blob: 29eba7219b01a17fd013109eff759b06d6fac34e [file] [log] [blame]
Kent Overstreetcafe5632013-03-23 16:11:31 -07001/*
2 * bcache journalling code, for btree insertions
3 *
4 * Copyright 2012 Google, Inc.
5 */
6
7#include "bcache.h"
8#include "btree.h"
9#include "debug.h"
Kent Overstreet9aa61a92014-04-10 17:58:49 -070010#include "extents.h"
Kent Overstreetcafe5632013-03-23 16:11:31 -070011
Kent Overstreetc37511b2013-04-26 15:39:55 -070012#include <trace/events/bcache.h>
13
Kent Overstreetcafe5632013-03-23 16:11:31 -070014/*
15 * Journal replay/recovery:
16 *
17 * This code is all driven from run_cache_set(); we first read the journal
18 * entries, do some other stuff, then we mark all the keys in the journal
19 * entries (same as garbage collection would), then we replay them - reinserting
20 * them into the cache in precisely the same order as they appear in the
21 * journal.
22 *
23 * We only journal keys that go in leaf nodes, which simplifies things quite a
24 * bit.
25 */
26
Christoph Hellwig4246a0b2015-07-20 15:29:37 +020027static void journal_read_endio(struct bio *bio)
Kent Overstreetcafe5632013-03-23 16:11:31 -070028{
29 struct closure *cl = bio->bi_private;
30 closure_put(cl);
31}
32
33static int journal_read_bucket(struct cache *ca, struct list_head *list,
Kent Overstreetc18536a2013-07-24 17:44:17 -070034 unsigned bucket_index)
Kent Overstreetcafe5632013-03-23 16:11:31 -070035{
36 struct journal_device *ja = &ca->journal;
37 struct bio *bio = &ja->bio;
38
39 struct journal_replay *i;
40 struct jset *j, *data = ca->set->journal.w[0].data;
Kent Overstreetc18536a2013-07-24 17:44:17 -070041 struct closure cl;
Kent Overstreetcafe5632013-03-23 16:11:31 -070042 unsigned len, left, offset = 0;
43 int ret = 0;
44 sector_t bucket = bucket_to_sector(ca->set, ca->sb.d[bucket_index]);
45
Kent Overstreetc18536a2013-07-24 17:44:17 -070046 closure_init_stack(&cl);
47
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070048 pr_debug("reading %u", bucket_index);
Kent Overstreetcafe5632013-03-23 16:11:31 -070049
50 while (offset < ca->sb.bucket_size) {
51reread: left = ca->sb.bucket_size - offset;
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070052 len = min_t(unsigned, left, PAGE_SECTORS << JSET_BITS);
Kent Overstreetcafe5632013-03-23 16:11:31 -070053
54 bio_reset(bio);
Kent Overstreet4f024f32013-10-11 15:44:27 -070055 bio->bi_iter.bi_sector = bucket + offset;
Kent Overstreetcafe5632013-03-23 16:11:31 -070056 bio->bi_bdev = ca->bdev;
57 bio->bi_rw = READ;
Kent Overstreet4f024f32013-10-11 15:44:27 -070058 bio->bi_iter.bi_size = len << 9;
Kent Overstreetcafe5632013-03-23 16:11:31 -070059
60 bio->bi_end_io = journal_read_endio;
Kent Overstreetc18536a2013-07-24 17:44:17 -070061 bio->bi_private = &cl;
Kent Overstreet169ef1c2013-03-28 12:50:55 -060062 bch_bio_map(bio, data);
Kent Overstreetcafe5632013-03-23 16:11:31 -070063
Kent Overstreet749b61d2013-11-23 23:11:25 -080064 closure_bio_submit(bio, &cl);
Kent Overstreetc18536a2013-07-24 17:44:17 -070065 closure_sync(&cl);
Kent Overstreetcafe5632013-03-23 16:11:31 -070066
67 /* This function could be simpler now since we no longer write
68 * journal entries that overlap bucket boundaries; this means
69 * the start of a bucket will always have a valid journal entry
70 * if it has any journal entries at all.
71 */
72
73 j = data;
74 while (len) {
75 struct list_head *where;
76 size_t blocks, bytes = set_bytes(j);
77
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070078 if (j->magic != jset_magic(&ca->sb)) {
79 pr_debug("%u: bad magic", bucket_index);
Kent Overstreetcafe5632013-03-23 16:11:31 -070080 return ret;
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070081 }
Kent Overstreetcafe5632013-03-23 16:11:31 -070082
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070083 if (bytes > left << 9 ||
84 bytes > PAGE_SIZE << JSET_BITS) {
85 pr_info("%u: too big, %zu bytes, offset %u",
86 bucket_index, bytes, offset);
Kent Overstreetcafe5632013-03-23 16:11:31 -070087 return ret;
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070088 }
Kent Overstreetcafe5632013-03-23 16:11:31 -070089
90 if (bytes > len << 9)
91 goto reread;
92
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070093 if (j->csum != csum_set(j)) {
94 pr_info("%u: bad csum, %zu bytes, offset %u",
95 bucket_index, bytes, offset);
Kent Overstreetcafe5632013-03-23 16:11:31 -070096 return ret;
Kent Overstreetb3fa7e72013-08-05 14:04:06 -070097 }
Kent Overstreetcafe5632013-03-23 16:11:31 -070098
Kent Overstreetee811282013-12-17 23:49:49 -080099 blocks = set_blocks(j, block_bytes(ca->set));
Kent Overstreetcafe5632013-03-23 16:11:31 -0700100
101 while (!list_empty(list)) {
102 i = list_first_entry(list,
103 struct journal_replay, list);
104 if (i->j.seq >= j->last_seq)
105 break;
106 list_del(&i->list);
107 kfree(i);
108 }
109
110 list_for_each_entry_reverse(i, list, list) {
111 if (j->seq == i->j.seq)
112 goto next_set;
113
114 if (j->seq < i->j.last_seq)
115 goto next_set;
116
117 if (j->seq > i->j.seq) {
118 where = &i->list;
119 goto add;
120 }
121 }
122
123 where = list;
124add:
125 i = kmalloc(offsetof(struct journal_replay, j) +
126 bytes, GFP_KERNEL);
127 if (!i)
128 return -ENOMEM;
129 memcpy(&i->j, j, bytes);
130 list_add(&i->list, where);
131 ret = 1;
132
133 ja->seq[bucket_index] = j->seq;
134next_set:
135 offset += blocks * ca->sb.block_size;
136 len -= blocks * ca->sb.block_size;
137 j = ((void *) j) + blocks * block_bytes(ca);
138 }
139 }
140
141 return ret;
142}
143
Kent Overstreetc18536a2013-07-24 17:44:17 -0700144int bch_journal_read(struct cache_set *c, struct list_head *list)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700145{
146#define read_bucket(b) \
147 ({ \
Kent Overstreetc18536a2013-07-24 17:44:17 -0700148 int ret = journal_read_bucket(ca, list, b); \
Kent Overstreetcafe5632013-03-23 16:11:31 -0700149 __set_bit(b, bitmap); \
150 if (ret < 0) \
151 return ret; \
152 ret; \
153 })
154
155 struct cache *ca;
156 unsigned iter;
157
158 for_each_cache(ca, c, iter) {
159 struct journal_device *ja = &ca->journal;
Joe Perchesd1aa1ab2015-06-30 14:59:54 -0700160 DECLARE_BITMAP(bitmap, SB_JOURNAL_BUCKETS);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700161 unsigned i, l, r, m;
162 uint64_t seq;
163
164 bitmap_zero(bitmap, SB_JOURNAL_BUCKETS);
165 pr_debug("%u journal buckets", ca->sb.njournal_buckets);
166
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700167 /*
168 * Read journal buckets ordered by golden ratio hash to quickly
Kent Overstreetcafe5632013-03-23 16:11:31 -0700169 * find a sequence of buckets with valid journal entries
170 */
171 for (i = 0; i < ca->sb.njournal_buckets; i++) {
172 l = (i * 2654435769U) % ca->sb.njournal_buckets;
173
174 if (test_bit(l, bitmap))
175 break;
176
177 if (read_bucket(l))
178 goto bsearch;
179 }
180
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700181 /*
182 * If that fails, check all the buckets we haven't checked
Kent Overstreetcafe5632013-03-23 16:11:31 -0700183 * already
184 */
185 pr_debug("falling back to linear search");
186
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700187 for (l = find_first_zero_bit(bitmap, ca->sb.njournal_buckets);
188 l < ca->sb.njournal_buckets;
189 l = find_next_zero_bit(bitmap, ca->sb.njournal_buckets, l + 1))
Kent Overstreetcafe5632013-03-23 16:11:31 -0700190 if (read_bucket(l))
191 goto bsearch;
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700192
Kent Overstreet6b708de2014-06-02 15:39:44 -0700193 /* no journal entries on this device? */
194 if (l == ca->sb.njournal_buckets)
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700195 continue;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700196bsearch:
Kent Overstreet6b708de2014-06-02 15:39:44 -0700197 BUG_ON(list_empty(list));
198
Kent Overstreetcafe5632013-03-23 16:11:31 -0700199 /* Binary search */
Surbhi Palandedbd810a2014-04-10 16:09:51 -0700200 m = l;
201 r = find_next_bit(bitmap, ca->sb.njournal_buckets, l + 1);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700202 pr_debug("starting binary search, l %u r %u", l, r);
203
204 while (l + 1 < r) {
Kent Overstreetfaa56732013-07-11 22:42:14 -0700205 seq = list_entry(list->prev, struct journal_replay,
206 list)->j.seq;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700207
Kent Overstreetfaa56732013-07-11 22:42:14 -0700208 m = (l + r) >> 1;
209 read_bucket(m);
210
211 if (seq != list_entry(list->prev, struct journal_replay,
212 list)->j.seq)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700213 l = m;
214 else
215 r = m;
216 }
217
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700218 /*
219 * Read buckets in reverse order until we stop finding more
Kent Overstreetcafe5632013-03-23 16:11:31 -0700220 * journal entries
221 */
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700222 pr_debug("finishing up: m %u njournal_buckets %u",
223 m, ca->sb.njournal_buckets);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700224 l = m;
225
226 while (1) {
227 if (!l--)
228 l = ca->sb.njournal_buckets - 1;
229
230 if (l == m)
231 break;
232
233 if (test_bit(l, bitmap))
234 continue;
235
236 if (!read_bucket(l))
237 break;
238 }
239
240 seq = 0;
241
242 for (i = 0; i < ca->sb.njournal_buckets; i++)
243 if (ja->seq[i] > seq) {
244 seq = ja->seq[i];
Kent Overstreet27201cfd2014-03-13 13:44:21 -0700245 /*
246 * When journal_reclaim() goes to allocate for
247 * the first time, it'll use the bucket after
248 * ja->cur_idx
249 */
250 ja->cur_idx = i;
251 ja->last_idx = ja->discard_idx = (i + 1) %
252 ca->sb.njournal_buckets;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700253
254 }
255 }
256
Kent Overstreetc426c4f2013-09-23 23:17:29 -0700257 if (!list_empty(list))
258 c->journal.seq = list_entry(list->prev,
259 struct journal_replay,
260 list)->j.seq;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700261
262 return 0;
263#undef read_bucket
264}
265
266void bch_journal_mark(struct cache_set *c, struct list_head *list)
267{
268 atomic_t p = { 0 };
269 struct bkey *k;
270 struct journal_replay *i;
271 struct journal *j = &c->journal;
272 uint64_t last = j->seq;
273
274 /*
275 * journal.pin should never fill up - we never write a journal
276 * entry when it would fill up. But if for some reason it does, we
277 * iterate over the list in reverse order so that we can just skip that
278 * refcount instead of bugging.
279 */
280
281 list_for_each_entry_reverse(i, list, list) {
282 BUG_ON(last < i->j.seq);
283 i->pin = NULL;
284
285 while (last-- != i->j.seq)
286 if (fifo_free(&j->pin) > 1) {
287 fifo_push_front(&j->pin, p);
288 atomic_set(&fifo_front(&j->pin), 0);
289 }
290
291 if (fifo_free(&j->pin) > 1) {
292 fifo_push_front(&j->pin, p);
293 i->pin = &fifo_front(&j->pin);
294 atomic_set(i->pin, 1);
295 }
296
297 for (k = i->j.start;
Kent Overstreetfafff812013-12-17 21:56:21 -0800298 k < bset_bkey_last(&i->j);
Kent Overstreet9aa61a92014-04-10 17:58:49 -0700299 k = bkey_next(k))
300 if (!__bch_extent_invalid(c, k)) {
301 unsigned j;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700302
Kent Overstreet9aa61a92014-04-10 17:58:49 -0700303 for (j = 0; j < KEY_PTRS(k); j++)
304 if (ptr_available(c, k, j))
305 atomic_inc(&PTR_BUCKET(c, k, j)->pin);
Kent Overstreet65ddf452014-02-24 19:55:28 -0800306
Kent Overstreet9aa61a92014-04-10 17:58:49 -0700307 bch_initial_mark_key(c, 0, k);
308 }
Kent Overstreetcafe5632013-03-23 16:11:31 -0700309 }
310}
311
Kent Overstreetc18536a2013-07-24 17:44:17 -0700312int bch_journal_replay(struct cache_set *s, struct list_head *list)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700313{
314 int ret = 0, keys = 0, entries = 0;
315 struct bkey *k;
316 struct journal_replay *i =
317 list_entry(list->prev, struct journal_replay, list);
318
319 uint64_t start = i->j.last_seq, end = i->j.seq, n = start;
Kent Overstreet0b932072013-07-24 17:26:51 -0700320 struct keylist keylist;
321
Kent Overstreetcafe5632013-03-23 16:11:31 -0700322 list_for_each_entry(i, list, list) {
323 BUG_ON(i->pin && atomic_read(i->pin) != 1);
324
Kent Overstreet77c320e2013-07-11 19:42:51 -0700325 cache_set_err_on(n != i->j.seq, s,
326"bcache: journal entries %llu-%llu missing! (replaying %llu-%llu)",
327 n, i->j.seq - 1, start, end);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700328
329 for (k = i->j.start;
Kent Overstreetfafff812013-12-17 21:56:21 -0800330 k < bset_bkey_last(&i->j);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700331 k = bkey_next(k)) {
Kent Overstreetc37511b2013-04-26 15:39:55 -0700332 trace_bcache_journal_replay_key(k);
333
Kent Overstreetc13f3af2014-01-08 21:22:02 -0800334 bch_keylist_init_single(&keylist, k);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700335
Kent Overstreetcc7b8812013-07-24 18:07:22 -0700336 ret = bch_btree_insert(s, &keylist, i->pin, NULL);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700337 if (ret)
338 goto err;
339
Kent Overstreet0b932072013-07-24 17:26:51 -0700340 BUG_ON(!bch_keylist_empty(&keylist));
Kent Overstreetcafe5632013-03-23 16:11:31 -0700341 keys++;
342
343 cond_resched();
344 }
345
346 if (i->pin)
347 atomic_dec(i->pin);
348 n = i->j.seq + 1;
349 entries++;
350 }
351
352 pr_info("journal replay done, %i keys in %i entries, seq %llu",
353 keys, entries, end);
Kent Overstreetb54d6932013-07-24 18:04:18 -0700354err:
Kent Overstreetcafe5632013-03-23 16:11:31 -0700355 while (!list_empty(list)) {
356 i = list_first_entry(list, struct journal_replay, list);
357 list_del(&i->list);
358 kfree(i);
359 }
Kent Overstreetb54d6932013-07-24 18:04:18 -0700360
Kent Overstreetcafe5632013-03-23 16:11:31 -0700361 return ret;
362}
363
364/* Journalling */
365
366static void btree_flush_write(struct cache_set *c)
367{
368 /*
369 * Try to find the btree node with that references the oldest journal
370 * entry, best is our current candidate and is locked if non NULL:
371 */
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700372 struct btree *b, *best;
373 unsigned i;
374retry:
375 best = NULL;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700376
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700377 for_each_cached_btree(b, c, i)
378 if (btree_current_write(b)->journal) {
379 if (!best)
380 best = b;
381 else if (journal_pin_cmp(c,
Kent Overstreetc18536a2013-07-24 17:44:17 -0700382 btree_current_write(best)->journal,
383 btree_current_write(b)->journal)) {
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700384 best = b;
385 }
Kent Overstreetcafe5632013-03-23 16:11:31 -0700386 }
387
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700388 b = best;
389 if (b) {
Kent Overstreet2a285682014-03-04 16:42:42 -0800390 mutex_lock(&b->write_lock);
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700391 if (!btree_current_write(b)->journal) {
Kent Overstreet2a285682014-03-04 16:42:42 -0800392 mutex_unlock(&b->write_lock);
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700393 /* We raced */
394 goto retry;
395 }
396
Kent Overstreet2a285682014-03-04 16:42:42 -0800397 __bch_btree_node_write(b, NULL);
398 mutex_unlock(&b->write_lock);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700399 }
Kent Overstreetcafe5632013-03-23 16:11:31 -0700400}
401
402#define last_seq(j) ((j)->seq - fifo_used(&(j)->pin) + 1)
403
Christoph Hellwig4246a0b2015-07-20 15:29:37 +0200404static void journal_discard_endio(struct bio *bio)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700405{
406 struct journal_device *ja =
407 container_of(bio, struct journal_device, discard_bio);
408 struct cache *ca = container_of(ja, struct cache, journal);
409
410 atomic_set(&ja->discard_in_flight, DISCARD_DONE);
411
412 closure_wake_up(&ca->set->journal.wait);
413 closure_put(&ca->set->cl);
414}
415
416static void journal_discard_work(struct work_struct *work)
417{
418 struct journal_device *ja =
419 container_of(work, struct journal_device, discard_work);
420
421 submit_bio(0, &ja->discard_bio);
422}
423
424static void do_journal_discard(struct cache *ca)
425{
426 struct journal_device *ja = &ca->journal;
427 struct bio *bio = &ja->discard_bio;
428
429 if (!ca->discard) {
430 ja->discard_idx = ja->last_idx;
431 return;
432 }
433
Kent Overstreet6d9d21e2013-09-23 23:17:27 -0700434 switch (atomic_read(&ja->discard_in_flight)) {
Kent Overstreetcafe5632013-03-23 16:11:31 -0700435 case DISCARD_IN_FLIGHT:
436 return;
437
438 case DISCARD_DONE:
439 ja->discard_idx = (ja->discard_idx + 1) %
440 ca->sb.njournal_buckets;
441
442 atomic_set(&ja->discard_in_flight, DISCARD_READY);
443 /* fallthrough */
444
445 case DISCARD_READY:
446 if (ja->discard_idx == ja->last_idx)
447 return;
448
449 atomic_set(&ja->discard_in_flight, DISCARD_IN_FLIGHT);
450
451 bio_init(bio);
Kent Overstreet4f024f32013-10-11 15:44:27 -0700452 bio->bi_iter.bi_sector = bucket_to_sector(ca->set,
Kent Overstreetb1a67b02013-03-25 11:46:44 -0700453 ca->sb.d[ja->discard_idx]);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700454 bio->bi_bdev = ca->bdev;
455 bio->bi_rw = REQ_WRITE|REQ_DISCARD;
456 bio->bi_max_vecs = 1;
457 bio->bi_io_vec = bio->bi_inline_vecs;
Kent Overstreet4f024f32013-10-11 15:44:27 -0700458 bio->bi_iter.bi_size = bucket_bytes(ca);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700459 bio->bi_end_io = journal_discard_endio;
460
461 closure_get(&ca->set->cl);
462 INIT_WORK(&ja->discard_work, journal_discard_work);
463 schedule_work(&ja->discard_work);
464 }
465}
466
467static void journal_reclaim(struct cache_set *c)
468{
469 struct bkey *k = &c->journal.key;
470 struct cache *ca;
471 uint64_t last_seq;
472 unsigned iter, n = 0;
473 atomic_t p;
474
475 while (!atomic_read(&fifo_front(&c->journal.pin)))
476 fifo_pop(&c->journal.pin, p);
477
478 last_seq = last_seq(&c->journal);
479
480 /* Update last_idx */
481
482 for_each_cache(ca, c, iter) {
483 struct journal_device *ja = &ca->journal;
484
485 while (ja->last_idx != ja->cur_idx &&
486 ja->seq[ja->last_idx] < last_seq)
487 ja->last_idx = (ja->last_idx + 1) %
488 ca->sb.njournal_buckets;
489 }
490
491 for_each_cache(ca, c, iter)
492 do_journal_discard(ca);
493
494 if (c->journal.blocks_free)
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700495 goto out;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700496
497 /*
498 * Allocate:
499 * XXX: Sort by free journal space
500 */
501
502 for_each_cache(ca, c, iter) {
503 struct journal_device *ja = &ca->journal;
504 unsigned next = (ja->cur_idx + 1) % ca->sb.njournal_buckets;
505
506 /* No space available on this device */
507 if (next == ja->discard_idx)
508 continue;
509
510 ja->cur_idx = next;
511 k->ptr[n++] = PTR(0,
512 bucket_to_sector(c, ca->sb.d[ja->cur_idx]),
513 ca->sb.nr_this_dev);
514 }
515
516 bkey_init(k);
517 SET_KEY_PTRS(k, n);
518
519 if (n)
520 c->journal.blocks_free = c->sb.bucket_size >> c->block_bits;
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700521out:
Kent Overstreetcafe5632013-03-23 16:11:31 -0700522 if (!journal_full(&c->journal))
523 __closure_wake_up(&c->journal.wait);
524}
525
526void bch_journal_next(struct journal *j)
527{
528 atomic_t p = { 1 };
529
530 j->cur = (j->cur == j->w)
531 ? &j->w[1]
532 : &j->w[0];
533
534 /*
535 * The fifo_push() needs to happen at the same time as j->seq is
536 * incremented for last_seq() to be calculated correctly
537 */
538 BUG_ON(!fifo_push(&j->pin, p));
539 atomic_set(&fifo_back(&j->pin), 1);
540
541 j->cur->data->seq = ++j->seq;
Kent Overstreetdabb4432014-02-19 19:48:26 -0800542 j->cur->dirty = false;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700543 j->cur->need_write = false;
544 j->cur->data->keys = 0;
545
546 if (fifo_full(&j->pin))
547 pr_debug("journal_pin full (%zu)", fifo_used(&j->pin));
548}
549
Christoph Hellwig4246a0b2015-07-20 15:29:37 +0200550static void journal_write_endio(struct bio *bio)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700551{
552 struct journal_write *w = bio->bi_private;
553
Christoph Hellwig4246a0b2015-07-20 15:29:37 +0200554 cache_set_err_on(bio->bi_error, w->c, "journal io error");
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700555 closure_put(&w->c->journal.io);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700556}
557
558static void journal_write(struct closure *);
559
560static void journal_write_done(struct closure *cl)
561{
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700562 struct journal *j = container_of(cl, struct journal, io);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700563 struct journal_write *w = (j->cur == j->w)
564 ? &j->w[1]
565 : &j->w[0];
566
567 __closure_wake_up(&w->wait);
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700568 continue_at_nobarrier(cl, journal_write, system_wq);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700569}
570
Kent Overstreetcb7a5832013-12-16 15:27:25 -0800571static void journal_write_unlock(struct closure *cl)
572{
573 struct cache_set *c = container_of(cl, struct cache_set, journal.io);
574
575 c->journal.io_in_flight = 0;
576 spin_unlock(&c->journal.lock);
577}
578
Kent Overstreetcafe5632013-03-23 16:11:31 -0700579static void journal_write_unlocked(struct closure *cl)
Kent Overstreetc19ed232013-03-26 13:49:02 -0700580 __releases(c->journal.lock)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700581{
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700582 struct cache_set *c = container_of(cl, struct cache_set, journal.io);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700583 struct cache *ca;
584 struct journal_write *w = c->journal.cur;
585 struct bkey *k = &c->journal.key;
Kent Overstreetee811282013-12-17 23:49:49 -0800586 unsigned i, sectors = set_blocks(w->data, block_bytes(c)) *
587 c->sb.block_size;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700588
589 struct bio *bio;
590 struct bio_list list;
591 bio_list_init(&list);
592
593 if (!w->need_write) {
Kent Overstreetcb7a5832013-12-16 15:27:25 -0800594 closure_return_with_destructor(cl, journal_write_unlock);
Jens Axboe77b5a082015-03-06 08:37:46 -0700595 return;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700596 } else if (journal_full(&c->journal)) {
597 journal_reclaim(c);
598 spin_unlock(&c->journal.lock);
599
600 btree_flush_write(c);
601 continue_at(cl, journal_write, system_wq);
Jens Axboe77b5a082015-03-06 08:37:46 -0700602 return;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700603 }
604
Kent Overstreetee811282013-12-17 23:49:49 -0800605 c->journal.blocks_free -= set_blocks(w->data, block_bytes(c));
Kent Overstreetcafe5632013-03-23 16:11:31 -0700606
607 w->data->btree_level = c->root->level;
608
609 bkey_copy(&w->data->btree_root, &c->root->key);
610 bkey_copy(&w->data->uuid_bucket, &c->uuid_bucket);
611
612 for_each_cache(ca, c, i)
613 w->data->prio_bucket[ca->sb.nr_this_dev] = ca->prio_buckets[0];
614
Kent Overstreet81ab4192013-10-31 15:46:42 -0700615 w->data->magic = jset_magic(&c->sb);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700616 w->data->version = BCACHE_JSET_VERSION;
617 w->data->last_seq = last_seq(&c->journal);
618 w->data->csum = csum_set(w->data);
619
620 for (i = 0; i < KEY_PTRS(k); i++) {
621 ca = PTR_CACHE(c, k, i);
622 bio = &ca->journal.bio;
623
624 atomic_long_add(sectors, &ca->meta_sectors_written);
625
626 bio_reset(bio);
Kent Overstreet4f024f32013-10-11 15:44:27 -0700627 bio->bi_iter.bi_sector = PTR_OFFSET(k, i);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700628 bio->bi_bdev = ca->bdev;
Kent Overstreete49c7c32013-06-26 17:25:38 -0700629 bio->bi_rw = REQ_WRITE|REQ_SYNC|REQ_META|REQ_FLUSH|REQ_FUA;
Kent Overstreet4f024f32013-10-11 15:44:27 -0700630 bio->bi_iter.bi_size = sectors << 9;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700631
632 bio->bi_end_io = journal_write_endio;
633 bio->bi_private = w;
Kent Overstreet169ef1c2013-03-28 12:50:55 -0600634 bch_bio_map(bio, w->data);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700635
636 trace_bcache_journal_write(bio);
637 bio_list_add(&list, bio);
638
639 SET_PTR_OFFSET(k, i, PTR_OFFSET(k, i) + sectors);
640
641 ca->journal.seq[ca->journal.cur_idx] = w->data->seq;
642 }
643
644 atomic_dec_bug(&fifo_back(&c->journal.pin));
645 bch_journal_next(&c->journal);
646 journal_reclaim(c);
647
648 spin_unlock(&c->journal.lock);
649
650 while ((bio = bio_list_pop(&list)))
Kent Overstreet749b61d2013-11-23 23:11:25 -0800651 closure_bio_submit(bio, cl);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700652
653 continue_at(cl, journal_write_done, NULL);
654}
655
656static void journal_write(struct closure *cl)
657{
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700658 struct cache_set *c = container_of(cl, struct cache_set, journal.io);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700659
660 spin_lock(&c->journal.lock);
661 journal_write_unlocked(cl);
662}
663
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700664static void journal_try_write(struct cache_set *c)
Kent Overstreetc19ed232013-03-26 13:49:02 -0700665 __releases(c->journal.lock)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700666{
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700667 struct closure *cl = &c->journal.io;
668 struct journal_write *w = c->journal.cur;
669
670 w->need_write = true;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700671
Kent Overstreetcb7a5832013-12-16 15:27:25 -0800672 if (!c->journal.io_in_flight) {
673 c->journal.io_in_flight = 1;
674 closure_call(cl, journal_write_unlocked, NULL, &c->cl);
675 } else {
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700676 spin_unlock(&c->journal.lock);
Kent Overstreetcb7a5832013-12-16 15:27:25 -0800677 }
Kent Overstreetcafe5632013-03-23 16:11:31 -0700678}
679
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700680static struct journal_write *journal_wait_for_write(struct cache_set *c,
681 unsigned nkeys)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700682{
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700683 size_t sectors;
684 struct closure cl;
Kent Overstreet5775e212013-12-10 16:10:46 -0800685 bool wait = false;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700686
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700687 closure_init_stack(&cl);
688
689 spin_lock(&c->journal.lock);
690
691 while (1) {
692 struct journal_write *w = c->journal.cur;
693
694 sectors = __set_blocks(w->data, w->data->keys + nkeys,
Kent Overstreetee811282013-12-17 23:49:49 -0800695 block_bytes(c)) * c->sb.block_size;
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700696
697 if (sectors <= min_t(size_t,
698 c->journal.blocks_free * c->sb.block_size,
699 PAGE_SECTORS << JSET_BITS))
700 return w;
701
Kent Overstreet5775e212013-12-10 16:10:46 -0800702 if (wait)
703 closure_wait(&c->journal.wait, &cl);
704
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700705 if (!journal_full(&c->journal)) {
Kent Overstreet5775e212013-12-10 16:10:46 -0800706 if (wait)
707 trace_bcache_journal_entry_full(c);
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700708
709 /*
710 * XXX: If we were inserting so many keys that they
711 * won't fit in an _empty_ journal write, we'll
712 * deadlock. For now, handle this in
713 * bch_keylist_realloc() - but something to think about.
714 */
715 BUG_ON(!w->data->keys);
716
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700717 journal_try_write(c); /* unlocks */
718 } else {
Kent Overstreet5775e212013-12-10 16:10:46 -0800719 if (wait)
720 trace_bcache_journal_full(c);
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700721
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700722 journal_reclaim(c);
723 spin_unlock(&c->journal.lock);
724
725 btree_flush_write(c);
726 }
727
728 closure_sync(&cl);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700729 spin_lock(&c->journal.lock);
Kent Overstreet5775e212013-12-10 16:10:46 -0800730 wait = true;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700731 }
732}
733
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700734static void journal_write_work(struct work_struct *work)
735{
736 struct cache_set *c = container_of(to_delayed_work(work),
737 struct cache_set,
738 journal.work);
739 spin_lock(&c->journal.lock);
Kent Overstreetdabb4432014-02-19 19:48:26 -0800740 if (c->journal.cur->dirty)
741 journal_try_write(c);
742 else
743 spin_unlock(&c->journal.lock);
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700744}
745
Kent Overstreetcafe5632013-03-23 16:11:31 -0700746/*
747 * Entry point to the journalling code - bio_insert() and btree_invalidate()
748 * pass bch_journal() a list of keys to be journalled, and then
749 * bch_journal() hands those same keys off to btree_insert_async()
750 */
751
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700752atomic_t *bch_journal(struct cache_set *c,
753 struct keylist *keys,
754 struct closure *parent)
Kent Overstreetcafe5632013-03-23 16:11:31 -0700755{
Kent Overstreetcafe5632013-03-23 16:11:31 -0700756 struct journal_write *w;
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700757 atomic_t *ret;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700758
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700759 if (!CACHE_SYNC(&c->sb))
760 return NULL;
Kent Overstreetcafe5632013-03-23 16:11:31 -0700761
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700762 w = journal_wait_for_write(c, bch_keylist_nkeys(keys));
Kent Overstreetcafe5632013-03-23 16:11:31 -0700763
Kent Overstreetfafff812013-12-17 21:56:21 -0800764 memcpy(bset_bkey_last(w->data), keys->keys, bch_keylist_bytes(keys));
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700765 w->data->keys += bch_keylist_nkeys(keys);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700766
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700767 ret = &fifo_back(&c->journal.pin);
768 atomic_inc(ret);
Kent Overstreetc37511b2013-04-26 15:39:55 -0700769
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700770 if (parent) {
771 closure_wait(&w->wait, parent);
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700772 journal_try_write(c);
Kent Overstreetdabb4432014-02-19 19:48:26 -0800773 } else if (!w->dirty) {
774 w->dirty = true;
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700775 schedule_delayed_work(&c->journal.work,
776 msecs_to_jiffies(c->journal_delay_ms));
777 spin_unlock(&c->journal.lock);
778 } else {
779 spin_unlock(&c->journal.lock);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700780 }
Kent Overstreeta34a8bf2013-10-24 17:07:04 -0700781
782
783 return ret;
784}
785
786void bch_journal_meta(struct cache_set *c, struct closure *cl)
787{
788 struct keylist keys;
789 atomic_t *ref;
790
791 bch_keylist_init(&keys);
792
793 ref = bch_journal(c, &keys, cl);
794 if (ref)
795 atomic_dec_bug(ref);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700796}
797
798void bch_journal_free(struct cache_set *c)
799{
800 free_pages((unsigned long) c->journal.w[1].data, JSET_BITS);
801 free_pages((unsigned long) c->journal.w[0].data, JSET_BITS);
802 free_fifo(&c->journal.pin);
803}
804
805int bch_journal_alloc(struct cache_set *c)
806{
807 struct journal *j = &c->journal;
808
Kent Overstreetcafe5632013-03-23 16:11:31 -0700809 spin_lock_init(&j->lock);
Kent Overstreet7857d5d42013-10-08 15:50:46 -0700810 INIT_DELAYED_WORK(&j->work, journal_write_work);
Kent Overstreetcafe5632013-03-23 16:11:31 -0700811
812 c->journal_delay_ms = 100;
813
814 j->w[0].c = c;
815 j->w[1].c = c;
816
817 if (!(init_fifo(&j->pin, JOURNAL_PIN, GFP_KERNEL)) ||
818 !(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)) ||
819 !(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)))
820 return -ENOMEM;
821
822 return 0;
823}