-c3e4d892b77b9361c88854b0192223f68947b6b0
+eb83f1f842bb95e1f61489b48854240777368763
static int invalidate_one_bucket(struct btree_trans *trans,
struct btree_iter *lru_iter,
struct bkey_s_c lru_k,
- struct bpos *last_flushed_pos,
s64 *nr_to_invalidate)
{
struct bch_fs *c = trans->c;
+ struct btree_iter alloc_iter = { NULL };
+ struct bkey_i_alloc_v4 *a = NULL;
+ struct printbuf buf = PRINTBUF;
+ struct bpos bucket = u64_to_bucket(lru_k.k->p.offset);
+ unsigned cached_sectors;
int ret = 0;
if (*nr_to_invalidate <= 0)
return 1;
- ret = bch2_check_lru_key(trans, lru_iter, lru_k, last_flushed_pos);
- if (ret)
- return ret < 0 ? ret : 0;
+ if (!bch2_dev_bucket_exists(c, bucket)) {
+ prt_str(&buf, "lru entry points to invalid bucket");
+ goto err;
+ }
- struct bpos bucket = u64_to_bucket(lru_k.k->p.offset);
if (bch2_bucket_is_open_safe(c, bucket.inode, bucket.offset))
return 0;
- struct btree_iter alloc_iter;
- struct bkey_i_alloc_v4 *a = bch2_trans_start_alloc_update(trans, &alloc_iter, bucket);
+ a = bch2_trans_start_alloc_update(trans, &alloc_iter, bucket);
ret = PTR_ERR_OR_ZERO(a);
if (ret)
goto out;
if (!a->v.cached_sectors)
bch_err(c, "invalidating empty bucket, confused");
- unsigned cached_sectors = a->v.cached_sectors;
+ cached_sectors = a->v.cached_sectors;
SET_BCH_ALLOC_V4_NEED_INC_GEN(&a->v, false);
a->v.gen++;
--*nr_to_invalidate;
out:
bch2_trans_iter_exit(trans, &alloc_iter);
+ printbuf_exit(&buf);
return ret;
+err:
+ prt_str(&buf, "\n lru key: ");
+ bch2_bkey_val_to_text(&buf, c, lru_k);
+
+ prt_str(&buf, "\n lru entry: ");
+ bch2_lru_pos_to_text(&buf, lru_iter->pos);
+
+ prt_str(&buf, "\n alloc key: ");
+ if (!a)
+ bch2_bpos_to_text(&buf, bucket);
+ else
+ bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&a->k_i));
+
+ bch_err(c, "%s", buf.buf);
+ if (c->curr_recovery_pass > BCH_RECOVERY_PASS_check_lrus) {
+ bch2_inconsistent_error(c);
+ ret = -EINVAL;
+ }
+
+ goto out;
}
static void bch2_do_invalidates_work(struct work_struct *work)
struct btree_trans *trans = bch2_trans_get(c);
struct btree_iter iter;
struct bkey_s_c k;
- struct bpos last_flushed_pos = POS_MIN;
unsigned i;
int ret = 0;
lru_pos(ca->dev_idx, 0, 0),
lru_pos(ca->dev_idx, U64_MAX, LRU_TIME_MAX),
BTREE_ITER_INTENT, k,
- invalidate_one_bucket(trans, &iter, k, &last_flushed_pos,
- &nr_to_invalidate));
+ invalidate_one_bucket(trans, &iter, k, &nr_to_invalidate));
if (ret < 0) {
percpu_ref_put(&ca->ref);
memcmp(bkey_s_c_to_backpointer(bp_k).v, &bp, sizeof(bp))) {
if (last_flushed->level != bp.level ||
!bpos_eq(last_flushed->pos, orig_k.k->p)) {
+ ret = bch2_btree_write_buffer_flush_sync(trans);
+ if (ret)
+ goto err;
+
last_flushed->level = bp.level;
last_flushed->pos = orig_k.k->p;
-
- ret = bch2_btree_write_buffer_flush_sync(trans) ?:
- -BCH_ERR_transaction_restart_write_buffer_flush;
+ ret = -BCH_ERR_transaction_restart_write_buffer_flush;
goto out;
}
goto missing;
struct btree_iter iter;
enum btree_id btree_id;
struct bkey_s_c k;
- struct bpos_level last_flushed = { UINT_MAX, POS_MIN };
int ret = 0;
for (btree_id = 0; btree_id < btree_id_nr_alive(c); btree_id++) {
+ struct bpos_level last_flushed = { UINT_MAX, POS_MIN };
int level, depth = btree_type_has_ptrs(btree_id) ? 0 : 1;
ret = commit_do(trans, NULL, NULL,
bch2_trans_node_iter_init(trans, &iter, btree_id, POS_MIN, 0,
level,
BTREE_ITER_PREFETCH);
- for_each_btree_key_continue(trans, iter, BTREE_ITER_PREFETCH, k, ret) {
- ret = commit_do(trans, NULL, NULL,
- BCH_TRANS_COMMIT_no_enospc,
+ while (1) {
+ bch2_trans_begin(trans);
+ k = bch2_btree_iter_peek(&iter);
+ if (!k.k)
+ break;
+ ret = bkey_err(k) ?:
check_extent_to_backpointers(trans, btree_id, level,
bucket_start, bucket_end,
- &last_flushed, k));
+ &last_flushed, k) ?:
+ bch2_trans_commit(trans, NULL, NULL,
+ BCH_TRANS_COMMIT_no_enospc);
+ if (bch2_err_matches(ret, BCH_ERR_transaction_restart)) {
+ ret = 0;
+ continue;
+ }
if (ret)
break;
-
- if (bpos_eq(iter.pos, SPOS_MAX))
- break;
+ bch2_btree_iter_advance(&iter);
}
bch2_trans_iter_exit(trans, &iter);
u64_to_bucket(lru.offset).offset);
}
-static inline int __bch2_lru_set(struct btree_trans *trans, u16 lru_id,
- u64 dev_bucket, u64 time, bool set)
+static int __bch2_lru_set(struct btree_trans *trans, u16 lru_id,
+ u64 dev_bucket, u64 time, bool set)
{
return time
? bch2_btree_bit_mod(trans, BTREE_ID_lru,
int bch2_lru_del(struct btree_trans *trans, u16 lru_id, u64 dev_bucket, u64 time)
{
- return __bch2_lru_set(trans, lru_id, dev_bucket, time, false);
+ return __bch2_lru_set(trans, lru_id, dev_bucket, time, KEY_TYPE_deleted);
}
int bch2_lru_set(struct btree_trans *trans, u16 lru_id, u64 dev_bucket, u64 time)
{
- return __bch2_lru_set(trans, lru_id, dev_bucket, time, true);
+ return __bch2_lru_set(trans, lru_id, dev_bucket, time, KEY_TYPE_set);
}
int bch2_lru_change(struct btree_trans *trans,
if (old_time == new_time)
return 0;
- return __bch2_lru_set(trans, lru_id, dev_bucket, old_time, false) ?:
- __bch2_lru_set(trans, lru_id, dev_bucket, new_time, true);
+ return bch2_lru_del(trans, lru_id, dev_bucket, old_time) ?:
+ bch2_lru_set(trans, lru_id, dev_bucket, new_time);
}
static const char * const bch2_lru_types[] = {
NULL
};
-/* Returns 1 if key has been deleted */
-int bch2_check_lru_key(struct btree_trans *trans,
- struct btree_iter *lru_iter,
- struct bkey_s_c lru_k,
- struct bpos *last_flushed_pos)
+static int bch2_check_lru_key(struct btree_trans *trans,
+ struct btree_iter *lru_iter,
+ struct bkey_s_c lru_k,
+ struct bpos *last_flushed_pos)
{
struct bch_fs *c = trans->c;
struct btree_iter iter;
const struct bch_alloc_v4 *a;
struct printbuf buf1 = PRINTBUF;
struct printbuf buf2 = PRINTBUF;
+ enum bch_lru_type type = lru_type(lru_k);
struct bpos alloc_pos = u64_to_bucket(lru_k.k->p.offset);
u64 idx;
int ret;
lru_entry_to_invalid_bucket,
"lru key points to nonexistent device:bucket %llu:%llu",
alloc_pos.inode, alloc_pos.offset))
- goto delete;
+ return bch2_btree_delete_at(trans, lru_iter, 0);
k = bch2_bkey_get_iter(trans, &iter, BTREE_ID_alloc, alloc_pos, 0);
ret = bkey_err(k);
a = bch2_alloc_to_v4(k, &a_convert);
- enum bch_lru_type type = lru_type(lru_k);
switch (type) {
case BCH_LRU_read:
idx = alloc_lru_idx_read(*a);
case BCH_LRU_fragmentation:
idx = a->fragmentation_lru;
break;
- default:
- /* unknown LRU type, don't check: */
- goto out;
}
if (lru_k.k->type != KEY_TYPE_set ||
lru_pos_time(lru_k.k->p) != idx) {
if (!bpos_eq(*last_flushed_pos, lru_k.k->p)) {
- ret = bch2_btree_write_buffer_flush_sync(trans);
- if (!ret) {
- *last_flushed_pos = lru_k.k->p;
- ret = -BCH_ERR_transaction_restart_write_buffer_flush;
- }
+ *last_flushed_pos = lru_k.k->p;
+ ret = bch2_btree_write_buffer_flush_sync(trans) ?:
+ -BCH_ERR_transaction_restart_write_buffer_flush;
goto out;
}
- if ((c->opts.reconstruct_alloc &&
- c->curr_recovery_pass <= BCH_RECOVERY_PASS_check_lrus) ||
+ if (c->opts.reconstruct_alloc ||
fsck_err(c, lru_entry_bad,
"incorrect lru entry: lru %s time %llu\n"
" %s\n"
- "for\n"
- " %s",
+ " for %s",
bch2_lru_types[type],
lru_pos_time(lru_k.k->p),
(bch2_bkey_val_to_text(&buf1, c, lru_k), buf1.buf),
(bch2_bkey_val_to_text(&buf2, c, k), buf2.buf)))
- goto delete;
+ ret = bch2_btree_delete_at(trans, lru_iter, 0);
}
out:
err:
printbuf_exit(&buf2);
printbuf_exit(&buf1);
return ret;
-delete:
- ret = bch2_btree_delete_at(trans, lru_iter, 0) ?:
- bch2_trans_commit(trans, NULL, NULL,
- BCH_WATERMARK_btree|
- BCH_TRANS_COMMIT_lazy_rw|
- BCH_TRANS_COMMIT_no_enospc) ?:
- 1;
- goto out;
}
int bch2_check_lrus(struct bch_fs *c)
struct btree_iter iter;
struct bkey_s_c k;
struct bpos last_flushed_pos = POS_MIN;
+ int ret = 0;
- int ret = bch2_trans_run(c,
- for_each_btree_key2(trans, iter,
- BTREE_ID_lru, POS_MIN, BTREE_ITER_PREFETCH, k, ({
- int ret2 = bch2_check_lru_key(trans, &iter, k, &last_flushed_pos);
-
- ret2 < 0 ? ret2 : 0;
- })));
- bch_err_fn(c, ret);
+ ret = bch2_trans_run(c,
+ for_each_btree_key_commit(trans, iter,
+ BTREE_ID_lru, POS_MIN, BTREE_ITER_PREFETCH, k,
+ NULL, NULL, BCH_TRANS_COMMIT_no_enospc|BCH_TRANS_COMMIT_lazy_rw,
+ bch2_check_lru_key(trans, &iter, k, &last_flushed_pos)));
+ if (ret)
+ bch_err_fn(c, ret);
return ret;
+
}
int bch2_lru_set(struct btree_trans *, u16, u64, u64);
int bch2_lru_change(struct btree_trans *, u16, u64, u64, u64);
-int bch2_check_lru_key(struct btree_trans *, struct btree_iter *,
- struct bkey_s_c, struct bpos *);
int bch2_check_lrus(struct bch_fs *);
#endif /* _BCACHEFS_LRU_H */
atomic_read(&ctxt->write_sectors) != sectors_pending);
}
-static void bch2_moving_ctxt_flush_all(struct moving_context *ctxt)
+void bch2_moving_ctxt_flush_all(struct moving_context *ctxt)
{
move_ctxt_wait_event(ctxt, list_empty(&ctxt->reads));
bch2_trans_unlock_long(ctxt->trans);
return ret;
}
-int __bch2_evacuate_bucket(struct moving_context *ctxt,
+int bch2_evacuate_bucket(struct moving_context *ctxt,
struct move_bucket_in_flight *bucket_in_flight,
struct bpos bucket, int gen,
struct data_update_opts _data_opts)
return ret;
}
-int bch2_evacuate_bucket(struct bch_fs *c,
- struct bpos bucket, int gen,
- struct data_update_opts data_opts,
- struct bch_ratelimit *rate,
- struct bch_move_stats *stats,
- struct write_point_specifier wp,
- bool wait_on_copygc)
-{
- struct moving_context ctxt;
- int ret;
-
- bch2_moving_ctxt_init(&ctxt, c, rate, stats, wp, wait_on_copygc);
- ret = __bch2_evacuate_bucket(&ctxt, NULL, bucket, gen, data_opts);
- bch2_moving_ctxt_exit(&ctxt);
-
- return ret;
-}
-
typedef bool (*move_btree_pred)(struct bch_fs *, void *,
struct btree *, struct bch_io_opts *,
struct data_update_opts *);
struct write_point_specifier, bool);
struct moving_io *bch2_moving_ctxt_next_pending_write(struct moving_context *);
void bch2_moving_ctxt_do_pending_writes(struct moving_context *);
+void bch2_moving_ctxt_flush_all(struct moving_context *);
void bch2_move_ctxt_wait_for_io(struct moving_context *);
int bch2_move_ratelimit(struct moving_context *);
bool,
move_pred_fn, void *);
-int __bch2_evacuate_bucket(struct moving_context *,
+int bch2_evacuate_bucket(struct moving_context *,
struct move_bucket_in_flight *,
struct bpos, int,
struct data_update_opts);
-int bch2_evacuate_bucket(struct bch_fs *, struct bpos, int,
- struct data_update_opts,
- struct bch_ratelimit *,
- struct bch_move_stats *,
- struct write_point_specifier,
- bool);
int bch2_data_job(struct bch_fs *,
struct bch_move_stats *,
struct bch_ioctl_data);
struct bkey_s_c k;
size_t nr_to_get = max_t(size_t, 16U, buckets_in_flight->nr / 4);
size_t saw = 0, in_flight = 0, not_movable = 0, sectors = 0;
- struct bpos last_flushed_pos = POS_MIN;
int ret;
move_buckets_wait(ctxt, buckets_in_flight, false);
lru_pos(BCH_LRU_FRAGMENTATION_START, 0, 0),
lru_pos(BCH_LRU_FRAGMENTATION_START, U64_MAX, LRU_TIME_MAX),
0, k, ({
- int ret2 = bch2_check_lru_key(trans, &iter, k, &last_flushed_pos);
- if (ret2) {
- ret2 = ret2 < 0 ? ret2 : 0;
- goto next;
- }
+ struct move_bucket b = { .k.bucket = u64_to_bucket(k.k->p.offset) };
+ int ret2 = 0;
saw++;
- struct move_bucket b = { .k.bucket = u64_to_bucket(k.k->p.offset) };
-
if (!bch2_bucket_is_movable(trans, &b, lru_pos_time(k.k->p)))
not_movable++;
else if (bucket_in_flight(buckets_in_flight, b.k))
if (ret2 >= 0)
sectors += b.sectors;
}
-next:
ret2;
}));
break;
}
- ret = __bch2_evacuate_bucket(ctxt, f, f->bucket.k.bucket,
+ ret = bch2_evacuate_bucket(ctxt, f, f->bucket.k.bucket,
f->bucket.k.gen, data_opts);
if (ret)
goto err;
if (!c->copy_gc_enabled) {
move_buckets_wait(&ctxt, buckets, true);
- kthread_wait_freezable(c->copy_gc_enabled);
+ kthread_wait_freezable(c->copy_gc_enabled ||
+ kthread_should_stop());
}
if (unlikely(freezing(current))) {
BTREE_ID_rebalance_work, POS_MIN,
BTREE_ITER_ALL_SNAPSHOTS);
- while (!bch2_move_ratelimit(ctxt) &&
- !kthread_wait_freezable(r->enabled)) {
+ while (!bch2_move_ratelimit(ctxt)) {
+ if (!r->enabled) {
+ bch2_moving_ctxt_flush_all(ctxt);
+ kthread_wait_freezable(c->copy_gc_enabled ||
+ kthread_should_stop());
+ }
+
+ if (kthread_should_stop())
+ break;
+
bch2_trans_begin(trans);
ret = bkey_err(k = next_rebalance_entry(trans, &rebalance_work_iter));
projects / bcachefs-tools-debian / commitdiff