 289 fmd_xprt_class_hash_t *xch, const char *class) 290 { 291 uint_t h = fmd_strhash(class) % xch->xch_hashlen; 292 fmd_xprt_class_t *xcp, **pp; 293 294 ASSERT(MUTEX_HELD(&xip->xi_lock)); 295 pp = &xch->xch_hash[h]; 296 297 for (xcp = *pp; xcp != NULL; xcp = xcp->xc_next) { 298 if (strcmp(class, xcp->xc_class) == 0) 299 break; 300 else 301 pp = &xcp->xc_next; 302 } 303 304 if (xcp == NULL) 305 return (-1U); /* explicitly permit an invalid delete */ 306 307 if (--xcp->xc_refs != 0) 308 return (xcp->xc_refs); 309 310 ASSERT(xcp->xc_refs == 0); 311 *pp = xcp->xc_next; 312 313 fmd_strfree(xcp->xc_class); 314 fmd_free(xcp, sizeof (fmd_xprt_class_t)); 315 316 if (xch->xch_queue != NULL) 317 fmd_dispq_delete(fmd.d_disp, xch->xch_queue, class); 318 319 return (0); 320 } 321 322 /* 323 * Queue subscribe events for the specified transport corresponding to all of 324 * the active module subscriptions. This is an extremely heavyweight operation 325 * that we expect to take place rarely (i.e. when loading a transport module 326 * or when it establishes a connection). We lock all of the known modules to 327 * prevent them from adding or deleting subscriptions, then snapshot their 328 * subscriptions, and then unlock all of the modules. We hold the modhash 329 * lock for the duration of this operation to prevent new modules from loading. 330 */ 331 static void 332 fmd_xprt_subscribe_modhash(fmd_xprt_impl_t *xip, fmd_modhash_t *mhp) 333 { 334 fmd_xprt_t *xp = (fmd_xprt_t *)xip; 335 const fmd_conf_path_t *pap; 336 fmd_module_t *mp; 337 uint_t i, j; 338 339 (void) pthread_rwlock_rdlock(&mhp->mh_lock); 340 341 for (i = 0; i < mhp->mh_hashlen; i++) { 342 for (mp = mhp->mh_hash[i]; mp != NULL; mp = mp->mod_next) 343 fmd_module_lock(mp); 344 } 345 346 (void) pthread_mutex_lock(&xip->xi_lock); 347 ASSERT(!(xip->xi_flags & FMD_XPRT_SUBSCRIBER)); 348 xip->xi_flags |= FMD_XPRT_SUBSCRIBER; 349 (void) pthread_mutex_unlock(&xip->xi_lock); 350 351 for (i = 0; i < mhp->mh_hashlen; i++) { 352 for (mp = mhp->mh_hash[i]; mp != NULL; mp = mp->mod_next) { 353 (void) fmd_conf_getprop(mp->mod_conf, 354 FMD_PROP_SUBSCRIPTIONS, &pap); 355 for (j = 0; j < pap->cpa_argc; j++) 356 fmd_xprt_subscribe(xp, pap->cpa_argv[j]); 357 } 358 } 359 360 for (i = 0; i < mhp->mh_hashlen; i++) { 361 for (mp = mhp->mh_hash[i]; mp != NULL; mp = mp->mod_next) 362 fmd_module_unlock(mp); 363 } 364 365 (void) pthread_rwlock_unlock(&mhp->mh_lock); 366 } 367 368 static void 369 fmd_xprt_transition(fmd_xprt_impl_t *xip, 370 const fmd_xprt_rule_t *state, const char *tag) 371 { 372 fmd_event_t *e; 373 nvlist_t *nvl; 374 char *s; 375 376 TRACE((FMD_DBG_XPRT, "xprt %u -> %s\n", xip->xi_id, tag)); 377 378 xip->xi_state = state; 379 s = fmd_strdup(tag, FMD_SLEEP); 380 381 (void) pthread_mutex_lock(&xip->xi_stats_lock); 382 fmd_strfree(xip->xi_stats->xs_state.fmds_value.str); 383 xip->xi_stats->xs_state.fmds_value.str = s; 384 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 385 386 /* 387 * If we've reached the SUB state, take out the big hammer and snapshot 388 * all of the subscriptions of all of the loaded modules. Then queue a 389 * run event for our remote peer indicating that it can enter RUN. 390 */ 391 if (state == _fmd_xprt_state_sub) { 392 fmd_xprt_subscribe_modhash(xip, fmd.d_mod_hash); 393 394 nvl = fmd_protocol_xprt_ctl(xip->xi_queue->eq_mod, 395 "resource.fm.xprt.run", xip->xi_version); 396 397 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 398 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 399 fmd_eventq_insert_at_time(xip->xi_queue, e); 400 } 401 } 402 403 static void 404 fmd_xprt_authupdate(fmd_xprt_impl_t *xip) 405 { 406 char *s = fmd_fmri_auth2str(xip->xi_auth); 407 408 (void) pthread_mutex_lock(&xip->xi_stats_lock); 409 fmd_strfree(xip->xi_stats->xs_authority.fmds_value.str); 410 xip->xi_stats->xs_authority.fmds_value.str = s; 411 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 412 } 413 414 static int 415 fmd_xprt_vmismatch(fmd_xprt_impl_t *xip, nvlist_t *nvl, uint_t *rversionp) 416 { 417 uint8_t rversion; 418 419 if (nvlist_lookup_uint8(nvl, FM_VERSION, &rversion) != 0) { 420 (void) pthread_mutex_lock(&xip->xi_stats_lock); 421 xip->xi_stats->xs_discarded.fmds_value.ui64++; 422 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 423 424 fmd_xprt_transition(xip, _fmd_xprt_state_err, "ERR"); 425 return (1); 426 } 427 428 if (rversion > xip->xi_version) { 429 fmd_dprintf(FMD_DBG_XPRT, "xprt %u protocol mismatch: %u>%u\n", 430 xip->xi_id, rversion, xip->xi_version); 431 432 (void) pthread_mutex_lock(&xip->xi_stats_lock); 433 xip->xi_stats->xs_discarded.fmds_value.ui64++; 434 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 435 436 fmd_xprt_transition(xip, _fmd_xprt_state_err, "ERR"); 437 return (1); 438 } 439 440 if (rversionp != NULL) 441 *rversionp = rversion; 442 443 return (0); 444 } 445 446 void 447 fmd_xprt_event_syn(fmd_xprt_impl_t *xip, nvlist_t *nvl) 448 { 449 fmd_event_t *e; 450 uint_t vers; 451 char *class; 452 453 if (fmd_xprt_vmismatch(xip, nvl, &vers)) 454 return; /* transitioned to error state */ 455 456 /* 457 * If the transport module didn't specify an authority, extract the 458 * one that is passed along with the xprt.syn event and use that. 459 */ 460 if (xip->xi_auth == NULL && 461 nvlist_lookup_nvlist(nvl, FM_RSRC_RESOURCE, &nvl) == 0 && 462 nvlist_lookup_nvlist(nvl, FM_FMRI_AUTHORITY, &nvl) == 0) { 463 (void) nvlist_xdup(nvl, &xip->xi_auth, &fmd.d_nva); 464 fmd_xprt_authupdate(xip); 465 } 466 467 nvl = fmd_protocol_xprt_ctl(xip->xi_queue->eq_mod, 468 "resource.fm.xprt.ack", xip->xi_version); 469 470 (void) nvlist_lookup_string(nvl, FM_CLASS, &class); 471 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, class); 472 fmd_eventq_insert_at_time(xip->xi_queue, e); 473 474 xip->xi_version = MIN(FM_RSRC_XPRT_VERSION, vers); 475 fmd_xprt_transition(xip, _fmd_xprt_state_sub, "SUB"); 476 } 477 478 void 479 fmd_xprt_event_ack(fmd_xprt_impl_t *xip, nvlist_t *nvl) 480 { 481 uint_t vers; 482 483 if (fmd_xprt_vmismatch(xip, nvl, &vers)) 484 return; /* transitioned to error state */ 485 486 /* 487 * If the transport module didn't specify an authority, extract the 488 * one that is passed along with the xprt.syn event and use that. 489 */ 490 if (xip->xi_auth == NULL && 491 nvlist_lookup_nvlist(nvl, FM_RSRC_RESOURCE, &nvl) == 0 && 492 nvlist_lookup_nvlist(nvl, FM_FMRI_AUTHORITY, &nvl) == 0) { 493 (void) nvlist_xdup(nvl, &xip->xi_auth, &fmd.d_nva); 494 fmd_xprt_authupdate(xip); 495 } 496 497 xip->xi_version = MIN(FM_RSRC_XPRT_VERSION, vers); 498 fmd_xprt_transition(xip, _fmd_xprt_state_sub, "SUB"); 499 } 500 501 /* 502 * Upon transition to RUN, we take every solved case and resend a list.suspect 503 * event for it to our remote peer. If a case transitions from solved to a 504 * future state (CLOSE_WAIT, CLOSED, or REPAIRED) while we are iterating over 505 * the case hash, we will get it as part of examining the resource cache, next. 506 */ 507 static void 508 fmd_xprt_send_case(fmd_case_t *cp, void *arg) 509 { 510 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 511 fmd_xprt_impl_t *xip = arg; 512 513 fmd_event_t *e; 514 nvlist_t *nvl; 515 char *class; 516 517 if (cip->ci_state == FMD_CASE_UNSOLVED) 518 return; 519 520 nvl = fmd_case_mkevent(cp, FM_LIST_SUSPECT_CLASS); 521 (void) nvlist_lookup_string(nvl, FM_CLASS, &class); 522 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, class); 523 524 fmd_dprintf(FMD_DBG_XPRT, "re-send %s for %s to transport %u\n", 525 FM_LIST_SUSPECT_CLASS, cip->ci_uuid, xip->xi_id); 526 527 fmd_dispq_dispatch_gid(fmd.d_disp, e, class, xip->xi_queue->eq_sgid); 528 } 529 530 void 531 fmd_xprt_event_run(fmd_xprt_impl_t *xip, nvlist_t *nvl) 532 { 533 if (!fmd_xprt_vmismatch(xip, nvl, NULL)) { 534 fmd_xprt_transition(xip, _fmd_xprt_state_run, "RUN"); 535 fmd_case_hash_apply(fmd.d_cases, fmd_xprt_send_case, xip); 536 } 537 } 538 539 void 540 fmd_xprt_event_sub(fmd_xprt_impl_t *xip, nvlist_t *nvl) 541 { 542 char *class; 543 544 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 545 return; /* transitioned to error state */ 546 547 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_SUBCLASS, &class) != 0) 548 return; /* malformed protocol event */ 549 550 (void) pthread_mutex_lock(&xip->xi_lock); 551 (void) fmd_xprt_class_hash_insert(xip, &xip->xi_lsub, class); 552 (void) pthread_mutex_unlock(&xip->xi_lock); 553 554 (void) pthread_mutex_lock(&xip->xi_stats_lock); 555 xip->xi_stats->xs_subscriptions.fmds_value.ui64++; 556 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 557 } 558 559 void 560 fmd_xprt_event_unsub(fmd_xprt_impl_t *xip, nvlist_t *nvl) 561 { 562 fmd_event_t *e; 563 char *class; 564 565 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 566 return; /* transitioned to error state */ 567 568 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_SUBCLASS, &class) != 0) 569 return; /* malformed protocol event */ 570 571 (void) pthread_mutex_lock(&xip->xi_lock); 572 (void) fmd_xprt_class_hash_delete(xip, &xip->xi_lsub, class); 573 (void) pthread_mutex_unlock(&xip->xi_lock); 574 575 (void) pthread_mutex_lock(&xip->xi_stats_lock); 576 xip->xi_stats->xs_subscriptions.fmds_value.ui64--; 577 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 578 579 nvl = fmd_protocol_xprt_sub(xip->xi_queue->eq_mod, 580 "resource.fm.xprt.unsuback", xip->xi_version, class); 581 582 (void) nvlist_lookup_string(nvl, FM_CLASS, &class); 583 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, class); 584 fmd_eventq_insert_at_time(xip->xi_queue, e); 585 } 586 587 void 588 fmd_xprt_event_unsuback(fmd_xprt_impl_t *xip, nvlist_t *nvl) 589 { 590 char *class; 591 592 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 593 return; /* transitioned to error state */ 594 595 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_SUBCLASS, &class) != 0) 596 return; /* malformed protocol event */ 597 598 (void) pthread_mutex_lock(&xip->xi_lock); 599 (void) fmd_xprt_class_hash_delete(xip, &xip->xi_usub, class); 600 (void) pthread_mutex_unlock(&xip->xi_lock); 601 } 602 603 /* 604 * on diagnosing side, receive a uuclose from the proxy. 605 */ 606 void 607 fmd_xprt_event_uuclose(fmd_xprt_impl_t *xip, nvlist_t *nvl) 608 { 609 fmd_case_t *cp; 610 char *uuid; 611 612 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 613 return; /* transitioned to error state */ 614 615 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_UUID, &uuid) == 0 && 616 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 617 /* 618 * update resource cache status and transition case 619 */ 620 fmd_case_close_status(cp); 621 fmd_case_transition(cp, FMD_CASE_CLOSE_WAIT, FMD_CF_ISOLATED); 622 fmd_case_rele(cp); 623 } 624 } 625 626 /* 627 * on diagnosing side, receive a uuresolved from the proxy. 628 */ 629 void 630 fmd_xprt_event_uuresolved(fmd_xprt_impl_t *xip, nvlist_t *nvl) 631 { 632 fmd_case_t *cp; 633 char *uuid; 634 635 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 636 return; /* transitioned to error state */ 637 638 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_UUID, &uuid) == 0 && 639 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 640 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 641 642 fmd_case_transition(cp, (cip->ci_state == FMD_CASE_REPAIRED) ? 643 FMD_CASE_RESOLVED : (cip->ci_state == FMD_CASE_CLOSED) ? 644 FMD_CASE_REPAIRED : FMD_CASE_CLOSE_WAIT, FMD_CF_RESOLVED); 645 fmd_case_rele(cp); 646 } 647 } 648 649 /* 650 * on diagnosing side, receive a repair/acquit from the proxy. 651 */ 652 void 653 fmd_xprt_event_updated(fmd_xprt_impl_t *xip, nvlist_t *nvl) 654 { 655 fmd_case_t *cp; 656 char *uuid; 657 658 if (fmd_xprt_vmismatch(xip, nvl, NULL)) 659 return; /* transitioned to error state */ 660 661 if (nvlist_lookup_string(nvl, FM_RSRC_XPRT_UUID, &uuid) == 0 && 662 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 663 uint8_t *statusp, *proxy_asrup = NULL; 664 uint_t nelem = 0; 665 666 /* 667 * Only update status with new repairs if "no remote repair" 668 * is not set. Do the case_update anyway though (as this will 669 * refresh the status on the proxy side). 670 */ 671 if (!(xip->xi_flags & FMD_XPRT_NO_REMOTE_REPAIR)) { 672 if (nvlist_lookup_uint8_array(nvl, 673 FM_RSRC_XPRT_FAULT_STATUS, &statusp, &nelem) == 0 && 674 nelem != 0) { 675 (void) nvlist_lookup_uint8_array(nvl, 676 FM_RSRC_XPRT_FAULT_HAS_ASRU, &proxy_asrup, 677 &nelem); 678 fmd_case_update_status(cp, statusp, 679 proxy_asrup, NULL); 680 } 681 fmd_case_update_containees(cp); 682 } 683 fmd_case_update(cp); 684 fmd_case_rele(cp); 685 } 686 } 687 688 void 689 fmd_xprt_event_error(fmd_xprt_impl_t *xip, nvlist_t *nvl) 690 { 691 char *class = "<unknown>"; 692 693 (void) pthread_mutex_lock(&xip->xi_stats_lock); 694 xip->xi_stats->xs_discarded.fmds_value.ui64++; 695 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 696 697 (void) nvlist_lookup_string(nvl, FM_CLASS, &class); 698 TRACE((FMD_DBG_XPRT, "xprt %u bad event %s\n", xip->xi_id, class)); 699 700 fmd_xprt_transition(xip, _fmd_xprt_state_err, "ERR"); 701 } 702 703 void 704 fmd_xprt_event_drop(fmd_xprt_impl_t *xip, nvlist_t *nvl) 705 { 706 char *class = "<unknown>"; 707 708 (void) pthread_mutex_lock(&xip->xi_stats_lock); 709 xip->xi_stats->xs_discarded.fmds_value.ui64++; 710 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 711 712 (void) nvlist_lookup_string(nvl, FM_CLASS, &class); 713 TRACE((FMD_DBG_XPRT, "xprt %u drop event %s\n", xip->xi_id, class)); 714 715 } 716 717 fmd_xprt_t * 718 fmd_xprt_create(fmd_module_t *mp, uint_t flags, nvlist_t *auth, void *data) 719 { 720 fmd_xprt_impl_t *xip = fmd_zalloc(sizeof (fmd_xprt_impl_t), FMD_SLEEP); 721 fmd_stat_t *statv; 722 uint_t i, statc; 723 724 char buf[PATH_MAX]; 725 fmd_event_t *e; 726 nvlist_t *nvl; 727 char *s; 728 729 (void) pthread_mutex_init(&xip->xi_lock, NULL); 730 (void) pthread_cond_init(&xip->xi_cv, NULL); 731 (void) pthread_mutex_init(&xip->xi_stats_lock, NULL); 732 733 xip->xi_auth = auth; 734 xip->xi_data = data; 735 xip->xi_version = FM_RSRC_XPRT_VERSION; 736 xip->xi_flags = flags; 737 738 /* 739 * Grab fmd.d_xprt_lock to block fmd_xprt_suspend_all() and then create 740 * a transport ID and make it visible in fmd.d_xprt_ids. If transports 741 * were previously suspended, set the FMD_XPRT_DSUSPENDED flag on us to 742 * ensure that this transport will not run until fmd_xprt_resume_all(). 743 */ 744 (void) pthread_mutex_lock(&fmd.d_xprt_lock); 745 xip->xi_id = fmd_idspace_alloc(fmd.d_xprt_ids, xip); 746 747 if (fmd.d_xprt_suspend != 0) 748 xip->xi_flags |= FMD_XPRT_DSUSPENDED; 749 750 (void) pthread_mutex_unlock(&fmd.d_xprt_lock); 751 752 /* 753 * If the module has not yet finished _fmd_init(), set the ISUSPENDED 754 * bit so that fmdo_send() is not called until _fmd_init() completes. 755 */ 756 if (!(mp->mod_flags & FMD_MOD_INIT)) 757 xip->xi_flags |= FMD_XPRT_ISUSPENDED; 758 759 /* 760 * Initialize the transport statistics that we keep on behalf of fmd. 761 * These are set up using a template defined at the top of this file. 762 * We rename each statistic with a prefix ensuring its uniqueness. 763 */ 764 statc = sizeof (_fmd_xprt_stat_tmpl) / sizeof (fmd_stat_t); 765 statv = fmd_alloc(sizeof (_fmd_xprt_stat_tmpl), FMD_SLEEP); 766 bcopy(&_fmd_xprt_stat_tmpl, statv, sizeof (_fmd_xprt_stat_tmpl)); 767 768 for (i = 0; i < statc; i++) { 769 (void) snprintf(statv[i].fmds_name, 770 sizeof (statv[i].fmds_name), "fmd.xprt.%u.%s", xip->xi_id, 771 ((fmd_stat_t *)&_fmd_xprt_stat_tmpl + i)->fmds_name); 772 } 773 774 xip->xi_stats = (fmd_xprt_stat_t *)fmd_ustat_insert( 775 mp->mod_ustat, FMD_USTAT_NOALLOC, statc, statv, NULL); 776 777 if (xip->xi_stats == NULL) 778 fmd_panic("failed to create xi_stats (%p)\n", (void *)statv); 779 780 xip->xi_stats->xs_module.fmds_value.str = 781 fmd_strdup(mp->mod_name, FMD_SLEEP); 782 783 if (xip->xi_auth != NULL) 784 fmd_xprt_authupdate(xip); 785 786 /* 787 * Create the outbound eventq for this transport and link to its stats. 788 * If any suspend bits were set above, suspend the eventq immediately. 789 */ 790 xip->xi_queue = fmd_eventq_create(mp, &xip->xi_stats->xs_evqstat, 791 &xip->xi_stats_lock, mp->mod_stats->ms_xprtqlimit.fmds_value.ui32); 792 793 if (xip->xi_flags & FMD_XPRT_SMASK) 794 fmd_eventq_suspend(xip->xi_queue); 795 796 /* 797 * Create our subscription hashes: local subscriptions go to xi_queue, 798 * remote subscriptions are tracked only for protocol requests, and 799 * pending unsubscriptions are associated with the /dev/null eventq. 800 */ 801 fmd_xprt_class_hash_create(&xip->xi_lsub, xip->xi_queue); 802 fmd_xprt_class_hash_create(&xip->xi_rsub, NULL); 803 fmd_xprt_class_hash_create(&xip->xi_usub, fmd.d_rmod->mod_queue); 804 805 /* 806 * Determine our initial state based upon the creation flags. If we're 807 * read-only, go directly to RUN. If we're accepting a new connection, 808 * wait for a SYN. Otherwise send a SYN and wait for an ACK. 809 */ 810 if ((flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 811 fmd_xprt_transition(xip, _fmd_xprt_state_run, "RUN"); 812 else if (flags & FMD_XPRT_ACCEPT) 813 fmd_xprt_transition(xip, _fmd_xprt_state_syn, "SYN"); 814 else 815 fmd_xprt_transition(xip, _fmd_xprt_state_ack, "ACK"); 816 817 /* 818 * If client.xprtlog is set to TRUE, create a debugging log for the 819 * events received by the transport in var/fm/fmd/xprt/. 820 */ 821 (void) fmd_conf_getprop(fmd.d_conf, "client.xprtlog", &i); 822 (void) fmd_conf_getprop(fmd.d_conf, "log.xprt", &s); 823 824 if (i) { 825 (void) snprintf(buf, sizeof (buf), "%s/%u.log", s, xip->xi_id); 826 xip->xi_log = fmd_log_open(fmd.d_rootdir, buf, FMD_LOG_XPRT); 827 } 828 829 ASSERT(fmd_module_locked(mp)); 830 fmd_list_append(&mp->mod_transports, xip); 831 832 (void) pthread_mutex_lock(&mp->mod_stats_lock); 833 mp->mod_stats->ms_xprtopen.fmds_value.ui32++; 834 (void) pthread_mutex_unlock(&mp->mod_stats_lock); 835 836 /* 837 * If this is a read-only transport, return without creating a send 838 * queue thread and setting up any connection events in our queue. 839 */ 840 if ((flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 841 goto out; 842 843 /* 844 * Once the transport is fully initialized, create a send queue thread 845 * and start any connect events flowing to complete our initialization. 846 */ 847 if ((xip->xi_thread = fmd_thread_create(mp, 848 (fmd_thread_f *)fmd_xprt_send, xip)) == NULL) { 849 850 fmd_error(EFMD_XPRT_THR, 851 "failed to create thread for transport %u", xip->xi_id); 852 853 fmd_xprt_destroy((fmd_xprt_t *)xip); 854 (void) fmd_set_errno(EFMD_XPRT_THR); 855 return (NULL); 856 } 857 858 /* 859 * If the transport is not being opened to accept an inbound connect, 860 * start an outbound connection by enqueuing a SYN event for our peer. 861 */ 862 if (!(flags & FMD_XPRT_ACCEPT)) { 863 nvl = fmd_protocol_xprt_ctl(mp, 864 "resource.fm.xprt.syn", FM_RSRC_XPRT_VERSION); 865 866 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 867 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 868 fmd_eventq_insert_at_time(xip->xi_queue, e); 869 } 870 out: 871 fmd_dprintf(FMD_DBG_XPRT, "opened transport %u\n", xip->xi_id); 872 return ((fmd_xprt_t *)xip); 873 } 874 875 void 876 fmd_xprt_destroy(fmd_xprt_t *xp) 877 { 878 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 879 fmd_module_t *mp = xip->xi_queue->eq_mod; 880 uint_t id = xip->xi_id; 881 882 fmd_case_impl_t *cip, *nip; 883 fmd_stat_t *sp; 884 uint_t i, n; 885 886 ASSERT(fmd_module_locked(mp)); 887 fmd_list_delete(&mp->mod_transports, xip); 888 889 (void) pthread_mutex_lock(&mp->mod_stats_lock); 890 mp->mod_stats->ms_xprtopen.fmds_value.ui32--; 891 (void) pthread_mutex_unlock(&mp->mod_stats_lock); 892 893 (void) pthread_mutex_lock(&xip->xi_lock); 894 895 while (xip->xi_busy != 0) 896 (void) pthread_cond_wait(&xip->xi_cv, &xip->xi_lock); 897 898 /* 899 * Remove the transport from global visibility, cancel its send-side 900 * thread, join with it, and then remove the transport from module 901 * visibility. Once all this is done, destroy and free the transport. 902 */ 903 (void) fmd_idspace_free(fmd.d_xprt_ids, xip->xi_id); 904 905 if (xip->xi_thread != NULL) { 906 fmd_eventq_abort(xip->xi_queue); 907 fmd_module_unlock(mp); 908 fmd_thread_destroy(xip->xi_thread, FMD_THREAD_JOIN); 909 fmd_module_lock(mp); 910 } 911 912 if (xip->xi_log != NULL) 913 fmd_log_rele(xip->xi_log); 914 915 /* 916 * Release every case handle in the module that was cached by this 917 * transport. This will result in these cases disappearing from the 918 * local case hash so that fmd_case_uuclose() and fmd_case_repaired() 919 * etc can no longer be used. 920 */ 921 for (cip = fmd_list_next(&mp->mod_cases); cip != NULL; cip = nip) { 922 nip = fmd_list_next(cip); 923 if (cip->ci_xprt == xp) 924 fmd_case_discard((fmd_case_t *)cip, B_TRUE); 925 } 926 927 /* 928 * Destroy every class in the various subscription hashes and remove 929 * any corresponding subscriptions from the event dispatch queue. 930 */ 931 fmd_xprt_class_hash_destroy(&xip->xi_lsub); 932 fmd_xprt_class_hash_destroy(&xip->xi_rsub); 933 fmd_xprt_class_hash_destroy(&xip->xi_usub); 934 935 /* 936 * Uniquify the stat names exactly as was done in fmd_xprt_create() 937 * before calling fmd_ustat_insert(), otherwise fmd_ustat_delete() 938 * won't find the entries in the hash table. 939 */ 940 n = sizeof (_fmd_xprt_stat_tmpl) / sizeof (fmd_stat_t); 941 sp = fmd_alloc(sizeof (_fmd_xprt_stat_tmpl), FMD_SLEEP); 942 bcopy(&_fmd_xprt_stat_tmpl, sp, sizeof (_fmd_xprt_stat_tmpl)); 943 for (i = 0; i < n; i++) { 944 (void) snprintf(sp[i].fmds_name, 945 sizeof (sp[i].fmds_name), "fmd.xprt.%u.%s", xip->xi_id, 946 ((fmd_stat_t *)&_fmd_xprt_stat_tmpl + i)->fmds_name); 947 } 948 fmd_ustat_delete(mp->mod_ustat, n, sp); 949 fmd_free(sp, sizeof (_fmd_xprt_stat_tmpl)); 950 951 fmd_free(xip->xi_stats, sizeof (fmd_xprt_stat_t)); 952 fmd_eventq_destroy(xip->xi_queue); 953 nvlist_free(xip->xi_auth); 954 fmd_free(xip, sizeof (fmd_xprt_impl_t)); 955 956 fmd_dprintf(FMD_DBG_XPRT, "closed transport %u\n", id); 957 } 958 959 void 960 fmd_xprt_xsuspend(fmd_xprt_t *xp, uint_t flags) 961 { 962 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 963 uint_t oflags; 964 965 ASSERT((flags & ~FMD_XPRT_SMASK) == 0); 966 (void) pthread_mutex_lock(&xip->xi_lock); 967 968 oflags = xip->xi_flags; 969 xip->xi_flags |= flags; 970 971 if (!(oflags & FMD_XPRT_SMASK) && (xip->xi_flags & FMD_XPRT_SMASK) != 0) 972 fmd_eventq_suspend(xip->xi_queue); 973 974 (void) pthread_cond_broadcast(&xip->xi_cv); 975 976 while (xip->xi_busy != 0) 977 (void) pthread_cond_wait(&xip->xi_cv, &xip->xi_lock); 978 979 (void) pthread_mutex_unlock(&xip->xi_lock); 980 } 981 982 void 983 fmd_xprt_xresume(fmd_xprt_t *xp, uint_t flags) 984 { 985 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 986 uint_t oflags; 987 988 ASSERT((flags & ~FMD_XPRT_SMASK) == 0); 989 (void) pthread_mutex_lock(&xip->xi_lock); 990 991 oflags = xip->xi_flags; 992 xip->xi_flags &= ~flags; 993 994 if ((oflags & FMD_XPRT_SMASK) != 0 && !(xip->xi_flags & FMD_XPRT_SMASK)) 995 fmd_eventq_resume(xip->xi_queue); 996 997 (void) pthread_cond_broadcast(&xip->xi_cv); 998 (void) pthread_mutex_unlock(&xip->xi_lock); 999 } 1000 1001 void 1002 fmd_xprt_send(fmd_xprt_t *xp) 1003 { 1004 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1005 fmd_module_t *mp = xip->xi_queue->eq_mod; 1006 fmd_event_t *ep; 1007 int err; 1008 1009 while ((ep = fmd_eventq_delete(xip->xi_queue)) != NULL) { 1010 if (FMD_EVENT_TTL(ep) == 0) { 1011 fmd_event_rele(ep); 1012 continue; 1013 } 1014 1015 fmd_dprintf(FMD_DBG_XPRT, "xprt %u sending %s\n", 1016 xip->xi_id, (char *)FMD_EVENT_DATA(ep)); 1017 1018 err = mp->mod_ops->mop_transport(mp, xp, ep); 1019 fmd_eventq_done(xip->xi_queue); 1020 1021 if (err == FMD_SEND_RETRY) { 1022 fmd_eventq_insert_at_time(xip->xi_queue, ep); 1023 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1024 xip->xi_stats->xs_retried.fmds_value.ui64++; 1025 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1026 } 1027 1028 if (err != FMD_SEND_SUCCESS && err != FMD_SEND_RETRY) { 1029 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1030 xip->xi_stats->xs_lost.fmds_value.ui64++; 1031 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1032 } 1033 1034 fmd_event_rele(ep); 1035 } 1036 } 1037 1038 /* 1039 * This function creates a local suspect list. This is used when a suspect list 1040 * is created directly by an external source like fminject. 1041 */ 1042 static void 1043 fmd_xprt_list_suspect_local(fmd_xprt_t *xp, nvlist_t *nvl) 1044 { 1045 nvlist_t **nvlp; 1046 nvlist_t *de_fmri, *de_fmri_dup = NULL; 1047 int64_t *diag_time; 1048 char *code = NULL; 1049 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1050 fmd_case_t *cp; 1051 uint_t nelem = 0, nelem2 = 0, i; 1052 1053 fmd_module_lock(xip->xi_queue->eq_mod); 1054 cp = fmd_case_create(xip->xi_queue->eq_mod, NULL); 1055 if (cp == NULL) { 1056 fmd_module_unlock(xip->xi_queue->eq_mod); 1057 return; 1058 } 1059 1060 /* 1061 * copy diag_code if present 1062 */ 1063 (void) nvlist_lookup_string(nvl, FM_SUSPECT_DIAG_CODE, &code); 1064 if (code != NULL) { 1065 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 1066 1067 cip->ci_precanned = 1; 1068 fmd_case_setcode(cp, code); 1069 } 1070 1071 /* 1072 * copy suspects 1073 */ 1074 (void) nvlist_lookup_nvlist_array(nvl, FM_SUSPECT_FAULT_LIST, &nvlp, 1075 &nelem); 1076 for (i = 0; i < nelem; i++) { 1077 nvlist_t *flt_copy, *asru = NULL, *fru = NULL, *rsrc = NULL; 1078 topo_hdl_t *thp; 1079 char *loc = NULL; 1080 int err; 1081 1082 thp = fmd_fmri_topo_hold(TOPO_VERSION); 1083 (void) nvlist_xdup(nvlp[i], &flt_copy, &fmd.d_nva); 1084 (void) nvlist_lookup_nvlist(nvlp[i], FM_FAULT_RESOURCE, &rsrc); 1085 1086 /* 1087 * If no fru specified, get it from topo 1088 */ 1089 if (nvlist_lookup_nvlist(nvlp[i], FM_FAULT_FRU, &fru) != 0 && 1090 rsrc && topo_fmri_fru(thp, rsrc, &fru, &err) == 0) 1091 (void) nvlist_add_nvlist(flt_copy, FM_FAULT_FRU, fru); 1092 /* 1093 * If no asru specified, get it from topo 1094 */ 1095 if (nvlist_lookup_nvlist(nvlp[i], FM_FAULT_ASRU, &asru) != 0 && 1096 rsrc && topo_fmri_asru(thp, rsrc, &asru, &err) == 0) 1097 (void) nvlist_add_nvlist(flt_copy, FM_FAULT_ASRU, asru); 1098 /* 1099 * If no location specified, get it from topo 1100 */ 1101 if (nvlist_lookup_string(nvlp[i], FM_FAULT_LOCATION, 1102 &loc) != 0) { 1103 if (fru && topo_fmri_label(thp, fru, &loc, &err) == 0) 1104 (void) nvlist_add_string(flt_copy, 1105 FM_FAULT_LOCATION, loc); 1106 else if (rsrc && topo_fmri_label(thp, rsrc, &loc, 1107 &err) == 0) 1108 (void) nvlist_add_string(flt_copy, 1109 FM_FAULT_LOCATION, loc); 1110 if (loc) 1111 topo_hdl_strfree(thp, loc); 1112 } 1113 if (fru) 1114 nvlist_free(fru); 1115 if (asru) 1116 nvlist_free(asru); 1117 if (rsrc) 1118 nvlist_free(rsrc); 1119 fmd_fmri_topo_rele(thp); 1120 fmd_case_insert_suspect(cp, flt_copy); 1121 } 1122 1123 /* 1124 * copy diag_time if present 1125 */ 1126 if (nvlist_lookup_int64_array(nvl, FM_SUSPECT_DIAG_TIME, &diag_time, 1127 &nelem2) == 0 && nelem2 >= 2) 1128 fmd_case_settime(cp, diag_time[0], diag_time[1]); 1129 1130 /* 1131 * copy DE fmri if present 1132 */ 1133 if (nvlist_lookup_nvlist(nvl, FM_SUSPECT_DE, &de_fmri) == 0) { 1134 (void) nvlist_xdup(de_fmri, &de_fmri_dup, &fmd.d_nva); 1135 fmd_case_set_de_fmri(cp, de_fmri_dup); 1136 } 1137 1138 fmd_case_transition(cp, FMD_CASE_SOLVED, FMD_CF_SOLVED); 1139 fmd_module_unlock(xip->xi_queue->eq_mod); 1140 } 1141 1142 /* 1143 * This function is called to create a proxy case on receipt of a list.suspect 1144 * from the diagnosing side of the transport. 1145 */ 1146 static void 1147 fmd_xprt_list_suspect(fmd_xprt_t *xp, nvlist_t *nvl) 1148 { 1149 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1150 nvlist_t **nvlp; 1151 uint_t nelem = 0, nelem2 = 0, i; 1152 int64_t *diag_time; 1153 topo_hdl_t *thp; 1154 char *class; 1155 nvlist_t *rsrc, *asru, *de_fmri, *de_fmri_dup = NULL; 1156 nvlist_t *flt_copy; 1157 int err; 1158 nvlist_t **asrua; 1159 uint8_t *proxy_asru = NULL; 1160 int got_proxy_asru = 0; 1161 int got_hc_rsrc = 0; 1162 int got_present_rsrc = 0; 1163 uint8_t *diag_asru = NULL; 1164 char *scheme; 1165 uint8_t *statusp; 1166 char *uuid, *code; 1167 fmd_case_t *cp; 1168 fmd_case_impl_t *cip; 1169 int need_update = 0; 1170 1171 if (nvlist_lookup_string(nvl, FM_SUSPECT_UUID, &uuid) != 0) 1172 return; 1173 if (nvlist_lookup_string(nvl, FM_SUSPECT_DIAG_CODE, &code) != 0) 1174 return; 1175 (void) nvlist_lookup_nvlist_array(nvl, FM_SUSPECT_FAULT_LIST, &nvlp, 1176 &nelem); 1177 1178 /* 1179 * In order to implement FMD_XPRT_HCONLY and FMD_XPRT_HC_PRESENT_ONLY 1180 * etc we first scan the suspects to see if 1181 * - there was an asru in the received fault 1182 * - there was an hc-scheme resource in the received fault 1183 * - any hc-scheme resource in the received fault is present in the 1184 * local topology 1185 * - any hc-scheme resource in the received fault has an asru in the 1186 * local topology 1187 */ 1188 if (nelem > 0) { 1189 asrua = fmd_zalloc(sizeof (nvlist_t *) * nelem, FMD_SLEEP); 1190 proxy_asru = fmd_zalloc(sizeof (uint8_t) * nelem, FMD_SLEEP); 1191 diag_asru = fmd_zalloc(sizeof (uint8_t) * nelem, FMD_SLEEP); 1192 thp = fmd_fmri_topo_hold(TOPO_VERSION); 1193 for (i = 0; i < nelem; i++) { 1194 if (nvlist_lookup_nvlist(nvlp[i], FM_FAULT_ASRU, 1195 &asru) == 0 && asru != NULL) 1196 diag_asru[i] = 1; 1197 if (nvlist_lookup_string(nvlp[i], FM_CLASS, 1198 &class) != 0 || strncmp(class, "fault", 5) != 0) 1199 continue; 1200 /* 1201 * If there is an hc-scheme asru, use that to find the 1202 * real asru. Otherwise if there is an hc-scheme 1203 * resource, work out the old asru from that. 1204 * This order is to allow a two stage evaluation 1205 * of the asru where a fault in the diagnosing side 1206 * is in a component not visible to the proxy side, 1207 * but prevents a component that is visible from 1208 * working. So the diagnosing side sets the asru to 1209 * the latter component (in hc-scheme as the diagnosing 1210 * side doesn't know about the proxy side's virtual 1211 * schemes), and then the proxy side can convert that 1212 * to a suitable virtual scheme asru. 1213 */ 1214 if (nvlist_lookup_nvlist(nvlp[i], FM_FAULT_ASRU, 1215 &asru) == 0 && asru != NULL && 1216 nvlist_lookup_string(asru, FM_FMRI_SCHEME, 1217 &scheme) == 0 && 1218 strcmp(scheme, FM_FMRI_SCHEME_HC) == 0) { 1219 got_hc_rsrc = 1; 1220 if (xip->xi_flags & FMD_XPRT_EXTERNAL) 1221 continue; 1222 if (topo_fmri_present(thp, asru, &err) == 0) 1223 got_present_rsrc = 1; 1224 if (topo_fmri_asru(thp, asru, &asrua[i], 1225 &err) == 0) { 1226 proxy_asru[i] = 1227 FMD_PROXY_ASRU_FROM_ASRU; 1228 got_proxy_asru = 1; 1229 } 1230 } else if (nvlist_lookup_nvlist(nvlp[i], 1231 FM_FAULT_RESOURCE, &rsrc) == 0 && rsrc != NULL && 1232 nvlist_lookup_string(rsrc, FM_FMRI_SCHEME, 1233 &scheme) == 0 && 1234 strcmp(scheme, FM_FMRI_SCHEME_HC) == 0) { 1235 got_hc_rsrc = 1; 1236 if (xip->xi_flags & FMD_XPRT_EXTERNAL) 1237 continue; 1238 if (topo_fmri_present(thp, rsrc, &err) == 0) 1239 got_present_rsrc = 1; 1240 if (topo_fmri_asru(thp, rsrc, &asrua[i], 1241 &err) == 0) { 1242 proxy_asru[i] = 1243 FMD_PROXY_ASRU_FROM_RSRC; 1244 got_proxy_asru = 1; 1245 } 1246 } 1247 } 1248 fmd_fmri_topo_rele(thp); 1249 } 1250 1251 /* 1252 * If we're set up only to report hc-scheme faults, and 1253 * there aren't any, then just drop the event. 1254 */ 1255 if (got_hc_rsrc == 0 && (xip->xi_flags & FMD_XPRT_HCONLY)) { 1256 if (nelem > 0) { 1257 fmd_free(proxy_asru, sizeof (uint8_t) * nelem); 1258 fmd_free(diag_asru, sizeof (uint8_t) * nelem); 1259 fmd_free(asrua, sizeof (nvlist_t *) * nelem); 1260 } 1261 return; 1262 } 1263 1264 /* 1265 * If we're set up only to report locally present hc-scheme 1266 * faults, and there aren't any, then just drop the event. 1267 */ 1268 if (got_present_rsrc == 0 && 1269 (xip->xi_flags & FMD_XPRT_HC_PRESENT_ONLY)) { 1270 if (nelem > 0) { 1271 for (i = 0; i < nelem; i++) 1272 if (asrua[i]) 1273 nvlist_free(asrua[i]); 1274 fmd_free(proxy_asru, sizeof (uint8_t) * nelem); 1275 fmd_free(diag_asru, sizeof (uint8_t) * nelem); 1276 fmd_free(asrua, sizeof (nvlist_t *) * nelem); 1277 } 1278 return; 1279 } 1280 1281 /* 1282 * If fmd_case_recreate() returns NULL, UUID is already known. 1283 */ 1284 fmd_module_lock(xip->xi_queue->eq_mod); 1285 if ((cp = fmd_case_recreate(xip->xi_queue->eq_mod, xp, 1286 FMD_CASE_UNSOLVED, uuid, code)) == NULL) { 1287 if (nelem > 0) { 1288 for (i = 0; i < nelem; i++) 1289 if (asrua[i]) 1290 nvlist_free(asrua[i]); 1291 fmd_free(proxy_asru, sizeof (uint8_t) * nelem); 1292 fmd_free(diag_asru, sizeof (uint8_t) * nelem); 1293 fmd_free(asrua, sizeof (nvlist_t *) * nelem); 1294 } 1295 fmd_module_unlock(xip->xi_queue->eq_mod); 1296 return; 1297 } 1298 1299 cip = (fmd_case_impl_t *)cp; 1300 cip->ci_diag_asru = diag_asru; 1301 cip->ci_proxy_asru = proxy_asru; 1302 for (i = 0; i < nelem; i++) { 1303 (void) nvlist_xdup(nvlp[i], &flt_copy, &fmd.d_nva); 1304 if (proxy_asru[i] != FMD_PROXY_ASRU_NOT_NEEDED) { 1305 /* 1306 * Copy suspects, but remove/replace asru first. Also if 1307 * the original asru was hc-scheme use that as resource. 1308 */ 1309 if (proxy_asru[i] == FMD_PROXY_ASRU_FROM_ASRU) { 1310 (void) nvlist_remove(flt_copy, 1311 FM_FAULT_RESOURCE, DATA_TYPE_NVLIST); 1312 (void) nvlist_lookup_nvlist(flt_copy, 1313 FM_FAULT_ASRU, &asru); 1314 (void) nvlist_add_nvlist(flt_copy, 1315 FM_FAULT_RESOURCE, asru); 1316 } 1317 (void) nvlist_remove(flt_copy, FM_FAULT_ASRU, 1318 DATA_TYPE_NVLIST); 1319 (void) nvlist_add_nvlist(flt_copy, FM_FAULT_ASRU, 1320 asrua[i]); 1321 nvlist_free(asrua[i]); 1322 } else if (nvlist_lookup_nvlist(flt_copy, FM_FAULT_ASRU, 1323 &asru) == 0 && asru != NULL) { 1324 /* 1325 * keep asru from diag side, but but mark as no retire 1326 */ 1327 (void) nvlist_add_boolean_value(flt_copy, 1328 FM_SUSPECT_RETIRE, B_FALSE); 1329 } 1330 fmd_case_insert_suspect(cp, flt_copy); 1331 } 1332 /* 1333 * copy diag_time 1334 */ 1335 if (nvlist_lookup_int64_array(nvl, FM_SUSPECT_DIAG_TIME, &diag_time, 1336 &nelem2) == 0 && nelem2 >= 2) 1337 fmd_case_settime(cp, diag_time[0], diag_time[1]); 1338 /* 1339 * copy DE fmri 1340 */ 1341 if (nvlist_lookup_nvlist(nvl, FM_SUSPECT_DE, &de_fmri) == 0) { 1342 (void) nvlist_xdup(de_fmri, &de_fmri_dup, &fmd.d_nva); 1343 fmd_case_set_de_fmri(cp, de_fmri_dup); 1344 } 1345 1346 /* 1347 * Transition to solved. This will log the suspect list and create 1348 * the resource cache entries. 1349 */ 1350 fmd_case_transition(cp, FMD_CASE_SOLVED, FMD_CF_SOLVED); 1351 1352 /* 1353 * Update status if it is not simply "all faulty" (can happen if 1354 * list.suspects are being re-sent when the transport has reconnected). 1355 */ 1356 (void) nvlist_lookup_uint8_array(nvl, FM_SUSPECT_FAULT_STATUS, &statusp, 1357 &nelem); 1358 for (i = 0; i < nelem; i++) { 1359 if ((statusp[i] & (FM_SUSPECT_FAULTY | FM_SUSPECT_UNUSABLE | 1360 FM_SUSPECT_NOT_PRESENT | FM_SUSPECT_DEGRADED)) != 1361 FM_SUSPECT_FAULTY) 1362 need_update = 1; 1363 } 1364 if (need_update) { 1365 fmd_case_update_status(cp, statusp, cip->ci_proxy_asru, 1366 cip->ci_diag_asru); 1367 fmd_case_update_containees(cp); 1368 fmd_case_update(cp); 1369 } 1370 1371 /* 1372 * if asru on proxy side, send an update back to the diagnosing side to 1373 * update UNUSABLE/DEGRADED. 1374 */ 1375 if (got_proxy_asru) 1376 fmd_case_xprt_updated(cp); 1377 1378 if (nelem > 0) 1379 fmd_free(asrua, sizeof (nvlist_t *) * nelem); 1380 fmd_module_unlock(xip->xi_queue->eq_mod); 1381 } 1382 1383 void 1384 fmd_xprt_recv(fmd_xprt_t *xp, nvlist_t *nvl, hrtime_t hrt, boolean_t logonly) 1385 { 1386 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1387 const fmd_xprt_rule_t *xrp; 1388 fmd_t *dp = &fmd; 1389 1390 fmd_event_t *e; 1391 char *class, *uuid; 1392 boolean_t isproto, isereport; 1393 1394 uint64_t *tod; 1395 uint8_t ttl; 1396 uint_t n; 1397 fmd_case_t *cp; 1398 1399 /* 1400 * Grab the transport lock and set the busy flag to indicate we are 1401 * busy receiving an event. If [DI]SUSPEND is pending, wait until fmd 1402 * resumes the transport before continuing on with the receive. 1403 */ 1404 (void) pthread_mutex_lock(&xip->xi_lock); 1405 1406 while (xip->xi_flags & (FMD_XPRT_DSUSPENDED | FMD_XPRT_ISUSPENDED)) { 1407 1408 if (fmd.d_signal != 0) { 1409 (void) pthread_mutex_unlock(&xip->xi_lock); 1410 return; /* fmd_destroy() is in progress */ 1411 } 1412 1413 (void) pthread_cond_wait(&xip->xi_cv, &xip->xi_lock); 1414 } 1415 1416 xip->xi_busy++; 1417 ASSERT(xip->xi_busy != 0); 1418 1419 (void) pthread_mutex_unlock(&xip->xi_lock); 1420 1421 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1422 xip->xi_stats->xs_received.fmds_value.ui64++; 1423 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1424 1425 if (nvlist_lookup_string(nvl, FM_CLASS, &class) != 0) { 1426 fmd_error(EFMD_XPRT_PAYLOAD, "discarding nvlist %p: missing " 1427 "required \"%s\" payload element", (void *)nvl, FM_CLASS); 1428 1429 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1430 xip->xi_stats->xs_discarded.fmds_value.ui64++; 1431 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1432 1433 nvlist_free(nvl); 1434 goto done; 1435 } 1436 1437 fmd_dprintf(FMD_DBG_XPRT, "xprt %u %s %s\n", xip->xi_id, 1438 ((logonly == FMD_B_TRUE) ? "logging" : "posting"), class); 1439 1440 isereport = (strncmp(class, FM_EREPORT_CLASS, 1441 sizeof (FM_EREPORT_CLASS - 1)) == 0) ? FMD_B_TRUE : FMD_B_FALSE; 1442 1443 /* 1444 * The logonly flag should only be set for ereports. 1445 */ 1446 if ((logonly == FMD_B_TRUE) && (isereport == FMD_B_FALSE)) { 1447 fmd_error(EFMD_XPRT_INVAL, "discarding nvlist %p: " 1448 "logonly flag is not valid for class %s", 1449 (void *)nvl, class); 1450 1451 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1452 xip->xi_stats->xs_discarded.fmds_value.ui64++; 1453 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1454 1455 nvlist_free(nvl); 1456 goto done; 1457 } 1458 1459 /* 1460 * If a time-to-live value is present in the event and is zero, drop 1461 * the event and bump xs_timeouts. Otherwise decrement the TTL value. 1462 */ 1463 if (nvlist_lookup_uint8(nvl, FMD_EVN_TTL, &ttl) == 0) { 1464 if (ttl == 0) { 1465 fmd_dprintf(FMD_DBG_XPRT, "xprt %u nvlist %p (%s) " 1466 "timeout: event received with ttl=0\n", 1467 xip->xi_id, (void *)nvl, class); 1468 1469 (void) pthread_mutex_lock(&xip->xi_stats_lock); 1470 xip->xi_stats->xs_timeouts.fmds_value.ui64++; 1471 (void) pthread_mutex_unlock(&xip->xi_stats_lock); 1472 1473 nvlist_free(nvl); 1474 goto done; 1475 } 1476 (void) nvlist_remove(nvl, FMD_EVN_TTL, DATA_TYPE_UINT8); 1477 (void) nvlist_add_uint8(nvl, FMD_EVN_TTL, ttl - 1); 1478 } 1479 1480 /* 1481 * If we are using the native system clock, the underlying transport 1482 * code can provide a tighter event time bound by telling us when the 1483 * event was enqueued. If we're using simulated clocks, this time 1484 * has no meaning to us, so just reset the value to use HRT_NOW. 1485 */ 1486 if (dp->d_clockops != &fmd_timeops_native) 1487 hrt = FMD_HRT_NOW; 1488 1489 /* 1490 * If an event's class is in the FMD_CTL_CLASS family, then create a 1491 * control event. If a FMD_EVN_TOD member is found, create a protocol 1492 * event using this time. Otherwise create a protocol event using hrt. 1493 */ 1494 isproto = (strncmp(class, FMD_CTL_CLASS, FMD_CTL_CLASS_LEN) == 0) ? 1495 FMD_B_FALSE : FMD_B_TRUE; 1496 if (isproto == FMD_B_FALSE) 1497 e = fmd_event_create(FMD_EVT_CTL, hrt, nvl, fmd_ctl_init(nvl)); 1498 else if (nvlist_lookup_uint64_array(nvl, FMD_EVN_TOD, &tod, &n) != 0) 1499 e = fmd_event_create(FMD_EVT_PROTOCOL, hrt, nvl, class); 1500 else { 1501 e = fmd_event_recreate(FMD_EVT_PROTOCOL, 1502 NULL, nvl, class, NULL, 0, 0); 1503 } 1504 1505 /* 1506 * If the debug log is enabled, create a temporary event, log it to the 1507 * debug log, and then reset the underlying state of the event. 1508 */ 1509 if (xip->xi_log != NULL) { 1510 fmd_event_impl_t *ep = (fmd_event_impl_t *)e; 1511 1512 fmd_log_append(xip->xi_log, e, NULL); 1513 1514 ep->ev_flags |= FMD_EVF_VOLATILE; 1515 ep->ev_off = 0; 1516 ep->ev_len = 0; 1517 1518 if (ep->ev_log != NULL) { 1519 fmd_log_rele(ep->ev_log); 1520 ep->ev_log = NULL; 1521 } 1522 } 1523 1524 /* 1525 * Iterate over the rules for the current state trying to match the 1526 * event class to one of our special rules. If a rule is matched, the 1527 * event is consumed and not dispatched to other modules. If the rule 1528 * set ends without matching an event, we fall through to dispatching. 1529 */ 1530 for (xrp = xip->xi_state; xrp->xr_class != NULL; xrp++) { 1531 if (fmd_event_match(e, FMD_EVT_PROTOCOL, xrp->xr_class)) { 1532 fmd_event_hold(e); 1533 xrp->xr_func(xip, nvl); 1534 fmd_event_rele(e); 1535 goto done; 1536 } 1537 } 1538 1539 /* 1540 * Record the event in the errlog if it is an ereport. This code will 1541 * be replaced later with a per-transport intent log instead. 1542 */ 1543 if (isereport == FMD_B_TRUE) { 1544 (void) pthread_rwlock_rdlock(&dp->d_log_lock); 1545 fmd_log_append(dp->d_errlog, e, NULL); 1546 (void) pthread_rwlock_unlock(&dp->d_log_lock); 1547 } 1548 1549 /* 1550 * If a list.suspect event is received, create a case for the specified 1551 * UUID in the case hash, with the transport module as its owner. 1552 */ 1553 if (fmd_event_match(e, FMD_EVT_PROTOCOL, FM_LIST_SUSPECT_CLASS)) { 1554 if (xip->xi_flags & FMD_XPRT_CACHE_AS_LOCAL) 1555 fmd_xprt_list_suspect_local(xp, nvl); 1556 else 1557 fmd_xprt_list_suspect(xp, nvl); 1558 fmd_event_hold(e); 1559 fmd_event_rele(e); 1560 goto done; 1561 } 1562 1563 /* 1564 * If a list.updated or list.repaired event is received, update the 1565 * resource cache status and the local case. 1566 */ 1567 if (fmd_event_match(e, FMD_EVT_PROTOCOL, FM_LIST_REPAIRED_CLASS) || 1568 fmd_event_match(e, FMD_EVT_PROTOCOL, FM_LIST_UPDATED_CLASS)) { 1569 uint8_t *statusp; 1570 uint_t nelem = 0; 1571 1572 (void) nvlist_lookup_uint8_array(nvl, FM_SUSPECT_FAULT_STATUS, 1573 &statusp, &nelem); 1574 fmd_module_lock(xip->xi_queue->eq_mod); 1575 if (nvlist_lookup_string(nvl, FM_SUSPECT_UUID, &uuid) == 0 && 1576 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 1577 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 1578 if (cip->ci_xprt != NULL) { 1579 fmd_case_update_status(cp, statusp, 1580 cip->ci_proxy_asru, cip->ci_diag_asru); 1581 fmd_case_update_containees(cp); 1582 fmd_case_update(cp); 1583 } 1584 fmd_case_rele(cp); 1585 } 1586 fmd_module_unlock(xip->xi_queue->eq_mod); 1587 fmd_event_hold(e); 1588 fmd_event_rele(e); 1589 goto done; 1590 } 1591 1592 /* 1593 * If a list.isolated event is received, update resource cache status 1594 */ 1595 if (fmd_event_match(e, FMD_EVT_PROTOCOL, FM_LIST_ISOLATED_CLASS)) { 1596 uint8_t *statusp; 1597 uint_t nelem = 0; 1598 1599 (void) nvlist_lookup_uint8_array(nvl, FM_SUSPECT_FAULT_STATUS, 1600 &statusp, &nelem); 1601 fmd_module_lock(xip->xi_queue->eq_mod); 1602 if (nvlist_lookup_string(nvl, FM_SUSPECT_UUID, &uuid) == 0 && 1603 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 1604 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 1605 if (cip->ci_xprt != NULL) 1606 fmd_case_update_status(cp, statusp, 1607 cip->ci_proxy_asru, cip->ci_diag_asru); 1608 fmd_case_rele(cp); 1609 } 1610 fmd_module_unlock(xip->xi_queue->eq_mod); 1611 fmd_event_hold(e); 1612 fmd_event_rele(e); 1613 goto done; 1614 } 1615 1616 /* 1617 * If a list.resolved event is received, resolve the local case. 1618 */ 1619 if (fmd_event_match(e, FMD_EVT_PROTOCOL, FM_LIST_RESOLVED_CLASS)) { 1620 fmd_module_lock(xip->xi_queue->eq_mod); 1621 if (nvlist_lookup_string(nvl, FM_SUSPECT_UUID, &uuid) == 0 && 1622 (cp = fmd_case_hash_lookup(fmd.d_cases, uuid)) != NULL) { 1623 fmd_case_impl_t *cip = (fmd_case_impl_t *)cp; 1624 if (cip->ci_xprt != NULL) 1625 fmd_case_transition(cp, (cip->ci_state == 1626 FMD_CASE_REPAIRED) ? FMD_CASE_RESOLVED : 1627 (cip->ci_state == FMD_CASE_CLOSED) ? 1628 FMD_CASE_REPAIRED : FMD_CASE_CLOSE_WAIT, 1629 FMD_CF_RESOLVED); 1630 fmd_case_rele(cp); 1631 } 1632 fmd_module_unlock(xip->xi_queue->eq_mod); 1633 fmd_event_hold(e); 1634 fmd_event_rele(e); 1635 goto done; 1636 } 1637 1638 if (logonly == FMD_B_TRUE || (xip->xi_flags & FMD_XPRT_EXTERNAL)) { 1639 /* 1640 * Don't proxy ereports on an EXTERNAL transport - we won't 1641 * know how to diagnose them with the wrong topology. Note 1642 * that here (and above) we have to hold/release the event in 1643 * order for it to be freed. 1644 */ 1645 fmd_event_hold(e); 1646 fmd_event_rele(e); 1647 } else if (isproto == FMD_B_TRUE) 1648 fmd_dispq_dispatch(dp->d_disp, e, class); 1649 else 1650 fmd_modhash_dispatch(dp->d_mod_hash, e); 1651 done: 1652 (void) pthread_mutex_lock(&xip->xi_lock); 1653 1654 ASSERT(xip->xi_busy != 0); 1655 xip->xi_busy--; 1656 1657 (void) pthread_cond_broadcast(&xip->xi_cv); 1658 (void) pthread_mutex_unlock(&xip->xi_lock); 1659 } 1660 1661 void 1662 fmd_xprt_uuclose(fmd_xprt_t *xp, const char *uuid) 1663 { 1664 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1665 1666 fmd_event_t *e; 1667 nvlist_t *nvl; 1668 char *s; 1669 1670 if ((xip->xi_flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 1671 return; /* read-only transports do not proxy uuclose */ 1672 1673 TRACE((FMD_DBG_XPRT, "xprt %u closing case %s\n", xip->xi_id, uuid)); 1674 1675 nvl = fmd_protocol_xprt_uuclose(xip->xi_queue->eq_mod, 1676 "resource.fm.xprt.uuclose", xip->xi_version, uuid); 1677 1678 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 1679 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 1680 fmd_eventq_insert_at_time(xip->xi_queue, e); 1681 } 1682 1683 /* 1684 * On proxy side, send back uuresolved request to diagnosing side 1685 */ 1686 void 1687 fmd_xprt_uuresolved(fmd_xprt_t *xp, const char *uuid) 1688 { 1689 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1690 1691 fmd_event_t *e; 1692 nvlist_t *nvl; 1693 char *s; 1694 1695 if ((xip->xi_flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 1696 return; /* read-only transports do not proxy uuresolved */ 1697 1698 TRACE((FMD_DBG_XPRT, "xprt %u resolving case %s\n", xip->xi_id, uuid)); 1699 1700 nvl = fmd_protocol_xprt_uuresolved(xip->xi_queue->eq_mod, 1701 "resource.fm.xprt.uuresolved", xip->xi_version, uuid); 1702 1703 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 1704 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 1705 fmd_eventq_insert_at_time(xip->xi_queue, e); 1706 } 1707 1708 /* 1709 * On proxy side, send back repair/acquit/etc request to diagnosing side 1710 */ 1711 void 1712 fmd_xprt_updated(fmd_xprt_t *xp, const char *uuid, uint8_t *statusp, 1713 uint8_t *has_asrup, uint_t nelem) 1714 { 1715 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1716 1717 fmd_event_t *e; 1718 nvlist_t *nvl; 1719 char *s; 1720 1721 if ((xip->xi_flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 1722 return; /* read-only transports do not support remote repairs */ 1723 1724 TRACE((FMD_DBG_XPRT, "xprt %u updating case %s\n", xip->xi_id, uuid)); 1725 1726 nvl = fmd_protocol_xprt_updated(xip->xi_queue->eq_mod, 1727 "resource.fm.xprt.updated", xip->xi_version, uuid, statusp, 1728 has_asrup, nelem); 1729 1730 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 1731 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 1732 fmd_eventq_insert_at_time(xip->xi_queue, e); 1733 } 1734 1735 /* 1736 * Insert the specified class into our remote subscription hash. If the class 1737 * is already present, bump the reference count; otherwise add it to the hash 1738 * and then enqueue an event for our remote peer to proxy our subscription. 1739 */ 1740 void 1741 fmd_xprt_subscribe(fmd_xprt_t *xp, const char *class) 1742 { 1743 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1744 1745 uint_t refs; 1746 nvlist_t *nvl; 1747 fmd_event_t *e; 1748 char *s; 1749 1750 if ((xip->xi_flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 1751 return; /* read-only transports do not proxy subscriptions */ 1752 1753 if (!(xip->xi_flags & FMD_XPRT_SUBSCRIBER)) 1754 return; /* transport is not yet an active subscriber */ 1755 1756 (void) pthread_mutex_lock(&xip->xi_lock); 1757 refs = fmd_xprt_class_hash_insert(xip, &xip->xi_rsub, class); 1758 (void) pthread_mutex_unlock(&xip->xi_lock); 1759 1760 if (refs > 1) 1761 return; /* we've already asked our peer for this subscription */ 1762 1763 fmd_dprintf(FMD_DBG_XPRT, 1764 "xprt %u subscribing to %s\n", xip->xi_id, class); 1765 1766 nvl = fmd_protocol_xprt_sub(xip->xi_queue->eq_mod, 1767 "resource.fm.xprt.subscribe", xip->xi_version, class); 1768 1769 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 1770 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 1771 fmd_eventq_insert_at_time(xip->xi_queue, e); 1772 } 1773 1774 /* 1775 * Delete the specified class from the remote subscription hash. If the 1776 * reference count drops to zero, ask our remote peer to unsubscribe by proxy. 1777 */ 1778 void 1779 fmd_xprt_unsubscribe(fmd_xprt_t *xp, const char *class) 1780 { 1781 fmd_xprt_impl_t *xip = (fmd_xprt_impl_t *)xp; 1782 1783 uint_t refs; 1784 nvlist_t *nvl; 1785 fmd_event_t *e; 1786 char *s; 1787 1788 if ((xip->xi_flags & FMD_XPRT_RDWR) == FMD_XPRT_RDONLY) 1789 return; /* read-only transports do not proxy subscriptions */ 1790 1791 if (!(xip->xi_flags & FMD_XPRT_SUBSCRIBER)) 1792 return; /* transport is not yet an active subscriber */ 1793 1794 /* 1795 * If the subscription reference count drops to zero in xi_rsub, insert 1796 * an entry into the xi_usub hash indicating we await an unsuback event. 1797 */ 1798 (void) pthread_mutex_lock(&xip->xi_lock); 1799 1800 if ((refs = fmd_xprt_class_hash_delete(xip, &xip->xi_rsub, class)) == 0) 1801 (void) fmd_xprt_class_hash_insert(xip, &xip->xi_usub, class); 1802 1803 (void) pthread_mutex_unlock(&xip->xi_lock); 1804 1805 if (refs != 0) 1806 return; /* other subscriptions for this class still active */ 1807 1808 fmd_dprintf(FMD_DBG_XPRT, 1809 "xprt %u unsubscribing from %s\n", xip->xi_id, class); 1810 1811 nvl = fmd_protocol_xprt_sub(xip->xi_queue->eq_mod, 1812 "resource.fm.xprt.unsubscribe", xip->xi_version, class); 1813 1814 (void) nvlist_lookup_string(nvl, FM_CLASS, &s); 1815 e = fmd_event_create(FMD_EVT_PROTOCOL, FMD_HRT_NOW, nvl, s); 1816 fmd_eventq_insert_at_time(xip->xi_queue, e); 1817 } 1818 1819 static void 1820 fmd_xprt_subscribe_xid(fmd_idspace_t *ids, id_t id, void *class) 1821 { 1822 fmd_xprt_t *xp; 1823 1824 if ((xp = fmd_idspace_hold(ids, id)) != NULL) { 1825 fmd_xprt_subscribe(xp, class); 1826 fmd_idspace_rele(ids, id); 1827 } 1828 } 1829 1830 void 1831 fmd_xprt_subscribe_all(const char *class) 1832 { 1833 fmd_idspace_t *ids = fmd.d_xprt_ids; 1834 1835 if (ids->ids_count != 0) 1836 fmd_idspace_apply(ids, fmd_xprt_subscribe_xid, (void *)class); 1837 } 1838 1839 static void 1840 fmd_xprt_unsubscribe_xid(fmd_idspace_t *ids, id_t id, void *class) 1841 { 1842 fmd_xprt_t *xp; 1843 1844 if ((xp = fmd_idspace_hold(ids, id)) != NULL) { 1845 fmd_xprt_unsubscribe(xp, class); 1846 fmd_idspace_rele(ids, id); 1847 } 1848 } 1849 1850 void 1851 fmd_xprt_unsubscribe_all(const char *class) 1852 { 1853 fmd_idspace_t *ids = fmd.d_xprt_ids; 1854 1855 if (ids->ids_count != 0) 1856 fmd_idspace_apply(ids, fmd_xprt_unsubscribe_xid, (void *)class); 1857 } 1858 1859 /*ARGSUSED*/ 1860 static void 1861 fmd_xprt_suspend_xid(fmd_idspace_t *ids, id_t id, void *arg) 1862 { 1863 fmd_xprt_t *xp; 1864 1865 if ((xp = fmd_idspace_hold(ids, id)) != NULL) { 1866 fmd_xprt_xsuspend(xp, FMD_XPRT_DSUSPENDED); 1867 fmd_idspace_rele(ids, id); 1868 } 1869 } 1870 1871 void 1872 fmd_xprt_suspend_all(void) 1873 { 1874 fmd_idspace_t *ids = fmd.d_xprt_ids; 1875 1876 (void) pthread_mutex_lock(&fmd.d_xprt_lock); 1877 1878 if (fmd.d_xprt_suspend++ != 0) { 1879 (void) pthread_mutex_unlock(&fmd.d_xprt_lock); 1880 return; /* already suspended */ 1881 } 1882 1883 if (ids->ids_count != 0) 1884 fmd_idspace_apply(ids, fmd_xprt_suspend_xid, NULL); 1885 1886 (void) pthread_mutex_unlock(&fmd.d_xprt_lock); 1887 } 1888 1889 /*ARGSUSED*/ 1890 static void 1891 fmd_xprt_resume_xid(fmd_idspace_t *ids, id_t id, void *arg) 1892 { 1893 fmd_xprt_t *xp; 1894 1895 if ((xp = fmd_idspace_hold(ids, id)) != NULL) { 1896 fmd_xprt_xresume(xp, FMD_XPRT_DSUSPENDED); 1897 fmd_idspace_rele(ids, id); 1898 } 1899 } 1900 1901 void 1902 fmd_xprt_resume_all(void) 1903 { 1904 fmd_idspace_t *ids = fmd.d_xprt_ids; 1905 1906 (void) pthread_mutex_lock(&fmd.d_xprt_lock); 1907 1908 if (fmd.d_xprt_suspend == 0) 1909 fmd_panic("fmd_xprt_suspend/resume_all mismatch\n"); 1910 1911 if (--fmd.d_xprt_suspend != 0) { 1912 (void) pthread_mutex_unlock(&fmd.d_xprt_lock); 1913 return; /* not ready to be resumed */ 1914 } 1915 1916 if (ids->ids_count != 0) 1917 fmd_idspace_apply(ids, fmd_xprt_resume_xid, NULL); 1918 1919 (void) pthread_mutex_unlock(&fmd.d_xprt_lock); 1920 } 1921