]>
Commit | Line | Data |
---|---|---|
8e854e9c GR |
1 | /* |
2 | * Copyright (C) 2015, SUSE | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or modify | |
5 | * it under the terms of the GNU General Public License as published by | |
6 | * the Free Software Foundation; either version 2, or (at your option) | |
7 | * any later version. | |
8 | * | |
9 | */ | |
10 | ||
11 | ||
12 | #include <linux/module.h> | |
7bcda714 | 13 | #include <linux/kthread.h> |
47741b7c GR |
14 | #include <linux/dlm.h> |
15 | #include <linux/sched.h> | |
1aee41f6 | 16 | #include <linux/raid/md_p.h> |
47741b7c | 17 | #include "md.h" |
e94987db | 18 | #include "bitmap.h" |
edb39c9d | 19 | #include "md-cluster.h" |
47741b7c GR |
20 | |
21 | #define LVB_SIZE 64 | |
1aee41f6 | 22 | #define NEW_DEV_TIMEOUT 5000 |
47741b7c GR |
23 | |
24 | struct dlm_lock_resource { | |
25 | dlm_lockspace_t *ls; | |
26 | struct dlm_lksb lksb; | |
27 | char *name; /* lock name. */ | |
28 | uint32_t flags; /* flags to pass to dlm_lock() */ | |
fccb60a4 GJ |
29 | wait_queue_head_t sync_locking; /* wait queue for synchronized locking */ |
30 | bool sync_locking_done; | |
c4ce867f GR |
31 | void (*bast)(void *arg, int mode); /* blocking AST function pointer*/ |
32 | struct mddev *mddev; /* pointing back to mddev. */ | |
dbb64f86 | 33 | int mode; |
c4ce867f GR |
34 | }; |
35 | ||
96ae923a GR |
36 | struct suspend_info { |
37 | int slot; | |
38 | sector_t lo; | |
39 | sector_t hi; | |
40 | struct list_head list; | |
41 | }; | |
42 | ||
43 | struct resync_info { | |
44 | __le64 lo; | |
45 | __le64 hi; | |
46 | }; | |
47 | ||
fa8259da GR |
48 | /* md_cluster_info flags */ |
49 | #define MD_CLUSTER_WAITING_FOR_NEWDISK 1 | |
90382ed9 | 50 | #define MD_CLUSTER_SUSPEND_READ_BALANCING 2 |
eece075c | 51 | #define MD_CLUSTER_BEGIN_JOIN_CLUSTER 3 |
fa8259da | 52 | |
8b9277c8 GJ |
53 | /* Lock the send communication. This is done through |
54 | * bit manipulation as opposed to a mutex in order to | |
55 | * accomodate lock and hold. See next comment. | |
56 | */ | |
57 | #define MD_CLUSTER_SEND_LOCK 4 | |
e19508fa GJ |
58 | /* If cluster operations (such as adding a disk) must lock the |
59 | * communication channel, so as to perform extra operations | |
60 | * (update metadata) and no other operation is allowed on the | |
61 | * MD. Token needs to be locked and held until the operation | |
62 | * completes witha md_update_sb(), which would eventually release | |
63 | * the lock. | |
8b9277c8 GJ |
64 | */ |
65 | #define MD_CLUSTER_SEND_LOCKED_ALREADY 5 | |
51e453ae GJ |
66 | /* We should receive message after node joined cluster and |
67 | * set up all the related infos such as bitmap and personality */ | |
68 | #define MD_CLUSTER_ALREADY_IN_CLUSTER 6 | |
69 | #define MD_CLUSTER_PENDING_RECV_EVENT 7 | |
8b9277c8 | 70 | |
fa8259da | 71 | |
c4ce867f GR |
72 | struct md_cluster_info { |
73 | /* dlm lock space and resources for clustered raid. */ | |
74 | dlm_lockspace_t *lockspace; | |
cf921cc1 GR |
75 | int slot_number; |
76 | struct completion completion; | |
8b9277c8 | 77 | struct mutex recv_mutex; |
54519c5f | 78 | struct dlm_lock_resource *bitmap_lockres; |
f6a2dc64 | 79 | struct dlm_lock_resource **other_bitmap_lockres; |
c186b128 | 80 | struct dlm_lock_resource *resync_lockres; |
96ae923a GR |
81 | struct list_head suspend_list; |
82 | spinlock_t suspend_lock; | |
e94987db GR |
83 | struct md_thread *recovery_thread; |
84 | unsigned long recovery_map; | |
4664680c GR |
85 | /* communication loc resources */ |
86 | struct dlm_lock_resource *ack_lockres; | |
87 | struct dlm_lock_resource *message_lockres; | |
88 | struct dlm_lock_resource *token_lockres; | |
1aee41f6 | 89 | struct dlm_lock_resource *no_new_dev_lockres; |
4664680c | 90 | struct md_thread *recv_thread; |
1aee41f6 | 91 | struct completion newdisk_completion; |
8b9277c8 | 92 | wait_queue_head_t wait; |
fa8259da | 93 | unsigned long state; |
18c9ff7f GJ |
94 | /* record the region in RESYNCING message */ |
95 | sector_t sync_low; | |
96 | sector_t sync_hi; | |
4664680c GR |
97 | }; |
98 | ||
99 | enum msg_type { | |
100 | METADATA_UPDATED = 0, | |
101 | RESYNCING, | |
1aee41f6 | 102 | NEWDISK, |
88bcfef7 | 103 | REMOVE, |
97f6cd39 | 104 | RE_ADD, |
dc737d7c | 105 | BITMAP_NEEDS_SYNC, |
4664680c GR |
106 | }; |
107 | ||
108 | struct cluster_msg { | |
cf97a348 GJ |
109 | __le32 type; |
110 | __le32 slot; | |
1aee41f6 | 111 | /* TODO: Unionize this for smaller footprint */ |
cf97a348 GJ |
112 | __le64 low; |
113 | __le64 high; | |
1aee41f6 | 114 | char uuid[16]; |
cf97a348 | 115 | __le32 raid_slot; |
47741b7c GR |
116 | }; |
117 | ||
118 | static void sync_ast(void *arg) | |
119 | { | |
120 | struct dlm_lock_resource *res; | |
121 | ||
2e2a7cd9 | 122 | res = arg; |
fccb60a4 GJ |
123 | res->sync_locking_done = true; |
124 | wake_up(&res->sync_locking); | |
47741b7c GR |
125 | } |
126 | ||
127 | static int dlm_lock_sync(struct dlm_lock_resource *res, int mode) | |
128 | { | |
129 | int ret = 0; | |
130 | ||
47741b7c GR |
131 | ret = dlm_lock(res->ls, mode, &res->lksb, |
132 | res->flags, res->name, strlen(res->name), | |
133 | 0, sync_ast, res, res->bast); | |
134 | if (ret) | |
135 | return ret; | |
fccb60a4 GJ |
136 | wait_event(res->sync_locking, res->sync_locking_done); |
137 | res->sync_locking_done = false; | |
dbb64f86 GR |
138 | if (res->lksb.sb_status == 0) |
139 | res->mode = mode; | |
47741b7c GR |
140 | return res->lksb.sb_status; |
141 | } | |
142 | ||
143 | static int dlm_unlock_sync(struct dlm_lock_resource *res) | |
144 | { | |
145 | return dlm_lock_sync(res, DLM_LOCK_NL); | |
146 | } | |
147 | ||
7bcda714 GJ |
148 | /* |
149 | * An variation of dlm_lock_sync, which make lock request could | |
150 | * be interrupted | |
151 | */ | |
152 | static int dlm_lock_sync_interruptible(struct dlm_lock_resource *res, int mode, | |
153 | struct mddev *mddev) | |
154 | { | |
155 | int ret = 0; | |
156 | ||
157 | ret = dlm_lock(res->ls, mode, &res->lksb, | |
158 | res->flags, res->name, strlen(res->name), | |
159 | 0, sync_ast, res, res->bast); | |
160 | if (ret) | |
161 | return ret; | |
162 | ||
163 | wait_event(res->sync_locking, res->sync_locking_done | |
d6385db9 GJ |
164 | || kthread_should_stop() |
165 | || test_bit(MD_CLOSING, &mddev->flags)); | |
7bcda714 GJ |
166 | if (!res->sync_locking_done) { |
167 | /* | |
168 | * the convert queue contains the lock request when request is | |
169 | * interrupted, and sync_ast could still be run, so need to | |
170 | * cancel the request and reset completion | |
171 | */ | |
172 | ret = dlm_unlock(res->ls, res->lksb.sb_lkid, DLM_LKF_CANCEL, | |
173 | &res->lksb, res); | |
174 | res->sync_locking_done = false; | |
175 | if (unlikely(ret != 0)) | |
176 | pr_info("failed to cancel previous lock request " | |
177 | "%s return %d\n", res->name, ret); | |
178 | return -EPERM; | |
179 | } else | |
180 | res->sync_locking_done = false; | |
181 | if (res->lksb.sb_status == 0) | |
182 | res->mode = mode; | |
183 | return res->lksb.sb_status; | |
184 | } | |
185 | ||
c4ce867f | 186 | static struct dlm_lock_resource *lockres_init(struct mddev *mddev, |
47741b7c GR |
187 | char *name, void (*bastfn)(void *arg, int mode), int with_lvb) |
188 | { | |
189 | struct dlm_lock_resource *res = NULL; | |
190 | int ret, namelen; | |
c4ce867f | 191 | struct md_cluster_info *cinfo = mddev->cluster_info; |
47741b7c GR |
192 | |
193 | res = kzalloc(sizeof(struct dlm_lock_resource), GFP_KERNEL); | |
194 | if (!res) | |
195 | return NULL; | |
fccb60a4 GJ |
196 | init_waitqueue_head(&res->sync_locking); |
197 | res->sync_locking_done = false; | |
c4ce867f GR |
198 | res->ls = cinfo->lockspace; |
199 | res->mddev = mddev; | |
dbb64f86 | 200 | res->mode = DLM_LOCK_IV; |
47741b7c GR |
201 | namelen = strlen(name); |
202 | res->name = kzalloc(namelen + 1, GFP_KERNEL); | |
203 | if (!res->name) { | |
204 | pr_err("md-cluster: Unable to allocate resource name for resource %s\n", name); | |
205 | goto out_err; | |
206 | } | |
207 | strlcpy(res->name, name, namelen + 1); | |
208 | if (with_lvb) { | |
209 | res->lksb.sb_lvbptr = kzalloc(LVB_SIZE, GFP_KERNEL); | |
210 | if (!res->lksb.sb_lvbptr) { | |
211 | pr_err("md-cluster: Unable to allocate LVB for resource %s\n", name); | |
212 | goto out_err; | |
213 | } | |
214 | res->flags = DLM_LKF_VALBLK; | |
215 | } | |
216 | ||
217 | if (bastfn) | |
218 | res->bast = bastfn; | |
219 | ||
220 | res->flags |= DLM_LKF_EXPEDITE; | |
221 | ||
222 | ret = dlm_lock_sync(res, DLM_LOCK_NL); | |
223 | if (ret) { | |
224 | pr_err("md-cluster: Unable to lock NL on new lock resource %s\n", name); | |
225 | goto out_err; | |
226 | } | |
227 | res->flags &= ~DLM_LKF_EXPEDITE; | |
228 | res->flags |= DLM_LKF_CONVERT; | |
229 | ||
230 | return res; | |
231 | out_err: | |
232 | kfree(res->lksb.sb_lvbptr); | |
233 | kfree(res->name); | |
234 | kfree(res); | |
235 | return NULL; | |
236 | } | |
237 | ||
238 | static void lockres_free(struct dlm_lock_resource *res) | |
239 | { | |
400cb454 | 240 | int ret = 0; |
b5ef5678 | 241 | |
47741b7c GR |
242 | if (!res) |
243 | return; | |
244 | ||
400cb454 GJ |
245 | /* |
246 | * use FORCEUNLOCK flag, so we can unlock even the lock is on the | |
247 | * waiting or convert queue | |
248 | */ | |
249 | ret = dlm_unlock(res->ls, res->lksb.sb_lkid, DLM_LKF_FORCEUNLOCK, | |
250 | &res->lksb, res); | |
251 | if (unlikely(ret != 0)) | |
252 | pr_err("failed to unlock %s return %d\n", res->name, ret); | |
253 | else | |
fccb60a4 | 254 | wait_event(res->sync_locking, res->sync_locking_done); |
47741b7c GR |
255 | |
256 | kfree(res->name); | |
257 | kfree(res->lksb.sb_lvbptr); | |
258 | kfree(res); | |
259 | } | |
8e854e9c | 260 | |
30661b49 N |
261 | static void add_resync_info(struct dlm_lock_resource *lockres, |
262 | sector_t lo, sector_t hi) | |
96ae923a GR |
263 | { |
264 | struct resync_info *ri; | |
265 | ||
266 | ri = (struct resync_info *)lockres->lksb.sb_lvbptr; | |
267 | ri->lo = cpu_to_le64(lo); | |
268 | ri->hi = cpu_to_le64(hi); | |
269 | } | |
270 | ||
271 | static struct suspend_info *read_resync_info(struct mddev *mddev, struct dlm_lock_resource *lockres) | |
272 | { | |
273 | struct resync_info ri; | |
274 | struct suspend_info *s = NULL; | |
275 | sector_t hi = 0; | |
276 | ||
277 | dlm_lock_sync(lockres, DLM_LOCK_CR); | |
278 | memcpy(&ri, lockres->lksb.sb_lvbptr, sizeof(struct resync_info)); | |
279 | hi = le64_to_cpu(ri.hi); | |
cf97a348 | 280 | if (hi > 0) { |
96ae923a GR |
281 | s = kzalloc(sizeof(struct suspend_info), GFP_KERNEL); |
282 | if (!s) | |
283 | goto out; | |
284 | s->hi = hi; | |
285 | s->lo = le64_to_cpu(ri.lo); | |
286 | } | |
287 | dlm_unlock_sync(lockres); | |
288 | out: | |
289 | return s; | |
290 | } | |
291 | ||
6dc69c9c | 292 | static void recover_bitmaps(struct md_thread *thread) |
e94987db GR |
293 | { |
294 | struct mddev *mddev = thread->mddev; | |
295 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
296 | struct dlm_lock_resource *bm_lockres; | |
297 | char str[64]; | |
298 | int slot, ret; | |
299 | struct suspend_info *s, *tmp; | |
300 | sector_t lo, hi; | |
301 | ||
302 | while (cinfo->recovery_map) { | |
303 | slot = fls64((u64)cinfo->recovery_map) - 1; | |
304 | ||
305 | /* Clear suspend_area associated with the bitmap */ | |
306 | spin_lock_irq(&cinfo->suspend_lock); | |
307 | list_for_each_entry_safe(s, tmp, &cinfo->suspend_list, list) | |
308 | if (slot == s->slot) { | |
309 | list_del(&s->list); | |
310 | kfree(s); | |
311 | } | |
312 | spin_unlock_irq(&cinfo->suspend_lock); | |
313 | ||
314 | snprintf(str, 64, "bitmap%04d", slot); | |
315 | bm_lockres = lockres_init(mddev, str, NULL, 1); | |
316 | if (!bm_lockres) { | |
317 | pr_err("md-cluster: Cannot initialize bitmaps\n"); | |
318 | goto clear_bit; | |
319 | } | |
320 | ||
7bcda714 | 321 | ret = dlm_lock_sync_interruptible(bm_lockres, DLM_LOCK_PW, mddev); |
e94987db GR |
322 | if (ret) { |
323 | pr_err("md-cluster: Could not DLM lock %s: %d\n", | |
324 | str, ret); | |
325 | goto clear_bit; | |
326 | } | |
97f6cd39 | 327 | ret = bitmap_copy_from_slot(mddev, slot, &lo, &hi, true); |
4b26a08a | 328 | if (ret) { |
e94987db | 329 | pr_err("md-cluster: Could not copy data from bitmap %d\n", slot); |
e3f924d3 | 330 | goto clear_bit; |
4b26a08a GR |
331 | } |
332 | if (hi > 0) { | |
4b26a08a GR |
333 | if (lo < mddev->recovery_cp) |
334 | mddev->recovery_cp = lo; | |
eb315cd0 GJ |
335 | /* wake up thread to continue resync in case resync |
336 | * is not finished */ | |
337 | if (mddev->recovery_cp != MaxSector) { | |
338 | set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); | |
339 | md_wakeup_thread(mddev->thread); | |
340 | } | |
4b26a08a | 341 | } |
e94987db | 342 | clear_bit: |
4ac7a65f | 343 | lockres_free(bm_lockres); |
e94987db GR |
344 | clear_bit(slot, &cinfo->recovery_map); |
345 | } | |
346 | } | |
347 | ||
cf921cc1 GR |
348 | static void recover_prep(void *arg) |
349 | { | |
90382ed9 GR |
350 | struct mddev *mddev = arg; |
351 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
352 | set_bit(MD_CLUSTER_SUSPEND_READ_BALANCING, &cinfo->state); | |
cf921cc1 GR |
353 | } |
354 | ||
05cd0e51 | 355 | static void __recover_slot(struct mddev *mddev, int slot) |
cf921cc1 | 356 | { |
cf921cc1 GR |
357 | struct md_cluster_info *cinfo = mddev->cluster_info; |
358 | ||
05cd0e51 | 359 | set_bit(slot, &cinfo->recovery_map); |
e94987db GR |
360 | if (!cinfo->recovery_thread) { |
361 | cinfo->recovery_thread = md_register_thread(recover_bitmaps, | |
362 | mddev, "recover"); | |
363 | if (!cinfo->recovery_thread) { | |
364 | pr_warn("md-cluster: Could not create recovery thread\n"); | |
365 | return; | |
366 | } | |
367 | } | |
368 | md_wakeup_thread(cinfo->recovery_thread); | |
cf921cc1 GR |
369 | } |
370 | ||
05cd0e51 GJ |
371 | static void recover_slot(void *arg, struct dlm_slot *slot) |
372 | { | |
373 | struct mddev *mddev = arg; | |
374 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
375 | ||
376 | pr_info("md-cluster: %s Node %d/%d down. My slot: %d. Initiating recovery.\n", | |
377 | mddev->bitmap_info.cluster_name, | |
378 | slot->nodeid, slot->slot, | |
379 | cinfo->slot_number); | |
380 | /* deduct one since dlm slot starts from one while the num of | |
381 | * cluster-md begins with 0 */ | |
382 | __recover_slot(mddev, slot->slot - 1); | |
383 | } | |
384 | ||
cf921cc1 GR |
385 | static void recover_done(void *arg, struct dlm_slot *slots, |
386 | int num_slots, int our_slot, | |
387 | uint32_t generation) | |
388 | { | |
389 | struct mddev *mddev = arg; | |
390 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
391 | ||
392 | cinfo->slot_number = our_slot; | |
eece075c GJ |
393 | /* completion is only need to be complete when node join cluster, |
394 | * it doesn't need to run during another node's failure */ | |
395 | if (test_bit(MD_CLUSTER_BEGIN_JOIN_CLUSTER, &cinfo->state)) { | |
396 | complete(&cinfo->completion); | |
397 | clear_bit(MD_CLUSTER_BEGIN_JOIN_CLUSTER, &cinfo->state); | |
398 | } | |
90382ed9 | 399 | clear_bit(MD_CLUSTER_SUSPEND_READ_BALANCING, &cinfo->state); |
cf921cc1 GR |
400 | } |
401 | ||
eece075c GJ |
402 | /* the ops is called when node join the cluster, and do lock recovery |
403 | * if node failure occurs */ | |
cf921cc1 GR |
404 | static const struct dlm_lockspace_ops md_ls_ops = { |
405 | .recover_prep = recover_prep, | |
406 | .recover_slot = recover_slot, | |
407 | .recover_done = recover_done, | |
408 | }; | |
409 | ||
4664680c GR |
410 | /* |
411 | * The BAST function for the ack lock resource | |
412 | * This function wakes up the receive thread in | |
413 | * order to receive and process the message. | |
414 | */ | |
415 | static void ack_bast(void *arg, int mode) | |
416 | { | |
2e2a7cd9 | 417 | struct dlm_lock_resource *res = arg; |
4664680c GR |
418 | struct md_cluster_info *cinfo = res->mddev->cluster_info; |
419 | ||
51e453ae GJ |
420 | if (mode == DLM_LOCK_EX) { |
421 | if (test_bit(MD_CLUSTER_ALREADY_IN_CLUSTER, &cinfo->state)) | |
422 | md_wakeup_thread(cinfo->recv_thread); | |
423 | else | |
424 | set_bit(MD_CLUSTER_PENDING_RECV_EVENT, &cinfo->state); | |
425 | } | |
4664680c GR |
426 | } |
427 | ||
e59721cc GR |
428 | static void __remove_suspend_info(struct md_cluster_info *cinfo, int slot) |
429 | { | |
430 | struct suspend_info *s, *tmp; | |
431 | ||
432 | list_for_each_entry_safe(s, tmp, &cinfo->suspend_list, list) | |
433 | if (slot == s->slot) { | |
e59721cc GR |
434 | list_del(&s->list); |
435 | kfree(s); | |
436 | break; | |
437 | } | |
438 | } | |
439 | ||
b8ca846e | 440 | static void remove_suspend_info(struct mddev *mddev, int slot) |
e59721cc | 441 | { |
b8ca846e | 442 | struct md_cluster_info *cinfo = mddev->cluster_info; |
e59721cc GR |
443 | spin_lock_irq(&cinfo->suspend_lock); |
444 | __remove_suspend_info(cinfo, slot); | |
445 | spin_unlock_irq(&cinfo->suspend_lock); | |
b8ca846e | 446 | mddev->pers->quiesce(mddev, 2); |
e59721cc GR |
447 | } |
448 | ||
449 | ||
9ed38ff5 | 450 | static void process_suspend_info(struct mddev *mddev, |
e59721cc GR |
451 | int slot, sector_t lo, sector_t hi) |
452 | { | |
9ed38ff5 | 453 | struct md_cluster_info *cinfo = mddev->cluster_info; |
e59721cc GR |
454 | struct suspend_info *s; |
455 | ||
456 | if (!hi) { | |
b8ca846e | 457 | remove_suspend_info(mddev, slot); |
c186b128 GR |
458 | set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); |
459 | md_wakeup_thread(mddev->thread); | |
e59721cc GR |
460 | return; |
461 | } | |
18c9ff7f GJ |
462 | |
463 | /* | |
464 | * The bitmaps are not same for different nodes | |
465 | * if RESYNCING is happening in one node, then | |
466 | * the node which received the RESYNCING message | |
467 | * probably will perform resync with the region | |
468 | * [lo, hi] again, so we could reduce resync time | |
469 | * a lot if we can ensure that the bitmaps among | |
470 | * different nodes are match up well. | |
471 | * | |
472 | * sync_low/hi is used to record the region which | |
473 | * arrived in the previous RESYNCING message, | |
474 | * | |
475 | * Call bitmap_sync_with_cluster to clear | |
476 | * NEEDED_MASK and set RESYNC_MASK since | |
477 | * resync thread is running in another node, | |
478 | * so we don't need to do the resync again | |
479 | * with the same section */ | |
480 | bitmap_sync_with_cluster(mddev, cinfo->sync_low, | |
481 | cinfo->sync_hi, | |
482 | lo, hi); | |
483 | cinfo->sync_low = lo; | |
484 | cinfo->sync_hi = hi; | |
485 | ||
e59721cc GR |
486 | s = kzalloc(sizeof(struct suspend_info), GFP_KERNEL); |
487 | if (!s) | |
488 | return; | |
489 | s->slot = slot; | |
490 | s->lo = lo; | |
491 | s->hi = hi; | |
9ed38ff5 GR |
492 | mddev->pers->quiesce(mddev, 1); |
493 | mddev->pers->quiesce(mddev, 0); | |
e59721cc GR |
494 | spin_lock_irq(&cinfo->suspend_lock); |
495 | /* Remove existing entry (if exists) before adding */ | |
496 | __remove_suspend_info(cinfo, slot); | |
497 | list_add(&s->list, &cinfo->suspend_list); | |
498 | spin_unlock_irq(&cinfo->suspend_lock); | |
b8ca846e | 499 | mddev->pers->quiesce(mddev, 2); |
e59721cc GR |
500 | } |
501 | ||
1aee41f6 GR |
502 | static void process_add_new_disk(struct mddev *mddev, struct cluster_msg *cmsg) |
503 | { | |
504 | char disk_uuid[64]; | |
505 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
506 | char event_name[] = "EVENT=ADD_DEVICE"; | |
507 | char raid_slot[16]; | |
508 | char *envp[] = {event_name, disk_uuid, raid_slot, NULL}; | |
509 | int len; | |
510 | ||
511 | len = snprintf(disk_uuid, 64, "DEVICE_UUID="); | |
b89f704a | 512 | sprintf(disk_uuid + len, "%pU", cmsg->uuid); |
faeff83f | 513 | snprintf(raid_slot, 16, "RAID_DISK=%d", le32_to_cpu(cmsg->raid_slot)); |
1aee41f6 GR |
514 | pr_info("%s:%d Sending kobject change with %s and %s\n", __func__, __LINE__, disk_uuid, raid_slot); |
515 | init_completion(&cinfo->newdisk_completion); | |
fa8259da | 516 | set_bit(MD_CLUSTER_WAITING_FOR_NEWDISK, &cinfo->state); |
1aee41f6 GR |
517 | kobject_uevent_env(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE, envp); |
518 | wait_for_completion_timeout(&cinfo->newdisk_completion, | |
519 | NEW_DEV_TIMEOUT); | |
fa8259da | 520 | clear_bit(MD_CLUSTER_WAITING_FOR_NEWDISK, &cinfo->state); |
1aee41f6 GR |
521 | } |
522 | ||
523 | ||
524 | static void process_metadata_update(struct mddev *mddev, struct cluster_msg *msg) | |
525 | { | |
526 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
15858fa5 GJ |
527 | mddev->good_device_nr = le32_to_cpu(msg->raid_slot); |
528 | set_bit(MD_RELOAD_SB, &mddev->flags); | |
1aee41f6 | 529 | dlm_lock_sync(cinfo->no_new_dev_lockres, DLM_LOCK_CR); |
15858fa5 | 530 | md_wakeup_thread(mddev->thread); |
1aee41f6 GR |
531 | } |
532 | ||
88bcfef7 GR |
533 | static void process_remove_disk(struct mddev *mddev, struct cluster_msg *msg) |
534 | { | |
5f0aa21d | 535 | struct md_rdev *rdev; |
88bcfef7 | 536 | |
5f0aa21d GJ |
537 | rcu_read_lock(); |
538 | rdev = md_find_rdev_nr_rcu(mddev, le32_to_cpu(msg->raid_slot)); | |
659b254f GJ |
539 | if (rdev) { |
540 | set_bit(ClusterRemove, &rdev->flags); | |
541 | set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); | |
542 | md_wakeup_thread(mddev->thread); | |
543 | } | |
88bcfef7 | 544 | else |
faeff83f GJ |
545 | pr_warn("%s: %d Could not find disk(%d) to REMOVE\n", |
546 | __func__, __LINE__, le32_to_cpu(msg->raid_slot)); | |
5f0aa21d | 547 | rcu_read_unlock(); |
88bcfef7 GR |
548 | } |
549 | ||
97f6cd39 GR |
550 | static void process_readd_disk(struct mddev *mddev, struct cluster_msg *msg) |
551 | { | |
5f0aa21d | 552 | struct md_rdev *rdev; |
97f6cd39 | 553 | |
5f0aa21d GJ |
554 | rcu_read_lock(); |
555 | rdev = md_find_rdev_nr_rcu(mddev, le32_to_cpu(msg->raid_slot)); | |
97f6cd39 GR |
556 | if (rdev && test_bit(Faulty, &rdev->flags)) |
557 | clear_bit(Faulty, &rdev->flags); | |
558 | else | |
faeff83f GJ |
559 | pr_warn("%s: %d Could not find disk(%d) which is faulty", |
560 | __func__, __LINE__, le32_to_cpu(msg->raid_slot)); | |
5f0aa21d | 561 | rcu_read_unlock(); |
97f6cd39 GR |
562 | } |
563 | ||
1fa9a1ad | 564 | static int process_recvd_msg(struct mddev *mddev, struct cluster_msg *msg) |
4664680c | 565 | { |
1fa9a1ad GJ |
566 | int ret = 0; |
567 | ||
256f5b24 GJ |
568 | if (WARN(mddev->cluster_info->slot_number - 1 == le32_to_cpu(msg->slot), |
569 | "node %d received it's own msg\n", le32_to_cpu(msg->slot))) | |
1fa9a1ad | 570 | return -1; |
cf97a348 | 571 | switch (le32_to_cpu(msg->type)) { |
4664680c | 572 | case METADATA_UPDATED: |
1aee41f6 | 573 | process_metadata_update(mddev, msg); |
4664680c GR |
574 | break; |
575 | case RESYNCING: | |
cf97a348 GJ |
576 | process_suspend_info(mddev, le32_to_cpu(msg->slot), |
577 | le64_to_cpu(msg->low), | |
578 | le64_to_cpu(msg->high)); | |
4664680c | 579 | break; |
1aee41f6 | 580 | case NEWDISK: |
1aee41f6 | 581 | process_add_new_disk(mddev, msg); |
88bcfef7 GR |
582 | break; |
583 | case REMOVE: | |
88bcfef7 GR |
584 | process_remove_disk(mddev, msg); |
585 | break; | |
97f6cd39 | 586 | case RE_ADD: |
97f6cd39 GR |
587 | process_readd_disk(mddev, msg); |
588 | break; | |
dc737d7c | 589 | case BITMAP_NEEDS_SYNC: |
cf97a348 | 590 | __recover_slot(mddev, le32_to_cpu(msg->slot)); |
dc737d7c | 591 | break; |
88bcfef7 | 592 | default: |
1fa9a1ad | 593 | ret = -1; |
88bcfef7 GR |
594 | pr_warn("%s:%d Received unknown message from %d\n", |
595 | __func__, __LINE__, msg->slot); | |
09dd1af2 | 596 | } |
1fa9a1ad | 597 | return ret; |
4664680c GR |
598 | } |
599 | ||
600 | /* | |
601 | * thread for receiving message | |
602 | */ | |
603 | static void recv_daemon(struct md_thread *thread) | |
604 | { | |
605 | struct md_cluster_info *cinfo = thread->mddev->cluster_info; | |
606 | struct dlm_lock_resource *ack_lockres = cinfo->ack_lockres; | |
607 | struct dlm_lock_resource *message_lockres = cinfo->message_lockres; | |
608 | struct cluster_msg msg; | |
b5ef5678 | 609 | int ret; |
4664680c | 610 | |
8b9277c8 | 611 | mutex_lock(&cinfo->recv_mutex); |
4664680c GR |
612 | /*get CR on Message*/ |
613 | if (dlm_lock_sync(message_lockres, DLM_LOCK_CR)) { | |
614 | pr_err("md/raid1:failed to get CR on MESSAGE\n"); | |
8b9277c8 | 615 | mutex_unlock(&cinfo->recv_mutex); |
4664680c GR |
616 | return; |
617 | } | |
618 | ||
619 | /* read lvb and wake up thread to process this message_lockres */ | |
620 | memcpy(&msg, message_lockres->lksb.sb_lvbptr, sizeof(struct cluster_msg)); | |
1fa9a1ad GJ |
621 | ret = process_recvd_msg(thread->mddev, &msg); |
622 | if (ret) | |
623 | goto out; | |
4664680c GR |
624 | |
625 | /*release CR on ack_lockres*/ | |
b5ef5678 GJ |
626 | ret = dlm_unlock_sync(ack_lockres); |
627 | if (unlikely(ret != 0)) | |
628 | pr_info("unlock ack failed return %d\n", ret); | |
66099bb0 | 629 | /*up-convert to PR on message_lockres*/ |
b5ef5678 GJ |
630 | ret = dlm_lock_sync(message_lockres, DLM_LOCK_PR); |
631 | if (unlikely(ret != 0)) | |
632 | pr_info("lock PR on msg failed return %d\n", ret); | |
4664680c | 633 | /*get CR on ack_lockres again*/ |
b5ef5678 GJ |
634 | ret = dlm_lock_sync(ack_lockres, DLM_LOCK_CR); |
635 | if (unlikely(ret != 0)) | |
636 | pr_info("lock CR on ack failed return %d\n", ret); | |
1fa9a1ad | 637 | out: |
4664680c | 638 | /*release CR on message_lockres*/ |
b5ef5678 GJ |
639 | ret = dlm_unlock_sync(message_lockres); |
640 | if (unlikely(ret != 0)) | |
641 | pr_info("unlock msg failed return %d\n", ret); | |
8b9277c8 | 642 | mutex_unlock(&cinfo->recv_mutex); |
4664680c GR |
643 | } |
644 | ||
8b9277c8 | 645 | /* lock_token() |
601b515c GR |
646 | * Takes the lock on the TOKEN lock resource so no other |
647 | * node can communicate while the operation is underway. | |
648 | */ | |
8b9277c8 | 649 | static int lock_token(struct md_cluster_info *cinfo) |
601b515c GR |
650 | { |
651 | int error; | |
652 | ||
653 | error = dlm_lock_sync(cinfo->token_lockres, DLM_LOCK_EX); | |
654 | if (error) | |
655 | pr_err("md-cluster(%s:%d): failed to get EX on TOKEN (%d)\n", | |
656 | __func__, __LINE__, error); | |
8b9277c8 GJ |
657 | |
658 | /* Lock the receive sequence */ | |
659 | mutex_lock(&cinfo->recv_mutex); | |
601b515c GR |
660 | return error; |
661 | } | |
662 | ||
8b9277c8 GJ |
663 | /* lock_comm() |
664 | * Sets the MD_CLUSTER_SEND_LOCK bit to lock the send channel. | |
665 | */ | |
666 | static int lock_comm(struct md_cluster_info *cinfo) | |
667 | { | |
668 | wait_event(cinfo->wait, | |
669 | !test_and_set_bit(MD_CLUSTER_SEND_LOCK, &cinfo->state)); | |
670 | ||
671 | return lock_token(cinfo); | |
672 | } | |
673 | ||
601b515c GR |
674 | static void unlock_comm(struct md_cluster_info *cinfo) |
675 | { | |
dbb64f86 | 676 | WARN_ON(cinfo->token_lockres->mode != DLM_LOCK_EX); |
8b9277c8 | 677 | mutex_unlock(&cinfo->recv_mutex); |
601b515c | 678 | dlm_unlock_sync(cinfo->token_lockres); |
8b9277c8 GJ |
679 | clear_bit(MD_CLUSTER_SEND_LOCK, &cinfo->state); |
680 | wake_up(&cinfo->wait); | |
601b515c GR |
681 | } |
682 | ||
683 | /* __sendmsg() | |
684 | * This function performs the actual sending of the message. This function is | |
685 | * usually called after performing the encompassing operation | |
686 | * The function: | |
687 | * 1. Grabs the message lockresource in EX mode | |
688 | * 2. Copies the message to the message LVB | |
66099bb0 | 689 | * 3. Downconverts message lockresource to CW |
601b515c GR |
690 | * 4. Upconverts ack lock resource from CR to EX. This forces the BAST on other nodes |
691 | * and the other nodes read the message. The thread will wait here until all other | |
692 | * nodes have released ack lock resource. | |
693 | * 5. Downconvert ack lockresource to CR | |
694 | */ | |
695 | static int __sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg) | |
696 | { | |
697 | int error; | |
698 | int slot = cinfo->slot_number - 1; | |
699 | ||
700 | cmsg->slot = cpu_to_le32(slot); | |
701 | /*get EX on Message*/ | |
702 | error = dlm_lock_sync(cinfo->message_lockres, DLM_LOCK_EX); | |
703 | if (error) { | |
704 | pr_err("md-cluster: failed to get EX on MESSAGE (%d)\n", error); | |
705 | goto failed_message; | |
706 | } | |
707 | ||
708 | memcpy(cinfo->message_lockres->lksb.sb_lvbptr, (void *)cmsg, | |
709 | sizeof(struct cluster_msg)); | |
66099bb0 GJ |
710 | /*down-convert EX to CW on Message*/ |
711 | error = dlm_lock_sync(cinfo->message_lockres, DLM_LOCK_CW); | |
601b515c | 712 | if (error) { |
66099bb0 | 713 | pr_err("md-cluster: failed to convert EX to CW on MESSAGE(%d)\n", |
601b515c | 714 | error); |
66099bb0 | 715 | goto failed_ack; |
601b515c GR |
716 | } |
717 | ||
718 | /*up-convert CR to EX on Ack*/ | |
719 | error = dlm_lock_sync(cinfo->ack_lockres, DLM_LOCK_EX); | |
720 | if (error) { | |
721 | pr_err("md-cluster: failed to convert CR to EX on ACK(%d)\n", | |
722 | error); | |
723 | goto failed_ack; | |
724 | } | |
725 | ||
726 | /*down-convert EX to CR on Ack*/ | |
727 | error = dlm_lock_sync(cinfo->ack_lockres, DLM_LOCK_CR); | |
728 | if (error) { | |
729 | pr_err("md-cluster: failed to convert EX to CR on ACK(%d)\n", | |
730 | error); | |
731 | goto failed_ack; | |
732 | } | |
733 | ||
734 | failed_ack: | |
b5ef5678 GJ |
735 | error = dlm_unlock_sync(cinfo->message_lockres); |
736 | if (unlikely(error != 0)) { | |
737 | pr_err("md-cluster: failed convert to NL on MESSAGE(%d)\n", | |
738 | error); | |
739 | /* in case the message can't be released due to some reason */ | |
740 | goto failed_ack; | |
741 | } | |
601b515c GR |
742 | failed_message: |
743 | return error; | |
744 | } | |
745 | ||
746 | static int sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg) | |
747 | { | |
748 | int ret; | |
749 | ||
750 | lock_comm(cinfo); | |
751 | ret = __sendmsg(cinfo, cmsg); | |
752 | unlock_comm(cinfo); | |
753 | return ret; | |
754 | } | |
755 | ||
96ae923a GR |
756 | static int gather_all_resync_info(struct mddev *mddev, int total_slots) |
757 | { | |
758 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
759 | int i, ret = 0; | |
760 | struct dlm_lock_resource *bm_lockres; | |
761 | struct suspend_info *s; | |
762 | char str[64]; | |
abb9b22a | 763 | sector_t lo, hi; |
96ae923a GR |
764 | |
765 | ||
766 | for (i = 0; i < total_slots; i++) { | |
767 | memset(str, '\0', 64); | |
768 | snprintf(str, 64, "bitmap%04d", i); | |
769 | bm_lockres = lockres_init(mddev, str, NULL, 1); | |
770 | if (!bm_lockres) | |
771 | return -ENOMEM; | |
4ac7a65f SL |
772 | if (i == (cinfo->slot_number - 1)) { |
773 | lockres_free(bm_lockres); | |
96ae923a | 774 | continue; |
4ac7a65f | 775 | } |
96ae923a GR |
776 | |
777 | bm_lockres->flags |= DLM_LKF_NOQUEUE; | |
778 | ret = dlm_lock_sync(bm_lockres, DLM_LOCK_PW); | |
779 | if (ret == -EAGAIN) { | |
96ae923a GR |
780 | s = read_resync_info(mddev, bm_lockres); |
781 | if (s) { | |
782 | pr_info("%s:%d Resync[%llu..%llu] in progress on %d\n", | |
783 | __func__, __LINE__, | |
784 | (unsigned long long) s->lo, | |
785 | (unsigned long long) s->hi, i); | |
786 | spin_lock_irq(&cinfo->suspend_lock); | |
787 | s->slot = i; | |
788 | list_add(&s->list, &cinfo->suspend_list); | |
789 | spin_unlock_irq(&cinfo->suspend_lock); | |
790 | } | |
791 | ret = 0; | |
792 | lockres_free(bm_lockres); | |
793 | continue; | |
794 | } | |
6e6d9f2c GJ |
795 | if (ret) { |
796 | lockres_free(bm_lockres); | |
96ae923a | 797 | goto out; |
6e6d9f2c | 798 | } |
abb9b22a GJ |
799 | |
800 | /* Read the disk bitmap sb and check if it needs recovery */ | |
801 | ret = bitmap_copy_from_slot(mddev, i, &lo, &hi, false); | |
802 | if (ret) { | |
803 | pr_warn("md-cluster: Could not gather bitmaps from slot %d", i); | |
804 | lockres_free(bm_lockres); | |
805 | continue; | |
806 | } | |
807 | if ((hi > 0) && (lo < mddev->recovery_cp)) { | |
808 | set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); | |
809 | mddev->recovery_cp = lo; | |
810 | md_check_recovery(mddev); | |
811 | } | |
812 | ||
96ae923a GR |
813 | lockres_free(bm_lockres); |
814 | } | |
815 | out: | |
816 | return ret; | |
817 | } | |
818 | ||
edb39c9d GR |
819 | static int join(struct mddev *mddev, int nodes) |
820 | { | |
c4ce867f | 821 | struct md_cluster_info *cinfo; |
cf921cc1 | 822 | int ret, ops_rv; |
c4ce867f GR |
823 | char str[64]; |
824 | ||
c4ce867f GR |
825 | cinfo = kzalloc(sizeof(struct md_cluster_info), GFP_KERNEL); |
826 | if (!cinfo) | |
827 | return -ENOMEM; | |
828 | ||
9e3072e3 GJ |
829 | INIT_LIST_HEAD(&cinfo->suspend_list); |
830 | spin_lock_init(&cinfo->suspend_lock); | |
cf921cc1 | 831 | init_completion(&cinfo->completion); |
eece075c | 832 | set_bit(MD_CLUSTER_BEGIN_JOIN_CLUSTER, &cinfo->state); |
8b9277c8 GJ |
833 | init_waitqueue_head(&cinfo->wait); |
834 | mutex_init(&cinfo->recv_mutex); | |
cf921cc1 | 835 | |
cf921cc1 GR |
836 | mddev->cluster_info = cinfo; |
837 | ||
c4ce867f | 838 | memset(str, 0, 64); |
b89f704a | 839 | sprintf(str, "%pU", mddev->uuid); |
cf921cc1 GR |
840 | ret = dlm_new_lockspace(str, mddev->bitmap_info.cluster_name, |
841 | DLM_LSFL_FS, LVB_SIZE, | |
842 | &md_ls_ops, mddev, &ops_rv, &cinfo->lockspace); | |
c4ce867f GR |
843 | if (ret) |
844 | goto err; | |
cf921cc1 | 845 | wait_for_completion(&cinfo->completion); |
8c58f02e GJ |
846 | if (nodes < cinfo->slot_number) { |
847 | pr_err("md-cluster: Slot allotted(%d) is greater than available slots(%d).", | |
848 | cinfo->slot_number, nodes); | |
b97e9257 GR |
849 | ret = -ERANGE; |
850 | goto err; | |
851 | } | |
4664680c GR |
852 | /* Initiate the communication resources */ |
853 | ret = -ENOMEM; | |
854 | cinfo->recv_thread = md_register_thread(recv_daemon, mddev, "cluster_recv"); | |
855 | if (!cinfo->recv_thread) { | |
856 | pr_err("md-cluster: cannot allocate memory for recv_thread!\n"); | |
857 | goto err; | |
858 | } | |
859 | cinfo->message_lockres = lockres_init(mddev, "message", NULL, 1); | |
860 | if (!cinfo->message_lockres) | |
861 | goto err; | |
862 | cinfo->token_lockres = lockres_init(mddev, "token", NULL, 0); | |
863 | if (!cinfo->token_lockres) | |
864 | goto err; | |
1aee41f6 GR |
865 | cinfo->no_new_dev_lockres = lockres_init(mddev, "no-new-dev", NULL, 0); |
866 | if (!cinfo->no_new_dev_lockres) | |
867 | goto err; | |
868 | ||
1535212c GJ |
869 | ret = dlm_lock_sync(cinfo->token_lockres, DLM_LOCK_EX); |
870 | if (ret) { | |
871 | ret = -EAGAIN; | |
872 | pr_err("md-cluster: can't join cluster to avoid lock issue\n"); | |
873 | goto err; | |
874 | } | |
875 | cinfo->ack_lockres = lockres_init(mddev, "ack", ack_bast, 0); | |
0f6187db WY |
876 | if (!cinfo->ack_lockres) { |
877 | ret = -ENOMEM; | |
1535212c | 878 | goto err; |
0f6187db | 879 | } |
4664680c GR |
880 | /* get sync CR lock on ACK. */ |
881 | if (dlm_lock_sync(cinfo->ack_lockres, DLM_LOCK_CR)) | |
882 | pr_err("md-cluster: failed to get a sync CR lock on ACK!(%d)\n", | |
883 | ret); | |
1535212c | 884 | dlm_unlock_sync(cinfo->token_lockres); |
1aee41f6 GR |
885 | /* get sync CR lock on no-new-dev. */ |
886 | if (dlm_lock_sync(cinfo->no_new_dev_lockres, DLM_LOCK_CR)) | |
887 | pr_err("md-cluster: failed to get a sync CR lock on no-new-dev!(%d)\n", ret); | |
888 | ||
54519c5f GR |
889 | |
890 | pr_info("md-cluster: Joined cluster %s slot %d\n", str, cinfo->slot_number); | |
891 | snprintf(str, 64, "bitmap%04d", cinfo->slot_number - 1); | |
892 | cinfo->bitmap_lockres = lockres_init(mddev, str, NULL, 1); | |
0f6187db WY |
893 | if (!cinfo->bitmap_lockres) { |
894 | ret = -ENOMEM; | |
54519c5f | 895 | goto err; |
0f6187db | 896 | } |
54519c5f GR |
897 | if (dlm_lock_sync(cinfo->bitmap_lockres, DLM_LOCK_PW)) { |
898 | pr_err("Failed to get bitmap lock\n"); | |
899 | ret = -EINVAL; | |
900 | goto err; | |
901 | } | |
902 | ||
c186b128 | 903 | cinfo->resync_lockres = lockres_init(mddev, "resync", NULL, 0); |
0f6187db WY |
904 | if (!cinfo->resync_lockres) { |
905 | ret = -ENOMEM; | |
c186b128 | 906 | goto err; |
0f6187db | 907 | } |
c186b128 | 908 | |
edb39c9d | 909 | return 0; |
c4ce867f | 910 | err: |
5b0fb33e GJ |
911 | md_unregister_thread(&cinfo->recovery_thread); |
912 | md_unregister_thread(&cinfo->recv_thread); | |
4664680c GR |
913 | lockres_free(cinfo->message_lockres); |
914 | lockres_free(cinfo->token_lockres); | |
915 | lockres_free(cinfo->ack_lockres); | |
1aee41f6 | 916 | lockres_free(cinfo->no_new_dev_lockres); |
c186b128 | 917 | lockres_free(cinfo->resync_lockres); |
96ae923a | 918 | lockres_free(cinfo->bitmap_lockres); |
c4ce867f GR |
919 | if (cinfo->lockspace) |
920 | dlm_release_lockspace(cinfo->lockspace, 2); | |
cf921cc1 | 921 | mddev->cluster_info = NULL; |
c4ce867f | 922 | kfree(cinfo); |
c4ce867f | 923 | return ret; |
edb39c9d GR |
924 | } |
925 | ||
51e453ae GJ |
926 | static void load_bitmaps(struct mddev *mddev, int total_slots) |
927 | { | |
928 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
929 | ||
930 | /* load all the node's bitmap info for resync */ | |
931 | if (gather_all_resync_info(mddev, total_slots)) | |
932 | pr_err("md-cluster: failed to gather all resyn infos\n"); | |
933 | set_bit(MD_CLUSTER_ALREADY_IN_CLUSTER, &cinfo->state); | |
934 | /* wake up recv thread in case something need to be handled */ | |
935 | if (test_and_clear_bit(MD_CLUSTER_PENDING_RECV_EVENT, &cinfo->state)) | |
936 | md_wakeup_thread(cinfo->recv_thread); | |
937 | } | |
938 | ||
09995411 GJ |
939 | static void resync_bitmap(struct mddev *mddev) |
940 | { | |
941 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
942 | struct cluster_msg cmsg = {0}; | |
943 | int err; | |
944 | ||
945 | cmsg.type = cpu_to_le32(BITMAP_NEEDS_SYNC); | |
946 | err = sendmsg(cinfo, &cmsg); | |
947 | if (err) | |
948 | pr_err("%s:%d: failed to send BITMAP_NEEDS_SYNC message (%d)\n", | |
949 | __func__, __LINE__, err); | |
950 | } | |
951 | ||
f6a2dc64 | 952 | static void unlock_all_bitmaps(struct mddev *mddev); |
edb39c9d GR |
953 | static int leave(struct mddev *mddev) |
954 | { | |
c4ce867f GR |
955 | struct md_cluster_info *cinfo = mddev->cluster_info; |
956 | ||
957 | if (!cinfo) | |
958 | return 0; | |
09995411 GJ |
959 | |
960 | /* BITMAP_NEEDS_SYNC message should be sent when node | |
961 | * is leaving the cluster with dirty bitmap, also we | |
962 | * can only deliver it when dlm connection is available */ | |
963 | if (cinfo->slot_number > 0 && mddev->recovery_cp != MaxSector) | |
964 | resync_bitmap(mddev); | |
965 | ||
e94987db | 966 | md_unregister_thread(&cinfo->recovery_thread); |
4664680c GR |
967 | md_unregister_thread(&cinfo->recv_thread); |
968 | lockres_free(cinfo->message_lockres); | |
969 | lockres_free(cinfo->token_lockres); | |
970 | lockres_free(cinfo->ack_lockres); | |
1aee41f6 | 971 | lockres_free(cinfo->no_new_dev_lockres); |
4ac7a65f | 972 | lockres_free(cinfo->resync_lockres); |
54519c5f | 973 | lockres_free(cinfo->bitmap_lockres); |
f6a2dc64 | 974 | unlock_all_bitmaps(mddev); |
c4ce867f | 975 | dlm_release_lockspace(cinfo->lockspace, 2); |
9c8043f3 | 976 | kfree(cinfo); |
edb39c9d GR |
977 | return 0; |
978 | } | |
979 | ||
cf921cc1 GR |
980 | /* slot_number(): Returns the MD slot number to use |
981 | * DLM starts the slot numbers from 1, wheras cluster-md | |
982 | * wants the number to be from zero, so we deduct one | |
983 | */ | |
984 | static int slot_number(struct mddev *mddev) | |
985 | { | |
986 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
987 | ||
988 | return cinfo->slot_number - 1; | |
989 | } | |
990 | ||
8b9277c8 GJ |
991 | /* |
992 | * Check if the communication is already locked, else lock the communication | |
993 | * channel. | |
994 | * If it is already locked, token is in EX mode, and hence lock_token() | |
995 | * should not be called. | |
996 | */ | |
293467aa GR |
997 | static int metadata_update_start(struct mddev *mddev) |
998 | { | |
8b9277c8 GJ |
999 | struct md_cluster_info *cinfo = mddev->cluster_info; |
1000 | ||
1001 | wait_event(cinfo->wait, | |
1002 | !test_and_set_bit(MD_CLUSTER_SEND_LOCK, &cinfo->state) || | |
1003 | test_and_clear_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state)); | |
1004 | ||
1005 | /* If token is already locked, return 0 */ | |
1006 | if (cinfo->token_lockres->mode == DLM_LOCK_EX) | |
1007 | return 0; | |
1008 | ||
1009 | return lock_token(cinfo); | |
293467aa GR |
1010 | } |
1011 | ||
1012 | static int metadata_update_finish(struct mddev *mddev) | |
1013 | { | |
1014 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1015 | struct cluster_msg cmsg; | |
70bcecdb GR |
1016 | struct md_rdev *rdev; |
1017 | int ret = 0; | |
ba2746b0 | 1018 | int raid_slot = -1; |
293467aa GR |
1019 | |
1020 | memset(&cmsg, 0, sizeof(cmsg)); | |
1021 | cmsg.type = cpu_to_le32(METADATA_UPDATED); | |
70bcecdb GR |
1022 | /* Pick up a good active device number to send. |
1023 | */ | |
1024 | rdev_for_each(rdev, mddev) | |
1025 | if (rdev->raid_disk > -1 && !test_bit(Faulty, &rdev->flags)) { | |
ba2746b0 | 1026 | raid_slot = rdev->desc_nr; |
70bcecdb GR |
1027 | break; |
1028 | } | |
ba2746b0 N |
1029 | if (raid_slot >= 0) { |
1030 | cmsg.raid_slot = cpu_to_le32(raid_slot); | |
70bcecdb | 1031 | ret = __sendmsg(cinfo, &cmsg); |
ba2746b0 | 1032 | } else |
70bcecdb | 1033 | pr_warn("md-cluster: No good device id found to send\n"); |
8b9277c8 | 1034 | clear_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state); |
293467aa GR |
1035 | unlock_comm(cinfo); |
1036 | return ret; | |
1037 | } | |
1038 | ||
dbb64f86 | 1039 | static void metadata_update_cancel(struct mddev *mddev) |
293467aa GR |
1040 | { |
1041 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
8b9277c8 | 1042 | clear_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state); |
dbb64f86 | 1043 | unlock_comm(cinfo); |
293467aa GR |
1044 | } |
1045 | ||
c186b128 GR |
1046 | static int resync_start(struct mddev *mddev) |
1047 | { | |
1048 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
d6385db9 | 1049 | return dlm_lock_sync_interruptible(cinfo->resync_lockres, DLM_LOCK_EX, mddev); |
c186b128 GR |
1050 | } |
1051 | ||
c40f341f | 1052 | static int resync_info_update(struct mddev *mddev, sector_t lo, sector_t hi) |
965400eb GR |
1053 | { |
1054 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
ac277c6a | 1055 | struct resync_info ri; |
aee177ac | 1056 | struct cluster_msg cmsg = {0}; |
965400eb | 1057 | |
ac277c6a GR |
1058 | /* do not send zero again, if we have sent before */ |
1059 | if (hi == 0) { | |
1060 | memcpy(&ri, cinfo->bitmap_lockres->lksb.sb_lvbptr, sizeof(struct resync_info)); | |
1061 | if (le64_to_cpu(ri.hi) == 0) | |
1062 | return 0; | |
1063 | } | |
1064 | ||
30661b49 | 1065 | add_resync_info(cinfo->bitmap_lockres, lo, hi); |
c40f341f GR |
1066 | /* Re-acquire the lock to refresh LVB */ |
1067 | dlm_lock_sync(cinfo->bitmap_lockres, DLM_LOCK_PW); | |
c40f341f | 1068 | cmsg.type = cpu_to_le32(RESYNCING); |
965400eb GR |
1069 | cmsg.low = cpu_to_le64(lo); |
1070 | cmsg.high = cpu_to_le64(hi); | |
c186b128 | 1071 | |
965400eb GR |
1072 | return sendmsg(cinfo, &cmsg); |
1073 | } | |
1074 | ||
c186b128 GR |
1075 | static int resync_finish(struct mddev *mddev) |
1076 | { | |
1077 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
c186b128 GR |
1078 | dlm_unlock_sync(cinfo->resync_lockres); |
1079 | return resync_info_update(mddev, 0, 0); | |
1080 | } | |
1081 | ||
90382ed9 GR |
1082 | static int area_resyncing(struct mddev *mddev, int direction, |
1083 | sector_t lo, sector_t hi) | |
589a1c49 GR |
1084 | { |
1085 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1086 | int ret = 0; | |
1087 | struct suspend_info *s; | |
1088 | ||
90382ed9 GR |
1089 | if ((direction == READ) && |
1090 | test_bit(MD_CLUSTER_SUSPEND_READ_BALANCING, &cinfo->state)) | |
1091 | return 1; | |
1092 | ||
589a1c49 GR |
1093 | spin_lock_irq(&cinfo->suspend_lock); |
1094 | if (list_empty(&cinfo->suspend_list)) | |
1095 | goto out; | |
1096 | list_for_each_entry(s, &cinfo->suspend_list, list) | |
1097 | if (hi > s->lo && lo < s->hi) { | |
1098 | ret = 1; | |
1099 | break; | |
1100 | } | |
1101 | out: | |
1102 | spin_unlock_irq(&cinfo->suspend_lock); | |
1103 | return ret; | |
1104 | } | |
1105 | ||
dbb64f86 GR |
1106 | /* add_new_disk() - initiates a disk add |
1107 | * However, if this fails before writing md_update_sb(), | |
1108 | * add_new_disk_cancel() must be called to release token lock | |
1109 | */ | |
1110 | static int add_new_disk(struct mddev *mddev, struct md_rdev *rdev) | |
1aee41f6 GR |
1111 | { |
1112 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1113 | struct cluster_msg cmsg; | |
1114 | int ret = 0; | |
1115 | struct mdp_superblock_1 *sb = page_address(rdev->sb_page); | |
1116 | char *uuid = sb->device_uuid; | |
1117 | ||
1118 | memset(&cmsg, 0, sizeof(cmsg)); | |
1119 | cmsg.type = cpu_to_le32(NEWDISK); | |
1120 | memcpy(cmsg.uuid, uuid, 16); | |
faeff83f | 1121 | cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); |
1aee41f6 GR |
1122 | lock_comm(cinfo); |
1123 | ret = __sendmsg(cinfo, &cmsg); | |
1124 | if (ret) | |
1125 | return ret; | |
1126 | cinfo->no_new_dev_lockres->flags |= DLM_LKF_NOQUEUE; | |
1127 | ret = dlm_lock_sync(cinfo->no_new_dev_lockres, DLM_LOCK_EX); | |
1128 | cinfo->no_new_dev_lockres->flags &= ~DLM_LKF_NOQUEUE; | |
1129 | /* Some node does not "see" the device */ | |
1130 | if (ret == -EAGAIN) | |
1131 | ret = -ENOENT; | |
dbb64f86 GR |
1132 | if (ret) |
1133 | unlock_comm(cinfo); | |
8b9277c8 | 1134 | else { |
1aee41f6 | 1135 | dlm_lock_sync(cinfo->no_new_dev_lockres, DLM_LOCK_CR); |
e19508fa GJ |
1136 | /* Since MD_CHANGE_DEVS will be set in add_bound_rdev which |
1137 | * will run soon after add_new_disk, the below path will be | |
1138 | * invoked: | |
1139 | * md_wakeup_thread(mddev->thread) | |
1140 | * -> conf->thread (raid1d) | |
1141 | * -> md_check_recovery -> md_update_sb | |
1142 | * -> metadata_update_start/finish | |
1143 | * MD_CLUSTER_SEND_LOCKED_ALREADY will be cleared eventually. | |
1144 | * | |
1145 | * For other failure cases, metadata_update_cancel and | |
1146 | * add_new_disk_cancel also clear below bit as well. | |
1147 | * */ | |
8b9277c8 GJ |
1148 | set_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state); |
1149 | wake_up(&cinfo->wait); | |
1150 | } | |
1aee41f6 GR |
1151 | return ret; |
1152 | } | |
1153 | ||
dbb64f86 | 1154 | static void add_new_disk_cancel(struct mddev *mddev) |
1aee41f6 | 1155 | { |
dbb64f86 | 1156 | struct md_cluster_info *cinfo = mddev->cluster_info; |
8b9277c8 | 1157 | clear_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state); |
dbb64f86 | 1158 | unlock_comm(cinfo); |
1aee41f6 GR |
1159 | } |
1160 | ||
fa8259da | 1161 | static int new_disk_ack(struct mddev *mddev, bool ack) |
1aee41f6 GR |
1162 | { |
1163 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1164 | ||
fa8259da GR |
1165 | if (!test_bit(MD_CLUSTER_WAITING_FOR_NEWDISK, &cinfo->state)) { |
1166 | pr_warn("md-cluster(%s): Spurious cluster confirmation\n", mdname(mddev)); | |
1167 | return -EINVAL; | |
1168 | } | |
1169 | ||
1aee41f6 GR |
1170 | if (ack) |
1171 | dlm_unlock_sync(cinfo->no_new_dev_lockres); | |
1172 | complete(&cinfo->newdisk_completion); | |
fa8259da | 1173 | return 0; |
1aee41f6 GR |
1174 | } |
1175 | ||
88bcfef7 GR |
1176 | static int remove_disk(struct mddev *mddev, struct md_rdev *rdev) |
1177 | { | |
aee177ac | 1178 | struct cluster_msg cmsg = {0}; |
88bcfef7 | 1179 | struct md_cluster_info *cinfo = mddev->cluster_info; |
faeff83f GJ |
1180 | cmsg.type = cpu_to_le32(REMOVE); |
1181 | cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); | |
54a88392 | 1182 | return sendmsg(cinfo, &cmsg); |
88bcfef7 GR |
1183 | } |
1184 | ||
f6a2dc64 GJ |
1185 | static int lock_all_bitmaps(struct mddev *mddev) |
1186 | { | |
1187 | int slot, my_slot, ret, held = 1, i = 0; | |
1188 | char str[64]; | |
1189 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1190 | ||
1191 | cinfo->other_bitmap_lockres = kzalloc((mddev->bitmap_info.nodes - 1) * | |
1192 | sizeof(struct dlm_lock_resource *), | |
1193 | GFP_KERNEL); | |
1194 | if (!cinfo->other_bitmap_lockres) { | |
1195 | pr_err("md: can't alloc mem for other bitmap locks\n"); | |
1196 | return 0; | |
1197 | } | |
1198 | ||
1199 | my_slot = slot_number(mddev); | |
1200 | for (slot = 0; slot < mddev->bitmap_info.nodes; slot++) { | |
1201 | if (slot == my_slot) | |
1202 | continue; | |
1203 | ||
1204 | memset(str, '\0', 64); | |
1205 | snprintf(str, 64, "bitmap%04d", slot); | |
1206 | cinfo->other_bitmap_lockres[i] = lockres_init(mddev, str, NULL, 1); | |
1207 | if (!cinfo->other_bitmap_lockres[i]) | |
1208 | return -ENOMEM; | |
1209 | ||
1210 | cinfo->other_bitmap_lockres[i]->flags |= DLM_LKF_NOQUEUE; | |
1211 | ret = dlm_lock_sync(cinfo->other_bitmap_lockres[i], DLM_LOCK_PW); | |
1212 | if (ret) | |
1213 | held = -1; | |
1214 | i++; | |
1215 | } | |
1216 | ||
1217 | return held; | |
1218 | } | |
1219 | ||
1220 | static void unlock_all_bitmaps(struct mddev *mddev) | |
1221 | { | |
1222 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1223 | int i; | |
1224 | ||
1225 | /* release other node's bitmap lock if they are existed */ | |
1226 | if (cinfo->other_bitmap_lockres) { | |
1227 | for (i = 0; i < mddev->bitmap_info.nodes - 1; i++) { | |
1228 | if (cinfo->other_bitmap_lockres[i]) { | |
f6a2dc64 GJ |
1229 | lockres_free(cinfo->other_bitmap_lockres[i]); |
1230 | } | |
1231 | } | |
1232 | kfree(cinfo->other_bitmap_lockres); | |
1233 | } | |
1234 | } | |
1235 | ||
97f6cd39 GR |
1236 | static int gather_bitmaps(struct md_rdev *rdev) |
1237 | { | |
1238 | int sn, err; | |
1239 | sector_t lo, hi; | |
aee177ac | 1240 | struct cluster_msg cmsg = {0}; |
97f6cd39 GR |
1241 | struct mddev *mddev = rdev->mddev; |
1242 | struct md_cluster_info *cinfo = mddev->cluster_info; | |
1243 | ||
faeff83f GJ |
1244 | cmsg.type = cpu_to_le32(RE_ADD); |
1245 | cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); | |
97f6cd39 GR |
1246 | err = sendmsg(cinfo, &cmsg); |
1247 | if (err) | |
1248 | goto out; | |
1249 | ||
1250 | for (sn = 0; sn < mddev->bitmap_info.nodes; sn++) { | |
1251 | if (sn == (cinfo->slot_number - 1)) | |
1252 | continue; | |
1253 | err = bitmap_copy_from_slot(mddev, sn, &lo, &hi, false); | |
1254 | if (err) { | |
1255 | pr_warn("md-cluster: Could not gather bitmaps from slot %d", sn); | |
1256 | goto out; | |
1257 | } | |
1258 | if ((hi > 0) && (lo < mddev->recovery_cp)) | |
1259 | mddev->recovery_cp = lo; | |
1260 | } | |
1261 | out: | |
1262 | return err; | |
1263 | } | |
1264 | ||
edb39c9d GR |
1265 | static struct md_cluster_operations cluster_ops = { |
1266 | .join = join, | |
1267 | .leave = leave, | |
cf921cc1 | 1268 | .slot_number = slot_number, |
c186b128 GR |
1269 | .resync_start = resync_start, |
1270 | .resync_finish = resync_finish, | |
96ae923a | 1271 | .resync_info_update = resync_info_update, |
293467aa GR |
1272 | .metadata_update_start = metadata_update_start, |
1273 | .metadata_update_finish = metadata_update_finish, | |
1274 | .metadata_update_cancel = metadata_update_cancel, | |
589a1c49 | 1275 | .area_resyncing = area_resyncing, |
dbb64f86 GR |
1276 | .add_new_disk = add_new_disk, |
1277 | .add_new_disk_cancel = add_new_disk_cancel, | |
1aee41f6 | 1278 | .new_disk_ack = new_disk_ack, |
88bcfef7 | 1279 | .remove_disk = remove_disk, |
51e453ae | 1280 | .load_bitmaps = load_bitmaps, |
97f6cd39 | 1281 | .gather_bitmaps = gather_bitmaps, |
f6a2dc64 GJ |
1282 | .lock_all_bitmaps = lock_all_bitmaps, |
1283 | .unlock_all_bitmaps = unlock_all_bitmaps, | |
edb39c9d GR |
1284 | }; |
1285 | ||
8e854e9c GR |
1286 | static int __init cluster_init(void) |
1287 | { | |
1288 | pr_warn("md-cluster: EXPERIMENTAL. Use with caution\n"); | |
1289 | pr_info("Registering Cluster MD functions\n"); | |
edb39c9d | 1290 | register_md_cluster_operations(&cluster_ops, THIS_MODULE); |
8e854e9c GR |
1291 | return 0; |
1292 | } | |
1293 | ||
1294 | static void cluster_exit(void) | |
1295 | { | |
edb39c9d | 1296 | unregister_md_cluster_operations(); |
8e854e9c GR |
1297 | } |
1298 | ||
1299 | module_init(cluster_init); | |
1300 | module_exit(cluster_exit); | |
86b57277 | 1301 | MODULE_AUTHOR("SUSE"); |
8e854e9c GR |
1302 | MODULE_LICENSE("GPL"); |
1303 | MODULE_DESCRIPTION("Clustering support for MD"); |