]>
Commit | Line | Data |
---|---|---|
b411b363 PR |
1 | /* |
2 | drbd_worker.c | |
3 | ||
4 | This file is part of DRBD by Philipp Reisner and Lars Ellenberg. | |
5 | ||
6 | Copyright (C) 2001-2008, LINBIT Information Technologies GmbH. | |
7 | Copyright (C) 1999-2008, Philipp Reisner <philipp.reisner@linbit.com>. | |
8 | Copyright (C) 2002-2008, Lars Ellenberg <lars.ellenberg@linbit.com>. | |
9 | ||
10 | drbd is free software; you can redistribute it and/or modify | |
11 | it under the terms of the GNU General Public License as published by | |
12 | the Free Software Foundation; either version 2, or (at your option) | |
13 | any later version. | |
14 | ||
15 | drbd is distributed in the hope that it will be useful, | |
16 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
17 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
18 | GNU General Public License for more details. | |
19 | ||
20 | You should have received a copy of the GNU General Public License | |
21 | along with drbd; see the file COPYING. If not, write to | |
22 | the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. | |
23 | ||
24 | */ | |
25 | ||
b411b363 | 26 | #include <linux/module.h> |
b411b363 PR |
27 | #include <linux/drbd.h> |
28 | #include <linux/sched.h> | |
29 | #include <linux/smp_lock.h> | |
30 | #include <linux/wait.h> | |
31 | #include <linux/mm.h> | |
32 | #include <linux/memcontrol.h> | |
33 | #include <linux/mm_inline.h> | |
34 | #include <linux/slab.h> | |
35 | #include <linux/random.h> | |
b411b363 PR |
36 | #include <linux/string.h> |
37 | #include <linux/scatterlist.h> | |
38 | ||
39 | #include "drbd_int.h" | |
40 | #include "drbd_req.h" | |
b411b363 PR |
41 | |
42 | #define SLEEP_TIME (HZ/10) | |
43 | ||
44 | static int w_make_ov_request(struct drbd_conf *mdev, struct drbd_work *w, int cancel); | |
45 | ||
46 | ||
47 | ||
48 | /* defined here: | |
49 | drbd_md_io_complete | |
45bb912b | 50 | drbd_endio_sec |
b411b363 PR |
51 | drbd_endio_pri |
52 | ||
53 | * more endio handlers: | |
54 | atodb_endio in drbd_actlog.c | |
55 | drbd_bm_async_io_complete in drbd_bitmap.c | |
56 | ||
57 | * For all these callbacks, note the following: | |
58 | * The callbacks will be called in irq context by the IDE drivers, | |
59 | * and in Softirqs/Tasklets/BH context by the SCSI drivers. | |
60 | * Try to get the locking right :) | |
61 | * | |
62 | */ | |
63 | ||
64 | ||
65 | /* About the global_state_lock | |
66 | Each state transition on an device holds a read lock. In case we have | |
67 | to evaluate the sync after dependencies, we grab a write lock, because | |
68 | we need stable states on all devices for that. */ | |
69 | rwlock_t global_state_lock; | |
70 | ||
71 | /* used for synchronous meta data and bitmap IO | |
72 | * submitted by drbd_md_sync_page_io() | |
73 | */ | |
74 | void drbd_md_io_complete(struct bio *bio, int error) | |
75 | { | |
76 | struct drbd_md_io *md_io; | |
77 | ||
78 | md_io = (struct drbd_md_io *)bio->bi_private; | |
79 | md_io->error = error; | |
80 | ||
b411b363 PR |
81 | complete(&md_io->event); |
82 | } | |
83 | ||
84 | /* reads on behalf of the partner, | |
85 | * "submitted" by the receiver | |
86 | */ | |
45bb912b | 87 | void drbd_endio_read_sec_final(struct drbd_epoch_entry *e) __releases(local) |
b411b363 PR |
88 | { |
89 | unsigned long flags = 0; | |
45bb912b | 90 | struct drbd_conf *mdev = e->mdev; |
b411b363 PR |
91 | |
92 | D_ASSERT(e->block_id != ID_VACANT); | |
93 | ||
b411b363 PR |
94 | spin_lock_irqsave(&mdev->req_lock, flags); |
95 | mdev->read_cnt += e->size >> 9; | |
96 | list_del(&e->w.list); | |
97 | if (list_empty(&mdev->read_ee)) | |
98 | wake_up(&mdev->ee_wait); | |
45bb912b LE |
99 | if (test_bit(__EE_WAS_ERROR, &e->flags)) |
100 | __drbd_chk_io_error(mdev, FALSE); | |
b411b363 PR |
101 | spin_unlock_irqrestore(&mdev->req_lock, flags); |
102 | ||
b411b363 PR |
103 | drbd_queue_work(&mdev->data.work, &e->w); |
104 | put_ldev(mdev); | |
b411b363 PR |
105 | } |
106 | ||
45bb912b LE |
107 | static int is_failed_barrier(int ee_flags) |
108 | { | |
109 | return (ee_flags & (EE_IS_BARRIER|EE_WAS_ERROR|EE_RESUBMITTED)) | |
110 | == (EE_IS_BARRIER|EE_WAS_ERROR); | |
111 | } | |
112 | ||
b411b363 | 113 | /* writes on behalf of the partner, or resync writes, |
45bb912b LE |
114 | * "submitted" by the receiver, final stage. */ |
115 | static void drbd_endio_write_sec_final(struct drbd_epoch_entry *e) __releases(local) | |
b411b363 PR |
116 | { |
117 | unsigned long flags = 0; | |
45bb912b | 118 | struct drbd_conf *mdev = e->mdev; |
b411b363 PR |
119 | sector_t e_sector; |
120 | int do_wake; | |
121 | int is_syncer_req; | |
122 | int do_al_complete_io; | |
b411b363 | 123 | |
45bb912b LE |
124 | /* if this is a failed barrier request, disable use of barriers, |
125 | * and schedule for resubmission */ | |
126 | if (is_failed_barrier(e->flags)) { | |
b411b363 PR |
127 | drbd_bump_write_ordering(mdev, WO_bdev_flush); |
128 | spin_lock_irqsave(&mdev->req_lock, flags); | |
129 | list_del(&e->w.list); | |
45bb912b | 130 | e->flags |= EE_RESUBMITTED; |
b411b363 PR |
131 | e->w.cb = w_e_reissue; |
132 | /* put_ldev actually happens below, once we come here again. */ | |
133 | __release(local); | |
134 | spin_unlock_irqrestore(&mdev->req_lock, flags); | |
135 | drbd_queue_work(&mdev->data.work, &e->w); | |
136 | return; | |
137 | } | |
138 | ||
139 | D_ASSERT(e->block_id != ID_VACANT); | |
140 | ||
b411b363 PR |
141 | /* after we moved e to done_ee, |
142 | * we may no longer access it, | |
143 | * it may be freed/reused already! | |
144 | * (as soon as we release the req_lock) */ | |
145 | e_sector = e->sector; | |
146 | do_al_complete_io = e->flags & EE_CALL_AL_COMPLETE_IO; | |
45bb912b | 147 | is_syncer_req = is_syncer_block_id(e->block_id); |
b411b363 | 148 | |
45bb912b LE |
149 | spin_lock_irqsave(&mdev->req_lock, flags); |
150 | mdev->writ_cnt += e->size >> 9; | |
b411b363 PR |
151 | list_del(&e->w.list); /* has been on active_ee or sync_ee */ |
152 | list_add_tail(&e->w.list, &mdev->done_ee); | |
153 | ||
b411b363 PR |
154 | /* No hlist_del_init(&e->colision) here, we did not send the Ack yet, |
155 | * neither did we wake possibly waiting conflicting requests. | |
156 | * done from "drbd_process_done_ee" within the appropriate w.cb | |
157 | * (e_end_block/e_end_resync_block) or from _drbd_clear_done_ee */ | |
158 | ||
159 | do_wake = is_syncer_req | |
160 | ? list_empty(&mdev->sync_ee) | |
161 | : list_empty(&mdev->active_ee); | |
162 | ||
45bb912b | 163 | if (test_bit(__EE_WAS_ERROR, &e->flags)) |
b411b363 PR |
164 | __drbd_chk_io_error(mdev, FALSE); |
165 | spin_unlock_irqrestore(&mdev->req_lock, flags); | |
166 | ||
167 | if (is_syncer_req) | |
168 | drbd_rs_complete_io(mdev, e_sector); | |
169 | ||
170 | if (do_wake) | |
171 | wake_up(&mdev->ee_wait); | |
172 | ||
173 | if (do_al_complete_io) | |
174 | drbd_al_complete_io(mdev, e_sector); | |
175 | ||
176 | wake_asender(mdev); | |
177 | put_ldev(mdev); | |
45bb912b | 178 | } |
b411b363 | 179 | |
45bb912b LE |
180 | /* writes on behalf of the partner, or resync writes, |
181 | * "submitted" by the receiver. | |
182 | */ | |
183 | void drbd_endio_sec(struct bio *bio, int error) | |
184 | { | |
185 | struct drbd_epoch_entry *e = bio->bi_private; | |
186 | struct drbd_conf *mdev = e->mdev; | |
187 | int uptodate = bio_flagged(bio, BIO_UPTODATE); | |
188 | int is_write = bio_data_dir(bio) == WRITE; | |
189 | ||
190 | if (error) | |
191 | dev_warn(DEV, "%s: error=%d s=%llus\n", | |
192 | is_write ? "write" : "read", error, | |
193 | (unsigned long long)e->sector); | |
194 | if (!error && !uptodate) { | |
195 | dev_warn(DEV, "%s: setting error to -EIO s=%llus\n", | |
196 | is_write ? "write" : "read", | |
197 | (unsigned long long)e->sector); | |
198 | /* strange behavior of some lower level drivers... | |
199 | * fail the request by clearing the uptodate flag, | |
200 | * but do not return any error?! */ | |
201 | error = -EIO; | |
202 | } | |
203 | ||
204 | if (error) | |
205 | set_bit(__EE_WAS_ERROR, &e->flags); | |
206 | ||
207 | bio_put(bio); /* no need for the bio anymore */ | |
208 | if (atomic_dec_and_test(&e->pending_bios)) { | |
209 | if (is_write) | |
210 | drbd_endio_write_sec_final(e); | |
211 | else | |
212 | drbd_endio_read_sec_final(e); | |
213 | } | |
b411b363 PR |
214 | } |
215 | ||
216 | /* read, readA or write requests on R_PRIMARY coming from drbd_make_request | |
217 | */ | |
218 | void drbd_endio_pri(struct bio *bio, int error) | |
219 | { | |
220 | unsigned long flags; | |
221 | struct drbd_request *req = bio->bi_private; | |
222 | struct drbd_conf *mdev = req->mdev; | |
223 | struct bio_and_error m; | |
224 | enum drbd_req_event what; | |
225 | int uptodate = bio_flagged(bio, BIO_UPTODATE); | |
226 | ||
227 | if (error) | |
228 | dev_warn(DEV, "p %s: error=%d\n", | |
229 | bio_data_dir(bio) == WRITE ? "write" : "read", error); | |
230 | if (!error && !uptodate) { | |
231 | dev_warn(DEV, "p %s: setting error to -EIO\n", | |
232 | bio_data_dir(bio) == WRITE ? "write" : "read"); | |
233 | /* strange behavior of some lower level drivers... | |
234 | * fail the request by clearing the uptodate flag, | |
235 | * but do not return any error?! */ | |
236 | error = -EIO; | |
237 | } | |
238 | ||
b411b363 PR |
239 | /* to avoid recursion in __req_mod */ |
240 | if (unlikely(error)) { | |
241 | what = (bio_data_dir(bio) == WRITE) | |
242 | ? write_completed_with_error | |
243 | : (bio_rw(bio) == READA) | |
244 | ? read_completed_with_error | |
245 | : read_ahead_completed_with_error; | |
246 | } else | |
247 | what = completed_ok; | |
248 | ||
249 | bio_put(req->private_bio); | |
250 | req->private_bio = ERR_PTR(error); | |
251 | ||
252 | spin_lock_irqsave(&mdev->req_lock, flags); | |
253 | __req_mod(req, what, &m); | |
254 | spin_unlock_irqrestore(&mdev->req_lock, flags); | |
255 | ||
256 | if (m.bio) | |
257 | complete_master_bio(mdev, &m); | |
258 | } | |
259 | ||
260 | int w_io_error(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
261 | { | |
262 | struct drbd_request *req = container_of(w, struct drbd_request, w); | |
263 | ||
264 | /* NOTE: mdev->ldev can be NULL by the time we get here! */ | |
265 | /* D_ASSERT(mdev->ldev->dc.on_io_error != EP_PASS_ON); */ | |
266 | ||
267 | /* the only way this callback is scheduled is from _req_may_be_done, | |
268 | * when it is done and had a local write error, see comments there */ | |
269 | drbd_req_free(req); | |
270 | ||
271 | return TRUE; | |
272 | } | |
273 | ||
274 | int w_read_retry_remote(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
275 | { | |
276 | struct drbd_request *req = container_of(w, struct drbd_request, w); | |
277 | ||
278 | /* We should not detach for read io-error, | |
279 | * but try to WRITE the P_DATA_REPLY to the failed location, | |
280 | * to give the disk the chance to relocate that block */ | |
281 | ||
282 | spin_lock_irq(&mdev->req_lock); | |
283 | if (cancel || | |
284 | mdev->state.conn < C_CONNECTED || | |
285 | mdev->state.pdsk <= D_INCONSISTENT) { | |
286 | _req_mod(req, send_canceled); | |
287 | spin_unlock_irq(&mdev->req_lock); | |
288 | dev_alert(DEV, "WE ARE LOST. Local IO failure, no peer.\n"); | |
289 | return 1; | |
290 | } | |
291 | spin_unlock_irq(&mdev->req_lock); | |
292 | ||
293 | return w_send_read_req(mdev, w, 0); | |
294 | } | |
295 | ||
296 | int w_resync_inactive(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
297 | { | |
298 | ERR_IF(cancel) return 1; | |
299 | dev_err(DEV, "resync inactive, but callback triggered??\n"); | |
300 | return 1; /* Simply ignore this! */ | |
301 | } | |
302 | ||
45bb912b LE |
303 | void drbd_csum_ee(struct drbd_conf *mdev, struct crypto_hash *tfm, struct drbd_epoch_entry *e, void *digest) |
304 | { | |
305 | struct hash_desc desc; | |
306 | struct scatterlist sg; | |
307 | struct page *page = e->pages; | |
308 | struct page *tmp; | |
309 | unsigned len; | |
310 | ||
311 | desc.tfm = tfm; | |
312 | desc.flags = 0; | |
313 | ||
314 | sg_init_table(&sg, 1); | |
315 | crypto_hash_init(&desc); | |
316 | ||
317 | while ((tmp = page_chain_next(page))) { | |
318 | /* all but the last page will be fully used */ | |
319 | sg_set_page(&sg, page, PAGE_SIZE, 0); | |
320 | crypto_hash_update(&desc, &sg, sg.length); | |
321 | page = tmp; | |
322 | } | |
323 | /* and now the last, possibly only partially used page */ | |
324 | len = e->size & (PAGE_SIZE - 1); | |
325 | sg_set_page(&sg, page, len ?: PAGE_SIZE, 0); | |
326 | crypto_hash_update(&desc, &sg, sg.length); | |
327 | crypto_hash_final(&desc, digest); | |
328 | } | |
329 | ||
330 | void drbd_csum_bio(struct drbd_conf *mdev, struct crypto_hash *tfm, struct bio *bio, void *digest) | |
b411b363 PR |
331 | { |
332 | struct hash_desc desc; | |
333 | struct scatterlist sg; | |
334 | struct bio_vec *bvec; | |
335 | int i; | |
336 | ||
337 | desc.tfm = tfm; | |
338 | desc.flags = 0; | |
339 | ||
340 | sg_init_table(&sg, 1); | |
341 | crypto_hash_init(&desc); | |
342 | ||
343 | __bio_for_each_segment(bvec, bio, i, 0) { | |
344 | sg_set_page(&sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); | |
345 | crypto_hash_update(&desc, &sg, sg.length); | |
346 | } | |
347 | crypto_hash_final(&desc, digest); | |
348 | } | |
349 | ||
350 | static int w_e_send_csum(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
351 | { | |
352 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
353 | int digest_size; | |
354 | void *digest; | |
355 | int ok; | |
356 | ||
357 | D_ASSERT(e->block_id == DRBD_MAGIC + 0xbeef); | |
358 | ||
359 | if (unlikely(cancel)) { | |
360 | drbd_free_ee(mdev, e); | |
361 | return 1; | |
362 | } | |
363 | ||
45bb912b | 364 | if (likely((e->flags & EE_WAS_ERROR) == 0)) { |
b411b363 PR |
365 | digest_size = crypto_hash_digestsize(mdev->csums_tfm); |
366 | digest = kmalloc(digest_size, GFP_NOIO); | |
367 | if (digest) { | |
45bb912b | 368 | drbd_csum_ee(mdev, mdev->csums_tfm, e, digest); |
b411b363 PR |
369 | |
370 | inc_rs_pending(mdev); | |
371 | ok = drbd_send_drequest_csum(mdev, | |
372 | e->sector, | |
373 | e->size, | |
374 | digest, | |
375 | digest_size, | |
376 | P_CSUM_RS_REQUEST); | |
377 | kfree(digest); | |
378 | } else { | |
379 | dev_err(DEV, "kmalloc() of digest failed.\n"); | |
380 | ok = 0; | |
381 | } | |
382 | } else | |
383 | ok = 1; | |
384 | ||
385 | drbd_free_ee(mdev, e); | |
386 | ||
387 | if (unlikely(!ok)) | |
388 | dev_err(DEV, "drbd_send_drequest(..., csum) failed\n"); | |
389 | return ok; | |
390 | } | |
391 | ||
392 | #define GFP_TRY (__GFP_HIGHMEM | __GFP_NOWARN) | |
393 | ||
394 | static int read_for_csum(struct drbd_conf *mdev, sector_t sector, int size) | |
395 | { | |
396 | struct drbd_epoch_entry *e; | |
397 | ||
398 | if (!get_ldev(mdev)) | |
399 | return 0; | |
400 | ||
401 | /* GFP_TRY, because if there is no memory available right now, this may | |
402 | * be rescheduled for later. It is "only" background resync, after all. */ | |
403 | e = drbd_alloc_ee(mdev, DRBD_MAGIC+0xbeef, sector, size, GFP_TRY); | |
45bb912b LE |
404 | if (!e) |
405 | goto fail; | |
b411b363 PR |
406 | |
407 | spin_lock_irq(&mdev->req_lock); | |
408 | list_add(&e->w.list, &mdev->read_ee); | |
409 | spin_unlock_irq(&mdev->req_lock); | |
410 | ||
b411b363 | 411 | e->w.cb = w_e_send_csum; |
45bb912b LE |
412 | if (drbd_submit_ee(mdev, e, READ, DRBD_FAULT_RS_RD) == 0) |
413 | return 1; | |
b411b363 | 414 | |
45bb912b LE |
415 | drbd_free_ee(mdev, e); |
416 | fail: | |
417 | put_ldev(mdev); | |
418 | return 2; | |
b411b363 PR |
419 | } |
420 | ||
421 | void resync_timer_fn(unsigned long data) | |
422 | { | |
423 | unsigned long flags; | |
424 | struct drbd_conf *mdev = (struct drbd_conf *) data; | |
425 | int queue; | |
426 | ||
427 | spin_lock_irqsave(&mdev->req_lock, flags); | |
428 | ||
429 | if (likely(!test_and_clear_bit(STOP_SYNC_TIMER, &mdev->flags))) { | |
430 | queue = 1; | |
431 | if (mdev->state.conn == C_VERIFY_S) | |
432 | mdev->resync_work.cb = w_make_ov_request; | |
433 | else | |
434 | mdev->resync_work.cb = w_make_resync_request; | |
435 | } else { | |
436 | queue = 0; | |
437 | mdev->resync_work.cb = w_resync_inactive; | |
438 | } | |
439 | ||
440 | spin_unlock_irqrestore(&mdev->req_lock, flags); | |
441 | ||
442 | /* harmless race: list_empty outside data.work.q_lock */ | |
443 | if (list_empty(&mdev->resync_work.list) && queue) | |
444 | drbd_queue_work(&mdev->data.work, &mdev->resync_work); | |
445 | } | |
446 | ||
cdd67a74 PR |
447 | static int calc_resync_rate(struct drbd_conf *mdev) |
448 | { | |
449 | int d = mdev->data_delay / 1000; /* us -> ms */ | |
450 | int td = mdev->sync_conf.throttle_th * 100; /* 0.1s -> ms */ | |
451 | int hd = mdev->sync_conf.hold_off_th * 100; /* 0.1s -> ms */ | |
452 | int cr = mdev->sync_conf.rate; | |
453 | ||
454 | return d <= td ? cr : | |
455 | d >= hd ? 0 : | |
456 | cr + (cr * (td - d) / (hd - td)); | |
457 | } | |
458 | ||
b411b363 PR |
459 | int w_make_resync_request(struct drbd_conf *mdev, |
460 | struct drbd_work *w, int cancel) | |
461 | { | |
462 | unsigned long bit; | |
463 | sector_t sector; | |
464 | const sector_t capacity = drbd_get_capacity(mdev->this_bdev); | |
bb3d000c | 465 | int max_segment_size; |
b411b363 PR |
466 | int number, i, size, pe, mx; |
467 | int align, queued, sndbuf; | |
468 | ||
469 | if (unlikely(cancel)) | |
470 | return 1; | |
471 | ||
472 | if (unlikely(mdev->state.conn < C_CONNECTED)) { | |
473 | dev_err(DEV, "Confused in w_make_resync_request()! cstate < Connected"); | |
474 | return 0; | |
475 | } | |
476 | ||
477 | if (mdev->state.conn != C_SYNC_TARGET) | |
478 | dev_err(DEV, "%s in w_make_resync_request\n", | |
479 | drbd_conn_str(mdev->state.conn)); | |
480 | ||
481 | if (!get_ldev(mdev)) { | |
482 | /* Since we only need to access mdev->rsync a | |
483 | get_ldev_if_state(mdev,D_FAILED) would be sufficient, but | |
484 | to continue resync with a broken disk makes no sense at | |
485 | all */ | |
486 | dev_err(DEV, "Disk broke down during resync!\n"); | |
487 | mdev->resync_work.cb = w_resync_inactive; | |
488 | return 1; | |
489 | } | |
490 | ||
bb3d000c LE |
491 | /* starting with drbd 8.3.8, we can handle multi-bio EEs, |
492 | * if it should be necessary */ | |
493 | max_segment_size = mdev->agreed_pro_version < 94 ? | |
494 | queue_max_segment_size(mdev->rq_queue) : DRBD_MAX_SEGMENT_SIZE; | |
495 | ||
cdd67a74 PR |
496 | mdev->c_sync_rate = calc_resync_rate(mdev); |
497 | number = SLEEP_TIME * mdev->c_sync_rate / ((BM_BLOCK_SIZE / 1024) * HZ); | |
b411b363 PR |
498 | pe = atomic_read(&mdev->rs_pending_cnt); |
499 | ||
500 | mutex_lock(&mdev->data.mutex); | |
501 | if (mdev->data.socket) | |
502 | mx = mdev->data.socket->sk->sk_rcvbuf / sizeof(struct p_block_req); | |
503 | else | |
504 | mx = 1; | |
505 | mutex_unlock(&mdev->data.mutex); | |
506 | ||
507 | /* For resync rates >160MB/sec, allow more pending RS requests */ | |
508 | if (number > mx) | |
509 | mx = number; | |
510 | ||
511 | /* Limit the number of pending RS requests to no more than the peer's receive buffer */ | |
512 | if ((pe + number) > mx) { | |
513 | number = mx - pe; | |
514 | } | |
515 | ||
516 | for (i = 0; i < number; i++) { | |
517 | /* Stop generating RS requests, when half of the send buffer is filled */ | |
518 | mutex_lock(&mdev->data.mutex); | |
519 | if (mdev->data.socket) { | |
520 | queued = mdev->data.socket->sk->sk_wmem_queued; | |
521 | sndbuf = mdev->data.socket->sk->sk_sndbuf; | |
522 | } else { | |
523 | queued = 1; | |
524 | sndbuf = 0; | |
525 | } | |
526 | mutex_unlock(&mdev->data.mutex); | |
527 | if (queued > sndbuf / 2) | |
528 | goto requeue; | |
529 | ||
530 | next_sector: | |
531 | size = BM_BLOCK_SIZE; | |
532 | bit = drbd_bm_find_next(mdev, mdev->bm_resync_fo); | |
533 | ||
534 | if (bit == -1UL) { | |
535 | mdev->bm_resync_fo = drbd_bm_bits(mdev); | |
536 | mdev->resync_work.cb = w_resync_inactive; | |
537 | put_ldev(mdev); | |
538 | return 1; | |
539 | } | |
540 | ||
541 | sector = BM_BIT_TO_SECT(bit); | |
542 | ||
543 | if (drbd_try_rs_begin_io(mdev, sector)) { | |
544 | mdev->bm_resync_fo = bit; | |
545 | goto requeue; | |
546 | } | |
547 | mdev->bm_resync_fo = bit + 1; | |
548 | ||
549 | if (unlikely(drbd_bm_test_bit(mdev, bit) == 0)) { | |
550 | drbd_rs_complete_io(mdev, sector); | |
551 | goto next_sector; | |
552 | } | |
553 | ||
554 | #if DRBD_MAX_SEGMENT_SIZE > BM_BLOCK_SIZE | |
555 | /* try to find some adjacent bits. | |
556 | * we stop if we have already the maximum req size. | |
557 | * | |
558 | * Additionally always align bigger requests, in order to | |
559 | * be prepared for all stripe sizes of software RAIDs. | |
b411b363 PR |
560 | */ |
561 | align = 1; | |
562 | for (;;) { | |
563 | if (size + BM_BLOCK_SIZE > max_segment_size) | |
564 | break; | |
565 | ||
566 | /* Be always aligned */ | |
567 | if (sector & ((1<<(align+3))-1)) | |
568 | break; | |
569 | ||
570 | /* do not cross extent boundaries */ | |
571 | if (((bit+1) & BM_BLOCKS_PER_BM_EXT_MASK) == 0) | |
572 | break; | |
573 | /* now, is it actually dirty, after all? | |
574 | * caution, drbd_bm_test_bit is tri-state for some | |
575 | * obscure reason; ( b == 0 ) would get the out-of-band | |
576 | * only accidentally right because of the "oddly sized" | |
577 | * adjustment below */ | |
578 | if (drbd_bm_test_bit(mdev, bit+1) != 1) | |
579 | break; | |
580 | bit++; | |
581 | size += BM_BLOCK_SIZE; | |
582 | if ((BM_BLOCK_SIZE << align) <= size) | |
583 | align++; | |
584 | i++; | |
585 | } | |
586 | /* if we merged some, | |
587 | * reset the offset to start the next drbd_bm_find_next from */ | |
588 | if (size > BM_BLOCK_SIZE) | |
589 | mdev->bm_resync_fo = bit + 1; | |
590 | #endif | |
591 | ||
592 | /* adjust very last sectors, in case we are oddly sized */ | |
593 | if (sector + (size>>9) > capacity) | |
594 | size = (capacity-sector)<<9; | |
595 | if (mdev->agreed_pro_version >= 89 && mdev->csums_tfm) { | |
596 | switch (read_for_csum(mdev, sector, size)) { | |
597 | case 0: /* Disk failure*/ | |
598 | put_ldev(mdev); | |
599 | return 0; | |
600 | case 2: /* Allocation failed */ | |
601 | drbd_rs_complete_io(mdev, sector); | |
602 | mdev->bm_resync_fo = BM_SECT_TO_BIT(sector); | |
603 | goto requeue; | |
604 | /* case 1: everything ok */ | |
605 | } | |
606 | } else { | |
607 | inc_rs_pending(mdev); | |
608 | if (!drbd_send_drequest(mdev, P_RS_DATA_REQUEST, | |
609 | sector, size, ID_SYNCER)) { | |
610 | dev_err(DEV, "drbd_send_drequest() failed, aborting...\n"); | |
611 | dec_rs_pending(mdev); | |
612 | put_ldev(mdev); | |
613 | return 0; | |
614 | } | |
615 | } | |
616 | } | |
617 | ||
618 | if (mdev->bm_resync_fo >= drbd_bm_bits(mdev)) { | |
619 | /* last syncer _request_ was sent, | |
620 | * but the P_RS_DATA_REPLY not yet received. sync will end (and | |
621 | * next sync group will resume), as soon as we receive the last | |
622 | * resync data block, and the last bit is cleared. | |
623 | * until then resync "work" is "inactive" ... | |
624 | */ | |
625 | mdev->resync_work.cb = w_resync_inactive; | |
626 | put_ldev(mdev); | |
627 | return 1; | |
628 | } | |
629 | ||
630 | requeue: | |
631 | mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME); | |
632 | put_ldev(mdev); | |
633 | return 1; | |
634 | } | |
635 | ||
636 | static int w_make_ov_request(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
637 | { | |
638 | int number, i, size; | |
639 | sector_t sector; | |
640 | const sector_t capacity = drbd_get_capacity(mdev->this_bdev); | |
641 | ||
642 | if (unlikely(cancel)) | |
643 | return 1; | |
644 | ||
645 | if (unlikely(mdev->state.conn < C_CONNECTED)) { | |
646 | dev_err(DEV, "Confused in w_make_ov_request()! cstate < Connected"); | |
647 | return 0; | |
648 | } | |
649 | ||
650 | number = SLEEP_TIME*mdev->sync_conf.rate / ((BM_BLOCK_SIZE/1024)*HZ); | |
651 | if (atomic_read(&mdev->rs_pending_cnt) > number) | |
652 | goto requeue; | |
653 | ||
654 | number -= atomic_read(&mdev->rs_pending_cnt); | |
655 | ||
656 | sector = mdev->ov_position; | |
657 | for (i = 0; i < number; i++) { | |
658 | if (sector >= capacity) { | |
659 | mdev->resync_work.cb = w_resync_inactive; | |
660 | return 1; | |
661 | } | |
662 | ||
663 | size = BM_BLOCK_SIZE; | |
664 | ||
665 | if (drbd_try_rs_begin_io(mdev, sector)) { | |
666 | mdev->ov_position = sector; | |
667 | goto requeue; | |
668 | } | |
669 | ||
670 | if (sector + (size>>9) > capacity) | |
671 | size = (capacity-sector)<<9; | |
672 | ||
673 | inc_rs_pending(mdev); | |
674 | if (!drbd_send_ov_request(mdev, sector, size)) { | |
675 | dec_rs_pending(mdev); | |
676 | return 0; | |
677 | } | |
678 | sector += BM_SECT_PER_BIT; | |
679 | } | |
680 | mdev->ov_position = sector; | |
681 | ||
682 | requeue: | |
683 | mod_timer(&mdev->resync_timer, jiffies + SLEEP_TIME); | |
684 | return 1; | |
685 | } | |
686 | ||
687 | ||
688 | int w_ov_finished(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
689 | { | |
690 | kfree(w); | |
691 | ov_oos_print(mdev); | |
692 | drbd_resync_finished(mdev); | |
693 | ||
694 | return 1; | |
695 | } | |
696 | ||
697 | static int w_resync_finished(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
698 | { | |
699 | kfree(w); | |
700 | ||
701 | drbd_resync_finished(mdev); | |
702 | ||
703 | return 1; | |
704 | } | |
705 | ||
706 | int drbd_resync_finished(struct drbd_conf *mdev) | |
707 | { | |
708 | unsigned long db, dt, dbdt; | |
709 | unsigned long n_oos; | |
710 | union drbd_state os, ns; | |
711 | struct drbd_work *w; | |
712 | char *khelper_cmd = NULL; | |
713 | ||
714 | /* Remove all elements from the resync LRU. Since future actions | |
715 | * might set bits in the (main) bitmap, then the entries in the | |
716 | * resync LRU would be wrong. */ | |
717 | if (drbd_rs_del_all(mdev)) { | |
718 | /* In case this is not possible now, most probably because | |
719 | * there are P_RS_DATA_REPLY Packets lingering on the worker's | |
720 | * queue (or even the read operations for those packets | |
721 | * is not finished by now). Retry in 100ms. */ | |
722 | ||
723 | drbd_kick_lo(mdev); | |
724 | __set_current_state(TASK_INTERRUPTIBLE); | |
725 | schedule_timeout(HZ / 10); | |
726 | w = kmalloc(sizeof(struct drbd_work), GFP_ATOMIC); | |
727 | if (w) { | |
728 | w->cb = w_resync_finished; | |
729 | drbd_queue_work(&mdev->data.work, w); | |
730 | return 1; | |
731 | } | |
732 | dev_err(DEV, "Warn failed to drbd_rs_del_all() and to kmalloc(w).\n"); | |
733 | } | |
734 | ||
735 | dt = (jiffies - mdev->rs_start - mdev->rs_paused) / HZ; | |
736 | if (dt <= 0) | |
737 | dt = 1; | |
738 | db = mdev->rs_total; | |
739 | dbdt = Bit2KB(db/dt); | |
740 | mdev->rs_paused /= HZ; | |
741 | ||
742 | if (!get_ldev(mdev)) | |
743 | goto out; | |
744 | ||
745 | spin_lock_irq(&mdev->req_lock); | |
746 | os = mdev->state; | |
747 | ||
748 | /* This protects us against multiple calls (that can happen in the presence | |
749 | of application IO), and against connectivity loss just before we arrive here. */ | |
750 | if (os.conn <= C_CONNECTED) | |
751 | goto out_unlock; | |
752 | ||
753 | ns = os; | |
754 | ns.conn = C_CONNECTED; | |
755 | ||
756 | dev_info(DEV, "%s done (total %lu sec; paused %lu sec; %lu K/sec)\n", | |
757 | (os.conn == C_VERIFY_S || os.conn == C_VERIFY_T) ? | |
758 | "Online verify " : "Resync", | |
759 | dt + mdev->rs_paused, mdev->rs_paused, dbdt); | |
760 | ||
761 | n_oos = drbd_bm_total_weight(mdev); | |
762 | ||
763 | if (os.conn == C_VERIFY_S || os.conn == C_VERIFY_T) { | |
764 | if (n_oos) { | |
765 | dev_alert(DEV, "Online verify found %lu %dk block out of sync!\n", | |
766 | n_oos, Bit2KB(1)); | |
767 | khelper_cmd = "out-of-sync"; | |
768 | } | |
769 | } else { | |
770 | D_ASSERT((n_oos - mdev->rs_failed) == 0); | |
771 | ||
772 | if (os.conn == C_SYNC_TARGET || os.conn == C_PAUSED_SYNC_T) | |
773 | khelper_cmd = "after-resync-target"; | |
774 | ||
775 | if (mdev->csums_tfm && mdev->rs_total) { | |
776 | const unsigned long s = mdev->rs_same_csum; | |
777 | const unsigned long t = mdev->rs_total; | |
778 | const int ratio = | |
779 | (t == 0) ? 0 : | |
780 | (t < 100000) ? ((s*100)/t) : (s/(t/100)); | |
781 | dev_info(DEV, "%u %% had equal check sums, eliminated: %luK; " | |
782 | "transferred %luK total %luK\n", | |
783 | ratio, | |
784 | Bit2KB(mdev->rs_same_csum), | |
785 | Bit2KB(mdev->rs_total - mdev->rs_same_csum), | |
786 | Bit2KB(mdev->rs_total)); | |
787 | } | |
788 | } | |
789 | ||
790 | if (mdev->rs_failed) { | |
791 | dev_info(DEV, " %lu failed blocks\n", mdev->rs_failed); | |
792 | ||
793 | if (os.conn == C_SYNC_TARGET || os.conn == C_PAUSED_SYNC_T) { | |
794 | ns.disk = D_INCONSISTENT; | |
795 | ns.pdsk = D_UP_TO_DATE; | |
796 | } else { | |
797 | ns.disk = D_UP_TO_DATE; | |
798 | ns.pdsk = D_INCONSISTENT; | |
799 | } | |
800 | } else { | |
801 | ns.disk = D_UP_TO_DATE; | |
802 | ns.pdsk = D_UP_TO_DATE; | |
803 | ||
804 | if (os.conn == C_SYNC_TARGET || os.conn == C_PAUSED_SYNC_T) { | |
805 | if (mdev->p_uuid) { | |
806 | int i; | |
807 | for (i = UI_BITMAP ; i <= UI_HISTORY_END ; i++) | |
808 | _drbd_uuid_set(mdev, i, mdev->p_uuid[i]); | |
809 | drbd_uuid_set(mdev, UI_BITMAP, mdev->ldev->md.uuid[UI_CURRENT]); | |
810 | _drbd_uuid_set(mdev, UI_CURRENT, mdev->p_uuid[UI_CURRENT]); | |
811 | } else { | |
812 | dev_err(DEV, "mdev->p_uuid is NULL! BUG\n"); | |
813 | } | |
814 | } | |
815 | ||
816 | drbd_uuid_set_bm(mdev, 0UL); | |
817 | ||
818 | if (mdev->p_uuid) { | |
819 | /* Now the two UUID sets are equal, update what we | |
820 | * know of the peer. */ | |
821 | int i; | |
822 | for (i = UI_CURRENT ; i <= UI_HISTORY_END ; i++) | |
823 | mdev->p_uuid[i] = mdev->ldev->md.uuid[i]; | |
824 | } | |
825 | } | |
826 | ||
827 | _drbd_set_state(mdev, ns, CS_VERBOSE, NULL); | |
828 | out_unlock: | |
829 | spin_unlock_irq(&mdev->req_lock); | |
830 | put_ldev(mdev); | |
831 | out: | |
832 | mdev->rs_total = 0; | |
833 | mdev->rs_failed = 0; | |
834 | mdev->rs_paused = 0; | |
835 | mdev->ov_start_sector = 0; | |
836 | ||
837 | if (test_and_clear_bit(WRITE_BM_AFTER_RESYNC, &mdev->flags)) { | |
838 | dev_warn(DEV, "Writing the whole bitmap, due to failed kmalloc\n"); | |
839 | drbd_queue_bitmap_io(mdev, &drbd_bm_write, NULL, "write from resync_finished"); | |
840 | } | |
841 | ||
842 | if (khelper_cmd) | |
843 | drbd_khelper(mdev, khelper_cmd); | |
844 | ||
845 | return 1; | |
846 | } | |
847 | ||
848 | /* helper */ | |
849 | static void move_to_net_ee_or_free(struct drbd_conf *mdev, struct drbd_epoch_entry *e) | |
850 | { | |
45bb912b | 851 | if (drbd_ee_has_active_page(e)) { |
b411b363 PR |
852 | /* This might happen if sendpage() has not finished */ |
853 | spin_lock_irq(&mdev->req_lock); | |
854 | list_add_tail(&e->w.list, &mdev->net_ee); | |
855 | spin_unlock_irq(&mdev->req_lock); | |
856 | } else | |
857 | drbd_free_ee(mdev, e); | |
858 | } | |
859 | ||
860 | /** | |
861 | * w_e_end_data_req() - Worker callback, to send a P_DATA_REPLY packet in response to a P_DATA_REQUEST | |
862 | * @mdev: DRBD device. | |
863 | * @w: work object. | |
864 | * @cancel: The connection will be closed anyways | |
865 | */ | |
866 | int w_e_end_data_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
867 | { | |
868 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
869 | int ok; | |
870 | ||
871 | if (unlikely(cancel)) { | |
872 | drbd_free_ee(mdev, e); | |
873 | dec_unacked(mdev); | |
874 | return 1; | |
875 | } | |
876 | ||
45bb912b | 877 | if (likely((e->flags & EE_WAS_ERROR) == 0)) { |
b411b363 PR |
878 | ok = drbd_send_block(mdev, P_DATA_REPLY, e); |
879 | } else { | |
880 | if (__ratelimit(&drbd_ratelimit_state)) | |
881 | dev_err(DEV, "Sending NegDReply. sector=%llus.\n", | |
882 | (unsigned long long)e->sector); | |
883 | ||
884 | ok = drbd_send_ack(mdev, P_NEG_DREPLY, e); | |
885 | } | |
886 | ||
887 | dec_unacked(mdev); | |
888 | ||
889 | move_to_net_ee_or_free(mdev, e); | |
890 | ||
891 | if (unlikely(!ok)) | |
892 | dev_err(DEV, "drbd_send_block() failed\n"); | |
893 | return ok; | |
894 | } | |
895 | ||
896 | /** | |
897 | * w_e_end_rsdata_req() - Worker callback to send a P_RS_DATA_REPLY packet in response to a P_RS_DATA_REQUESTRS | |
898 | * @mdev: DRBD device. | |
899 | * @w: work object. | |
900 | * @cancel: The connection will be closed anyways | |
901 | */ | |
902 | int w_e_end_rsdata_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
903 | { | |
904 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
905 | int ok; | |
906 | ||
907 | if (unlikely(cancel)) { | |
908 | drbd_free_ee(mdev, e); | |
909 | dec_unacked(mdev); | |
910 | return 1; | |
911 | } | |
912 | ||
913 | if (get_ldev_if_state(mdev, D_FAILED)) { | |
914 | drbd_rs_complete_io(mdev, e->sector); | |
915 | put_ldev(mdev); | |
916 | } | |
917 | ||
45bb912b | 918 | if (likely((e->flags & EE_WAS_ERROR) == 0)) { |
b411b363 PR |
919 | if (likely(mdev->state.pdsk >= D_INCONSISTENT)) { |
920 | inc_rs_pending(mdev); | |
921 | ok = drbd_send_block(mdev, P_RS_DATA_REPLY, e); | |
922 | } else { | |
923 | if (__ratelimit(&drbd_ratelimit_state)) | |
924 | dev_err(DEV, "Not sending RSDataReply, " | |
925 | "partner DISKLESS!\n"); | |
926 | ok = 1; | |
927 | } | |
928 | } else { | |
929 | if (__ratelimit(&drbd_ratelimit_state)) | |
930 | dev_err(DEV, "Sending NegRSDReply. sector %llus.\n", | |
931 | (unsigned long long)e->sector); | |
932 | ||
933 | ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e); | |
934 | ||
935 | /* update resync data with failure */ | |
936 | drbd_rs_failed_io(mdev, e->sector, e->size); | |
937 | } | |
938 | ||
939 | dec_unacked(mdev); | |
940 | ||
941 | move_to_net_ee_or_free(mdev, e); | |
942 | ||
943 | if (unlikely(!ok)) | |
944 | dev_err(DEV, "drbd_send_block() failed\n"); | |
945 | return ok; | |
946 | } | |
947 | ||
948 | int w_e_end_csum_rs_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
949 | { | |
950 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
951 | struct digest_info *di; | |
952 | int digest_size; | |
953 | void *digest = NULL; | |
954 | int ok, eq = 0; | |
955 | ||
956 | if (unlikely(cancel)) { | |
957 | drbd_free_ee(mdev, e); | |
958 | dec_unacked(mdev); | |
959 | return 1; | |
960 | } | |
961 | ||
962 | drbd_rs_complete_io(mdev, e->sector); | |
963 | ||
964 | di = (struct digest_info *)(unsigned long)e->block_id; | |
965 | ||
45bb912b | 966 | if (likely((e->flags & EE_WAS_ERROR) == 0)) { |
b411b363 PR |
967 | /* quick hack to try to avoid a race against reconfiguration. |
968 | * a real fix would be much more involved, | |
969 | * introducing more locking mechanisms */ | |
970 | if (mdev->csums_tfm) { | |
971 | digest_size = crypto_hash_digestsize(mdev->csums_tfm); | |
972 | D_ASSERT(digest_size == di->digest_size); | |
973 | digest = kmalloc(digest_size, GFP_NOIO); | |
974 | } | |
975 | if (digest) { | |
45bb912b | 976 | drbd_csum_ee(mdev, mdev->csums_tfm, e, digest); |
b411b363 PR |
977 | eq = !memcmp(digest, di->digest, digest_size); |
978 | kfree(digest); | |
979 | } | |
980 | ||
981 | if (eq) { | |
982 | drbd_set_in_sync(mdev, e->sector, e->size); | |
676396d5 LE |
983 | /* rs_same_csums unit is BM_BLOCK_SIZE */ |
984 | mdev->rs_same_csum += e->size >> BM_BLOCK_SHIFT; | |
b411b363 PR |
985 | ok = drbd_send_ack(mdev, P_RS_IS_IN_SYNC, e); |
986 | } else { | |
987 | inc_rs_pending(mdev); | |
988 | e->block_id = ID_SYNCER; | |
989 | ok = drbd_send_block(mdev, P_RS_DATA_REPLY, e); | |
990 | } | |
991 | } else { | |
992 | ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e); | |
993 | if (__ratelimit(&drbd_ratelimit_state)) | |
994 | dev_err(DEV, "Sending NegDReply. I guess it gets messy.\n"); | |
995 | } | |
996 | ||
997 | dec_unacked(mdev); | |
998 | ||
999 | kfree(di); | |
1000 | ||
1001 | move_to_net_ee_or_free(mdev, e); | |
1002 | ||
1003 | if (unlikely(!ok)) | |
1004 | dev_err(DEV, "drbd_send_block/ack() failed\n"); | |
1005 | return ok; | |
1006 | } | |
1007 | ||
1008 | int w_e_end_ov_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1009 | { | |
1010 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
1011 | int digest_size; | |
1012 | void *digest; | |
1013 | int ok = 1; | |
1014 | ||
1015 | if (unlikely(cancel)) | |
1016 | goto out; | |
1017 | ||
45bb912b | 1018 | if (unlikely((e->flags & EE_WAS_ERROR) != 0)) |
b411b363 PR |
1019 | goto out; |
1020 | ||
1021 | digest_size = crypto_hash_digestsize(mdev->verify_tfm); | |
1022 | /* FIXME if this allocation fails, online verify will not terminate! */ | |
1023 | digest = kmalloc(digest_size, GFP_NOIO); | |
1024 | if (digest) { | |
45bb912b | 1025 | drbd_csum_ee(mdev, mdev->verify_tfm, e, digest); |
b411b363 PR |
1026 | inc_rs_pending(mdev); |
1027 | ok = drbd_send_drequest_csum(mdev, e->sector, e->size, | |
1028 | digest, digest_size, P_OV_REPLY); | |
1029 | if (!ok) | |
1030 | dec_rs_pending(mdev); | |
1031 | kfree(digest); | |
1032 | } | |
1033 | ||
1034 | out: | |
1035 | drbd_free_ee(mdev, e); | |
1036 | ||
1037 | dec_unacked(mdev); | |
1038 | ||
1039 | return ok; | |
1040 | } | |
1041 | ||
1042 | void drbd_ov_oos_found(struct drbd_conf *mdev, sector_t sector, int size) | |
1043 | { | |
1044 | if (mdev->ov_last_oos_start + mdev->ov_last_oos_size == sector) { | |
1045 | mdev->ov_last_oos_size += size>>9; | |
1046 | } else { | |
1047 | mdev->ov_last_oos_start = sector; | |
1048 | mdev->ov_last_oos_size = size>>9; | |
1049 | } | |
1050 | drbd_set_out_of_sync(mdev, sector, size); | |
1051 | set_bit(WRITE_BM_AFTER_RESYNC, &mdev->flags); | |
1052 | } | |
1053 | ||
1054 | int w_e_end_ov_reply(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1055 | { | |
1056 | struct drbd_epoch_entry *e = container_of(w, struct drbd_epoch_entry, w); | |
1057 | struct digest_info *di; | |
1058 | int digest_size; | |
1059 | void *digest; | |
1060 | int ok, eq = 0; | |
1061 | ||
1062 | if (unlikely(cancel)) { | |
1063 | drbd_free_ee(mdev, e); | |
1064 | dec_unacked(mdev); | |
1065 | return 1; | |
1066 | } | |
1067 | ||
1068 | /* after "cancel", because after drbd_disconnect/drbd_rs_cancel_all | |
1069 | * the resync lru has been cleaned up already */ | |
1070 | drbd_rs_complete_io(mdev, e->sector); | |
1071 | ||
1072 | di = (struct digest_info *)(unsigned long)e->block_id; | |
1073 | ||
45bb912b | 1074 | if (likely((e->flags & EE_WAS_ERROR) == 0)) { |
b411b363 PR |
1075 | digest_size = crypto_hash_digestsize(mdev->verify_tfm); |
1076 | digest = kmalloc(digest_size, GFP_NOIO); | |
1077 | if (digest) { | |
45bb912b | 1078 | drbd_csum_ee(mdev, mdev->verify_tfm, e, digest); |
b411b363 PR |
1079 | |
1080 | D_ASSERT(digest_size == di->digest_size); | |
1081 | eq = !memcmp(digest, di->digest, digest_size); | |
1082 | kfree(digest); | |
1083 | } | |
1084 | } else { | |
1085 | ok = drbd_send_ack(mdev, P_NEG_RS_DREPLY, e); | |
1086 | if (__ratelimit(&drbd_ratelimit_state)) | |
1087 | dev_err(DEV, "Sending NegDReply. I guess it gets messy.\n"); | |
1088 | } | |
1089 | ||
1090 | dec_unacked(mdev); | |
1091 | ||
1092 | kfree(di); | |
1093 | ||
1094 | if (!eq) | |
1095 | drbd_ov_oos_found(mdev, e->sector, e->size); | |
1096 | else | |
1097 | ov_oos_print(mdev); | |
1098 | ||
1099 | ok = drbd_send_ack_ex(mdev, P_OV_RESULT, e->sector, e->size, | |
1100 | eq ? ID_IN_SYNC : ID_OUT_OF_SYNC); | |
1101 | ||
1102 | drbd_free_ee(mdev, e); | |
1103 | ||
1104 | if (--mdev->ov_left == 0) { | |
1105 | ov_oos_print(mdev); | |
1106 | drbd_resync_finished(mdev); | |
1107 | } | |
1108 | ||
1109 | return ok; | |
1110 | } | |
1111 | ||
1112 | int w_prev_work_done(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1113 | { | |
1114 | struct drbd_wq_barrier *b = container_of(w, struct drbd_wq_barrier, w); | |
1115 | complete(&b->done); | |
1116 | return 1; | |
1117 | } | |
1118 | ||
1119 | int w_send_barrier(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1120 | { | |
1121 | struct drbd_tl_epoch *b = container_of(w, struct drbd_tl_epoch, w); | |
1122 | struct p_barrier *p = &mdev->data.sbuf.barrier; | |
1123 | int ok = 1; | |
1124 | ||
1125 | /* really avoid racing with tl_clear. w.cb may have been referenced | |
1126 | * just before it was reassigned and re-queued, so double check that. | |
1127 | * actually, this race was harmless, since we only try to send the | |
1128 | * barrier packet here, and otherwise do nothing with the object. | |
1129 | * but compare with the head of w_clear_epoch */ | |
1130 | spin_lock_irq(&mdev->req_lock); | |
1131 | if (w->cb != w_send_barrier || mdev->state.conn < C_CONNECTED) | |
1132 | cancel = 1; | |
1133 | spin_unlock_irq(&mdev->req_lock); | |
1134 | if (cancel) | |
1135 | return 1; | |
1136 | ||
1137 | if (!drbd_get_data_sock(mdev)) | |
1138 | return 0; | |
1139 | p->barrier = b->br_number; | |
1140 | /* inc_ap_pending was done where this was queued. | |
1141 | * dec_ap_pending will be done in got_BarrierAck | |
1142 | * or (on connection loss) in w_clear_epoch. */ | |
1143 | ok = _drbd_send_cmd(mdev, mdev->data.socket, P_BARRIER, | |
1144 | (struct p_header *)p, sizeof(*p), 0); | |
1145 | drbd_put_data_sock(mdev); | |
1146 | ||
1147 | return ok; | |
1148 | } | |
1149 | ||
1150 | int w_send_write_hint(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1151 | { | |
1152 | if (cancel) | |
1153 | return 1; | |
1154 | return drbd_send_short_cmd(mdev, P_UNPLUG_REMOTE); | |
1155 | } | |
1156 | ||
1157 | /** | |
1158 | * w_send_dblock() - Worker callback to send a P_DATA packet in order to mirror a write request | |
1159 | * @mdev: DRBD device. | |
1160 | * @w: work object. | |
1161 | * @cancel: The connection will be closed anyways | |
1162 | */ | |
1163 | int w_send_dblock(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1164 | { | |
1165 | struct drbd_request *req = container_of(w, struct drbd_request, w); | |
1166 | int ok; | |
1167 | ||
1168 | if (unlikely(cancel)) { | |
1169 | req_mod(req, send_canceled); | |
1170 | return 1; | |
1171 | } | |
1172 | ||
1173 | ok = drbd_send_dblock(mdev, req); | |
1174 | req_mod(req, ok ? handed_over_to_network : send_failed); | |
1175 | ||
1176 | return ok; | |
1177 | } | |
1178 | ||
1179 | /** | |
1180 | * w_send_read_req() - Worker callback to send a read request (P_DATA_REQUEST) packet | |
1181 | * @mdev: DRBD device. | |
1182 | * @w: work object. | |
1183 | * @cancel: The connection will be closed anyways | |
1184 | */ | |
1185 | int w_send_read_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) | |
1186 | { | |
1187 | struct drbd_request *req = container_of(w, struct drbd_request, w); | |
1188 | int ok; | |
1189 | ||
1190 | if (unlikely(cancel)) { | |
1191 | req_mod(req, send_canceled); | |
1192 | return 1; | |
1193 | } | |
1194 | ||
1195 | ok = drbd_send_drequest(mdev, P_DATA_REQUEST, req->sector, req->size, | |
1196 | (unsigned long)req); | |
1197 | ||
1198 | if (!ok) { | |
1199 | /* ?? we set C_TIMEOUT or C_BROKEN_PIPE in drbd_send(); | |
1200 | * so this is probably redundant */ | |
1201 | if (mdev->state.conn >= C_CONNECTED) | |
1202 | drbd_force_state(mdev, NS(conn, C_NETWORK_FAILURE)); | |
1203 | } | |
1204 | req_mod(req, ok ? handed_over_to_network : send_failed); | |
1205 | ||
1206 | return ok; | |
1207 | } | |
1208 | ||
1209 | static int _drbd_may_sync_now(struct drbd_conf *mdev) | |
1210 | { | |
1211 | struct drbd_conf *odev = mdev; | |
1212 | ||
1213 | while (1) { | |
1214 | if (odev->sync_conf.after == -1) | |
1215 | return 1; | |
1216 | odev = minor_to_mdev(odev->sync_conf.after); | |
1217 | ERR_IF(!odev) return 1; | |
1218 | if ((odev->state.conn >= C_SYNC_SOURCE && | |
1219 | odev->state.conn <= C_PAUSED_SYNC_T) || | |
1220 | odev->state.aftr_isp || odev->state.peer_isp || | |
1221 | odev->state.user_isp) | |
1222 | return 0; | |
1223 | } | |
1224 | } | |
1225 | ||
1226 | /** | |
1227 | * _drbd_pause_after() - Pause resync on all devices that may not resync now | |
1228 | * @mdev: DRBD device. | |
1229 | * | |
1230 | * Called from process context only (admin command and after_state_ch). | |
1231 | */ | |
1232 | static int _drbd_pause_after(struct drbd_conf *mdev) | |
1233 | { | |
1234 | struct drbd_conf *odev; | |
1235 | int i, rv = 0; | |
1236 | ||
1237 | for (i = 0; i < minor_count; i++) { | |
1238 | odev = minor_to_mdev(i); | |
1239 | if (!odev) | |
1240 | continue; | |
1241 | if (odev->state.conn == C_STANDALONE && odev->state.disk == D_DISKLESS) | |
1242 | continue; | |
1243 | if (!_drbd_may_sync_now(odev)) | |
1244 | rv |= (__drbd_set_state(_NS(odev, aftr_isp, 1), CS_HARD, NULL) | |
1245 | != SS_NOTHING_TO_DO); | |
1246 | } | |
1247 | ||
1248 | return rv; | |
1249 | } | |
1250 | ||
1251 | /** | |
1252 | * _drbd_resume_next() - Resume resync on all devices that may resync now | |
1253 | * @mdev: DRBD device. | |
1254 | * | |
1255 | * Called from process context only (admin command and worker). | |
1256 | */ | |
1257 | static int _drbd_resume_next(struct drbd_conf *mdev) | |
1258 | { | |
1259 | struct drbd_conf *odev; | |
1260 | int i, rv = 0; | |
1261 | ||
1262 | for (i = 0; i < minor_count; i++) { | |
1263 | odev = minor_to_mdev(i); | |
1264 | if (!odev) | |
1265 | continue; | |
1266 | if (odev->state.conn == C_STANDALONE && odev->state.disk == D_DISKLESS) | |
1267 | continue; | |
1268 | if (odev->state.aftr_isp) { | |
1269 | if (_drbd_may_sync_now(odev)) | |
1270 | rv |= (__drbd_set_state(_NS(odev, aftr_isp, 0), | |
1271 | CS_HARD, NULL) | |
1272 | != SS_NOTHING_TO_DO) ; | |
1273 | } | |
1274 | } | |
1275 | return rv; | |
1276 | } | |
1277 | ||
1278 | void resume_next_sg(struct drbd_conf *mdev) | |
1279 | { | |
1280 | write_lock_irq(&global_state_lock); | |
1281 | _drbd_resume_next(mdev); | |
1282 | write_unlock_irq(&global_state_lock); | |
1283 | } | |
1284 | ||
1285 | void suspend_other_sg(struct drbd_conf *mdev) | |
1286 | { | |
1287 | write_lock_irq(&global_state_lock); | |
1288 | _drbd_pause_after(mdev); | |
1289 | write_unlock_irq(&global_state_lock); | |
1290 | } | |
1291 | ||
1292 | static int sync_after_error(struct drbd_conf *mdev, int o_minor) | |
1293 | { | |
1294 | struct drbd_conf *odev; | |
1295 | ||
1296 | if (o_minor == -1) | |
1297 | return NO_ERROR; | |
1298 | if (o_minor < -1 || minor_to_mdev(o_minor) == NULL) | |
1299 | return ERR_SYNC_AFTER; | |
1300 | ||
1301 | /* check for loops */ | |
1302 | odev = minor_to_mdev(o_minor); | |
1303 | while (1) { | |
1304 | if (odev == mdev) | |
1305 | return ERR_SYNC_AFTER_CYCLE; | |
1306 | ||
1307 | /* dependency chain ends here, no cycles. */ | |
1308 | if (odev->sync_conf.after == -1) | |
1309 | return NO_ERROR; | |
1310 | ||
1311 | /* follow the dependency chain */ | |
1312 | odev = minor_to_mdev(odev->sync_conf.after); | |
1313 | } | |
1314 | } | |
1315 | ||
1316 | int drbd_alter_sa(struct drbd_conf *mdev, int na) | |
1317 | { | |
1318 | int changes; | |
1319 | int retcode; | |
1320 | ||
1321 | write_lock_irq(&global_state_lock); | |
1322 | retcode = sync_after_error(mdev, na); | |
1323 | if (retcode == NO_ERROR) { | |
1324 | mdev->sync_conf.after = na; | |
1325 | do { | |
1326 | changes = _drbd_pause_after(mdev); | |
1327 | changes |= _drbd_resume_next(mdev); | |
1328 | } while (changes); | |
1329 | } | |
1330 | write_unlock_irq(&global_state_lock); | |
1331 | return retcode; | |
1332 | } | |
1333 | ||
309d1608 PR |
1334 | static void ping_peer(struct drbd_conf *mdev) |
1335 | { | |
1336 | clear_bit(GOT_PING_ACK, &mdev->flags); | |
1337 | request_ping(mdev); | |
1338 | wait_event(mdev->misc_wait, | |
1339 | test_bit(GOT_PING_ACK, &mdev->flags) || mdev->state.conn < C_CONNECTED); | |
1340 | } | |
1341 | ||
b411b363 PR |
1342 | /** |
1343 | * drbd_start_resync() - Start the resync process | |
1344 | * @mdev: DRBD device. | |
1345 | * @side: Either C_SYNC_SOURCE or C_SYNC_TARGET | |
1346 | * | |
1347 | * This function might bring you directly into one of the | |
1348 | * C_PAUSED_SYNC_* states. | |
1349 | */ | |
1350 | void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) | |
1351 | { | |
1352 | union drbd_state ns; | |
1353 | int r; | |
1354 | ||
1355 | if (mdev->state.conn >= C_SYNC_SOURCE) { | |
1356 | dev_err(DEV, "Resync already running!\n"); | |
1357 | return; | |
1358 | } | |
1359 | ||
b411b363 PR |
1360 | /* In case a previous resync run was aborted by an IO error/detach on the peer. */ |
1361 | drbd_rs_cancel_all(mdev); | |
1362 | ||
1363 | if (side == C_SYNC_TARGET) { | |
1364 | /* Since application IO was locked out during C_WF_BITMAP_T and | |
1365 | C_WF_SYNC_UUID we are still unmodified. Before going to C_SYNC_TARGET | |
1366 | we check that we might make the data inconsistent. */ | |
1367 | r = drbd_khelper(mdev, "before-resync-target"); | |
1368 | r = (r >> 8) & 0xff; | |
1369 | if (r > 0) { | |
1370 | dev_info(DEV, "before-resync-target handler returned %d, " | |
1371 | "dropping connection.\n", r); | |
1372 | drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); | |
1373 | return; | |
1374 | } | |
1375 | } | |
1376 | ||
1377 | drbd_state_lock(mdev); | |
1378 | ||
1379 | if (!get_ldev_if_state(mdev, D_NEGOTIATING)) { | |
1380 | drbd_state_unlock(mdev); | |
1381 | return; | |
1382 | } | |
1383 | ||
1384 | if (side == C_SYNC_TARGET) { | |
1385 | mdev->bm_resync_fo = 0; | |
1386 | } else /* side == C_SYNC_SOURCE */ { | |
1387 | u64 uuid; | |
1388 | ||
1389 | get_random_bytes(&uuid, sizeof(u64)); | |
1390 | drbd_uuid_set(mdev, UI_BITMAP, uuid); | |
1391 | drbd_send_sync_uuid(mdev, uuid); | |
1392 | ||
1393 | D_ASSERT(mdev->state.disk == D_UP_TO_DATE); | |
1394 | } | |
1395 | ||
1396 | write_lock_irq(&global_state_lock); | |
1397 | ns = mdev->state; | |
1398 | ||
1399 | ns.aftr_isp = !_drbd_may_sync_now(mdev); | |
1400 | ||
1401 | ns.conn = side; | |
1402 | ||
1403 | if (side == C_SYNC_TARGET) | |
1404 | ns.disk = D_INCONSISTENT; | |
1405 | else /* side == C_SYNC_SOURCE */ | |
1406 | ns.pdsk = D_INCONSISTENT; | |
1407 | ||
1408 | r = __drbd_set_state(mdev, ns, CS_VERBOSE, NULL); | |
1409 | ns = mdev->state; | |
1410 | ||
1411 | if (ns.conn < C_CONNECTED) | |
1412 | r = SS_UNKNOWN_ERROR; | |
1413 | ||
1414 | if (r == SS_SUCCESS) { | |
1415 | mdev->rs_total = | |
1416 | mdev->rs_mark_left = drbd_bm_total_weight(mdev); | |
1417 | mdev->rs_failed = 0; | |
1418 | mdev->rs_paused = 0; | |
1419 | mdev->rs_start = | |
1420 | mdev->rs_mark_time = jiffies; | |
1421 | mdev->rs_same_csum = 0; | |
1422 | _drbd_pause_after(mdev); | |
1423 | } | |
1424 | write_unlock_irq(&global_state_lock); | |
b411b363 PR |
1425 | put_ldev(mdev); |
1426 | ||
1427 | if (r == SS_SUCCESS) { | |
1428 | dev_info(DEV, "Began resync as %s (will sync %lu KB [%lu bits set]).\n", | |
1429 | drbd_conn_str(ns.conn), | |
1430 | (unsigned long) mdev->rs_total << (BM_BLOCK_SHIFT-10), | |
1431 | (unsigned long) mdev->rs_total); | |
1432 | ||
1433 | if (mdev->rs_total == 0) { | |
1434 | /* Peer still reachable? Beware of failing before-resync-target handlers! */ | |
309d1608 | 1435 | ping_peer(mdev); |
b411b363 | 1436 | drbd_resync_finished(mdev); |
b411b363 PR |
1437 | } |
1438 | ||
1439 | /* ns.conn may already be != mdev->state.conn, | |
1440 | * we may have been paused in between, or become paused until | |
1441 | * the timer triggers. | |
1442 | * No matter, that is handled in resync_timer_fn() */ | |
1443 | if (ns.conn == C_SYNC_TARGET) | |
1444 | mod_timer(&mdev->resync_timer, jiffies); | |
1445 | ||
1446 | drbd_md_sync(mdev); | |
1447 | } | |
d0c3f60f | 1448 | drbd_state_unlock(mdev); |
b411b363 PR |
1449 | } |
1450 | ||
1451 | int drbd_worker(struct drbd_thread *thi) | |
1452 | { | |
1453 | struct drbd_conf *mdev = thi->mdev; | |
1454 | struct drbd_work *w = NULL; | |
1455 | LIST_HEAD(work_list); | |
1456 | int intr = 0, i; | |
1457 | ||
1458 | sprintf(current->comm, "drbd%d_worker", mdev_to_minor(mdev)); | |
1459 | ||
1460 | while (get_t_state(thi) == Running) { | |
1461 | drbd_thread_current_set_cpu(mdev); | |
1462 | ||
1463 | if (down_trylock(&mdev->data.work.s)) { | |
1464 | mutex_lock(&mdev->data.mutex); | |
1465 | if (mdev->data.socket && !mdev->net_conf->no_cork) | |
1466 | drbd_tcp_uncork(mdev->data.socket); | |
1467 | mutex_unlock(&mdev->data.mutex); | |
1468 | ||
1469 | intr = down_interruptible(&mdev->data.work.s); | |
1470 | ||
1471 | mutex_lock(&mdev->data.mutex); | |
1472 | if (mdev->data.socket && !mdev->net_conf->no_cork) | |
1473 | drbd_tcp_cork(mdev->data.socket); | |
1474 | mutex_unlock(&mdev->data.mutex); | |
1475 | } | |
1476 | ||
1477 | if (intr) { | |
1478 | D_ASSERT(intr == -EINTR); | |
1479 | flush_signals(current); | |
1480 | ERR_IF (get_t_state(thi) == Running) | |
1481 | continue; | |
1482 | break; | |
1483 | } | |
1484 | ||
1485 | if (get_t_state(thi) != Running) | |
1486 | break; | |
1487 | /* With this break, we have done a down() but not consumed | |
1488 | the entry from the list. The cleanup code takes care of | |
1489 | this... */ | |
1490 | ||
1491 | w = NULL; | |
1492 | spin_lock_irq(&mdev->data.work.q_lock); | |
1493 | ERR_IF(list_empty(&mdev->data.work.q)) { | |
1494 | /* something terribly wrong in our logic. | |
1495 | * we were able to down() the semaphore, | |
1496 | * but the list is empty... doh. | |
1497 | * | |
1498 | * what is the best thing to do now? | |
1499 | * try again from scratch, restarting the receiver, | |
1500 | * asender, whatnot? could break even more ugly, | |
1501 | * e.g. when we are primary, but no good local data. | |
1502 | * | |
1503 | * I'll try to get away just starting over this loop. | |
1504 | */ | |
1505 | spin_unlock_irq(&mdev->data.work.q_lock); | |
1506 | continue; | |
1507 | } | |
1508 | w = list_entry(mdev->data.work.q.next, struct drbd_work, list); | |
1509 | list_del_init(&w->list); | |
1510 | spin_unlock_irq(&mdev->data.work.q_lock); | |
1511 | ||
1512 | if (!w->cb(mdev, w, mdev->state.conn < C_CONNECTED)) { | |
1513 | /* dev_warn(DEV, "worker: a callback failed! \n"); */ | |
1514 | if (mdev->state.conn >= C_CONNECTED) | |
1515 | drbd_force_state(mdev, | |
1516 | NS(conn, C_NETWORK_FAILURE)); | |
1517 | } | |
1518 | } | |
1519 | D_ASSERT(test_bit(DEVICE_DYING, &mdev->flags)); | |
1520 | D_ASSERT(test_bit(CONFIG_PENDING, &mdev->flags)); | |
1521 | ||
1522 | spin_lock_irq(&mdev->data.work.q_lock); | |
1523 | i = 0; | |
1524 | while (!list_empty(&mdev->data.work.q)) { | |
1525 | list_splice_init(&mdev->data.work.q, &work_list); | |
1526 | spin_unlock_irq(&mdev->data.work.q_lock); | |
1527 | ||
1528 | while (!list_empty(&work_list)) { | |
1529 | w = list_entry(work_list.next, struct drbd_work, list); | |
1530 | list_del_init(&w->list); | |
1531 | w->cb(mdev, w, 1); | |
1532 | i++; /* dead debugging code */ | |
1533 | } | |
1534 | ||
1535 | spin_lock_irq(&mdev->data.work.q_lock); | |
1536 | } | |
1537 | sema_init(&mdev->data.work.s, 0); | |
1538 | /* DANGEROUS race: if someone did queue his work within the spinlock, | |
1539 | * but up() ed outside the spinlock, we could get an up() on the | |
1540 | * semaphore without corresponding list entry. | |
1541 | * So don't do that. | |
1542 | */ | |
1543 | spin_unlock_irq(&mdev->data.work.q_lock); | |
1544 | ||
1545 | D_ASSERT(mdev->state.disk == D_DISKLESS && mdev->state.conn == C_STANDALONE); | |
1546 | /* _drbd_set_state only uses stop_nowait. | |
1547 | * wait here for the Exiting receiver. */ | |
1548 | drbd_thread_stop(&mdev->receiver); | |
1549 | drbd_mdev_cleanup(mdev); | |
1550 | ||
1551 | dev_info(DEV, "worker terminated\n"); | |
1552 | ||
1553 | clear_bit(DEVICE_DYING, &mdev->flags); | |
1554 | clear_bit(CONFIG_PENDING, &mdev->flags); | |
1555 | wake_up(&mdev->state_wait); | |
1556 | ||
1557 | return 0; | |
1558 | } |