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 * Copyright 2009 Sun Microsystems, Inc. All rights reserved. 23 * Use is subject to license terms. 24 */ 25 26 #include <sys/types.h> 27 #include <fmadm.h> 28 #include <errno.h> 29 #include <limits.h> 30 #include <strings.h> 31 #include <stdio.h> 32 #include <unistd.h> 33 #include <sys/wait.h> 34 #include <sys/stat.h> 35 #include <fcntl.h> 36 #include <fm/fmd_log.h> 37 #include <sys/fm/protocol.h> 38 #include <fm/libtopo.h> 39 #include <fm/fmd_adm.h> 40 #include <fm/fmd_msg.h> 41 #include <dlfcn.h> 42 #include <sys/systeminfo.h> 43 #include <sys/utsname.h> 44 #include <libintl.h> 45 #include <locale.h> 46 #include <sys/smbios.h> 47 #include <libdevinfo.h> 48 #include <stdlib.h> 49 50 #define offsetof(s, m) ((size_t)(&(((s*)0)->m))) 51 52 /* 53 * Fault records are added to catalog by calling add_fault_record_to_catalog() 54 * records are stored in order of importance to the system. 55 * If -g flag is set or not_suppressed is not set and the class fru, fault, 56 * type are the same then details are merged into an existing record, with uuid 57 * records are stored in time order. 58 * For each record information is extracted from nvlist and merged into linked 59 * list each is checked for identical records for which percentage certainty are 60 * added together. 61 * print_catalog() is called to print out catalog and release external resources 62 * 63 * /---------------\ 64 * status_rec_list -> | | -| 65 * \---------------/ 66 * \/ 67 * /---------------\ /-------\ /-------\ 68 * status_fru_list | status_record | -> | uurec | -> | uurec | -| 69 * \/ | | |- | | <- | | 70 * /-------------\ | | \-------/ \-------/ 71 * | | -> | | \/ \/ 72 * \-------------/ | | /-------\ /-------\ 73 * \/ | | -> | asru | -> | asru | 74 * --- | | | | <- | | 75 * | | \-------/ \-------/ 76 * status_asru_list | class | 77 * \/ | resource | /-------\ /-------\ 78 * /-------------\ | fru | -> | list | -> | list | 79 * | | -> | serial | | | <- | | 80 * \-------------/ | | \-------/ \-------/ 81 * \/ \---------------/ 82 * --- \/ /\ 83 * /---------------\ 84 * | status_record | 85 * \---------------/ 86 * 87 * Fmadm faulty takes a number of options which affect the format of the 88 * output displayed. By default, the display reports the FRU and ASRU along 89 * with other information on per-case basis as in the example below. 90 * 91 * --------------- ------------------------------------ -------------- ------- 92 * TIME EVENT-ID MSG-ID SEVERITY 93 * --------------- ------------------------------------ -------------- ------- 94 * Sep 21 10:01:36 d482f935-5c8f-e9ab-9f25-d0aaafec1e6c AMD-8000-2F Major 95 * 96 * Fault class : fault.memory.dimm_sb 97 * Affects : mem:///motherboard=0/chip=0/memory-controller=0/dimm=0/rank=0 98 * faulted but still in service 99 * FRU : "CPU 0 DIMM 0" (hc://.../memory-controller=0/dimm=0) 100 * faulty 101 * 102 * Description : The number of errors associated with this memory module has 103 * exceeded acceptable levels. Refer to 104 * http://sun.com/msg/AMD-8000-2F for more information. 105 * 106 * Response : Pages of memory associated with this memory module are being 107 * removed from service as errors are reported. 108 * 109 * Impact : Total system memory capacity will be reduced as pages are 110 * retired. 111 * 112 * Action : Schedule a repair procedure to replace the affected memory 113 * module. Use fmdump -v -u <EVENT_ID> to identify the module. 114 * 115 * The -v flag is similar, but adds some additonal information such as the 116 * resource. The -s flag is also similar but just gives the top line summary. 117 * All these options (ie without the -f or -r flags) use the print_catalog() 118 * function to do the display. 119 * 120 * The -f flag changes the output so that it appears sorted on a per-fru basis. 121 * The output is somewhat cut down compared to the default output. If -f is 122 * used, then print_fru() is used to print the output. 123 * 124 * ----------------------------------------------------------------------------- 125 * "SLOT 2" (hc://.../hostbridge=3/pciexrc=3/pciexbus=4/pciexdev=0) faulty 126 * 5ca4aeb3-36...f6be-c2e8166dc484 2 suspects in this FRU total certainty 100% 127 * 128 * Description : A problem was detected for a PCI device. 129 * Refer to http://sun.com/msg/PCI-8000-7J for more information. 130 * 131 * Response : One or more device instances may be disabled 132 * 133 * Impact : Possible loss of services provided by the device instances 134 * associated with this fault 135 * 136 * Action : Schedule a repair procedure to replace the affected device. 137 * Use fmdump -v -u <EVENT_ID> to identify the device or contact 138 * Sun for support. 139 * 140 * The -r flag changes the output so that it appears sorted on a per-asru basis. 141 * The output is very much cut down compared to the default output, just giving 142 * the asru fmri and state. Here print_asru() is used to print the output. 143 * 144 * mem:///motherboard=0/chip=0/memory-controller=0/dimm=0/rank=0 degraded 145 * 146 * For all fmadm faulty options, the sequence of events is 147 * 148 * 1) Walk through all the cases in the system using fmd_adm_case_iter() and 149 * for each case call dfault_rec(). This will call add_fault_record_to_catalog() 150 * This will extract the data from the nvlist and call catalog_new_record() to 151 * save the data away in various linked lists in the catalogue. 152 * 153 * 2) Once this is done, the data can be supplemented by using 154 * fmd_adm_rsrc_iter(). However this is now only necessary for the -i option. 155 * 156 * 3) Finally print_catalog(), print_fru() or print_asru() are called as 157 * appropriate to display the information from the catalogue sorted in the 158 * requested way. 159 * 160 */ 161 162 typedef struct name_list { 163 struct name_list *next; 164 struct name_list *prev; 165 char *name; 166 uint8_t pct; 167 uint8_t max_pct; 168 ushort_t count; 169 int status; 170 char *label; 171 } name_list_t; 172 173 typedef struct ari_list { 174 char *ari_uuid; 175 struct ari_list *next; 176 } ari_list_t; 177 178 typedef struct uurec { 179 struct uurec *next; 180 struct uurec *prev; 181 char *uuid; 182 ari_list_t *ari_uuid_list; 183 name_list_t *asru; 184 uint64_t sec; 185 nvlist_t *event; 186 } uurec_t; 187 188 typedef struct uurec_select { 189 struct uurec_select *next; 190 char *uuid; 191 } uurec_select_t; 192 193 typedef struct host_id { 194 char *chassis; 195 char *server; 196 char *platform; 197 char *domain; 198 char *product_sn; 199 } hostid_t; 200 201 typedef struct host_id_list { 202 hostid_t hostid; 203 struct host_id_list *next; 204 } host_id_list_t; 205 206 typedef struct status_record { 207 hostid_t *host; 208 int nrecs; 209 uurec_t *uurec; 210 char *severity; /* in C locale */ 211 char *msgid; 212 name_list_t *class; 213 name_list_t *resource; 214 name_list_t *asru; 215 name_list_t *fru; 216 name_list_t *serial; 217 uint8_t not_suppressed; 218 uint8_t injected; 219 } status_record_t; 220 221 typedef struct sr_list { 222 struct sr_list *next; 223 struct sr_list *prev; 224 struct status_record *status_record; 225 } sr_list_t; 226 227 typedef struct resource_list { 228 struct resource_list *next; 229 struct resource_list *prev; 230 sr_list_t *status_rec_list; 231 char *resource; 232 uint8_t not_suppressed; 233 uint8_t injected; 234 uint8_t max_pct; 235 } resource_list_t; 236 237 typedef struct tgetlabel_data { 238 char *label; 239 char *fru; 240 } tgetlabel_data_t; 241 242 sr_list_t *status_rec_list; 243 resource_list_t *status_fru_list; 244 resource_list_t *status_asru_list; 245 246 static int max_display; 247 static int max_fault = 0; 248 static topo_hdl_t *topo_handle; 249 static char *topo_handle_uuid; 250 static host_id_list_t *host_list; 251 static int n_server; 252 static int opt_g; 253 static fmd_msg_hdl_t *fmadm_msghdl = NULL; /* handle for libfmd_msg calls */ 254 255 static char * 256 format_date(char *buf, size_t len, uint64_t sec) 257 { 258 if (sec > LONG_MAX) { 259 (void) fprintf(stderr, 260 "record time is too large for 32-bit utility\n"); 261 (void) snprintf(buf, len, "0x%llx", sec); 262 } else { 263 time_t tod = (time_t)sec; 264 time_t now = time(NULL); 265 if (tod > now+60 || 266 tod < now - 6L*30L*24L*60L*60L) { /* 6 months ago */ 267 (void) strftime(buf, len, "%b %d %Y ", 268 localtime(&tod)); 269 } else { 270 (void) strftime(buf, len, "%b %d %T", localtime(&tod)); 271 } 272 } 273 274 return (buf); 275 } 276 277 static hostid_t * 278 find_hostid_in_list(char *platform, char *chassis, char *server, char *domain, 279 char *product_sn) 280 { 281 hostid_t *rt = NULL; 282 host_id_list_t *hostp; 283 284 if (platform == NULL) 285 platform = "-"; 286 if (server == NULL) 287 server = "-"; 288 hostp = host_list; 289 while (hostp) { 290 if (hostp->hostid.platform && 291 strcmp(hostp->hostid.platform, platform) == 0 && 292 hostp->hostid.server && 293 strcmp(hostp->hostid.server, server) == 0 && 294 (chassis == NULL || hostp->hostid.chassis == NULL || 295 strcmp(chassis, hostp->hostid.chassis) == 0) && 296 (product_sn == NULL || hostp->hostid.product_sn == NULL || 297 strcmp(product_sn, hostp->hostid.product_sn) == 0) && 298 (domain == NULL || hostp->hostid.domain == NULL || 299 strcmp(domain, hostp->hostid.domain) == 0)) { 300 rt = &hostp->hostid; 301 break; 302 } 303 hostp = hostp->next; 304 } 305 if (rt == NULL) { 306 hostp = malloc(sizeof (host_id_list_t)); 307 hostp->hostid.platform = strdup(platform); 308 hostp->hostid.product_sn = 309 product_sn ? strdup(product_sn) : NULL; 310 hostp->hostid.server = strdup(server); 311 hostp->hostid.chassis = chassis ? strdup(chassis) : NULL; 312 hostp->hostid.domain = domain ? strdup(domain) : NULL; 313 hostp->next = host_list; 314 host_list = hostp; 315 rt = &hostp->hostid; 316 n_server++; 317 } 318 return (rt); 319 } 320 321 static hostid_t * 322 find_hostid(nvlist_t *nvl) 323 { 324 char *platform = NULL, *chassis = NULL, *server = NULL, *domain = NULL; 325 char *product_sn = NULL; 326 nvlist_t *auth, *fmri; 327 hostid_t *rt = NULL; 328 329 if (nvlist_lookup_nvlist(nvl, FM_SUSPECT_DE, &fmri) == 0 && 330 nvlist_lookup_nvlist(fmri, FM_FMRI_AUTHORITY, &auth) == 0) { 331 (void) nvlist_lookup_string(auth, FM_FMRI_AUTH_PRODUCT, 332 &platform); 333 (void) nvlist_lookup_string(auth, FM_FMRI_AUTH_PRODUCT_SN, 334 &product_sn); 335 (void) nvlist_lookup_string(auth, FM_FMRI_AUTH_SERVER, &server); 336 (void) nvlist_lookup_string(auth, FM_FMRI_AUTH_CHASSIS, 337 &chassis); 338 (void) nvlist_lookup_string(auth, FM_FMRI_AUTH_DOMAIN, &domain); 339 rt = find_hostid_in_list(platform, chassis, server, 340 domain, product_sn); 341 } 342 return (rt); 343 } 344 345 /* 346 * compare two fru strings which are made up of substrings seperated by '/' 347 * return true if every substring is the same in the two strings, or if a 348 * substring is null in one. 349 */ 350 351 static int 352 frucmp(char *f1, char *f2) 353 { 354 char c1, c2; 355 int i = 0; 356 357 for (;;) { 358 c1 = *f1; 359 c2 = *f2; 360 if (c1 == c2) { 361 i = (c1 == '/') ? 0 : i + 1; 362 } else if (i == 0) { 363 if (c1 == '/') { 364 do { 365 f2++; 366 } while ((c2 = *f2) != 0 && c2 != '/'); 367 if (c2 == NULL) 368 break; 369 } else if (c2 == '/') { 370 do { 371 f1++; 372 } while ((c1 = *f1) != 0 && c1 != '/'); 373 if (c1 == NULL) 374 break; 375 } else 376 break; 377 } else 378 break; 379 if (c1 == NULL) 380 return (0); 381 f1++; 382 f2++; 383 } 384 return (1); 385 } 386 387 static int 388 tgetlabel(topo_hdl_t *thp, tnode_t *node, void *arg) 389 { 390 int err; 391 char *fru_name, *lname; 392 nvlist_t *fru = NULL; 393 int rt = TOPO_WALK_NEXT; 394 tgetlabel_data_t *tdp = (tgetlabel_data_t *)arg; 395 396 if (topo_node_fru(node, &fru, NULL, &err) == 0) { 397 if (topo_fmri_nvl2str(thp, fru, &fru_name, &err) == 0) { 398 if (frucmp(tdp->fru, fru_name) == 0 && 399 topo_node_label(node, &lname, &err) == 0) { 400 tdp->label = strdup(lname); 401 topo_hdl_strfree(thp, lname); 402 rt = TOPO_WALK_TERMINATE; 403 } 404 topo_hdl_strfree(thp, fru_name); 405 } 406 nvlist_free(fru); 407 } 408 return (rt); 409 } 410 411 static void 412 label_get_topo(void) 413 { 414 int err; 415 416 topo_handle = topo_open(TOPO_VERSION, 0, &err); 417 if (topo_handle) { 418 topo_handle_uuid = topo_snap_hold(topo_handle, NULL, &err); 419 } 420 } 421 422 static void 423 label_release_topo(void) 424 { 425 if (topo_handle_uuid) 426 topo_hdl_strfree(topo_handle, topo_handle_uuid); 427 if (topo_handle) { 428 topo_snap_release(topo_handle); 429 topo_close(topo_handle); 430 } 431 } 432 433 static char * 434 get_fmri_label(char *fru) 435 { 436 topo_walk_t *twp; 437 tgetlabel_data_t td; 438 int err; 439 440 td.label = NULL; 441 td.fru = fru; 442 if (topo_handle == NULL) 443 label_get_topo(); 444 if (topo_handle_uuid) { 445 twp = topo_walk_init(topo_handle, FM_FMRI_SCHEME_HC, 446 tgetlabel, &td, &err); 447 if (twp) { 448 topo_walk_step(twp, TOPO_WALK_CHILD); 449 topo_walk_fini(twp); 450 } 451 } 452 return (td.label); 453 } 454 455 static char * 456 get_nvl2str_topo(nvlist_t *nvl) 457 { 458 char *name = NULL; 459 char *tname; 460 int err; 461 char *scheme = NULL; 462 char *mod_name = NULL; 463 char buf[128]; 464 465 if (topo_handle == NULL) 466 label_get_topo(); 467 if (topo_fmri_nvl2str(topo_handle, nvl, &tname, &err) == 0) { 468 name = strdup(tname); 469 topo_hdl_strfree(topo_handle, tname); 470 } else { 471 (void) nvlist_lookup_string(nvl, FM_FMRI_SCHEME, &scheme); 472 (void) nvlist_lookup_string(nvl, FM_FMRI_MOD_NAME, &mod_name); 473 if (scheme && strcmp(scheme, FM_FMRI_SCHEME_FMD) == 0 && 474 mod_name) { 475 (void) snprintf(buf, sizeof (buf), "%s:///module/%s", 476 scheme, mod_name); 477 name = strdup(buf); 478 } 479 } 480 return (name); 481 } 482 483 static int 484 set_priority(char *s) 485 { 486 int rt = 0; 487 488 if (s) { 489 if (strcmp(s, "Minor") == 0) 490 rt = 1; 491 else if (strcmp(s, "Major") == 0) 492 rt = 10; 493 else if (strcmp(s, "Critical") == 0) 494 rt = 100; 495 } 496 return (rt); 497 } 498 499 static int 500 cmp_priority(char *s1, char *s2, uint64_t t1, uint64_t t2, uint8_t p1, 501 uint8_t p2) 502 { 503 int r1, r2; 504 int rt; 505 506 r1 = set_priority(s1); 507 r2 = set_priority(s2); 508 rt = r1 - r2; 509 if (rt == 0) { 510 if (t1 > t2) 511 rt = 1; 512 else if (t1 < t2) 513 rt = -1; 514 else 515 rt = p1 - p2; 516 } 517 return (rt); 518 } 519 520 /* 521 * merge two lists into one, by comparing enties in new and moving into list if 522 * name is not there or free off memory for names which are already there 523 * add_pct indicates if pct is the sum or highest pct 524 */ 525 static name_list_t * 526 merge_name_list(name_list_t **list, name_list_t *new, int add_pct) 527 { 528 name_list_t *lp, *np, *sp, *rt = NULL; 529 int max_pct; 530 531 rt = *list; 532 np = new; 533 while (np) { 534 lp = *list; 535 while (lp) { 536 if (strcmp(lp->name, np->name) == 0) 537 break; 538 lp = lp->next; 539 if (lp == *list) 540 lp = NULL; 541 } 542 if (np->next == new) 543 sp = NULL; 544 else 545 sp = np->next; 546 if (lp) { 547 lp->status |= (np->status & FM_SUSPECT_FAULTY); 548 if (add_pct) { 549 lp->pct += np->pct; 550 lp->count += np->count; 551 } else if (np->pct > lp->pct) { 552 lp->pct = np->pct; 553 } 554 max_pct = np->max_pct; 555 if (np->label) 556 free(np->label); 557 free(np->name); 558 free(np); 559 np = NULL; 560 if (max_pct > lp->max_pct) { 561 lp->max_pct = max_pct; 562 if (lp->max_pct > lp->prev->max_pct && 563 lp != *list) { 564 lp->prev->next = lp->next; 565 lp->next->prev = lp->prev; 566 np = lp; 567 } 568 } 569 } 570 if (np) { 571 lp = *list; 572 if (lp) { 573 if (np->max_pct > lp->max_pct) { 574 np->next = lp; 575 np->prev = lp->prev; 576 lp->prev->next = np; 577 lp->prev = np; 578 *list = np; 579 rt = np; 580 } else { 581 lp = lp->next; 582 while (lp != *list && 583 np->max_pct < lp->max_pct) { 584 lp = lp->next; 585 } 586 np->next = lp; 587 np->prev = lp->prev; 588 lp->prev->next = np; 589 lp->prev = np; 590 } 591 } else { 592 *list = np; 593 np->next = np; 594 np->prev = np; 595 rt = np; 596 } 597 } 598 np = sp; 599 } 600 return (rt); 601 } 602 603 static name_list_t * 604 alloc_name_list(char *name, uint8_t pct) 605 { 606 name_list_t *nlp; 607 608 nlp = malloc(sizeof (*nlp)); 609 nlp->name = strdup(name); 610 nlp->pct = pct; 611 nlp->max_pct = pct; 612 nlp->count = 1; 613 nlp->next = nlp; 614 nlp->prev = nlp; 615 nlp->status = 0; 616 nlp->label = NULL; 617 return (nlp); 618 } 619 620 static status_record_t * 621 new_record_init(uurec_t *uurec_p, char *msgid, name_list_t *class, 622 name_list_t *fru, name_list_t *asru, name_list_t *resource, 623 name_list_t *serial, boolean_t not_suppressed, 624 hostid_t *hostid, boolean_t injected) 625 { 626 status_record_t *status_rec_p; 627 628 status_rec_p = (status_record_t *)malloc(sizeof (status_record_t)); 629 status_rec_p->nrecs = 1; 630 status_rec_p->host = hostid; 631 status_rec_p->uurec = uurec_p; 632 uurec_p->next = NULL; 633 uurec_p->prev = NULL; 634 uurec_p->asru = asru; 635 if ((status_rec_p->severity = fmd_msg_getitem_id(fmadm_msghdl, NULL, 636 msgid, FMD_MSG_ITEM_SEVERITY)) == NULL) 637 status_rec_p->severity = strdup("unknown"); 638 status_rec_p->class = class; 639 status_rec_p->fru = fru; 640 status_rec_p->asru = asru; 641 status_rec_p->resource = resource; 642 status_rec_p->serial = serial; 643 status_rec_p->msgid = strdup(msgid); 644 status_rec_p->not_suppressed = not_suppressed; 645 status_rec_p->injected = injected; 646 return (status_rec_p); 647 } 648 649 /* 650 * add record to given list maintaining order higher priority first. 651 */ 652 static void 653 add_rec_list(status_record_t *status_rec_p, sr_list_t **list_pp) 654 { 655 sr_list_t *tp, *np, *sp; 656 int order; 657 uint64_t sec; 658 659 np = malloc(sizeof (sr_list_t)); 660 np->status_record = status_rec_p; 661 sec = status_rec_p->uurec->sec; 662 if ((sp = *list_pp) == NULL) { 663 *list_pp = np; 664 np->next = np; 665 np->prev = np; 666 } else { 667 /* insert new record in front of lower priority */ 668 tp = sp; 669 order = cmp_priority(status_rec_p->severity, 670 sp->status_record->severity, sec, 671 tp->status_record->uurec->sec, 0, 0); 672 if (order > 0) { 673 *list_pp = np; 674 } else { 675 tp = sp->next; 676 while (tp != sp && 677 cmp_priority(status_rec_p->severity, 678 tp->status_record->severity, sec, 679 tp->status_record->uurec->sec, 0, 0)) { 680 tp = tp->next; 681 } 682 } 683 np->next = tp; 684 np->prev = tp->prev; 685 tp->prev->next = np; 686 tp->prev = np; 687 } 688 } 689 690 static void 691 add_resource(status_record_t *status_rec_p, resource_list_t **rp, 692 resource_list_t *np) 693 { 694 int order; 695 uint64_t sec; 696 resource_list_t *sp, *tp; 697 status_record_t *srp; 698 char *severity = status_rec_p->severity; 699 700 add_rec_list(status_rec_p, &np->status_rec_list); 701 if ((sp = *rp) == NULL) { 702 np->next = np; 703 np->prev = np; 704 *rp = np; 705 } else { 706 /* 707 * insert new record in front of lower priority 708 */ 709 tp = sp->next; 710 srp = sp->status_rec_list->status_record; 711 sec = status_rec_p->uurec->sec; 712 order = cmp_priority(severity, srp->severity, sec, 713 srp->uurec->sec, np->max_pct, sp->max_pct); 714 if (order > 0) { 715 *rp = np; 716 } else { 717 srp = tp->status_rec_list->status_record; 718 while (tp != sp && 719 cmp_priority(severity, srp->severity, sec, 720 srp->uurec->sec, np->max_pct, sp->max_pct) < 0) { 721 tp = tp->next; 722 srp = tp->status_rec_list->status_record; 723 } 724 } 725 np->next = tp; 726 np->prev = tp->prev; 727 tp->prev->next = np; 728 tp->prev = np; 729 } 730 } 731 732 static void 733 add_resource_list(status_record_t *status_rec_p, name_list_t *fp, 734 resource_list_t **rpp) 735 { 736 int order; 737 resource_list_t *np, *end; 738 status_record_t *srp; 739 740 np = *rpp; 741 end = np; 742 while (np) { 743 if (strcmp(fp->name, np->resource) == 0) { 744 np->not_suppressed |= status_rec_p->not_suppressed; 745 np->injected |= status_rec_p->injected; 746 srp = np->status_rec_list->status_record; 747 order = cmp_priority(status_rec_p->severity, 748 srp->severity, status_rec_p->uurec->sec, 749 srp->uurec->sec, fp->max_pct, np->max_pct); 750 if (order > 0 && np != end) { 751 /* 752 * remove from list and add again using 753 * new priority 754 */ 755 np->prev->next = np->next; 756 np->next->prev = np->prev; 757 add_resource(status_rec_p, 758 rpp, np); 759 } else { 760 add_rec_list(status_rec_p, 761 &np->status_rec_list); 762 } 763 break; 764 } 765 np = np->next; 766 if (np == end) { 767 np = NULL; 768 break; 769 } 770 } 771 if (np == NULL) { 772 np = malloc(sizeof (resource_list_t)); 773 np->resource = fp->name; 774 np->not_suppressed = status_rec_p->not_suppressed; 775 np->injected = status_rec_p->injected; 776 np->status_rec_list = NULL; 777 np->max_pct = fp->max_pct; 778 add_resource(status_rec_p, rpp, np); 779 } 780 } 781 782 static void 783 add_list(status_record_t *status_rec_p, name_list_t *listp, 784 resource_list_t **glistp) 785 { 786 name_list_t *fp, *end; 787 788 fp = listp; 789 end = fp; 790 while (fp) { 791 add_resource_list(status_rec_p, fp, glistp); 792 fp = fp->next; 793 if (fp == end) 794 break; 795 } 796 } 797 798 /* 799 * add record to rec, fru and asru lists. 800 */ 801 static void 802 catalog_new_record(uurec_t *uurec_p, char *msgid, name_list_t *class, 803 name_list_t *fru, name_list_t *asru, name_list_t *resource, 804 name_list_t *serial, boolean_t not_suppressed, 805 hostid_t *hostid, boolean_t injected) 806 { 807 status_record_t *status_rec_p; 808 809 status_rec_p = new_record_init(uurec_p, msgid, class, fru, asru, 810 resource, serial, not_suppressed, hostid, injected); 811 add_rec_list(status_rec_p, &status_rec_list); 812 if (status_rec_p->fru) 813 add_list(status_rec_p, status_rec_p->fru, &status_fru_list); 814 if (status_rec_p->asru) 815 add_list(status_rec_p, status_rec_p->asru, &status_asru_list); 816 } 817 818 static void 819 get_serial_no(nvlist_t *nvl, name_list_t **serial_p, uint8_t pct) 820 { 821 char *name; 822 char *serial = NULL; 823 char **lserial = NULL; 824 uint64_t serint; 825 name_list_t *nlp; 826 int j; 827 uint_t nelem; 828 char buf[64]; 829 830 if (nvlist_lookup_string(nvl, FM_FMRI_SCHEME, &name) == 0) { 831 if (strcmp(name, FM_FMRI_SCHEME_CPU) == 0) { 832 if (nvlist_lookup_uint64(nvl, FM_FMRI_CPU_SERIAL_ID, 833 &serint) == 0) { 834 (void) snprintf(buf, sizeof (buf), "%llX", 835 serint); 836 nlp = alloc_name_list(buf, pct); 837 (void) merge_name_list(serial_p, nlp, 1); 838 } 839 } else if (strcmp(name, FM_FMRI_SCHEME_MEM) == 0) { 840 if (nvlist_lookup_string_array(nvl, 841 FM_FMRI_MEM_SERIAL_ID, &lserial, &nelem) == 0) { 842 nlp = alloc_name_list(lserial[0], pct); 843 for (j = 1; j < nelem; j++) { 844 name_list_t *n1lp; 845 n1lp = alloc_name_list(lserial[j], pct); 846 (void) merge_name_list(&nlp, n1lp, 1); 847 } 848 (void) merge_name_list(serial_p, nlp, 1); 849 } 850 } else if (strcmp(name, FM_FMRI_SCHEME_HC) == 0) { 851 if (nvlist_lookup_string(nvl, FM_FMRI_HC_SERIAL_ID, 852 &serial) == 0) { 853 nlp = alloc_name_list(serial, pct); 854 (void) merge_name_list(serial_p, nlp, 1); 855 } 856 } 857 } 858 } 859 860 static void 861 extract_record_info(nvlist_t *nvl, name_list_t **class_p, 862 name_list_t **fru_p, name_list_t **serial_p, 863 name_list_t **resource_p, name_list_t **asru_p, uint8_t status) 864 { 865 nvlist_t *lfru, *lasru, *rsrc; 866 name_list_t *nlp; 867 char *name; 868 uint8_t lpct = 0; 869 char *lclass = NULL; 870 char *label; 871 872 (void) nvlist_lookup_uint8(nvl, FM_FAULT_CERTAINTY, &lpct); 873 if (nvlist_lookup_string(nvl, FM_CLASS, &lclass) == 0) { 874 nlp = alloc_name_list(lclass, lpct); 875 (void) merge_name_list(class_p, nlp, 1); 876 } 877 if (nvlist_lookup_nvlist(nvl, FM_FAULT_FRU, &lfru) == 0) { 878 name = get_nvl2str_topo(lfru); 879 if (name != NULL) { 880 nlp = alloc_name_list(name, lpct); 881 nlp->status = status & ~(FM_SUSPECT_UNUSABLE | 882 FM_SUSPECT_DEGRADED); 883 free(name); 884 if (nvlist_lookup_string(nvl, FM_FAULT_LOCATION, 885 &label) == 0) 886 nlp->label = strdup(label); 887 (void) merge_name_list(fru_p, nlp, 1); 888 } 889 get_serial_no(lfru, serial_p, lpct); 890 } else if (nvlist_lookup_nvlist(nvl, FM_FAULT_RESOURCE, &rsrc) != 0) { 891 /* 892 * No FRU or resource. But we want to display the repair status 893 * somehow, so create a dummy FRU field. 894 */ 895 nlp = alloc_name_list(dgettext("FMD", "None"), lpct); 896 nlp->status = status & ~(FM_SUSPECT_UNUSABLE | 897 FM_SUSPECT_DEGRADED); 898 (void) merge_name_list(fru_p, nlp, 1); 899 } 900 if (nvlist_lookup_nvlist(nvl, FM_FAULT_ASRU, &lasru) == 0) { 901 name = get_nvl2str_topo(lasru); 902 if (name != NULL) { 903 nlp = alloc_name_list(name, lpct); 904 nlp->status = status & ~(FM_SUSPECT_NOT_PRESENT | 905 FM_SUSPECT_REPAIRED | FM_SUSPECT_REPLACED | 906 FM_SUSPECT_ACQUITTED); 907 free(name); 908 (void) merge_name_list(asru_p, nlp, 1); 909 } 910 get_serial_no(lasru, serial_p, lpct); 911 } 912 if (nvlist_lookup_nvlist(nvl, FM_FAULT_RESOURCE, &rsrc) == 0) { 913 name = get_nvl2str_topo(rsrc); 914 if (name != NULL) { 915 nlp = alloc_name_list(name, lpct); 916 nlp->status = status; 917 free(name); 918 if (nvlist_lookup_string(nvl, FM_FAULT_LOCATION, 919 &label) == 0) 920 nlp->label = strdup(label); 921 (void) merge_name_list(resource_p, nlp, 1); 922 } 923 } 924 } 925 926 static void 927 add_fault_record_to_catalog(nvlist_t *nvl, uint64_t sec, char *uuid) 928 { 929 char *msgid = "-"; 930 uint_t i, size = 0; 931 name_list_t *class = NULL, *resource = NULL; 932 name_list_t *asru = NULL, *fru = NULL, *serial = NULL; 933 nvlist_t **nva; 934 uint8_t *ba; 935 uurec_t *uurec_p; 936 hostid_t *host; 937 boolean_t not_suppressed = 1; 938 boolean_t any_present = 0; 939 boolean_t injected = 0; 940 941 (void) nvlist_lookup_string(nvl, FM_SUSPECT_DIAG_CODE, &msgid); 942 (void) nvlist_lookup_uint32(nvl, FM_SUSPECT_FAULT_SZ, &size); 943 (void) nvlist_lookup_boolean_value(nvl, FM_SUSPECT_MESSAGE, 944 ¬_suppressed); 945 (void) nvlist_lookup_boolean_value(nvl, FM_SUSPECT_INJECTED, &injected); 946 947 if (size != 0) { 948 (void) nvlist_lookup_nvlist_array(nvl, FM_SUSPECT_FAULT_LIST, 949 &nva, &size); 950 (void) nvlist_lookup_uint8_array(nvl, FM_SUSPECT_FAULT_STATUS, 951 &ba, &size); 952 for (i = 0; i < size; i++) { 953 extract_record_info(nva[i], &class, &fru, &serial, 954 &resource, &asru, ba[i]); 955 if (!(ba[i] & FM_SUSPECT_NOT_PRESENT) && 956 (ba[i] & FM_SUSPECT_FAULTY)) 957 any_present = 1; 958 } 959 /* 960 * also suppress if no resources present 961 */ 962 if (any_present == 0) 963 not_suppressed = 0; 964 } 965 966 uurec_p = (uurec_t *)malloc(sizeof (uurec_t)); 967 uurec_p->uuid = strdup(uuid); 968 uurec_p->sec = sec; 969 uurec_p->ari_uuid_list = NULL; 970 uurec_p->event = NULL; 971 (void) nvlist_dup(nvl, &uurec_p->event, 0); 972 host = find_hostid(nvl); 973 catalog_new_record(uurec_p, msgid, class, fru, asru, 974 resource, serial, not_suppressed, host, injected); 975 } 976 977 static void 978 update_asru_state_in_catalog(const char *uuid, const char *ari_uuid) 979 { 980 sr_list_t *srp; 981 uurec_t *uurp; 982 ari_list_t *ari_list; 983 984 srp = status_rec_list; 985 if (srp) { 986 for (;;) { 987 uurp = srp->status_record->uurec; 988 while (uurp) { 989 if (strcmp(uuid, uurp->uuid) == 0) { 990 ari_list = (ari_list_t *) 991 malloc(sizeof (ari_list_t)); 992 ari_list->ari_uuid = strdup(ari_uuid); 993 ari_list->next = uurp->ari_uuid_list; 994 uurp->ari_uuid_list = ari_list; 995 return; 996 } 997 uurp = uurp->next; 998 } 999 if (srp->next == status_rec_list) 1000 break; 1001 srp = srp->next; 1002 } 1003 } 1004 } 1005 1006 static void 1007 print_line(char *label, char *buf) 1008 { 1009 char *cp, *ep, *wp; 1010 char c; 1011 int i; 1012 int lsz; 1013 char *padding; 1014 1015 lsz = strlen(label); 1016 padding = malloc(lsz + 1); 1017 for (i = 0; i < lsz; i++) 1018 padding[i] = ' '; 1019 padding[i] = 0; 1020 cp = buf; 1021 ep = buf; 1022 c = *ep; 1023 (void) printf("\n"); 1024 while (c) { 1025 i = lsz; 1026 wp = NULL; 1027 while ((c = *ep) != NULL && (wp == NULL || i < 80)) { 1028 if (c == ' ') 1029 wp = ep; 1030 else if (c == '\n') { 1031 i = 0; 1032 *ep = 0; 1033 do { 1034 ep++; 1035 } while ((c = *ep) != NULL && c == ' '); 1036 break; 1037 } 1038 ep++; 1039 i++; 1040 } 1041 if (i >= 80 && wp) { 1042 *wp = 0; 1043 ep = wp + 1; 1044 c = *ep; 1045 } 1046 (void) printf("%s%s\n", label, cp); 1047 cp = ep; 1048 label = padding; 1049 } 1050 free(padding); 1051 } 1052 1053 static void 1054 print_dict_info_line(nvlist_t *e, fmd_msg_item_t what, const char *linehdr) 1055 { 1056 char *cp = fmd_msg_getitem_nv(fmadm_msghdl, NULL, e, what); 1057 1058 if (cp) { 1059 print_line(dgettext("FMD", linehdr), cp); 1060 free(cp); 1061 } 1062 } 1063 1064 static void 1065 print_dict_info(nvlist_t *nvl) 1066 { 1067 print_dict_info_line(nvl, FMD_MSG_ITEM_DESC, "Description : "); 1068 print_dict_info_line(nvl, FMD_MSG_ITEM_RESPONSE, "Response : "); 1069 print_dict_info_line(nvl, FMD_MSG_ITEM_IMPACT, "Impact : "); 1070 print_dict_info_line(nvl, FMD_MSG_ITEM_ACTION, "Action : "); 1071 } 1072 1073 static void 1074 print_name(name_list_t *list, char *(func)(char *), char *padding, int *np, 1075 int pct, int full) 1076 { 1077 char *name, *fru_label = NULL; 1078 1079 name = list->name; 1080 if (list->label) { 1081 (void) printf("%s \"%s\" (%s)", padding, list->label, name); 1082 *np += 1; 1083 } else if (func && (fru_label = func(list->name)) != NULL) { 1084 (void) printf("%s \"%s\" (%s)", padding, fru_label, name); 1085 *np += 1; 1086 free(fru_label); 1087 } else { 1088 (void) printf("%s %s", padding, name); 1089 *np += 1; 1090 } 1091 if (list->pct && pct > 0 && pct < 100) { 1092 if (list->count > 1) { 1093 if (full) { 1094 (void) printf(" %d @ %s %d%%\n", list->count, 1095 dgettext("FMD", "max"), 1096 list->max_pct); 1097 } else { 1098 (void) printf(" %s %d%%\n", 1099 dgettext("FMD", "max"), 1100 list->max_pct); 1101 } 1102 } else { 1103 (void) printf(" %d%%\n", list->pct); 1104 } 1105 } else { 1106 (void) printf("\n"); 1107 } 1108 } 1109 1110 static void 1111 print_asru_status(int status, char *label) 1112 { 1113 char *msg = NULL; 1114 1115 switch (status) { 1116 case 0: 1117 msg = dgettext("FMD", "ok and in service"); 1118 break; 1119 case FM_SUSPECT_DEGRADED: 1120 msg = dgettext("FMD", "service degraded, " 1121 "but associated components no longer faulty"); 1122 break; 1123 case FM_SUSPECT_FAULTY | FM_SUSPECT_DEGRADED: 1124 msg = dgettext("FMD", "faulted but still " 1125 "providing degraded service"); 1126 break; 1127 case FM_SUSPECT_FAULTY: 1128 msg = dgettext("FMD", "faulted but still in service"); 1129 break; 1130 case FM_SUSPECT_UNUSABLE: 1131 msg = dgettext("FMD", "out of service, " 1132 "but associated components no longer faulty"); 1133 break; 1134 case FM_SUSPECT_FAULTY | FM_SUSPECT_UNUSABLE: 1135 msg = dgettext("FMD", "faulted and taken out of service"); 1136 break; 1137 default: 1138 break; 1139 } 1140 if (msg) { 1141 (void) printf("%s %s\n", label, msg); 1142 } 1143 } 1144 1145 static void 1146 print_fru_status(int status, char *label) 1147 { 1148 char *msg = NULL; 1149 1150 if (status & FM_SUSPECT_NOT_PRESENT) 1151 msg = dgettext("FMD", "not present"); 1152 else if (status & FM_SUSPECT_FAULTY) 1153 msg = dgettext("FMD", "faulty"); 1154 else if (status & FM_SUSPECT_REPLACED) 1155 msg = dgettext("FMD", "replaced"); 1156 else if (status & FM_SUSPECT_REPAIRED) 1157 msg = dgettext("FMD", "repair attempted"); 1158 else if (status & FM_SUSPECT_ACQUITTED) 1159 msg = dgettext("FMD", "acquitted"); 1160 else 1161 msg = dgettext("FMD", "removed"); 1162 (void) printf("%s %s\n", label, msg); 1163 } 1164 1165 static void 1166 print_rsrc_status(int status, char *label) 1167 { 1168 char *msg = ""; 1169 1170 if (status & FM_SUSPECT_NOT_PRESENT) 1171 msg = dgettext("FMD", "not present"); 1172 else if (status & FM_SUSPECT_FAULTY) { 1173 if (status & FM_SUSPECT_DEGRADED) 1174 msg = dgettext("FMD", 1175 "faulted but still providing degraded service"); 1176 else if (status & FM_SUSPECT_UNUSABLE) 1177 msg = dgettext("FMD", 1178 "faulted and taken out of service"); 1179 else 1180 msg = dgettext("FMD", "faulted but still in service"); 1181 } else if (status & FM_SUSPECT_REPLACED) 1182 msg = dgettext("FMD", "replaced"); 1183 else if (status & FM_SUSPECT_REPAIRED) 1184 msg = dgettext("FMD", "repair attempted"); 1185 else if (status & FM_SUSPECT_ACQUITTED) 1186 msg = dgettext("FMD", "acquitted"); 1187 else 1188 msg = dgettext("FMD", "removed"); 1189 (void) printf("%s %s\n", label, msg); 1190 } 1191 1192 static void 1193 print_name_list(name_list_t *list, char *label, char *(func)(char *), 1194 int limit, int pct, void (func1)(int, char *), int full) 1195 { 1196 char *name, *fru_label = NULL; 1197 char *padding; 1198 int i, j, l, n; 1199 name_list_t *end = list; 1200 1201 l = strlen(label); 1202 padding = malloc(l + 1); 1203 for (i = 0; i < l; i++) 1204 padding[i] = ' '; 1205 padding[l] = 0; 1206 (void) printf("%s", label); 1207 name = list->name; 1208 if (list->label) 1209 (void) printf(" \"%s\" (%s)", list->label, name); 1210 else if (func && (fru_label = func(list->name)) != NULL) { 1211 (void) printf(" \"%s\" (%s)", fru_label, name); 1212 free(fru_label); 1213 } else 1214 (void) printf(" %s", name); 1215 if (list->pct && pct > 0 && pct < 100) { 1216 if (list->count > 1) { 1217 if (full) { 1218 (void) printf(" %d @ %s %d%%\n", list->count, 1219 dgettext("FMD", "max"), list->max_pct); 1220 } else { 1221 (void) printf(" %s %d%%\n", 1222 dgettext("FMD", "max"), list->max_pct); 1223 } 1224 } else { 1225 (void) printf(" %d%%\n", list->pct); 1226 } 1227 } else { 1228 (void) printf("\n"); 1229 } 1230 if (func1) 1231 func1(list->status, padding); 1232 n = 1; 1233 j = 0; 1234 while ((list = list->next) != end) { 1235 if (limit == 0 || n < limit) { 1236 print_name(list, func, padding, &n, pct, full); 1237 if (func1) 1238 func1(list->status, padding); 1239 } else 1240 j++; 1241 } 1242 if (j == 1) { 1243 print_name(list->prev, func, padding, &n, pct, full); 1244 } else if (j > 1) { 1245 (void) printf("%s... %d %s\n", padding, j, 1246 dgettext("FMD", "more entries suppressed," 1247 " use -v option for full list")); 1248 } 1249 free(padding); 1250 } 1251 1252 static int 1253 asru_same_status(name_list_t *list) 1254 { 1255 name_list_t *end = list; 1256 int status = list->status; 1257 1258 while ((list = list->next) != end) { 1259 if (status == -1) { 1260 status = list->status; 1261 continue; 1262 } 1263 if (list->status != -1 && status != list->status) { 1264 status = -1; 1265 break; 1266 } 1267 } 1268 return (status); 1269 } 1270 1271 static int 1272 serial_in_fru(name_list_t *fru, name_list_t *serial) 1273 { 1274 name_list_t *sp = serial; 1275 name_list_t *fp; 1276 int nserial = 0; 1277 int found = 0; 1278 char buf[128]; 1279 1280 while (sp) { 1281 fp = fru; 1282 nserial++; 1283 (void) snprintf(buf, sizeof (buf), "serial=%s", sp->name); 1284 buf[sizeof (buf) - 1] = 0; 1285 while (fp) { 1286 if (strstr(fp->name, buf) != NULL) { 1287 found++; 1288 break; 1289 } 1290 fp = fp->next; 1291 if (fp == fru) 1292 break; 1293 } 1294 sp = sp->next; 1295 if (sp == serial) 1296 break; 1297 } 1298 return (found == nserial ? 1 : 0); 1299 } 1300 1301 static void 1302 print_sup_record(status_record_t *srp, int opt_i, int full) 1303 { 1304 char buf[32]; 1305 uurec_t *uurp = srp->uurec; 1306 int n, j, k, max; 1307 int status; 1308 ari_list_t *ari_list; 1309 1310 n = 0; 1311 max = max_fault; 1312 if (max < 0) { 1313 max = 0; 1314 } 1315 j = max / 2; 1316 max -= j; 1317 k = srp->nrecs - max; 1318 while ((uurp = uurp->next) != NULL) { 1319 if (full || n < j || n >= k || max_fault == 0 || 1320 srp->nrecs == max_fault+1) { 1321 if (opt_i) { 1322 ari_list = uurp->ari_uuid_list; 1323 while (ari_list) { 1324 (void) printf("%-15s %s\n", 1325 format_date(buf, sizeof (buf), 1326 uurp->sec), ari_list->ari_uuid); 1327 ari_list = ari_list->next; 1328 } 1329 } else { 1330 (void) printf("%-15s %s\n", 1331 format_date(buf, sizeof (buf), uurp->sec), 1332 uurp->uuid); 1333 } 1334 } else if (n == j) 1335 (void) printf("... %d %s\n", srp->nrecs - max_fault, 1336 dgettext("FMD", "more entries suppressed")); 1337 n++; 1338 } 1339 (void) printf("\n"); 1340 (void) printf("%s %s", dgettext("FMD", "Host :"), 1341 srp->host->server); 1342 if (srp->host->domain) 1343 (void) printf("\t%s %s", dgettext("FMD", "Domain :"), 1344 srp->host->domain); 1345 (void) printf("\n%s %s", dgettext("FMD", "Platform :"), 1346 srp->host->platform); 1347 (void) printf("\t%s %s", dgettext("FMD", "Chassis_id :"), 1348 srp->host->chassis ? srp->host->chassis : ""); 1349 (void) printf("\n%s %s\n\n", dgettext("FMD", "Product_sn :"), 1350 srp->host->product_sn? srp->host->product_sn : ""); 1351 if (srp->class) 1352 print_name_list(srp->class, 1353 dgettext("FMD", "Fault class :"), NULL, 0, srp->class->pct, 1354 NULL, full); 1355 if (srp->asru) { 1356 status = asru_same_status(srp->asru); 1357 if (status != -1) { 1358 print_name_list(srp->asru, 1359 dgettext("FMD", "Affects :"), NULL, 1360 full ? 0 : max_display, 0, NULL, full); 1361 print_asru_status(status, " "); 1362 } else 1363 print_name_list(srp->asru, 1364 dgettext("FMD", "Affects :"), NULL, 1365 full ? 0 : max_display, 0, print_asru_status, full); 1366 } 1367 if (full || srp->fru == NULL || srp->asru == NULL) { 1368 if (srp->resource) { 1369 status = asru_same_status(srp->resource); 1370 if (status != -1) { 1371 print_name_list(srp->resource, 1372 dgettext("FMD", "Problem in :"), NULL, 1373 full ? 0 : max_display, 0, NULL, full); 1374 print_rsrc_status(status, " "); 1375 } else 1376 print_name_list(srp->resource, 1377 dgettext("FMD", "Problem in :"), 1378 NULL, full ? 0 : max_display, 0, 1379 print_rsrc_status, full); 1380 } 1381 } 1382 if (srp->fru) { 1383 status = asru_same_status(srp->fru); 1384 if (status != -1) { 1385 print_name_list(srp->fru, dgettext("FMD", 1386 "FRU :"), get_fmri_label, 0, 1387 srp->fru->pct == 100 ? 100 : srp->fru->max_pct, 1388 NULL, full); 1389 print_fru_status(status, " "); 1390 } else 1391 print_name_list(srp->fru, dgettext("FMD", 1392 "FRU :"), get_fmri_label, 0, 1393 srp->fru->pct == 100 ? 100 : srp->fru->max_pct, 1394 print_fru_status, full); 1395 } 1396 if (srp->serial && !serial_in_fru(srp->fru, srp->serial) && 1397 !serial_in_fru(srp->asru, srp->serial)) { 1398 print_name_list(srp->serial, dgettext("FMD", "Serial ID. :"), 1399 NULL, 0, 0, NULL, full); 1400 } 1401 print_dict_info(srp->uurec->event); 1402 (void) printf("\n"); 1403 } 1404 1405 static void 1406 print_status_record(status_record_t *srp, int summary, int opt_i, int full) 1407 { 1408 char buf[32]; 1409 uurec_t *uurp = srp->uurec; 1410 static int header = 0; 1411 char *head; 1412 ari_list_t *ari_list; 1413 1414 if (!summary || !header) { 1415 if (opt_i) { 1416 head = "--------------- " 1417 "------------------------------------ " 1418 "-------------- ---------\n" 1419 "TIME CACHE-ID" 1420 " MSG-ID" 1421 " SEVERITY\n--------------- " 1422 "------------------------------------ " 1423 " -------------- ---------"; 1424 } else { 1425 head = "--------------- " 1426 "------------------------------------ " 1427 "-------------- ---------\n" 1428 "TIME EVENT-ID" 1429 " MSG-ID" 1430 " SEVERITY\n--------------- " 1431 "------------------------------------ " 1432 " -------------- ---------"; 1433 } 1434 (void) printf("%s\n", dgettext("FMD", head)); 1435 header = 1; 1436 } 1437 if (opt_i) { 1438 ari_list = uurp->ari_uuid_list; 1439 while (ari_list) { 1440 (void) printf("%-15s %-37s %-14s %-9s %s\n", 1441 format_date(buf, sizeof (buf), uurp->sec), 1442 ari_list->ari_uuid, srp->msgid, srp->severity, 1443 srp->injected ? dgettext("FMD", "injected") : ""); 1444 ari_list = ari_list->next; 1445 } 1446 } else { 1447 (void) printf("%-15s %-37s %-14s %-9s %s\n", 1448 format_date(buf, sizeof (buf), uurp->sec), 1449 uurp->uuid, srp->msgid, srp->severity, 1450 srp->injected ? dgettext("FMD", "injected") : ""); 1451 } 1452 1453 if (!summary) 1454 print_sup_record(srp, opt_i, full); 1455 } 1456 1457 static void 1458 print_catalog(int summary, int opt_a, int full, int opt_i, int page_feed) 1459 { 1460 status_record_t *srp; 1461 sr_list_t *slp; 1462 1463 slp = status_rec_list; 1464 if (slp) { 1465 for (;;) { 1466 srp = slp->status_record; 1467 if (opt_a || srp->not_suppressed) { 1468 if (page_feed) 1469 (void) printf("\f\n"); 1470 print_status_record(srp, summary, opt_i, full); 1471 } 1472 if (slp->next == status_rec_list) 1473 break; 1474 slp = slp->next; 1475 } 1476 } 1477 } 1478 1479 static name_list_t * 1480 find_fru(status_record_t *srp, char *resource) 1481 { 1482 name_list_t *rt = NULL; 1483 name_list_t *fru = srp->fru; 1484 1485 while (fru) { 1486 if (strcmp(resource, fru->name) == 0) { 1487 rt = fru; 1488 break; 1489 } 1490 fru = fru->next; 1491 if (fru == srp->fru) 1492 break; 1493 } 1494 return (rt); 1495 } 1496 1497 static void 1498 print_fru_line(name_list_t *fru, char *uuid) 1499 { 1500 if (fru->pct == 100) { 1501 (void) printf("%s %d %s %d%%\n", uuid, fru->count, 1502 dgettext("FMD", "suspects in this FRU total certainty"), 1503 100); 1504 } else { 1505 (void) printf("%s %d %s %d%%\n", uuid, fru->count, 1506 dgettext("FMD", "suspects in this FRU max certainty"), 1507 fru->max_pct); 1508 } 1509 } 1510 1511 static void 1512 print_fru(int summary, int opt_a, int opt_i, int page_feed) 1513 { 1514 resource_list_t *tp = status_fru_list; 1515 status_record_t *srp; 1516 sr_list_t *slp, *end; 1517 char *msgid, *fru_label; 1518 uurec_t *uurp; 1519 name_list_t *fru; 1520 int status; 1521 ari_list_t *ari_list; 1522 1523 while (tp) { 1524 if (opt_a || tp->not_suppressed) { 1525 if (page_feed) 1526 (void) printf("\f\n"); 1527 if (!summary) 1528 (void) printf("-----------------------------" 1529 "---------------------------------------" 1530 "----------\n"); 1531 slp = tp->status_rec_list; 1532 end = slp; 1533 do { 1534 srp = slp->status_record; 1535 fru = find_fru(srp, tp->resource); 1536 if (fru) { 1537 if (fru->label) 1538 (void) printf("\"%s\" (%s) ", 1539 fru->label, fru->name); 1540 else if ((fru_label = get_fmri_label( 1541 fru->name)) != NULL) { 1542 (void) printf("\"%s\" (%s) ", 1543 fru_label, fru->name); 1544 free(fru_label); 1545 } else 1546 (void) printf("%s ", 1547 fru->name); 1548 break; 1549 } 1550 slp = slp->next; 1551 } while (slp != end); 1552 1553 slp = tp->status_rec_list; 1554 end = slp; 1555 status = 0; 1556 do { 1557 srp = slp->status_record; 1558 fru = srp->fru; 1559 while (fru) { 1560 if (strcmp(tp->resource, 1561 fru->name) == 0) 1562 status |= fru->status; 1563 fru = fru->next; 1564 if (fru == srp->fru) 1565 break; 1566 } 1567 slp = slp->next; 1568 } while (slp != end); 1569 if (status & FM_SUSPECT_NOT_PRESENT) 1570 (void) printf(dgettext("FMD", "not present")); 1571 else if (status & FM_SUSPECT_FAULTY) 1572 (void) printf(dgettext("FMD", "faulty")); 1573 else if (status & FM_SUSPECT_REPLACED) 1574 (void) printf(dgettext("FMD", "replaced")); 1575 else if (status & FM_SUSPECT_REPAIRED) 1576 (void) printf(dgettext("FMD", 1577 "repair attempted")); 1578 else if (status & FM_SUSPECT_ACQUITTED) 1579 (void) printf(dgettext("FMD", "acquitted")); 1580 else 1581 (void) printf(dgettext("FMD", "removed")); 1582 1583 if (tp->injected) 1584 (void) printf(dgettext("FMD", " injected\n")); 1585 else 1586 (void) printf(dgettext("FMD", "\n")); 1587 1588 slp = tp->status_rec_list; 1589 end = slp; 1590 do { 1591 srp = slp->status_record; 1592 uurp = srp->uurec; 1593 fru = find_fru(srp, tp->resource); 1594 if (fru) { 1595 if (opt_i) { 1596 ari_list = uurp->ari_uuid_list; 1597 while (ari_list) { 1598 print_fru_line(fru, 1599 ari_list->ari_uuid); 1600 ari_list = 1601 ari_list->next; 1602 } 1603 } else { 1604 print_fru_line(fru, uurp->uuid); 1605 } 1606 } 1607 slp = slp->next; 1608 } while (slp != end); 1609 if (!summary) { 1610 slp = tp->status_rec_list; 1611 end = slp; 1612 srp = slp->status_record; 1613 if (srp->serial && 1614 !serial_in_fru(srp->fru, srp->serial)) { 1615 print_name_list(srp->serial, 1616 dgettext("FMD", "Serial ID. :"), 1617 NULL, 0, 0, NULL, 1); 1618 } 1619 msgid = NULL; 1620 do { 1621 if (msgid == NULL || 1622 strcmp(msgid, srp->msgid) != 0) { 1623 msgid = srp->msgid; 1624 print_dict_info(uurp->event); 1625 } 1626 slp = slp->next; 1627 } while (slp != end); 1628 } 1629 } 1630 tp = tp->next; 1631 if (tp == status_fru_list) 1632 break; 1633 } 1634 } 1635 1636 static void 1637 print_asru(int opt_a) 1638 { 1639 resource_list_t *tp = status_asru_list; 1640 status_record_t *srp; 1641 sr_list_t *slp, *end; 1642 char *msg; 1643 int status; 1644 name_list_t *asru; 1645 1646 while (tp) { 1647 if (opt_a || tp->not_suppressed) { 1648 status = 0; 1649 slp = tp->status_rec_list; 1650 end = slp; 1651 do { 1652 srp = slp->status_record; 1653 asru = srp->asru; 1654 while (asru) { 1655 if (strcmp(tp->resource, 1656 asru->name) == 0) 1657 status |= asru->status; 1658 asru = asru->next; 1659 if (asru == srp->asru) 1660 break; 1661 } 1662 slp = slp->next; 1663 } while (slp != end); 1664 switch (status) { 1665 case 0: 1666 msg = dgettext("FMD", "ok"); 1667 break; 1668 case FM_SUSPECT_DEGRADED: 1669 msg = dgettext("FMD", "degraded"); 1670 break; 1671 case FM_SUSPECT_FAULTY | FM_SUSPECT_DEGRADED: 1672 msg = dgettext("FMD", "degraded"); 1673 break; 1674 case FM_SUSPECT_FAULTY: 1675 msg = dgettext("FMD", "degraded"); 1676 break; 1677 case FM_SUSPECT_UNUSABLE: 1678 msg = dgettext("FMD", "unknown"); 1679 break; 1680 case FM_SUSPECT_FAULTY | FM_SUSPECT_UNUSABLE: 1681 msg = dgettext("FMD", "faulted"); 1682 break; 1683 default: 1684 msg = ""; 1685 break; 1686 } 1687 (void) printf("%-69s %s", tp->resource, msg); 1688 if (tp->injected) 1689 (void) printf(dgettext("FMD", " injected\n")); 1690 else 1691 (void) printf(dgettext("FMD", "\n")); 1692 } 1693 tp = tp->next; 1694 if (tp == status_asru_list) 1695 break; 1696 } 1697 } 1698 1699 static int 1700 uuid_in_list(char *uuid, uurec_select_t *uurecp) 1701 { 1702 while (uurecp) { 1703 if (strcmp(uuid, uurecp->uuid) == 0) 1704 return (1); 1705 uurecp = uurecp->next; 1706 } 1707 return (0); 1708 } 1709 1710 static int 1711 dfault_rec(const fmd_adm_caseinfo_t *acp, void *arg) 1712 { 1713 int64_t *diag_time; 1714 uint_t nelem; 1715 int rt = 0; 1716 char *uuid = "-"; 1717 uurec_select_t *uurecp = (uurec_select_t *)arg; 1718 1719 if (nvlist_lookup_int64_array(acp->aci_event, FM_SUSPECT_DIAG_TIME, 1720 &diag_time, &nelem) == 0 && nelem >= 2) { 1721 (void) nvlist_lookup_string(acp->aci_event, FM_SUSPECT_UUID, 1722 &uuid); 1723 if (uurecp == NULL || uuid_in_list(uuid, uurecp)) 1724 add_fault_record_to_catalog(acp->aci_event, *diag_time, 1725 uuid); 1726 } else { 1727 rt = -1; 1728 } 1729 return (rt); 1730 } 1731 1732 /*ARGSUSED*/ 1733 static int 1734 dstatus_rec(const fmd_adm_rsrcinfo_t *ari, void *unused) 1735 { 1736 update_asru_state_in_catalog(ari->ari_case, ari->ari_uuid); 1737 return (0); 1738 } 1739 1740 static int 1741 get_cases_from_fmd(fmd_adm_t *adm, uurec_select_t *uurecp, int opt_i) 1742 { 1743 int rt = FMADM_EXIT_SUCCESS; 1744 1745 /* 1746 * These calls may fail with Protocol error if message payload is 1747 * too big 1748 */ 1749 if (fmd_adm_case_iter(adm, NULL, dfault_rec, uurecp) != 0) 1750 die("failed to get case list from fmd"); 1751 if (opt_i && fmd_adm_rsrc_iter(adm, 1, dstatus_rec, NULL) != 0) 1752 die("failed to get case status from fmd"); 1753 return (rt); 1754 } 1755 1756 /* 1757 * fmadm faulty command 1758 * 1759 * -a show hidden fault records 1760 * -f show faulty fru's 1761 * -g force grouping of similar faults on the same fru 1762 * -n number of fault records to display 1763 * -p pipe output through pager 1764 * -r show faulty asru's 1765 * -s print summary of first fault 1766 * -u print listed uuid's only 1767 * -v full output 1768 */ 1769 1770 int 1771 cmd_faulty(fmd_adm_t *adm, int argc, char *argv[]) 1772 { 1773 int opt_a = 0, opt_v = 0, opt_p = 0, opt_s = 0, opt_r = 0, opt_f = 0; 1774 int opt_i = 0; 1775 char *pager; 1776 FILE *fp; 1777 int rt, c, stat; 1778 uurec_select_t *tp; 1779 uurec_select_t *uurecp = NULL; 1780 1781 while ((c = getopt(argc, argv, "afgin:prsu:v")) != EOF) { 1782 switch (c) { 1783 case 'a': 1784 opt_a++; 1785 break; 1786 case 'f': 1787 opt_f++; 1788 break; 1789 case 'g': 1790 opt_g++; 1791 break; 1792 case 'i': 1793 opt_i++; 1794 break; 1795 case 'n': 1796 max_fault = atoi(optarg); 1797 break; 1798 case 'p': 1799 opt_p++; 1800 break; 1801 case 'r': 1802 opt_r++; 1803 break; 1804 case 's': 1805 opt_s++; 1806 break; 1807 case 'u': 1808 tp = (uurec_select_t *)malloc(sizeof (uurec_select_t)); 1809 tp->uuid = optarg; 1810 tp->next = uurecp; 1811 uurecp = tp; 1812 opt_a = 1; 1813 break; 1814 case 'v': 1815 opt_v++; 1816 break; 1817 default: 1818 return (FMADM_EXIT_USAGE); 1819 } 1820 } 1821 if (optind < argc) 1822 return (FMADM_EXIT_USAGE); 1823 1824 if ((fmadm_msghdl = fmd_msg_init(NULL, FMD_MSG_VERSION)) == NULL) 1825 return (FMADM_EXIT_ERROR); 1826 rt = get_cases_from_fmd(adm, uurecp, opt_i); 1827 if (opt_p) { 1828 if ((pager = getenv("PAGER")) == NULL) 1829 pager = "/usr/bin/more"; 1830 fp = popen(pager, "w"); 1831 if (fp == NULL) { 1832 rt = FMADM_EXIT_ERROR; 1833 opt_p = 0; 1834 } else { 1835 dup2(fileno(fp), 1); 1836 setbuf(stdout, NULL); 1837 (void) fclose(fp); 1838 } 1839 } 1840 max_display = max_fault; 1841 if (opt_f) 1842 print_fru(opt_s, opt_a, opt_i, opt_p && !opt_s); 1843 if (opt_r) 1844 print_asru(opt_a); 1845 if (opt_f == 0 && opt_r == 0) 1846 print_catalog(opt_s, opt_a, opt_v, opt_i, opt_p && !opt_s); 1847 fmd_msg_fini(fmadm_msghdl); 1848 label_release_topo(); 1849 if (opt_p) { 1850 (void) fclose(stdout); 1851 (void) wait(&stat); 1852 } 1853 return (rt); 1854 } 1855 1856 int 1857 cmd_flush(fmd_adm_t *adm, int argc, char *argv[]) 1858 { 1859 int i, status = FMADM_EXIT_SUCCESS; 1860 1861 if (argc < 2 || (i = getopt(argc, argv, "")) != EOF) 1862 return (FMADM_EXIT_USAGE); 1863 1864 for (i = 1; i < argc; i++) { 1865 if (fmd_adm_rsrc_flush(adm, argv[i]) != 0) { 1866 warn("failed to flush %s", argv[i]); 1867 status = FMADM_EXIT_ERROR; 1868 } else 1869 note("flushed resource history for %s\n", argv[i]); 1870 } 1871 1872 return (status); 1873 } 1874 1875 int 1876 cmd_repair(fmd_adm_t *adm, int argc, char *argv[]) 1877 { 1878 int err; 1879 1880 if (getopt(argc, argv, "") != EOF) 1881 return (FMADM_EXIT_USAGE); 1882 1883 if (argc - optind != 1) 1884 return (FMADM_EXIT_USAGE); 1885 1886 /* 1887 * argument could be a uuid, an fmri (asru, fru or resource) 1888 * or a label. Try uuid first, If that fails try the others. 1889 */ 1890 err = fmd_adm_case_repair(adm, argv[optind]); 1891 if (err != 0) 1892 err = fmd_adm_rsrc_repaired(adm, argv[optind]); 1893 1894 if (err != 0) 1895 die("failed to record repair to %s", argv[optind]); 1896 1897 note("recorded repair to %s\n", argv[optind]); 1898 return (FMADM_EXIT_SUCCESS); 1899 } 1900 1901 int 1902 cmd_repaired(fmd_adm_t *adm, int argc, char *argv[]) 1903 { 1904 int err; 1905 1906 if (getopt(argc, argv, "") != EOF) 1907 return (FMADM_EXIT_USAGE); 1908 1909 if (argc - optind != 1) 1910 return (FMADM_EXIT_USAGE); 1911 1912 /* 1913 * argument could be an fmri (asru, fru or resource) or a label. 1914 */ 1915 err = fmd_adm_rsrc_repaired(adm, argv[optind]); 1916 if (err != 0) 1917 die("failed to record repair to %s", argv[optind]); 1918 1919 note("recorded repair to of %s\n", argv[optind]); 1920 return (FMADM_EXIT_SUCCESS); 1921 } 1922 1923 int 1924 cmd_replaced(fmd_adm_t *adm, int argc, char *argv[]) 1925 { 1926 int err; 1927 1928 if (getopt(argc, argv, "") != EOF) 1929 return (FMADM_EXIT_USAGE); 1930 1931 if (argc - optind != 1) 1932 return (FMADM_EXIT_USAGE); 1933 1934 /* 1935 * argument could be an fmri (asru, fru or resource) or a label. 1936 */ 1937 err = fmd_adm_rsrc_replaced(adm, argv[optind]); 1938 if (err != 0) 1939 die("failed to record replacement of %s", argv[optind]); 1940 1941 note("recorded replacement of %s\n", argv[optind]); 1942 return (FMADM_EXIT_SUCCESS); 1943 } 1944 1945 int 1946 cmd_acquit(fmd_adm_t *adm, int argc, char *argv[]) 1947 { 1948 int err; 1949 1950 if (getopt(argc, argv, "") != EOF) 1951 return (FMADM_EXIT_USAGE); 1952 1953 if (argc - optind != 1 && argc - optind != 2) 1954 return (FMADM_EXIT_USAGE); 1955 1956 /* 1957 * argument could be a uuid, an fmri (asru, fru or resource) 1958 * or a label. Or it could be a uuid and an fmri or label. 1959 */ 1960 if (argc - optind == 2) { 1961 err = fmd_adm_rsrc_acquit(adm, argv[optind], argv[optind + 1]); 1962 if (err != 0) 1963 err = fmd_adm_rsrc_acquit(adm, argv[optind + 1], 1964 argv[optind]); 1965 } else { 1966 err = fmd_adm_case_acquit(adm, argv[optind]); 1967 if (err != 0) 1968 err = fmd_adm_rsrc_acquit(adm, argv[optind], ""); 1969 } 1970 1971 if (err != 0) 1972 die("failed to record acquital of %s", argv[optind]); 1973 1974 note("recorded acquital of %s\n", argv[optind]); 1975 return (FMADM_EXIT_SUCCESS); 1976 } 1977