2013-03-24 06:11:31 +07:00
|
|
|
/*
|
|
|
|
* bcache journalling code, for btree insertions
|
|
|
|
*
|
|
|
|
* Copyright 2012 Google, Inc.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "bcache.h"
|
|
|
|
#include "btree.h"
|
|
|
|
#include "debug.h"
|
2014-04-11 07:58:49 +07:00
|
|
|
#include "extents.h"
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-04-27 05:39:55 +07:00
|
|
|
#include <trace/events/bcache.h>
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
/*
|
|
|
|
* Journal replay/recovery:
|
|
|
|
*
|
|
|
|
* This code is all driven from run_cache_set(); we first read the journal
|
|
|
|
* entries, do some other stuff, then we mark all the keys in the journal
|
|
|
|
* entries (same as garbage collection would), then we replay them - reinserting
|
|
|
|
* them into the cache in precisely the same order as they appear in the
|
|
|
|
* journal.
|
|
|
|
*
|
|
|
|
* We only journal keys that go in leaf nodes, which simplifies things quite a
|
|
|
|
* bit.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void journal_read_endio(struct bio *bio, int error)
|
|
|
|
{
|
|
|
|
struct closure *cl = bio->bi_private;
|
|
|
|
closure_put(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int journal_read_bucket(struct cache *ca, struct list_head *list,
|
2013-07-25 07:44:17 +07:00
|
|
|
unsigned bucket_index)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
|
|
|
struct journal_device *ja = &ca->journal;
|
|
|
|
struct bio *bio = &ja->bio;
|
|
|
|
|
|
|
|
struct journal_replay *i;
|
|
|
|
struct jset *j, *data = ca->set->journal.w[0].data;
|
2013-07-25 07:44:17 +07:00
|
|
|
struct closure cl;
|
2013-03-24 06:11:31 +07:00
|
|
|
unsigned len, left, offset = 0;
|
|
|
|
int ret = 0;
|
|
|
|
sector_t bucket = bucket_to_sector(ca->set, ca->sb.d[bucket_index]);
|
|
|
|
|
2013-07-25 07:44:17 +07:00
|
|
|
closure_init_stack(&cl);
|
|
|
|
|
2013-08-06 04:04:06 +07:00
|
|
|
pr_debug("reading %u", bucket_index);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
while (offset < ca->sb.bucket_size) {
|
|
|
|
reread: left = ca->sb.bucket_size - offset;
|
2013-08-06 04:04:06 +07:00
|
|
|
len = min_t(unsigned, left, PAGE_SECTORS << JSET_BITS);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
bio_reset(bio);
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_sector = bucket + offset;
|
2013-03-24 06:11:31 +07:00
|
|
|
bio->bi_bdev = ca->bdev;
|
|
|
|
bio->bi_rw = READ;
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_size = len << 9;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
bio->bi_end_io = journal_read_endio;
|
2013-07-25 07:44:17 +07:00
|
|
|
bio->bi_private = &cl;
|
2013-03-29 01:50:55 +07:00
|
|
|
bch_bio_map(bio, data);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-07-25 07:44:17 +07:00
|
|
|
closure_bio_submit(bio, &cl, ca);
|
|
|
|
closure_sync(&cl);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
/* This function could be simpler now since we no longer write
|
|
|
|
* journal entries that overlap bucket boundaries; this means
|
|
|
|
* the start of a bucket will always have a valid journal entry
|
|
|
|
* if it has any journal entries at all.
|
|
|
|
*/
|
|
|
|
|
|
|
|
j = data;
|
|
|
|
while (len) {
|
|
|
|
struct list_head *where;
|
|
|
|
size_t blocks, bytes = set_bytes(j);
|
|
|
|
|
2013-08-06 04:04:06 +07:00
|
|
|
if (j->magic != jset_magic(&ca->sb)) {
|
|
|
|
pr_debug("%u: bad magic", bucket_index);
|
2013-03-24 06:11:31 +07:00
|
|
|
return ret;
|
2013-08-06 04:04:06 +07:00
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-08-06 04:04:06 +07:00
|
|
|
if (bytes > left << 9 ||
|
|
|
|
bytes > PAGE_SIZE << JSET_BITS) {
|
|
|
|
pr_info("%u: too big, %zu bytes, offset %u",
|
|
|
|
bucket_index, bytes, offset);
|
2013-03-24 06:11:31 +07:00
|
|
|
return ret;
|
2013-08-06 04:04:06 +07:00
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
if (bytes > len << 9)
|
|
|
|
goto reread;
|
|
|
|
|
2013-08-06 04:04:06 +07:00
|
|
|
if (j->csum != csum_set(j)) {
|
|
|
|
pr_info("%u: bad csum, %zu bytes, offset %u",
|
|
|
|
bucket_index, bytes, offset);
|
2013-03-24 06:11:31 +07:00
|
|
|
return ret;
|
2013-08-06 04:04:06 +07:00
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-12-18 14:49:49 +07:00
|
|
|
blocks = set_blocks(j, block_bytes(ca->set));
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
while (!list_empty(list)) {
|
|
|
|
i = list_first_entry(list,
|
|
|
|
struct journal_replay, list);
|
|
|
|
if (i->j.seq >= j->last_seq)
|
|
|
|
break;
|
|
|
|
list_del(&i->list);
|
|
|
|
kfree(i);
|
|
|
|
}
|
|
|
|
|
|
|
|
list_for_each_entry_reverse(i, list, list) {
|
|
|
|
if (j->seq == i->j.seq)
|
|
|
|
goto next_set;
|
|
|
|
|
|
|
|
if (j->seq < i->j.last_seq)
|
|
|
|
goto next_set;
|
|
|
|
|
|
|
|
if (j->seq > i->j.seq) {
|
|
|
|
where = &i->list;
|
|
|
|
goto add;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
where = list;
|
|
|
|
add:
|
|
|
|
i = kmalloc(offsetof(struct journal_replay, j) +
|
|
|
|
bytes, GFP_KERNEL);
|
|
|
|
if (!i)
|
|
|
|
return -ENOMEM;
|
|
|
|
memcpy(&i->j, j, bytes);
|
|
|
|
list_add(&i->list, where);
|
|
|
|
ret = 1;
|
|
|
|
|
|
|
|
ja->seq[bucket_index] = j->seq;
|
|
|
|
next_set:
|
|
|
|
offset += blocks * ca->sb.block_size;
|
|
|
|
len -= blocks * ca->sb.block_size;
|
|
|
|
j = ((void *) j) + blocks * block_bytes(ca);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2013-07-25 07:44:17 +07:00
|
|
|
int bch_journal_read(struct cache_set *c, struct list_head *list)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
|
|
|
#define read_bucket(b) \
|
|
|
|
({ \
|
2013-07-25 07:44:17 +07:00
|
|
|
int ret = journal_read_bucket(ca, list, b); \
|
2013-03-24 06:11:31 +07:00
|
|
|
__set_bit(b, bitmap); \
|
|
|
|
if (ret < 0) \
|
|
|
|
return ret; \
|
|
|
|
ret; \
|
|
|
|
})
|
|
|
|
|
|
|
|
struct cache *ca;
|
|
|
|
unsigned iter;
|
|
|
|
|
|
|
|
for_each_cache(ca, c, iter) {
|
|
|
|
struct journal_device *ja = &ca->journal;
|
|
|
|
unsigned long bitmap[SB_JOURNAL_BUCKETS / BITS_PER_LONG];
|
|
|
|
unsigned i, l, r, m;
|
|
|
|
uint64_t seq;
|
|
|
|
|
|
|
|
bitmap_zero(bitmap, SB_JOURNAL_BUCKETS);
|
|
|
|
pr_debug("%u journal buckets", ca->sb.njournal_buckets);
|
|
|
|
|
2013-09-24 13:17:29 +07:00
|
|
|
/*
|
|
|
|
* Read journal buckets ordered by golden ratio hash to quickly
|
2013-03-24 06:11:31 +07:00
|
|
|
* find a sequence of buckets with valid journal entries
|
|
|
|
*/
|
|
|
|
for (i = 0; i < ca->sb.njournal_buckets; i++) {
|
|
|
|
l = (i * 2654435769U) % ca->sb.njournal_buckets;
|
|
|
|
|
|
|
|
if (test_bit(l, bitmap))
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (read_bucket(l))
|
|
|
|
goto bsearch;
|
|
|
|
}
|
|
|
|
|
2013-09-24 13:17:29 +07:00
|
|
|
/*
|
|
|
|
* If that fails, check all the buckets we haven't checked
|
2013-03-24 06:11:31 +07:00
|
|
|
* already
|
|
|
|
*/
|
|
|
|
pr_debug("falling back to linear search");
|
|
|
|
|
2013-09-24 13:17:29 +07:00
|
|
|
for (l = find_first_zero_bit(bitmap, ca->sb.njournal_buckets);
|
|
|
|
l < ca->sb.njournal_buckets;
|
|
|
|
l = find_next_zero_bit(bitmap, ca->sb.njournal_buckets, l + 1))
|
2013-03-24 06:11:31 +07:00
|
|
|
if (read_bucket(l))
|
|
|
|
goto bsearch;
|
2013-09-24 13:17:29 +07:00
|
|
|
|
2014-06-03 05:39:44 +07:00
|
|
|
/* no journal entries on this device? */
|
|
|
|
if (l == ca->sb.njournal_buckets)
|
2013-09-24 13:17:29 +07:00
|
|
|
continue;
|
2013-03-24 06:11:31 +07:00
|
|
|
bsearch:
|
2014-06-03 05:39:44 +07:00
|
|
|
BUG_ON(list_empty(list));
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
/* Binary search */
|
2014-04-11 06:09:51 +07:00
|
|
|
m = l;
|
|
|
|
r = find_next_bit(bitmap, ca->sb.njournal_buckets, l + 1);
|
2013-03-24 06:11:31 +07:00
|
|
|
pr_debug("starting binary search, l %u r %u", l, r);
|
|
|
|
|
|
|
|
while (l + 1 < r) {
|
2013-07-12 12:42:14 +07:00
|
|
|
seq = list_entry(list->prev, struct journal_replay,
|
|
|
|
list)->j.seq;
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
m = (l + r) >> 1;
|
2013-07-12 12:42:14 +07:00
|
|
|
read_bucket(m);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-07-12 12:42:14 +07:00
|
|
|
if (seq != list_entry(list->prev, struct journal_replay,
|
|
|
|
list)->j.seq)
|
2013-03-24 06:11:31 +07:00
|
|
|
l = m;
|
|
|
|
else
|
|
|
|
r = m;
|
|
|
|
}
|
|
|
|
|
2013-09-24 13:17:29 +07:00
|
|
|
/*
|
|
|
|
* Read buckets in reverse order until we stop finding more
|
2013-03-24 06:11:31 +07:00
|
|
|
* journal entries
|
|
|
|
*/
|
2013-09-24 13:17:29 +07:00
|
|
|
pr_debug("finishing up: m %u njournal_buckets %u",
|
|
|
|
m, ca->sb.njournal_buckets);
|
2013-03-24 06:11:31 +07:00
|
|
|
l = m;
|
|
|
|
|
|
|
|
while (1) {
|
|
|
|
if (!l--)
|
|
|
|
l = ca->sb.njournal_buckets - 1;
|
|
|
|
|
|
|
|
if (l == m)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (test_bit(l, bitmap))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (!read_bucket(l))
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
seq = 0;
|
|
|
|
|
|
|
|
for (i = 0; i < ca->sb.njournal_buckets; i++)
|
|
|
|
if (ja->seq[i] > seq) {
|
|
|
|
seq = ja->seq[i];
|
2014-03-14 03:44:21 +07:00
|
|
|
/*
|
|
|
|
* When journal_reclaim() goes to allocate for
|
|
|
|
* the first time, it'll use the bucket after
|
|
|
|
* ja->cur_idx
|
|
|
|
*/
|
|
|
|
ja->cur_idx = i;
|
|
|
|
ja->last_idx = ja->discard_idx = (i + 1) %
|
|
|
|
ca->sb.njournal_buckets;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-09-24 13:17:29 +07:00
|
|
|
if (!list_empty(list))
|
|
|
|
c->journal.seq = list_entry(list->prev,
|
|
|
|
struct journal_replay,
|
|
|
|
list)->j.seq;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
return 0;
|
|
|
|
#undef read_bucket
|
|
|
|
}
|
|
|
|
|
|
|
|
void bch_journal_mark(struct cache_set *c, struct list_head *list)
|
|
|
|
{
|
|
|
|
atomic_t p = { 0 };
|
|
|
|
struct bkey *k;
|
|
|
|
struct journal_replay *i;
|
|
|
|
struct journal *j = &c->journal;
|
|
|
|
uint64_t last = j->seq;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* journal.pin should never fill up - we never write a journal
|
|
|
|
* entry when it would fill up. But if for some reason it does, we
|
|
|
|
* iterate over the list in reverse order so that we can just skip that
|
|
|
|
* refcount instead of bugging.
|
|
|
|
*/
|
|
|
|
|
|
|
|
list_for_each_entry_reverse(i, list, list) {
|
|
|
|
BUG_ON(last < i->j.seq);
|
|
|
|
i->pin = NULL;
|
|
|
|
|
|
|
|
while (last-- != i->j.seq)
|
|
|
|
if (fifo_free(&j->pin) > 1) {
|
|
|
|
fifo_push_front(&j->pin, p);
|
|
|
|
atomic_set(&fifo_front(&j->pin), 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (fifo_free(&j->pin) > 1) {
|
|
|
|
fifo_push_front(&j->pin, p);
|
|
|
|
i->pin = &fifo_front(&j->pin);
|
|
|
|
atomic_set(i->pin, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
for (k = i->j.start;
|
2013-12-18 12:56:21 +07:00
|
|
|
k < bset_bkey_last(&i->j);
|
2014-04-11 07:58:49 +07:00
|
|
|
k = bkey_next(k))
|
|
|
|
if (!__bch_extent_invalid(c, k)) {
|
|
|
|
unsigned j;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2014-04-11 07:58:49 +07:00
|
|
|
for (j = 0; j < KEY_PTRS(k); j++)
|
|
|
|
if (ptr_available(c, k, j))
|
|
|
|
atomic_inc(&PTR_BUCKET(c, k, j)->pin);
|
2014-02-25 10:55:28 +07:00
|
|
|
|
2014-04-11 07:58:49 +07:00
|
|
|
bch_initial_mark_key(c, 0, k);
|
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-07-25 07:44:17 +07:00
|
|
|
int bch_journal_replay(struct cache_set *s, struct list_head *list)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
|
|
|
int ret = 0, keys = 0, entries = 0;
|
|
|
|
struct bkey *k;
|
|
|
|
struct journal_replay *i =
|
|
|
|
list_entry(list->prev, struct journal_replay, list);
|
|
|
|
|
|
|
|
uint64_t start = i->j.last_seq, end = i->j.seq, n = start;
|
2013-07-25 07:26:51 +07:00
|
|
|
struct keylist keylist;
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
list_for_each_entry(i, list, list) {
|
|
|
|
BUG_ON(i->pin && atomic_read(i->pin) != 1);
|
|
|
|
|
2013-07-12 09:42:51 +07:00
|
|
|
cache_set_err_on(n != i->j.seq, s,
|
|
|
|
"bcache: journal entries %llu-%llu missing! (replaying %llu-%llu)",
|
|
|
|
n, i->j.seq - 1, start, end);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
for (k = i->j.start;
|
2013-12-18 12:56:21 +07:00
|
|
|
k < bset_bkey_last(&i->j);
|
2013-03-24 06:11:31 +07:00
|
|
|
k = bkey_next(k)) {
|
2013-04-27 05:39:55 +07:00
|
|
|
trace_bcache_journal_replay_key(k);
|
|
|
|
|
2014-01-09 12:22:02 +07:00
|
|
|
bch_keylist_init_single(&keylist, k);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-07-25 08:07:22 +07:00
|
|
|
ret = bch_btree_insert(s, &keylist, i->pin, NULL);
|
2013-03-24 06:11:31 +07:00
|
|
|
if (ret)
|
|
|
|
goto err;
|
|
|
|
|
2013-07-25 07:26:51 +07:00
|
|
|
BUG_ON(!bch_keylist_empty(&keylist));
|
2013-03-24 06:11:31 +07:00
|
|
|
keys++;
|
|
|
|
|
|
|
|
cond_resched();
|
|
|
|
}
|
|
|
|
|
|
|
|
if (i->pin)
|
|
|
|
atomic_dec(i->pin);
|
|
|
|
n = i->j.seq + 1;
|
|
|
|
entries++;
|
|
|
|
}
|
|
|
|
|
|
|
|
pr_info("journal replay done, %i keys in %i entries, seq %llu",
|
|
|
|
keys, entries, end);
|
2013-07-25 08:04:18 +07:00
|
|
|
err:
|
2013-03-24 06:11:31 +07:00
|
|
|
while (!list_empty(list)) {
|
|
|
|
i = list_first_entry(list, struct journal_replay, list);
|
|
|
|
list_del(&i->list);
|
|
|
|
kfree(i);
|
|
|
|
}
|
2013-07-25 08:04:18 +07:00
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Journalling */
|
|
|
|
|
|
|
|
static void btree_flush_write(struct cache_set *c)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Try to find the btree node with that references the oldest journal
|
|
|
|
* entry, best is our current candidate and is locked if non NULL:
|
|
|
|
*/
|
2013-10-25 07:07:04 +07:00
|
|
|
struct btree *b, *best;
|
|
|
|
unsigned i;
|
|
|
|
retry:
|
|
|
|
best = NULL;
|
|
|
|
|
|
|
|
for_each_cached_btree(b, c, i)
|
|
|
|
if (btree_current_write(b)->journal) {
|
|
|
|
if (!best)
|
|
|
|
best = b;
|
|
|
|
else if (journal_pin_cmp(c,
|
2013-07-25 07:44:17 +07:00
|
|
|
btree_current_write(best)->journal,
|
|
|
|
btree_current_write(b)->journal)) {
|
2013-10-25 07:07:04 +07:00
|
|
|
best = b;
|
|
|
|
}
|
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
b = best;
|
|
|
|
if (b) {
|
2014-03-05 07:42:42 +07:00
|
|
|
mutex_lock(&b->write_lock);
|
2013-10-25 07:07:04 +07:00
|
|
|
if (!btree_current_write(b)->journal) {
|
2014-03-05 07:42:42 +07:00
|
|
|
mutex_unlock(&b->write_lock);
|
2013-10-25 07:07:04 +07:00
|
|
|
/* We raced */
|
|
|
|
goto retry;
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
|
2014-03-05 07:42:42 +07:00
|
|
|
__bch_btree_node_write(b, NULL);
|
|
|
|
mutex_unlock(&b->write_lock);
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#define last_seq(j) ((j)->seq - fifo_used(&(j)->pin) + 1)
|
|
|
|
|
|
|
|
static void journal_discard_endio(struct bio *bio, int error)
|
|
|
|
{
|
|
|
|
struct journal_device *ja =
|
|
|
|
container_of(bio, struct journal_device, discard_bio);
|
|
|
|
struct cache *ca = container_of(ja, struct cache, journal);
|
|
|
|
|
|
|
|
atomic_set(&ja->discard_in_flight, DISCARD_DONE);
|
|
|
|
|
|
|
|
closure_wake_up(&ca->set->journal.wait);
|
|
|
|
closure_put(&ca->set->cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void journal_discard_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct journal_device *ja =
|
|
|
|
container_of(work, struct journal_device, discard_work);
|
|
|
|
|
|
|
|
submit_bio(0, &ja->discard_bio);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void do_journal_discard(struct cache *ca)
|
|
|
|
{
|
|
|
|
struct journal_device *ja = &ca->journal;
|
|
|
|
struct bio *bio = &ja->discard_bio;
|
|
|
|
|
|
|
|
if (!ca->discard) {
|
|
|
|
ja->discard_idx = ja->last_idx;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2013-09-24 13:17:27 +07:00
|
|
|
switch (atomic_read(&ja->discard_in_flight)) {
|
2013-03-24 06:11:31 +07:00
|
|
|
case DISCARD_IN_FLIGHT:
|
|
|
|
return;
|
|
|
|
|
|
|
|
case DISCARD_DONE:
|
|
|
|
ja->discard_idx = (ja->discard_idx + 1) %
|
|
|
|
ca->sb.njournal_buckets;
|
|
|
|
|
|
|
|
atomic_set(&ja->discard_in_flight, DISCARD_READY);
|
|
|
|
/* fallthrough */
|
|
|
|
|
|
|
|
case DISCARD_READY:
|
|
|
|
if (ja->discard_idx == ja->last_idx)
|
|
|
|
return;
|
|
|
|
|
|
|
|
atomic_set(&ja->discard_in_flight, DISCARD_IN_FLIGHT);
|
|
|
|
|
|
|
|
bio_init(bio);
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_sector = bucket_to_sector(ca->set,
|
2013-03-26 01:46:44 +07:00
|
|
|
ca->sb.d[ja->discard_idx]);
|
2013-03-24 06:11:31 +07:00
|
|
|
bio->bi_bdev = ca->bdev;
|
|
|
|
bio->bi_rw = REQ_WRITE|REQ_DISCARD;
|
|
|
|
bio->bi_max_vecs = 1;
|
|
|
|
bio->bi_io_vec = bio->bi_inline_vecs;
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_size = bucket_bytes(ca);
|
2013-03-24 06:11:31 +07:00
|
|
|
bio->bi_end_io = journal_discard_endio;
|
|
|
|
|
|
|
|
closure_get(&ca->set->cl);
|
|
|
|
INIT_WORK(&ja->discard_work, journal_discard_work);
|
|
|
|
schedule_work(&ja->discard_work);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void journal_reclaim(struct cache_set *c)
|
|
|
|
{
|
|
|
|
struct bkey *k = &c->journal.key;
|
|
|
|
struct cache *ca;
|
|
|
|
uint64_t last_seq;
|
|
|
|
unsigned iter, n = 0;
|
|
|
|
atomic_t p;
|
|
|
|
|
|
|
|
while (!atomic_read(&fifo_front(&c->journal.pin)))
|
|
|
|
fifo_pop(&c->journal.pin, p);
|
|
|
|
|
|
|
|
last_seq = last_seq(&c->journal);
|
|
|
|
|
|
|
|
/* Update last_idx */
|
|
|
|
|
|
|
|
for_each_cache(ca, c, iter) {
|
|
|
|
struct journal_device *ja = &ca->journal;
|
|
|
|
|
|
|
|
while (ja->last_idx != ja->cur_idx &&
|
|
|
|
ja->seq[ja->last_idx] < last_seq)
|
|
|
|
ja->last_idx = (ja->last_idx + 1) %
|
|
|
|
ca->sb.njournal_buckets;
|
|
|
|
}
|
|
|
|
|
|
|
|
for_each_cache(ca, c, iter)
|
|
|
|
do_journal_discard(ca);
|
|
|
|
|
|
|
|
if (c->journal.blocks_free)
|
2013-10-25 07:07:04 +07:00
|
|
|
goto out;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate:
|
|
|
|
* XXX: Sort by free journal space
|
|
|
|
*/
|
|
|
|
|
|
|
|
for_each_cache(ca, c, iter) {
|
|
|
|
struct journal_device *ja = &ca->journal;
|
|
|
|
unsigned next = (ja->cur_idx + 1) % ca->sb.njournal_buckets;
|
|
|
|
|
|
|
|
/* No space available on this device */
|
|
|
|
if (next == ja->discard_idx)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
ja->cur_idx = next;
|
|
|
|
k->ptr[n++] = PTR(0,
|
|
|
|
bucket_to_sector(c, ca->sb.d[ja->cur_idx]),
|
|
|
|
ca->sb.nr_this_dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
bkey_init(k);
|
|
|
|
SET_KEY_PTRS(k, n);
|
|
|
|
|
|
|
|
if (n)
|
|
|
|
c->journal.blocks_free = c->sb.bucket_size >> c->block_bits;
|
2013-10-25 07:07:04 +07:00
|
|
|
out:
|
2013-03-24 06:11:31 +07:00
|
|
|
if (!journal_full(&c->journal))
|
|
|
|
__closure_wake_up(&c->journal.wait);
|
|
|
|
}
|
|
|
|
|
|
|
|
void bch_journal_next(struct journal *j)
|
|
|
|
{
|
|
|
|
atomic_t p = { 1 };
|
|
|
|
|
|
|
|
j->cur = (j->cur == j->w)
|
|
|
|
? &j->w[1]
|
|
|
|
: &j->w[0];
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The fifo_push() needs to happen at the same time as j->seq is
|
|
|
|
* incremented for last_seq() to be calculated correctly
|
|
|
|
*/
|
|
|
|
BUG_ON(!fifo_push(&j->pin, p));
|
|
|
|
atomic_set(&fifo_back(&j->pin), 1);
|
|
|
|
|
|
|
|
j->cur->data->seq = ++j->seq;
|
2014-02-20 10:48:26 +07:00
|
|
|
j->cur->dirty = false;
|
2013-03-24 06:11:31 +07:00
|
|
|
j->cur->need_write = false;
|
|
|
|
j->cur->data->keys = 0;
|
|
|
|
|
|
|
|
if (fifo_full(&j->pin))
|
|
|
|
pr_debug("journal_pin full (%zu)", fifo_used(&j->pin));
|
|
|
|
}
|
|
|
|
|
|
|
|
static void journal_write_endio(struct bio *bio, int error)
|
|
|
|
{
|
|
|
|
struct journal_write *w = bio->bi_private;
|
|
|
|
|
|
|
|
cache_set_err_on(error, w->c, "journal io error");
|
2013-10-09 05:50:46 +07:00
|
|
|
closure_put(&w->c->journal.io);
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static void journal_write(struct closure *);
|
|
|
|
|
|
|
|
static void journal_write_done(struct closure *cl)
|
|
|
|
{
|
2013-10-09 05:50:46 +07:00
|
|
|
struct journal *j = container_of(cl, struct journal, io);
|
2013-03-24 06:11:31 +07:00
|
|
|
struct journal_write *w = (j->cur == j->w)
|
|
|
|
? &j->w[1]
|
|
|
|
: &j->w[0];
|
|
|
|
|
|
|
|
__closure_wake_up(&w->wait);
|
2013-10-09 05:50:46 +07:00
|
|
|
continue_at_nobarrier(cl, journal_write, system_wq);
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
|
2013-12-17 06:27:25 +07:00
|
|
|
static void journal_write_unlock(struct closure *cl)
|
|
|
|
{
|
|
|
|
struct cache_set *c = container_of(cl, struct cache_set, journal.io);
|
|
|
|
|
|
|
|
c->journal.io_in_flight = 0;
|
|
|
|
spin_unlock(&c->journal.lock);
|
|
|
|
}
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
static void journal_write_unlocked(struct closure *cl)
|
2013-03-27 03:49:02 +07:00
|
|
|
__releases(c->journal.lock)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
2013-10-09 05:50:46 +07:00
|
|
|
struct cache_set *c = container_of(cl, struct cache_set, journal.io);
|
2013-03-24 06:11:31 +07:00
|
|
|
struct cache *ca;
|
|
|
|
struct journal_write *w = c->journal.cur;
|
|
|
|
struct bkey *k = &c->journal.key;
|
2013-12-18 14:49:49 +07:00
|
|
|
unsigned i, sectors = set_blocks(w->data, block_bytes(c)) *
|
|
|
|
c->sb.block_size;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
struct bio *bio;
|
|
|
|
struct bio_list list;
|
|
|
|
bio_list_init(&list);
|
|
|
|
|
|
|
|
if (!w->need_write) {
|
2013-12-17 06:27:25 +07:00
|
|
|
closure_return_with_destructor(cl, journal_write_unlock);
|
2013-03-24 06:11:31 +07:00
|
|
|
} else if (journal_full(&c->journal)) {
|
|
|
|
journal_reclaim(c);
|
|
|
|
spin_unlock(&c->journal.lock);
|
|
|
|
|
|
|
|
btree_flush_write(c);
|
|
|
|
continue_at(cl, journal_write, system_wq);
|
|
|
|
}
|
|
|
|
|
2013-12-18 14:49:49 +07:00
|
|
|
c->journal.blocks_free -= set_blocks(w->data, block_bytes(c));
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
w->data->btree_level = c->root->level;
|
|
|
|
|
|
|
|
bkey_copy(&w->data->btree_root, &c->root->key);
|
|
|
|
bkey_copy(&w->data->uuid_bucket, &c->uuid_bucket);
|
|
|
|
|
|
|
|
for_each_cache(ca, c, i)
|
|
|
|
w->data->prio_bucket[ca->sb.nr_this_dev] = ca->prio_buckets[0];
|
|
|
|
|
2013-11-01 05:46:42 +07:00
|
|
|
w->data->magic = jset_magic(&c->sb);
|
2013-03-24 06:11:31 +07:00
|
|
|
w->data->version = BCACHE_JSET_VERSION;
|
|
|
|
w->data->last_seq = last_seq(&c->journal);
|
|
|
|
w->data->csum = csum_set(w->data);
|
|
|
|
|
|
|
|
for (i = 0; i < KEY_PTRS(k); i++) {
|
|
|
|
ca = PTR_CACHE(c, k, i);
|
|
|
|
bio = &ca->journal.bio;
|
|
|
|
|
|
|
|
atomic_long_add(sectors, &ca->meta_sectors_written);
|
|
|
|
|
|
|
|
bio_reset(bio);
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_sector = PTR_OFFSET(k, i);
|
2013-03-24 06:11:31 +07:00
|
|
|
bio->bi_bdev = ca->bdev;
|
2013-06-27 07:25:38 +07:00
|
|
|
bio->bi_rw = REQ_WRITE|REQ_SYNC|REQ_META|REQ_FLUSH|REQ_FUA;
|
2013-10-12 05:44:27 +07:00
|
|
|
bio->bi_iter.bi_size = sectors << 9;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
bio->bi_end_io = journal_write_endio;
|
|
|
|
bio->bi_private = w;
|
2013-03-29 01:50:55 +07:00
|
|
|
bch_bio_map(bio, w->data);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
trace_bcache_journal_write(bio);
|
|
|
|
bio_list_add(&list, bio);
|
|
|
|
|
|
|
|
SET_PTR_OFFSET(k, i, PTR_OFFSET(k, i) + sectors);
|
|
|
|
|
|
|
|
ca->journal.seq[ca->journal.cur_idx] = w->data->seq;
|
|
|
|
}
|
|
|
|
|
|
|
|
atomic_dec_bug(&fifo_back(&c->journal.pin));
|
|
|
|
bch_journal_next(&c->journal);
|
|
|
|
journal_reclaim(c);
|
|
|
|
|
|
|
|
spin_unlock(&c->journal.lock);
|
|
|
|
|
|
|
|
while ((bio = bio_list_pop(&list)))
|
|
|
|
closure_bio_submit(bio, cl, c->cache[0]);
|
|
|
|
|
|
|
|
continue_at(cl, journal_write_done, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void journal_write(struct closure *cl)
|
|
|
|
{
|
2013-10-09 05:50:46 +07:00
|
|
|
struct cache_set *c = container_of(cl, struct cache_set, journal.io);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
spin_lock(&c->journal.lock);
|
|
|
|
journal_write_unlocked(cl);
|
|
|
|
}
|
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
static void journal_try_write(struct cache_set *c)
|
2013-03-27 03:49:02 +07:00
|
|
|
__releases(c->journal.lock)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
2013-10-09 05:50:46 +07:00
|
|
|
struct closure *cl = &c->journal.io;
|
|
|
|
struct journal_write *w = c->journal.cur;
|
|
|
|
|
|
|
|
w->need_write = true;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-12-17 06:27:25 +07:00
|
|
|
if (!c->journal.io_in_flight) {
|
|
|
|
c->journal.io_in_flight = 1;
|
|
|
|
closure_call(cl, journal_write_unlocked, NULL, &c->cl);
|
|
|
|
} else {
|
2013-10-25 07:07:04 +07:00
|
|
|
spin_unlock(&c->journal.lock);
|
2013-12-17 06:27:25 +07:00
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
static struct journal_write *journal_wait_for_write(struct cache_set *c,
|
|
|
|
unsigned nkeys)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
2013-10-25 07:07:04 +07:00
|
|
|
size_t sectors;
|
|
|
|
struct closure cl;
|
2013-12-11 07:10:46 +07:00
|
|
|
bool wait = false;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
closure_init_stack(&cl);
|
|
|
|
|
|
|
|
spin_lock(&c->journal.lock);
|
|
|
|
|
|
|
|
while (1) {
|
|
|
|
struct journal_write *w = c->journal.cur;
|
|
|
|
|
|
|
|
sectors = __set_blocks(w->data, w->data->keys + nkeys,
|
2013-12-18 14:49:49 +07:00
|
|
|
block_bytes(c)) * c->sb.block_size;
|
2013-10-25 07:07:04 +07:00
|
|
|
|
|
|
|
if (sectors <= min_t(size_t,
|
|
|
|
c->journal.blocks_free * c->sb.block_size,
|
|
|
|
PAGE_SECTORS << JSET_BITS))
|
|
|
|
return w;
|
|
|
|
|
2013-12-11 07:10:46 +07:00
|
|
|
if (wait)
|
|
|
|
closure_wait(&c->journal.wait, &cl);
|
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
if (!journal_full(&c->journal)) {
|
2013-12-11 07:10:46 +07:00
|
|
|
if (wait)
|
|
|
|
trace_bcache_journal_entry_full(c);
|
2013-10-25 07:07:04 +07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* XXX: If we were inserting so many keys that they
|
|
|
|
* won't fit in an _empty_ journal write, we'll
|
|
|
|
* deadlock. For now, handle this in
|
|
|
|
* bch_keylist_realloc() - but something to think about.
|
|
|
|
*/
|
|
|
|
BUG_ON(!w->data->keys);
|
|
|
|
|
|
|
|
journal_try_write(c); /* unlocks */
|
|
|
|
} else {
|
2013-12-11 07:10:46 +07:00
|
|
|
if (wait)
|
|
|
|
trace_bcache_journal_full(c);
|
2013-10-25 07:07:04 +07:00
|
|
|
|
|
|
|
journal_reclaim(c);
|
|
|
|
spin_unlock(&c->journal.lock);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
btree_flush_write(c);
|
|
|
|
}
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
closure_sync(&cl);
|
|
|
|
spin_lock(&c->journal.lock);
|
2013-12-11 07:10:46 +07:00
|
|
|
wait = true;
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-10-09 05:50:46 +07:00
|
|
|
static void journal_write_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct cache_set *c = container_of(to_delayed_work(work),
|
|
|
|
struct cache_set,
|
|
|
|
journal.work);
|
|
|
|
spin_lock(&c->journal.lock);
|
2014-02-20 10:48:26 +07:00
|
|
|
if (c->journal.cur->dirty)
|
|
|
|
journal_try_write(c);
|
|
|
|
else
|
|
|
|
spin_unlock(&c->journal.lock);
|
2013-10-09 05:50:46 +07:00
|
|
|
}
|
|
|
|
|
2013-03-24 06:11:31 +07:00
|
|
|
/*
|
|
|
|
* Entry point to the journalling code - bio_insert() and btree_invalidate()
|
|
|
|
* pass bch_journal() a list of keys to be journalled, and then
|
|
|
|
* bch_journal() hands those same keys off to btree_insert_async()
|
|
|
|
*/
|
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
atomic_t *bch_journal(struct cache_set *c,
|
|
|
|
struct keylist *keys,
|
|
|
|
struct closure *parent)
|
2013-03-24 06:11:31 +07:00
|
|
|
{
|
|
|
|
struct journal_write *w;
|
2013-10-25 07:07:04 +07:00
|
|
|
atomic_t *ret;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
if (!CACHE_SYNC(&c->sb))
|
|
|
|
return NULL;
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
w = journal_wait_for_write(c, bch_keylist_nkeys(keys));
|
2013-04-27 05:39:55 +07:00
|
|
|
|
2013-12-18 12:56:21 +07:00
|
|
|
memcpy(bset_bkey_last(w->data), keys->keys, bch_keylist_bytes(keys));
|
2013-10-25 07:07:04 +07:00
|
|
|
w->data->keys += bch_keylist_nkeys(keys);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
ret = &fifo_back(&c->journal.pin);
|
|
|
|
atomic_inc(ret);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
2013-10-25 07:07:04 +07:00
|
|
|
if (parent) {
|
|
|
|
closure_wait(&w->wait, parent);
|
2013-10-09 05:50:46 +07:00
|
|
|
journal_try_write(c);
|
2014-02-20 10:48:26 +07:00
|
|
|
} else if (!w->dirty) {
|
|
|
|
w->dirty = true;
|
2013-10-09 05:50:46 +07:00
|
|
|
schedule_delayed_work(&c->journal.work,
|
|
|
|
msecs_to_jiffies(c->journal_delay_ms));
|
|
|
|
spin_unlock(&c->journal.lock);
|
|
|
|
} else {
|
|
|
|
spin_unlock(&c->journal.lock);
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
2013-10-25 07:07:04 +07:00
|
|
|
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void bch_journal_meta(struct cache_set *c, struct closure *cl)
|
|
|
|
{
|
|
|
|
struct keylist keys;
|
|
|
|
atomic_t *ref;
|
|
|
|
|
|
|
|
bch_keylist_init(&keys);
|
|
|
|
|
|
|
|
ref = bch_journal(c, &keys, cl);
|
|
|
|
if (ref)
|
|
|
|
atomic_dec_bug(ref);
|
2013-03-24 06:11:31 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
void bch_journal_free(struct cache_set *c)
|
|
|
|
{
|
|
|
|
free_pages((unsigned long) c->journal.w[1].data, JSET_BITS);
|
|
|
|
free_pages((unsigned long) c->journal.w[0].data, JSET_BITS);
|
|
|
|
free_fifo(&c->journal.pin);
|
|
|
|
}
|
|
|
|
|
|
|
|
int bch_journal_alloc(struct cache_set *c)
|
|
|
|
{
|
|
|
|
struct journal *j = &c->journal;
|
|
|
|
|
|
|
|
spin_lock_init(&j->lock);
|
2013-10-09 05:50:46 +07:00
|
|
|
INIT_DELAYED_WORK(&j->work, journal_write_work);
|
2013-03-24 06:11:31 +07:00
|
|
|
|
|
|
|
c->journal_delay_ms = 100;
|
|
|
|
|
|
|
|
j->w[0].c = c;
|
|
|
|
j->w[1].c = c;
|
|
|
|
|
|
|
|
if (!(init_fifo(&j->pin, JOURNAL_PIN, GFP_KERNEL)) ||
|
|
|
|
!(j->w[0].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)) ||
|
|
|
|
!(j->w[1].data = (void *) __get_free_pages(GFP_KERNEL, JSET_BITS)))
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|