1 // SPDX-License-Identifier: GPL-2.0-only 2 /****************************************************************************** 3 ******************************************************************************* 4 ** 5 ** Copyright (C) 2005-2011 Red Hat, Inc. All rights reserved. 6 ** 7 ** 8 ******************************************************************************* 9 ******************************************************************************/ 10 11 #include "dlm_internal.h" 12 #include "lockspace.h" 13 #include "member.h" 14 #include "recoverd.h" 15 #include "recover.h" 16 #include "rcom.h" 17 #include "config.h" 18 #include "midcomms.h" 19 #include "lowcomms.h" 20 21 int dlm_slots_version(struct dlm_header *h) 22 { 23 if ((h->h_version & 0x0000FFFF) < DLM_HEADER_SLOTS) 24 return 0; 25 return 1; 26 } 27 28 void dlm_slot_save(struct dlm_ls *ls, struct dlm_rcom *rc, 29 struct dlm_member *memb) 30 { 31 struct rcom_config *rf = (struct rcom_config *)rc->rc_buf; 32 33 if (!dlm_slots_version(&rc->rc_header)) 34 return; 35 36 memb->slot = le16_to_cpu(rf->rf_our_slot); 37 memb->generation = le32_to_cpu(rf->rf_generation); 38 } 39 40 void dlm_slots_copy_out(struct dlm_ls *ls, struct dlm_rcom *rc) 41 { 42 struct dlm_slot *slot; 43 struct rcom_slot *ro; 44 int i; 45 46 ro = (struct rcom_slot *)(rc->rc_buf + sizeof(struct rcom_config)); 47 48 /* ls_slots array is sparse, but not rcom_slots */ 49 50 for (i = 0; i < ls->ls_slots_size; i++) { 51 slot = &ls->ls_slots[i]; 52 if (!slot->nodeid) 53 continue; 54 ro->ro_nodeid = cpu_to_le32(slot->nodeid); 55 ro->ro_slot = cpu_to_le16(slot->slot); 56 ro++; 57 } 58 } 59 60 #define SLOT_DEBUG_LINE 128 61 62 static void log_slots(struct dlm_ls *ls, uint32_t gen, int num_slots, 63 struct rcom_slot *ro0, struct dlm_slot *array, 64 int array_size) 65 { 66 char line[SLOT_DEBUG_LINE]; 67 int len = SLOT_DEBUG_LINE - 1; 68 int pos = 0; 69 int ret, i; 70 71 memset(line, 0, sizeof(line)); 72 73 if (array) { 74 for (i = 0; i < array_size; i++) { 75 if (!array[i].nodeid) 76 continue; 77 78 ret = snprintf(line + pos, len - pos, " %d:%d", 79 array[i].slot, array[i].nodeid); 80 if (ret >= len - pos) 81 break; 82 pos += ret; 83 } 84 } else if (ro0) { 85 for (i = 0; i < num_slots; i++) { 86 ret = snprintf(line + pos, len - pos, " %d:%d", 87 ro0[i].ro_slot, ro0[i].ro_nodeid); 88 if (ret >= len - pos) 89 break; 90 pos += ret; 91 } 92 } 93 94 log_rinfo(ls, "generation %u slots %d%s", gen, num_slots, line); 95 } 96 97 int dlm_slots_copy_in(struct dlm_ls *ls) 98 { 99 struct dlm_member *memb; 100 struct dlm_rcom *rc = ls->ls_recover_buf; 101 struct rcom_config *rf = (struct rcom_config *)rc->rc_buf; 102 struct rcom_slot *ro0, *ro; 103 int our_nodeid = dlm_our_nodeid(); 104 int i, num_slots; 105 uint32_t gen; 106 107 if (!dlm_slots_version(&rc->rc_header)) 108 return -1; 109 110 gen = le32_to_cpu(rf->rf_generation); 111 if (gen <= ls->ls_generation) { 112 log_error(ls, "dlm_slots_copy_in gen %u old %u", 113 gen, ls->ls_generation); 114 } 115 ls->ls_generation = gen; 116 117 num_slots = le16_to_cpu(rf->rf_num_slots); 118 if (!num_slots) 119 return -1; 120 121 ro0 = (struct rcom_slot *)(rc->rc_buf + sizeof(struct rcom_config)); 122 123 for (i = 0, ro = ro0; i < num_slots; i++, ro++) { 124 ro->ro_nodeid = le32_to_cpu(ro->ro_nodeid); 125 ro->ro_slot = le16_to_cpu(ro->ro_slot); 126 } 127 128 log_slots(ls, gen, num_slots, ro0, NULL, 0); 129 130 list_for_each_entry(memb, &ls->ls_nodes, list) { 131 for (i = 0, ro = ro0; i < num_slots; i++, ro++) { 132 if (ro->ro_nodeid != memb->nodeid) 133 continue; 134 memb->slot = ro->ro_slot; 135 memb->slot_prev = memb->slot; 136 break; 137 } 138 139 if (memb->nodeid == our_nodeid) { 140 if (ls->ls_slot && ls->ls_slot != memb->slot) { 141 log_error(ls, "dlm_slots_copy_in our slot " 142 "changed %d %d", ls->ls_slot, 143 memb->slot); 144 return -1; 145 } 146 147 if (!ls->ls_slot) 148 ls->ls_slot = memb->slot; 149 } 150 151 if (!memb->slot) { 152 log_error(ls, "dlm_slots_copy_in nodeid %d no slot", 153 memb->nodeid); 154 return -1; 155 } 156 } 157 158 return 0; 159 } 160 161 /* for any nodes that do not support slots, we will not have set memb->slot 162 in wait_status_all(), so memb->slot will remain -1, and we will not 163 assign slots or set ls_num_slots here */ 164 165 int dlm_slots_assign(struct dlm_ls *ls, int *num_slots, int *slots_size, 166 struct dlm_slot **slots_out, uint32_t *gen_out) 167 { 168 struct dlm_member *memb; 169 struct dlm_slot *array; 170 int our_nodeid = dlm_our_nodeid(); 171 int array_size, max_slots, i; 172 int need = 0; 173 int max = 0; 174 int num = 0; 175 uint32_t gen = 0; 176 177 /* our own memb struct will have slot -1 gen 0 */ 178 179 list_for_each_entry(memb, &ls->ls_nodes, list) { 180 if (memb->nodeid == our_nodeid) { 181 memb->slot = ls->ls_slot; 182 memb->generation = ls->ls_generation; 183 break; 184 } 185 } 186 187 list_for_each_entry(memb, &ls->ls_nodes, list) { 188 if (memb->generation > gen) 189 gen = memb->generation; 190 191 /* node doesn't support slots */ 192 193 if (memb->slot == -1) 194 return -1; 195 196 /* node needs a slot assigned */ 197 198 if (!memb->slot) 199 need++; 200 201 /* node has a slot assigned */ 202 203 num++; 204 205 if (!max || max < memb->slot) 206 max = memb->slot; 207 208 /* sanity check, once slot is assigned it shouldn't change */ 209 210 if (memb->slot_prev && memb->slot && memb->slot_prev != memb->slot) { 211 log_error(ls, "nodeid %d slot changed %d %d", 212 memb->nodeid, memb->slot_prev, memb->slot); 213 return -1; 214 } 215 memb->slot_prev = memb->slot; 216 } 217 218 array_size = max + need; 219 array = kcalloc(array_size, sizeof(*array), GFP_NOFS); 220 if (!array) 221 return -ENOMEM; 222 223 num = 0; 224 225 /* fill in slots (offsets) that are used */ 226 227 list_for_each_entry(memb, &ls->ls_nodes, list) { 228 if (!memb->slot) 229 continue; 230 231 if (memb->slot > array_size) { 232 log_error(ls, "invalid slot number %d", memb->slot); 233 kfree(array); 234 return -1; 235 } 236 237 array[memb->slot - 1].nodeid = memb->nodeid; 238 array[memb->slot - 1].slot = memb->slot; 239 num++; 240 } 241 242 /* assign new slots from unused offsets */ 243 244 list_for_each_entry(memb, &ls->ls_nodes, list) { 245 if (memb->slot) 246 continue; 247 248 for (i = 0; i < array_size; i++) { 249 if (array[i].nodeid) 250 continue; 251 252 memb->slot = i + 1; 253 memb->slot_prev = memb->slot; 254 array[i].nodeid = memb->nodeid; 255 array[i].slot = memb->slot; 256 num++; 257 258 if (!ls->ls_slot && memb->nodeid == our_nodeid) 259 ls->ls_slot = memb->slot; 260 break; 261 } 262 263 if (!memb->slot) { 264 log_error(ls, "no free slot found"); 265 kfree(array); 266 return -1; 267 } 268 } 269 270 gen++; 271 272 log_slots(ls, gen, num, NULL, array, array_size); 273 274 max_slots = (DLM_MAX_APP_BUFSIZE - sizeof(struct dlm_rcom) - 275 sizeof(struct rcom_config)) / sizeof(struct rcom_slot); 276 277 if (num > max_slots) { 278 log_error(ls, "num_slots %d exceeds max_slots %d", 279 num, max_slots); 280 kfree(array); 281 return -1; 282 } 283 284 *gen_out = gen; 285 *slots_out = array; 286 *slots_size = array_size; 287 *num_slots = num; 288 return 0; 289 } 290 291 static void add_ordered_member(struct dlm_ls *ls, struct dlm_member *new) 292 { 293 struct dlm_member *memb = NULL; 294 struct list_head *tmp; 295 struct list_head *newlist = &new->list; 296 struct list_head *head = &ls->ls_nodes; 297 298 list_for_each(tmp, head) { 299 memb = list_entry(tmp, struct dlm_member, list); 300 if (new->nodeid < memb->nodeid) 301 break; 302 } 303 304 if (!memb) 305 list_add_tail(newlist, head); 306 else { 307 /* FIXME: can use list macro here */ 308 newlist->prev = tmp->prev; 309 newlist->next = tmp; 310 tmp->prev->next = newlist; 311 tmp->prev = newlist; 312 } 313 } 314 315 static int dlm_add_member(struct dlm_ls *ls, struct dlm_config_node *node) 316 { 317 struct dlm_member *memb; 318 int error; 319 320 memb = kzalloc(sizeof(*memb), GFP_NOFS); 321 if (!memb) 322 return -ENOMEM; 323 324 error = dlm_lowcomms_connect_node(node->nodeid); 325 if (error < 0) { 326 kfree(memb); 327 return error; 328 } 329 330 memb->nodeid = node->nodeid; 331 memb->weight = node->weight; 332 memb->comm_seq = node->comm_seq; 333 dlm_midcomms_add_member(node->nodeid); 334 add_ordered_member(ls, memb); 335 ls->ls_num_nodes++; 336 return 0; 337 } 338 339 static struct dlm_member *find_memb(struct list_head *head, int nodeid) 340 { 341 struct dlm_member *memb; 342 343 list_for_each_entry(memb, head, list) { 344 if (memb->nodeid == nodeid) 345 return memb; 346 } 347 return NULL; 348 } 349 350 int dlm_is_member(struct dlm_ls *ls, int nodeid) 351 { 352 if (find_memb(&ls->ls_nodes, nodeid)) 353 return 1; 354 return 0; 355 } 356 357 int dlm_is_removed(struct dlm_ls *ls, int nodeid) 358 { 359 if (find_memb(&ls->ls_nodes_gone, nodeid)) 360 return 1; 361 return 0; 362 } 363 364 static void clear_memb_list(struct list_head *head, 365 void (*after_del)(int nodeid)) 366 { 367 struct dlm_member *memb; 368 369 while (!list_empty(head)) { 370 memb = list_entry(head->next, struct dlm_member, list); 371 list_del(&memb->list); 372 if (after_del) 373 after_del(memb->nodeid); 374 kfree(memb); 375 } 376 } 377 378 static void clear_members_cb(int nodeid) 379 { 380 dlm_midcomms_remove_member(nodeid); 381 } 382 383 void dlm_clear_members(struct dlm_ls *ls) 384 { 385 clear_memb_list(&ls->ls_nodes, clear_members_cb); 386 ls->ls_num_nodes = 0; 387 } 388 389 void dlm_clear_members_gone(struct dlm_ls *ls) 390 { 391 clear_memb_list(&ls->ls_nodes_gone, NULL); 392 } 393 394 static void make_member_array(struct dlm_ls *ls) 395 { 396 struct dlm_member *memb; 397 int i, w, x = 0, total = 0, all_zero = 0, *array; 398 399 kfree(ls->ls_node_array); 400 ls->ls_node_array = NULL; 401 402 list_for_each_entry(memb, &ls->ls_nodes, list) { 403 if (memb->weight) 404 total += memb->weight; 405 } 406 407 /* all nodes revert to weight of 1 if all have weight 0 */ 408 409 if (!total) { 410 total = ls->ls_num_nodes; 411 all_zero = 1; 412 } 413 414 ls->ls_total_weight = total; 415 array = kmalloc_array(total, sizeof(*array), GFP_NOFS); 416 if (!array) 417 return; 418 419 list_for_each_entry(memb, &ls->ls_nodes, list) { 420 if (!all_zero && !memb->weight) 421 continue; 422 423 if (all_zero) 424 w = 1; 425 else 426 w = memb->weight; 427 428 DLM_ASSERT(x < total, printk("total %d x %d\n", total, x);); 429 430 for (i = 0; i < w; i++) 431 array[x++] = memb->nodeid; 432 } 433 434 ls->ls_node_array = array; 435 } 436 437 /* send a status request to all members just to establish comms connections */ 438 439 static int ping_members(struct dlm_ls *ls) 440 { 441 struct dlm_member *memb; 442 int error = 0; 443 444 list_for_each_entry(memb, &ls->ls_nodes, list) { 445 error = dlm_recovery_stopped(ls); 446 if (error) { 447 error = -EINTR; 448 break; 449 } 450 error = dlm_rcom_status(ls, memb->nodeid, 0); 451 if (error) 452 break; 453 } 454 if (error) 455 log_rinfo(ls, "ping_members aborted %d last nodeid %d", 456 error, ls->ls_recover_nodeid); 457 return error; 458 } 459 460 static void dlm_lsop_recover_prep(struct dlm_ls *ls) 461 { 462 if (!ls->ls_ops || !ls->ls_ops->recover_prep) 463 return; 464 ls->ls_ops->recover_prep(ls->ls_ops_arg); 465 } 466 467 static void dlm_lsop_recover_slot(struct dlm_ls *ls, struct dlm_member *memb) 468 { 469 struct dlm_slot slot; 470 uint32_t seq; 471 int error; 472 473 if (!ls->ls_ops || !ls->ls_ops->recover_slot) 474 return; 475 476 /* if there is no comms connection with this node 477 or the present comms connection is newer 478 than the one when this member was added, then 479 we consider the node to have failed (versus 480 being removed due to dlm_release_lockspace) */ 481 482 error = dlm_comm_seq(memb->nodeid, &seq); 483 484 if (!error && seq == memb->comm_seq) 485 return; 486 487 slot.nodeid = memb->nodeid; 488 slot.slot = memb->slot; 489 490 ls->ls_ops->recover_slot(ls->ls_ops_arg, &slot); 491 } 492 493 void dlm_lsop_recover_done(struct dlm_ls *ls) 494 { 495 struct dlm_member *memb; 496 struct dlm_slot *slots; 497 int i, num; 498 499 if (!ls->ls_ops || !ls->ls_ops->recover_done) 500 return; 501 502 num = ls->ls_num_nodes; 503 slots = kcalloc(num, sizeof(*slots), GFP_KERNEL); 504 if (!slots) 505 return; 506 507 i = 0; 508 list_for_each_entry(memb, &ls->ls_nodes, list) { 509 if (i == num) { 510 log_error(ls, "dlm_lsop_recover_done bad num %d", num); 511 goto out; 512 } 513 slots[i].nodeid = memb->nodeid; 514 slots[i].slot = memb->slot; 515 i++; 516 } 517 518 ls->ls_ops->recover_done(ls->ls_ops_arg, slots, num, 519 ls->ls_slot, ls->ls_generation); 520 out: 521 kfree(slots); 522 } 523 524 static struct dlm_config_node *find_config_node(struct dlm_recover *rv, 525 int nodeid) 526 { 527 int i; 528 529 for (i = 0; i < rv->nodes_count; i++) { 530 if (rv->nodes[i].nodeid == nodeid) 531 return &rv->nodes[i]; 532 } 533 return NULL; 534 } 535 536 int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out) 537 { 538 struct dlm_member *memb, *safe; 539 struct dlm_config_node *node; 540 int i, error, neg = 0, low = -1; 541 542 /* previously removed members that we've not finished removing need to 543 count as a negative change so the "neg" recovery steps will happen */ 544 545 list_for_each_entry(memb, &ls->ls_nodes_gone, list) { 546 log_rinfo(ls, "prev removed member %d", memb->nodeid); 547 neg++; 548 } 549 550 /* move departed members from ls_nodes to ls_nodes_gone */ 551 552 list_for_each_entry_safe(memb, safe, &ls->ls_nodes, list) { 553 node = find_config_node(rv, memb->nodeid); 554 if (node && !node->new) 555 continue; 556 557 if (!node) { 558 log_rinfo(ls, "remove member %d", memb->nodeid); 559 } else { 560 /* removed and re-added */ 561 log_rinfo(ls, "remove member %d comm_seq %u %u", 562 memb->nodeid, memb->comm_seq, node->comm_seq); 563 } 564 565 neg++; 566 list_move(&memb->list, &ls->ls_nodes_gone); 567 dlm_midcomms_remove_member(memb->nodeid); 568 ls->ls_num_nodes--; 569 dlm_lsop_recover_slot(ls, memb); 570 } 571 572 /* add new members to ls_nodes */ 573 574 for (i = 0; i < rv->nodes_count; i++) { 575 node = &rv->nodes[i]; 576 if (dlm_is_member(ls, node->nodeid)) 577 continue; 578 dlm_add_member(ls, node); 579 log_rinfo(ls, "add member %d", node->nodeid); 580 } 581 582 list_for_each_entry(memb, &ls->ls_nodes, list) { 583 if (low == -1 || memb->nodeid < low) 584 low = memb->nodeid; 585 } 586 ls->ls_low_nodeid = low; 587 588 make_member_array(ls); 589 *neg_out = neg; 590 591 error = ping_members(ls); 592 /* error -EINTR means that a new recovery action is triggered. 593 * We ignore this recovery action and let run the new one which might 594 * have new member configuration. 595 */ 596 if (error == -EINTR) 597 error = 0; 598 599 /* new_lockspace() may be waiting to know if the config 600 * is good or bad 601 */ 602 ls->ls_members_result = error; 603 complete(&ls->ls_members_done); 604 605 log_rinfo(ls, "dlm_recover_members %d nodes", ls->ls_num_nodes); 606 return error; 607 } 608 609 /* Userspace guarantees that dlm_ls_stop() has completed on all nodes before 610 dlm_ls_start() is called on any of them to start the new recovery. */ 611 612 int dlm_ls_stop(struct dlm_ls *ls) 613 { 614 int new; 615 616 /* 617 * Prevent dlm_recv from being in the middle of something when we do 618 * the stop. This includes ensuring dlm_recv isn't processing a 619 * recovery message (rcom), while dlm_recoverd is aborting and 620 * resetting things from an in-progress recovery. i.e. we want 621 * dlm_recoverd to abort its recovery without worrying about dlm_recv 622 * processing an rcom at the same time. Stopping dlm_recv also makes 623 * it easy for dlm_receive_message() to check locking stopped and add a 624 * message to the requestqueue without races. 625 */ 626 627 down_write(&ls->ls_recv_active); 628 629 /* 630 * Abort any recovery that's in progress (see RECOVER_STOP, 631 * dlm_recovery_stopped()) and tell any other threads running in the 632 * dlm to quit any processing (see RUNNING, dlm_locking_stopped()). 633 */ 634 635 spin_lock(&ls->ls_recover_lock); 636 set_bit(LSFL_RECOVER_STOP, &ls->ls_flags); 637 new = test_and_clear_bit(LSFL_RUNNING, &ls->ls_flags); 638 ls->ls_recover_seq++; 639 spin_unlock(&ls->ls_recover_lock); 640 641 /* 642 * Let dlm_recv run again, now any normal messages will be saved on the 643 * requestqueue for later. 644 */ 645 646 up_write(&ls->ls_recv_active); 647 648 /* 649 * This in_recovery lock does two things: 650 * 1) Keeps this function from returning until all threads are out 651 * of locking routines and locking is truly stopped. 652 * 2) Keeps any new requests from being processed until it's unlocked 653 * when recovery is complete. 654 */ 655 656 if (new) { 657 set_bit(LSFL_RECOVER_DOWN, &ls->ls_flags); 658 wake_up_process(ls->ls_recoverd_task); 659 wait_event(ls->ls_recover_lock_wait, 660 test_bit(LSFL_RECOVER_LOCK, &ls->ls_flags)); 661 } 662 663 /* 664 * The recoverd suspend/resume makes sure that dlm_recoverd (if 665 * running) has noticed RECOVER_STOP above and quit processing the 666 * previous recovery. 667 */ 668 669 dlm_recoverd_suspend(ls); 670 671 spin_lock(&ls->ls_recover_lock); 672 kfree(ls->ls_slots); 673 ls->ls_slots = NULL; 674 ls->ls_num_slots = 0; 675 ls->ls_slots_size = 0; 676 ls->ls_recover_status = 0; 677 spin_unlock(&ls->ls_recover_lock); 678 679 dlm_recoverd_resume(ls); 680 681 if (!ls->ls_recover_begin) 682 ls->ls_recover_begin = jiffies; 683 684 dlm_lsop_recover_prep(ls); 685 return 0; 686 } 687 688 int dlm_ls_start(struct dlm_ls *ls) 689 { 690 struct dlm_recover *rv, *rv_old; 691 struct dlm_config_node *nodes = NULL; 692 int error, count; 693 694 rv = kzalloc(sizeof(*rv), GFP_NOFS); 695 if (!rv) 696 return -ENOMEM; 697 698 error = dlm_config_nodes(ls->ls_name, &nodes, &count); 699 if (error < 0) 700 goto fail_rv; 701 702 spin_lock(&ls->ls_recover_lock); 703 704 /* the lockspace needs to be stopped before it can be started */ 705 706 if (!dlm_locking_stopped(ls)) { 707 spin_unlock(&ls->ls_recover_lock); 708 log_error(ls, "start ignored: lockspace running"); 709 error = -EINVAL; 710 goto fail; 711 } 712 713 rv->nodes = nodes; 714 rv->nodes_count = count; 715 rv->seq = ++ls->ls_recover_seq; 716 rv_old = ls->ls_recover_args; 717 ls->ls_recover_args = rv; 718 spin_unlock(&ls->ls_recover_lock); 719 720 if (rv_old) { 721 log_error(ls, "unused recovery %llx %d", 722 (unsigned long long)rv_old->seq, rv_old->nodes_count); 723 kfree(rv_old->nodes); 724 kfree(rv_old); 725 } 726 727 set_bit(LSFL_RECOVER_WORK, &ls->ls_flags); 728 wake_up_process(ls->ls_recoverd_task); 729 return 0; 730 731 fail: 732 kfree(nodes); 733 fail_rv: 734 kfree(rv); 735 return error; 736 } 737 738