1 /******************************************************************************
2 *******************************************************************************
4 ** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
5 ** Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
7 ** This copyrighted material is made available to anyone wishing to use,
8 ** modify, copy, or redistribute it subject to the terms and conditions
9 ** of the GNU General Public License v.2.
11 *******************************************************************************
12 ******************************************************************************/
14 #include "dlm_internal.h"
15 #include "lockspace.h"
25 #include "requestqueue.h"
29 static struct mutex ls_lock;
30 static struct list_head lslist;
31 static spinlock_t lslist_lock;
32 static struct task_struct * scand_task;
35 static ssize_t dlm_control_store(struct dlm_ls *ls, const char *buf, size_t len)
38 int n = simple_strtol(buf, NULL, 0);
40 ls = dlm_find_lockspace_local(ls->ls_local_handle);
54 dlm_put_lockspace(ls);
58 static ssize_t dlm_event_store(struct dlm_ls *ls, const char *buf, size_t len)
60 ls->ls_uevent_result = simple_strtol(buf, NULL, 0);
61 set_bit(LSFL_UEVENT_WAIT, &ls->ls_flags);
62 wake_up(&ls->ls_uevent_wait);
66 static ssize_t dlm_id_show(struct dlm_ls *ls, char *buf)
68 return snprintf(buf, PAGE_SIZE, "%u\n", ls->ls_global_id);
71 static ssize_t dlm_id_store(struct dlm_ls *ls, const char *buf, size_t len)
73 ls->ls_global_id = simple_strtoul(buf, NULL, 0);
77 static ssize_t dlm_recover_status_show(struct dlm_ls *ls, char *buf)
79 uint32_t status = dlm_recover_status(ls);
80 return snprintf(buf, PAGE_SIZE, "%x\n", status);
83 static ssize_t dlm_recover_nodeid_show(struct dlm_ls *ls, char *buf)
85 return snprintf(buf, PAGE_SIZE, "%d\n", ls->ls_recover_nodeid);
89 struct attribute attr;
90 ssize_t (*show)(struct dlm_ls *, char *);
91 ssize_t (*store)(struct dlm_ls *, const char *, size_t);
94 static struct dlm_attr dlm_attr_control = {
95 .attr = {.name = "control", .mode = S_IWUSR},
96 .store = dlm_control_store
99 static struct dlm_attr dlm_attr_event = {
100 .attr = {.name = "event_done", .mode = S_IWUSR},
101 .store = dlm_event_store
104 static struct dlm_attr dlm_attr_id = {
105 .attr = {.name = "id", .mode = S_IRUGO | S_IWUSR},
107 .store = dlm_id_store
110 static struct dlm_attr dlm_attr_recover_status = {
111 .attr = {.name = "recover_status", .mode = S_IRUGO},
112 .show = dlm_recover_status_show
115 static struct dlm_attr dlm_attr_recover_nodeid = {
116 .attr = {.name = "recover_nodeid", .mode = S_IRUGO},
117 .show = dlm_recover_nodeid_show
120 static struct attribute *dlm_attrs[] = {
121 &dlm_attr_control.attr,
122 &dlm_attr_event.attr,
124 &dlm_attr_recover_status.attr,
125 &dlm_attr_recover_nodeid.attr,
129 static ssize_t dlm_attr_show(struct kobject *kobj, struct attribute *attr,
132 struct dlm_ls *ls = container_of(kobj, struct dlm_ls, ls_kobj);
133 struct dlm_attr *a = container_of(attr, struct dlm_attr, attr);
134 return a->show ? a->show(ls, buf) : 0;
137 static ssize_t dlm_attr_store(struct kobject *kobj, struct attribute *attr,
138 const char *buf, size_t len)
140 struct dlm_ls *ls = container_of(kobj, struct dlm_ls, ls_kobj);
141 struct dlm_attr *a = container_of(attr, struct dlm_attr, attr);
142 return a->store ? a->store(ls, buf, len) : len;
145 static void lockspace_kobj_release(struct kobject *k)
147 struct dlm_ls *ls = container_of(k, struct dlm_ls, ls_kobj);
151 static struct sysfs_ops dlm_attr_ops = {
152 .show = dlm_attr_show,
153 .store = dlm_attr_store,
156 static struct kobj_type dlm_ktype = {
157 .default_attrs = dlm_attrs,
158 .sysfs_ops = &dlm_attr_ops,
159 .release = lockspace_kobj_release,
162 static struct kset *dlm_kset;
164 static int do_uevent(struct dlm_ls *ls, int in)
169 kobject_uevent(&ls->ls_kobj, KOBJ_ONLINE);
171 kobject_uevent(&ls->ls_kobj, KOBJ_OFFLINE);
173 log_debug(ls, "%s the lockspace group...", in ? "joining" : "leaving");
175 /* dlm_controld will see the uevent, do the necessary group management
176 and then write to sysfs to wake us */
178 error = wait_event_interruptible(ls->ls_uevent_wait,
179 test_and_clear_bit(LSFL_UEVENT_WAIT, &ls->ls_flags));
181 log_debug(ls, "group event done %d %d", error, ls->ls_uevent_result);
186 error = ls->ls_uevent_result;
189 log_error(ls, "group %s failed %d %d", in ? "join" : "leave",
190 error, ls->ls_uevent_result);
195 int __init dlm_lockspace_init(void)
198 mutex_init(&ls_lock);
199 INIT_LIST_HEAD(&lslist);
200 spin_lock_init(&lslist_lock);
202 dlm_kset = kset_create_and_add("dlm", NULL, kernel_kobj);
204 printk(KERN_WARNING "%s: can not create kset\n", __func__);
210 void dlm_lockspace_exit(void)
212 kset_unregister(dlm_kset);
215 static struct dlm_ls *find_ls_to_scan(void)
219 spin_lock(&lslist_lock);
220 list_for_each_entry(ls, &lslist, ls_list) {
221 if (time_after_eq(jiffies, ls->ls_scan_time +
222 dlm_config.ci_scan_secs * HZ)) {
223 spin_unlock(&lslist_lock);
227 spin_unlock(&lslist_lock);
231 static int dlm_scand(void *data)
234 int timeout_jiffies = dlm_config.ci_scan_secs * HZ;
236 while (!kthread_should_stop()) {
237 ls = find_ls_to_scan();
239 if (dlm_lock_recovery_try(ls)) {
240 ls->ls_scan_time = jiffies;
242 dlm_scan_timeout(ls);
243 dlm_unlock_recovery(ls);
245 ls->ls_scan_time += HZ;
248 schedule_timeout_interruptible(timeout_jiffies);
254 static int dlm_scand_start(void)
256 struct task_struct *p;
259 p = kthread_run(dlm_scand, NULL, "dlm_scand");
267 static void dlm_scand_stop(void)
269 kthread_stop(scand_task);
272 struct dlm_ls *dlm_find_lockspace_global(uint32_t id)
276 spin_lock(&lslist_lock);
278 list_for_each_entry(ls, &lslist, ls_list) {
279 if (ls->ls_global_id == id) {
286 spin_unlock(&lslist_lock);
290 struct dlm_ls *dlm_find_lockspace_local(dlm_lockspace_t *lockspace)
294 spin_lock(&lslist_lock);
295 list_for_each_entry(ls, &lslist, ls_list) {
296 if (ls->ls_local_handle == lockspace) {
303 spin_unlock(&lslist_lock);
307 struct dlm_ls *dlm_find_lockspace_device(int minor)
311 spin_lock(&lslist_lock);
312 list_for_each_entry(ls, &lslist, ls_list) {
313 if (ls->ls_device.minor == minor) {
320 spin_unlock(&lslist_lock);
324 void dlm_put_lockspace(struct dlm_ls *ls)
326 spin_lock(&lslist_lock);
328 spin_unlock(&lslist_lock);
331 static void remove_lockspace(struct dlm_ls *ls)
334 spin_lock(&lslist_lock);
335 if (ls->ls_count == 0) {
336 WARN_ON(ls->ls_create_count != 0);
337 list_del(&ls->ls_list);
338 spin_unlock(&lslist_lock);
341 spin_unlock(&lslist_lock);
346 static int threads_start(void)
350 /* Thread which process lock requests for all lockspace's */
351 error = dlm_astd_start();
353 log_print("cannot start dlm_astd thread %d", error);
357 error = dlm_scand_start();
359 log_print("cannot start dlm_scand thread %d", error);
363 /* Thread for sending/receiving messages for all lockspace's */
364 error = dlm_lowcomms_start();
366 log_print("cannot start dlm lowcomms %d", error);
380 static void threads_stop(void)
387 static int new_lockspace(char *name, int namelen, void **lockspace,
388 uint32_t flags, int lvblen)
394 if (namelen > DLM_LOCKSPACE_LEN)
397 if (!lvblen || (lvblen % 8))
400 if (!try_module_get(THIS_MODULE))
403 if (!dlm_user_daemon_available()) {
404 module_put(THIS_MODULE);
410 spin_lock(&lslist_lock);
411 list_for_each_entry(ls, &lslist, ls_list) {
412 WARN_ON(ls->ls_create_count <= 0);
413 if (ls->ls_namelen != namelen)
415 if (memcmp(ls->ls_name, name, namelen))
417 if (flags & DLM_LSFL_NEWEXCL) {
421 ls->ls_create_count++;
422 module_put(THIS_MODULE);
423 error = 1; /* not an error, return 0 */
426 spin_unlock(&lslist_lock);
435 ls = kzalloc(sizeof(struct dlm_ls) + namelen, GFP_KERNEL);
438 memcpy(ls->ls_name, name, namelen);
439 ls->ls_namelen = namelen;
440 ls->ls_lvblen = lvblen;
443 ls->ls_scan_time = jiffies;
445 if (flags & DLM_LSFL_TIMEWARN)
446 set_bit(LSFL_TIMEWARN, &ls->ls_flags);
448 if (flags & DLM_LSFL_FS)
449 ls->ls_allocation = GFP_NOFS;
451 ls->ls_allocation = GFP_KERNEL;
453 /* ls_exflags are forced to match among nodes, and we don't
454 need to require all nodes to have some flags set */
455 ls->ls_exflags = (flags & ~(DLM_LSFL_TIMEWARN | DLM_LSFL_FS |
458 size = dlm_config.ci_rsbtbl_size;
459 ls->ls_rsbtbl_size = size;
461 ls->ls_rsbtbl = kmalloc(sizeof(struct dlm_rsbtable) * size, GFP_KERNEL);
464 for (i = 0; i < size; i++) {
465 INIT_LIST_HEAD(&ls->ls_rsbtbl[i].list);
466 INIT_LIST_HEAD(&ls->ls_rsbtbl[i].toss);
467 spin_lock_init(&ls->ls_rsbtbl[i].lock);
470 size = dlm_config.ci_lkbtbl_size;
471 ls->ls_lkbtbl_size = size;
473 ls->ls_lkbtbl = kmalloc(sizeof(struct dlm_lkbtable) * size, GFP_KERNEL);
476 for (i = 0; i < size; i++) {
477 INIT_LIST_HEAD(&ls->ls_lkbtbl[i].list);
478 rwlock_init(&ls->ls_lkbtbl[i].lock);
479 ls->ls_lkbtbl[i].counter = 1;
482 size = dlm_config.ci_dirtbl_size;
483 ls->ls_dirtbl_size = size;
485 ls->ls_dirtbl = kmalloc(sizeof(struct dlm_dirtable) * size, GFP_KERNEL);
488 for (i = 0; i < size; i++) {
489 INIT_LIST_HEAD(&ls->ls_dirtbl[i].list);
490 spin_lock_init(&ls->ls_dirtbl[i].lock);
493 INIT_LIST_HEAD(&ls->ls_waiters);
494 mutex_init(&ls->ls_waiters_mutex);
495 INIT_LIST_HEAD(&ls->ls_orphans);
496 mutex_init(&ls->ls_orphans_mutex);
497 INIT_LIST_HEAD(&ls->ls_timeout);
498 mutex_init(&ls->ls_timeout_mutex);
500 INIT_LIST_HEAD(&ls->ls_nodes);
501 INIT_LIST_HEAD(&ls->ls_nodes_gone);
502 ls->ls_num_nodes = 0;
503 ls->ls_low_nodeid = 0;
504 ls->ls_total_weight = 0;
505 ls->ls_node_array = NULL;
507 memset(&ls->ls_stub_rsb, 0, sizeof(struct dlm_rsb));
508 ls->ls_stub_rsb.res_ls = ls;
510 ls->ls_debug_rsb_dentry = NULL;
511 ls->ls_debug_waiters_dentry = NULL;
513 init_waitqueue_head(&ls->ls_uevent_wait);
514 ls->ls_uevent_result = 0;
515 init_completion(&ls->ls_members_done);
516 ls->ls_members_result = -1;
518 ls->ls_recoverd_task = NULL;
519 mutex_init(&ls->ls_recoverd_active);
520 spin_lock_init(&ls->ls_recover_lock);
521 spin_lock_init(&ls->ls_rcom_spin);
522 get_random_bytes(&ls->ls_rcom_seq, sizeof(uint64_t));
523 ls->ls_recover_status = 0;
524 ls->ls_recover_seq = 0;
525 ls->ls_recover_args = NULL;
526 init_rwsem(&ls->ls_in_recovery);
527 init_rwsem(&ls->ls_recv_active);
528 INIT_LIST_HEAD(&ls->ls_requestqueue);
529 mutex_init(&ls->ls_requestqueue_mutex);
530 mutex_init(&ls->ls_clear_proc_locks);
532 ls->ls_recover_buf = kmalloc(dlm_config.ci_buffer_size, GFP_KERNEL);
533 if (!ls->ls_recover_buf)
536 INIT_LIST_HEAD(&ls->ls_recover_list);
537 spin_lock_init(&ls->ls_recover_list_lock);
538 ls->ls_recover_list_count = 0;
539 ls->ls_local_handle = ls;
540 init_waitqueue_head(&ls->ls_wait_general);
541 INIT_LIST_HEAD(&ls->ls_root_list);
542 init_rwsem(&ls->ls_root_sem);
544 down_write(&ls->ls_in_recovery);
546 spin_lock(&lslist_lock);
547 ls->ls_create_count = 1;
548 list_add(&ls->ls_list, &lslist);
549 spin_unlock(&lslist_lock);
551 /* needs to find ls in lslist */
552 error = dlm_recoverd_start(ls);
554 log_error(ls, "can't start dlm_recoverd %d", error);
558 ls->ls_kobj.kset = dlm_kset;
559 error = kobject_init_and_add(&ls->ls_kobj, &dlm_ktype, NULL,
563 kobject_uevent(&ls->ls_kobj, KOBJ_ADD);
565 /* let kobject handle freeing of ls if there's an error */
568 /* This uevent triggers dlm_controld in userspace to add us to the
569 group of nodes that are members of this lockspace (managed by the
570 cluster infrastructure.) Once it's done that, it tells us who the
571 current lockspace members are (via configfs) and then tells the
572 lockspace to start running (via sysfs) in dlm_ls_start(). */
574 error = do_uevent(ls, 1);
578 wait_for_completion(&ls->ls_members_done);
579 error = ls->ls_members_result;
583 dlm_create_debug_file(ls);
585 log_debug(ls, "join complete");
592 dlm_clear_members(ls);
593 kfree(ls->ls_node_array);
595 dlm_recoverd_stop(ls);
597 spin_lock(&lslist_lock);
598 list_del(&ls->ls_list);
599 spin_unlock(&lslist_lock);
600 kfree(ls->ls_recover_buf);
602 kfree(ls->ls_dirtbl);
604 kfree(ls->ls_lkbtbl);
606 kfree(ls->ls_rsbtbl);
609 kobject_put(&ls->ls_kobj);
613 module_put(THIS_MODULE);
617 int dlm_new_lockspace(char *name, int namelen, void **lockspace,
618 uint32_t flags, int lvblen)
622 mutex_lock(&ls_lock);
624 error = threads_start();
628 error = new_lockspace(name, namelen, lockspace, flags, lvblen);
634 mutex_unlock(&ls_lock);
638 /* Return 1 if the lockspace still has active remote locks,
639 * 2 if the lockspace still has active local locks.
641 static int lockspace_busy(struct dlm_ls *ls)
643 int i, lkb_found = 0;
646 /* NOTE: We check the lockidtbl here rather than the resource table.
647 This is because there may be LKBs queued as ASTs that have been
648 unlinked from their RSBs and are pending deletion once the AST has
651 for (i = 0; i < ls->ls_lkbtbl_size; i++) {
652 read_lock(&ls->ls_lkbtbl[i].lock);
653 if (!list_empty(&ls->ls_lkbtbl[i].list)) {
655 list_for_each_entry(lkb, &ls->ls_lkbtbl[i].list,
657 if (!lkb->lkb_nodeid) {
658 read_unlock(&ls->ls_lkbtbl[i].lock);
663 read_unlock(&ls->ls_lkbtbl[i].lock);
668 static int release_lockspace(struct dlm_ls *ls, int force)
672 struct list_head *head;
675 busy = lockspace_busy(ls);
677 spin_lock(&lslist_lock);
678 if (ls->ls_create_count == 1) {
682 /* remove_lockspace takes ls off lslist */
683 ls->ls_create_count = 0;
686 } else if (ls->ls_create_count > 1) {
687 rv = --ls->ls_create_count;
691 spin_unlock(&lslist_lock);
694 log_debug(ls, "release_lockspace no remove %d", rv);
698 dlm_device_deregister(ls);
700 if (force < 3 && dlm_user_daemon_available())
703 dlm_recoverd_stop(ls);
705 remove_lockspace(ls);
707 dlm_delete_debug_file(ls);
711 kfree(ls->ls_recover_buf);
714 * Free direntry structs.
718 kfree(ls->ls_dirtbl);
721 * Free all lkb's on lkbtbl[] lists.
724 for (i = 0; i < ls->ls_lkbtbl_size; i++) {
725 head = &ls->ls_lkbtbl[i].list;
726 while (!list_empty(head)) {
727 lkb = list_entry(head->next, struct dlm_lkb,
730 list_del(&lkb->lkb_idtbl_list);
734 if (lkb->lkb_lvbptr && lkb->lkb_flags & DLM_IFL_MSTCPY)
735 dlm_free_lvb(lkb->lkb_lvbptr);
742 kfree(ls->ls_lkbtbl);
745 * Free all rsb's on rsbtbl[] lists
748 for (i = 0; i < ls->ls_rsbtbl_size; i++) {
749 head = &ls->ls_rsbtbl[i].list;
750 while (!list_empty(head)) {
751 rsb = list_entry(head->next, struct dlm_rsb,
754 list_del(&rsb->res_hashchain);
758 head = &ls->ls_rsbtbl[i].toss;
759 while (!list_empty(head)) {
760 rsb = list_entry(head->next, struct dlm_rsb,
762 list_del(&rsb->res_hashchain);
767 kfree(ls->ls_rsbtbl);
770 * Free structures on any other lists
773 dlm_purge_requestqueue(ls);
774 kfree(ls->ls_recover_args);
775 dlm_clear_free_entries(ls);
776 dlm_clear_members(ls);
777 dlm_clear_members_gone(ls);
778 kfree(ls->ls_node_array);
779 log_debug(ls, "release_lockspace final free");
780 kobject_put(&ls->ls_kobj);
781 /* The ls structure will be freed when the kobject is done with */
783 module_put(THIS_MODULE);
788 * Called when a system has released all its locks and is not going to use the
789 * lockspace any longer. We free everything we're managing for this lockspace.
790 * Remaining nodes will go through the recovery process as if we'd died. The
791 * lockspace must continue to function as usual, participating in recoveries,
792 * until this returns.
794 * Force has 4 possible values:
795 * 0 - don't destroy locksapce if it has any LKBs
796 * 1 - destroy lockspace if it has remote LKBs but not if it has local LKBs
797 * 2 - destroy lockspace regardless of LKBs
798 * 3 - destroy lockspace as part of a forced shutdown
801 int dlm_release_lockspace(void *lockspace, int force)
806 ls = dlm_find_lockspace_local(lockspace);
809 dlm_put_lockspace(ls);
811 mutex_lock(&ls_lock);
812 error = release_lockspace(ls, force);
817 mutex_unlock(&ls_lock);
822 void dlm_stop_lockspaces(void)
827 spin_lock(&lslist_lock);
828 list_for_each_entry(ls, &lslist, ls_list) {
829 if (!test_bit(LSFL_RUNNING, &ls->ls_flags))
831 spin_unlock(&lslist_lock);
832 log_error(ls, "no userland control daemon, stopping lockspace");
836 spin_unlock(&lslist_lock);