Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit cb7a583e authored by Kent Overstreet's avatar Kent Overstreet
Browse files

bcache: kill closure locking usage 



Signed-off-by: default avatarKent Overstreet <kmo@daterainc.com>
parent a5ae4300

drivers/md/bcache/bcache.h

+6 −3
Original line number Diff line number Diff line
@@ -309,7 +309,8 @@ struct cached_dev { 
	struct cache_sb		sb;

	struct bio		sb_bio;

	struct bio_vec		sb_bv[1];

	struct closure_with_waitlist sb_write;

	struct closure		sb_write;

	struct semaphore	sb_write_mutex;



	/* Refcount on the cache set. Always nonzero when we're caching. */

	atomic_t		count;
@@ -514,7 +515,8 @@ struct cache_set { 
	uint64_t		cached_dev_sectors;

	struct closure		caching;



	struct closure_with_waitlist sb_write;

	struct closure		sb_write;

	struct semaphore	sb_write_mutex;



	mempool_t		*search;

	mempool_t		*bio_meta;
@@ -635,7 +637,8 @@ struct cache_set { 
	unsigned		nr_uuids;

	struct uuid_entry	*uuids;

	BKEY_PADDED(uuid_bucket);

	struct closure_with_waitlist uuid_write;

	struct closure		uuid_write;

	struct semaphore	uuid_write_mutex;



	/*

	 * A btree node on disk could have too many bsets for an iterator to fit

drivers/md/bcache/btree.c

+34 −18
Original line number Diff line number Diff line
@@ -340,9 +340,16 @@ static void btree_complete_write(struct btree *b, struct btree_write *w) 
	w->journal	= NULL;

}



static void btree_node_write_unlock(struct closure *cl)

{

	struct btree *b = container_of(cl, struct btree, io);



	up(&b->io_mutex);

}



static void __btree_node_write_done(struct closure *cl)

{

	struct btree *b = container_of(cl, struct btree, io.cl);

	struct btree *b = container_of(cl, struct btree, io);

	struct btree_write *w = btree_prev_write(b);



	bch_bbio_free(b->bio, b->c);
@@ -353,12 +360,12 @@ static void __btree_node_write_done(struct closure *cl) 
		queue_delayed_work(btree_io_wq, &b->work,

				   msecs_to_jiffies(30000));



	closure_return(cl);

	closure_return_with_destructor(cl, btree_node_write_unlock);

}



static void btree_node_write_done(struct closure *cl)

{

	struct btree *b = container_of(cl, struct btree, io.cl);

	struct btree *b = container_of(cl, struct btree, io);

	struct bio_vec *bv;

	int n;
@@ -371,7 +378,7 @@ static void btree_node_write_done(struct closure *cl) 
static void btree_node_write_endio(struct bio *bio, int error)

{

	struct closure *cl = bio->bi_private;

	struct btree *b = container_of(cl, struct btree, io.cl);

	struct btree *b = container_of(cl, struct btree, io);



	if (error)

		set_btree_node_io_error(b);
@@ -382,7 +389,7 @@ static void btree_node_write_endio(struct bio *bio, int error) 


static void do_btree_node_write(struct btree *b)

{

	struct closure *cl = &b->io.cl;

	struct closure *cl = &b->io;

	struct bset *i = b->sets[b->nsets].data;

	BKEY_PADDED(key) k;
@@ -435,7 +442,7 @@ static void do_btree_node_write(struct btree *b) 
		bch_submit_bbio(b->bio, b->c, &k.key, 0);



		closure_sync(cl);

		__btree_node_write_done(cl);

		continue_at_nobarrier(cl, __btree_node_write_done, NULL);

	}

}
@@ -454,7 +461,8 @@ void bch_btree_node_write(struct btree *b, struct closure *parent) 
	cancel_delayed_work(&b->work);



	/* If caller isn't waiting for write, parent refcount is cache set */

	closure_lock(&b->io, parent ?: &b->c->cl);

	down(&b->io_mutex);

	closure_init(&b->io, parent ?: &b->c->cl);



	clear_bit(BTREE_NODE_dirty,	 &b->flags);

	change_bit(BTREE_NODE_write_idx, &b->flags);
@@ -554,7 +562,8 @@ static void mca_reinit(struct btree *b) 
static void mca_data_free(struct btree *b)

{

	struct bset_tree *t = b->sets;

	BUG_ON(!closure_is_unlocked(&b->io.cl));



	BUG_ON(b->io_mutex.count != 1);



	if (bset_prev_bytes(b) < PAGE_SIZE)

		kfree(t->prev);
@@ -635,7 +644,7 @@ static struct btree *mca_bucket_alloc(struct cache_set *c, 
	INIT_LIST_HEAD(&b->list);

	INIT_DELAYED_WORK(&b->work, btree_node_write_work);

	b->c = c;

	closure_init_unlocked(&b->io);

	sema_init(&b->io_mutex, 1);



	mca_data_alloc(b, k, gfp);

	return b;
@@ -653,22 +662,29 @@ static int mca_reap(struct btree *b, unsigned min_order, bool flush) 


	BUG_ON(btree_node_dirty(b) && !b->sets[0].data);



	if (b->page_order < min_order ||

	    (!flush &&

	     (btree_node_dirty(b) ||

	      atomic_read(&b->io.cl.remaining) != -1))) {

		rw_unlock(true, b);

		return -ENOMEM;

	if (b->page_order < min_order)

		goto out_unlock;



	if (!flush) {

		if (btree_node_dirty(b))

			goto out_unlock;



		if (down_trylock(&b->io_mutex))

			goto out_unlock;

		up(&b->io_mutex);

	}



	if (btree_node_dirty(b))

		bch_btree_node_write_sync(b);



	/* wait for any in flight btree write */

	closure_wait_event(&b->io.wait, &cl,

			   atomic_read(&b->io.cl.remaining) == -1);

	down(&b->io_mutex);

	up(&b->io_mutex);



	return 0;

out_unlock:

	rw_unlock(true, b);

	return -ENOMEM;

}



static unsigned long bch_mca_scan(struct shrinker *shrink,
@@ -918,7 +934,7 @@ static struct btree *mca_alloc(struct cache_set *c, struct bkey *k, int level) 
	if (!b->sets->data)

		goto err;

out:

	BUG_ON(!closure_is_unlocked(&b->io.cl));

	BUG_ON(b->io_mutex.count != 1);



	bkey_copy(&b->key, k);

	list_move(&b->list, &c->btree_cache);

drivers/md/bcache/btree.h

+2 −1
Original line number Diff line number Diff line
@@ -143,7 +143,8 @@ struct btree { 
	struct bset_tree	sets[MAX_BSETS];



	/* For outstanding btree writes, used as a lock - protects write_idx */

	struct closure_with_waitlist	io;

	struct closure		io;

	struct semaphore	io_mutex;



	struct list_head	list;

	struct delayed_work	work;

drivers/md/bcache/debug.c

+2 −5
Original line number Diff line number Diff line
@@ -127,9 +127,7 @@ void bch_btree_verify(struct btree *b, struct bset *new) 
	if (!b->c->verify)

		return;



	closure_wait_event(&b->io.wait, &cl,

			   atomic_read(&b->io.cl.remaining) == -1);



	down(&b->io_mutex);

	mutex_lock(&b->c->verify_lock);



	bkey_copy(&v->key, &b->key);
@@ -137,8 +135,6 @@ void bch_btree_verify(struct btree *b, struct bset *new) 
	v->level = b->level;



	bch_btree_node_read(v);

	closure_wait_event(&v->io.wait, &cl,

			   atomic_read(&b->io.cl.remaining) == -1);



	if (new->keys != v->sets[0].data->keys ||

	    memcmp(new->start,
@@ -167,6 +163,7 @@ void bch_btree_verify(struct btree *b, struct bset *new) 
	}



	mutex_unlock(&b->c->verify_lock);

	up(&b->io_mutex);

}



void bch_data_verify(struct cached_dev *dc, struct bio *bio)

drivers/md/bcache/journal.c

+14 −13
Original line number Diff line number Diff line
@@ -564,6 +564,14 @@ static void journal_write_done(struct closure *cl) 
	continue_at_nobarrier(cl, journal_write, system_wq);

}



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);

}



static void journal_write_unlocked(struct closure *cl)

	__releases(c->journal.lock)

{
@@ -578,15 +586,7 @@ static void journal_write_unlocked(struct closure *cl) 
	bio_list_init(&list);



	if (!w->need_write) {

		/*

		 * XXX: have to unlock closure before we unlock journal lock,

		 * else we race with bch_journal(). But this way we race

		 * against cache set unregister. Doh.

		 */

		set_closure_fn(cl, NULL, NULL);

		closure_sub(cl, CLOSURE_RUNNING + 1);

		spin_unlock(&c->journal.lock);

		return;

		closure_return_with_destructor(cl, journal_write_unlock);

	} else if (journal_full(&c->journal)) {

		journal_reclaim(c);

		spin_unlock(&c->journal.lock);
@@ -662,11 +662,13 @@ static void journal_try_write(struct cache_set *c) 


	w->need_write = true;



	if (closure_trylock(cl, &c->cl))

		journal_write_unlocked(cl);

	else

	if (!c->journal.io_in_flight) {

		c->journal.io_in_flight = 1;

		closure_call(cl, journal_write_unlocked, NULL, &c->cl);

	} else {

		spin_unlock(&c->journal.lock);

	}

}



static struct journal_write *journal_wait_for_write(struct cache_set *c,

						    unsigned nkeys)
@@ -793,7 +795,6 @@ int bch_journal_alloc(struct cache_set *c) 
{

	struct journal *j = &c->journal;



	closure_init_unlocked(&j->io);

	spin_lock_init(&j->lock);

	INIT_DELAYED_WORK(&j->work, journal_write_work);