Merge branch 'for-3.11/drivers' of git://git.kernel.dk/linux-block

Pull block IO driver bits from Jens Axboe:
 "As I mentioned in the core block pull request, due to real life
  circumstances the driver pull request would be late.  Now it looks
  like -rc2 late...  On the plus side, apart form the rsxx update, these
  are all things that I could argue could go in later in the cycle as
  they are fixes and not features.  So even though things are late, it's
  not ALL bad.

  The pull request contains:

   - Updates to bcache, all bug fixes, from Kent.

   - A pile of drbd bug fixes (no big features this time!).

   - xen blk front/back fixes.

   - rsxx driver updates, some of them deferred form 3.10.  So should be
     well cooked by now"

* 'for-3.11/drivers' of git://git.kernel.dk/linux-block: (63 commits)
  bcache: Allocation kthread fixes
  bcache: Fix GC_SECTORS_USED() calculation
  bcache: Journal replay fix
  bcache: Shutdown fix
  bcache: Fix a sysfs splat on shutdown
  bcache: Advertise that flushes are supported
  bcache: check for allocation failures
  bcache: Fix a dumb race
  bcache: Use standard utility code
  bcache: Update email address
  bcache: Delete fuzz tester
  bcache: Document shrinker reserve better
  bcache: FUA fixes
  drbd: Allow online change of al-stripes and al-stripe-size
  drbd: Constants should be UPPERCASE
  drbd: Ignore the exit code of a fence-peer handler if it returns too late
  drbd: Fix rcu_read_lock balance on error path
  drbd: fix error return code in drbd_init()
  drbd: Do not sleep inside rcu
  bcache: Refresh usage docs
  ...

21 files changed, 869 insertions, 801 deletions

diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c
index 048f2947e08..e45f5575fd4 100644
--- a/drivers/md/bcache/alloc.c
+++ b/drivers/md/bcache/alloc.c
@@ -63,7 +63,10 @@
 #include "bcache.h"
 #include "btree.h"
 
+#include <linux/freezer.h>
+#include <linux/kthread.h>
 #include <linux/random.h>
+#include <trace/events/bcache.h>
 
 #define MAX_IN_FLIGHT_DISCARDS		8U
 
@@ -151,7 +154,7 @@ static void discard_finish(struct work_struct *w)
 	mutex_unlock(&ca->set->bucket_lock);
 
 	closure_wake_up(&ca->set->bucket_wait);
-	wake_up(&ca->set->alloc_wait);
+	wake_up_process(ca->alloc_thread);
 
 	closure_put(&ca->set->cl);
 }
@@ -350,38 +353,30 @@ static void invalidate_buckets(struct cache *ca)
 		break;
 	}
 
-	pr_debug("free %zu/%zu free_inc %zu/%zu unused %zu/%zu",
-		 fifo_used(&ca->free), ca->free.size,
-		 fifo_used(&ca->free_inc), ca->free_inc.size,
-		 fifo_used(&ca->unused), ca->unused.size);
+	trace_bcache_alloc_invalidate(ca);
 }
 
 #define allocator_wait(ca, cond)					\
 do {									\
-	DEFINE_WAIT(__wait);						\
-									\
 	while (1) {							\
-		prepare_to_wait(&ca->set->alloc_wait,			\
-				&__wait, TASK_INTERRUPTIBLE);		\
+		set_current_state(TASK_INTERRUPTIBLE);			\
 		if (cond)						\
 			break;						\
 									\
 		mutex_unlock(&(ca)->set->bucket_lock);			\
-		if (test_bit(CACHE_SET_STOPPING_2, &ca->set->flags)) {	\
-			finish_wait(&ca->set->alloc_wait, &__wait);	\
-			closure_return(cl);				\
-		}							\
+		if (kthread_should_stop())				\
+			return 0;					\
 									\
+		try_to_freeze();					\
 		schedule();						\
 		mutex_lock(&(ca)->set->bucket_lock);			\
 	}								\
-									\
-	finish_wait(&ca->set->alloc_wait, &__wait);			\
+	__set_current_state(TASK_RUNNING);				\
 } while (0)
 
-void bch_allocator_thread(struct closure *cl)
+static int bch_allocator_thread(void *arg)
 {
-	struct cache *ca = container_of(cl, struct cache, alloc);
+	struct cache *ca = arg;
 
 	mutex_lock(&ca->set->bucket_lock);
 
@@ -442,7 +437,7 @@ long bch_bucket_alloc(struct cache *ca, unsigned watermark, struct closure *cl)
 {
 	long r = -1;
 again:
-	wake_up(&ca->set->alloc_wait);
+	wake_up_process(ca->alloc_thread);
 
 	if (fifo_used(&ca->free) > ca->watermark[watermark] &&
 	    fifo_pop(&ca->free, r)) {
@@ -476,9 +471,7 @@ again:
 		return r;
 	}
 
-	pr_debug("alloc failure: blocked %i free %zu free_inc %zu unused %zu",
-		 atomic_read(&ca->set->prio_blocked), fifo_used(&ca->free),
-		 fifo_used(&ca->free_inc), fifo_used(&ca->unused));
+	trace_bcache_alloc_fail(ca);
 
 	if (cl) {
 		closure_wait(&ca->set->bucket_wait, cl);
@@ -552,6 +545,17 @@ int bch_bucket_alloc_set(struct cache_set *c, unsigned watermark,
 
 /* Init */
 
+int bch_cache_allocator_start(struct cache *ca)
+{
+	struct task_struct *k = kthread_run(bch_allocator_thread,
+					    ca, "bcache_allocator");
+	if (IS_ERR(k))
+		return PTR_ERR(k);
+
+	ca->alloc_thread = k;
+	return 0;
+}
+
 void bch_cache_allocator_exit(struct cache *ca)
 {
 	struct discard *d;
diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index d3e15b42a4a..b39f6f0b45f 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -178,7 +178,6 @@
 #define pr_fmt(fmt) "bcache: %s() " fmt "\n", __func__
 
 #include <linux/bio.h>
-#include <linux/blktrace_api.h>
 #include <linux/kobject.h>
 #include <linux/list.h>
 #include <linux/mutex.h>
@@ -388,8 +387,6 @@ struct keybuf_key {
 typedef bool (keybuf_pred_fn)(struct keybuf *, struct bkey *);
 
 struct keybuf {
-	keybuf_pred_fn		*key_predicate;
-
 	struct bkey		last_scanned;
 	spinlock_t		lock;
 
@@ -437,9 +434,12 @@ struct bcache_device {
 
 	/* If nonzero, we're detaching/unregistering from cache set */
 	atomic_t		detaching;
+	int			flush_done;
+
+	uint64_t		nr_stripes;
+	unsigned		stripe_size_bits;
+	atomic_t		*stripe_sectors_dirty;
 
-	atomic_long_t		sectors_dirty;
-	unsigned long		sectors_dirty_gc;
 	unsigned long		sectors_dirty_last;
 	long			sectors_dirty_derivative;
 
@@ -531,6 +531,7 @@ struct cached_dev {
 	unsigned		sequential_merge:1;
 	unsigned		verify:1;
 
+	unsigned		partial_stripes_expensive:1;
 	unsigned		writeback_metadata:1;
 	unsigned		writeback_running:1;
 	unsigned char		writeback_percent;
@@ -565,8 +566,7 @@ struct cache {
 
 	unsigned		watermark[WATERMARK_MAX];
 
-	struct closure		alloc;
-	struct workqueue_struct	*alloc_workqueue;
+	struct task_struct	*alloc_thread;
 
 	struct closure		prio;
 	struct prio_set		*disk_buckets;
@@ -664,13 +664,9 @@ struct gc_stat {
  * CACHE_SET_STOPPING always gets set first when we're closing down a cache set;
  * we'll continue to run normally for awhile with CACHE_SET_STOPPING set (i.e.
  * flushing dirty data).
- *
- * CACHE_SET_STOPPING_2 gets set at the last phase, when it's time to shut down
- * the allocation thread.
  */
 #define CACHE_SET_UNREGISTERING		0
 #define	CACHE_SET_STOPPING		1
-#define	CACHE_SET_STOPPING_2		2
 
 struct cache_set {
 	struct closure		cl;
@@ -703,9 +699,6 @@ struct cache_set {
 	/* For the btree cache */
 	struct shrinker		shrink;
 
-	/* For the allocator itself */
-	wait_queue_head_t	alloc_wait;
-
 	/* For the btree cache and anything allocation related */
 	struct mutex		bucket_lock;
 
@@ -823,10 +816,9 @@ struct cache_set {
 
 	/*
 	 * A btree node on disk could have too many bsets for an iterator to fit
-	 * on the stack - this is a single element mempool for btree_read_work()
+	 * on the stack - have to dynamically allocate them
 	 */
-	struct mutex		fill_lock;
-	struct btree_iter	*fill_iter;
+	mempool_t		*fill_iter;
 
 	/*
 	 * btree_sort() is a merge sort and requires temporary space - single
@@ -834,6 +826,7 @@ struct cache_set {
 	 */
 	struct mutex		sort_lock;
 	struct bset		*sort;
+	unsigned		sort_crit_factor;
 
 	/* List of buckets we're currently writing data to */
 	struct list_head	data_buckets;
@@ -906,8 +899,6 @@ static inline unsigned local_clock_us(void)
 	return local_clock() >> 10;
 }
 
-#define MAX_BSETS		4U
-
 #define BTREE_PRIO		USHRT_MAX
 #define INITIAL_PRIO		32768
 
@@ -1112,23 +1103,6 @@ static inline void __bkey_put(struct cache_set *c, struct bkey *k)
 		atomic_dec_bug(&PTR_BUCKET(c, k, i)->pin);
 }
 
-/* Blktrace macros */
-
-#define blktrace_msg(c, fmt, ...)					\
-do {									\
-	struct request_queue *q = bdev_get_queue(c->bdev);		\
-	if (q)								\
-		blk_add_trace_msg(q, fmt, ##__VA_ARGS__);		\
-} while (0)
-
-#define blktrace_msg_all(s, fmt, ...)					\
-do {									\
-	struct cache *_c;						\
-	unsigned i;							\
-	for_each_cache(_c, (s), i)					\
-		blktrace_msg(_c, fmt, ##__VA_ARGS__);			\
-} while (0)
-
 static inline void cached_dev_put(struct cached_dev *dc)
 {
 	if (atomic_dec_and_test(&dc->count))
@@ -1173,10 +1147,16 @@ static inline uint8_t bucket_disk_gen(struct bucket *b)
 	static struct kobj_attribute ksysfs_##n =			\
 		__ATTR(n, S_IWUSR|S_IRUSR, show, store)
 
-/* Forward declarations */
+static inline void wake_up_allocators(struct cache_set *c)
+{
+	struct cache *ca;
+	unsigned i;
+
+	for_each_cache(ca, c, i)
+		wake_up_process(ca->alloc_thread);
+}
 
-void bch_writeback_queue(struct cached_dev *);
-void bch_writeback_add(struct cached_dev *, unsigned);
+/* Forward declarations */
 
 void bch_count_io_errors(struct cache *, int, const char *);
 void bch_bbio_count_io_errors(struct cache_set *, struct bio *,
@@ -1193,7 +1173,6 @@ void bch_submit_bbio(struct bio *, struct cache_set *, struct bkey *, unsigned);
 uint8_t bch_inc_gen(struct cache *, struct bucket *);
 void bch_rescale_priorities(struct cache_set *, int);
 bool bch_bucket_add_unused(struct cache *, struct bucket *);
-void bch_allocator_thread(struct closure *);
 
 long bch_bucket_alloc(struct cache *, unsigned, struct closure *);
 void bch_bucket_free(struct cache_set *, struct bkey *);
@@ -1241,9 +1220,9 @@ void bch_cache_set_stop(struct cache_set *);
 struct cache_set *bch_cache_set_alloc(struct cache_sb *);
 void bch_btree_cache_free(struct cache_set *);
 int bch_btree_cache_alloc(struct cache_set *);
-void bch_cached_dev_writeback_init(struct cached_dev *);
 void bch_moving_init_cache_set(struct cache_set *);
 
+int bch_cache_allocator_start(struct cache *ca);
 void bch_cache_allocator_exit(struct cache *ca);
 int bch_cache_allocator_init(struct cache *ca);
 
diff --git a/drivers/md/bcache/bset.c b/drivers/md/bcache/bset.c
index 1d27d3af325..8010eed06a5 100644
--- a/drivers/md/bcache/bset.c
+++ b/drivers/md/bcache/bset.c
@@ -78,6 +78,7 @@ struct bkey *bch_keylist_pop(struct keylist *l)
 bool __bch_ptr_invalid(struct cache_set *c, int level, const struct bkey *k)
 {
 	unsigned i;
+	char buf[80];
 
 	if (level && (!KEY_PTRS(k) || !KEY_SIZE(k) || KEY_DIRTY(k)))
 		goto bad;
@@ -102,7 +103,8 @@ bool __bch_ptr_invalid(struct cache_set *c, int level, const struct bkey *k)
 
 	return false;
 bad:
-	cache_bug(c, "spotted bad key %s: %s", pkey(k), bch_ptr_status(c, k));
+	bch_bkey_to_text(buf, sizeof(buf), k);
+	cache_bug(c, "spotted bad key %s: %s", buf, bch_ptr_status(c, k));
 	return true;
 }
 
@@ -162,10 +164,16 @@ bool bch_ptr_bad(struct btree *b, const struct bkey *k)
 #ifdef CONFIG_BCACHE_EDEBUG
 bug:
 	mutex_unlock(&b->c->bucket_lock);
-	btree_bug(b,
+
+	{
+		char buf[80];
+
+		bch_bkey_to_text(buf, sizeof(buf), k);
+		btree_bug(b,
 "inconsistent pointer %s: bucket %zu pin %i prio %i gen %i last_gc %i mark %llu gc_gen %i",
-		  pkey(k), PTR_BUCKET_NR(b->c, k, i), atomic_read(&g->pin),
-		  g->prio, g->gen, g->last_gc, GC_MARK(g), g->gc_gen);
+			  buf, PTR_BUCKET_NR(b->c, k, i), atomic_read(&g->pin),
+			  g->prio, g->gen, g->last_gc, GC_MARK(g), g->gc_gen);
+	}
 	return true;
 #endif
 }
@@ -1084,33 +1092,39 @@ void bch_btree_sort_into(struct btree *b, struct btree *new)
 	new->sets->size = 0;
 }
 
+#define SORT_CRIT	(4096 / sizeof(uint64_t))
+
 void bch_btree_sort_lazy(struct btree *b)
 {
-	if (b->nsets) {
-		unsigned i, j, keys = 0, total;
+	unsigned crit = SORT_CRIT;
+	int i;
 
-		for (i = 0; i <= b->nsets; i++)
-			keys += b->sets[i].data->keys;
-
-		total = keys;
+	/* Don't sort if nothing to do */
+	if (!b->nsets)
+		goto out;
 
-		for (j = 0; j < b->nsets; j++) {
-			if (keys * 2 < total ||
-			    keys < 1000) {
-				bch_btree_sort_partial(b, j);
-				return;
-			}
+	/* If not a leaf node, always sort */
+	if (b->level) {
+		bch_btree_sort(b);
+		return;
+	}
 
-			keys -= b->sets[j].data->keys;
-		}
+	for (i = b->nsets - 1; i >= 0; --i) {
+		crit *= b->c->sort_crit_factor;
 
-		/* Must sort if b->nsets == 3 or we'll overflow */
-		if (b->nsets >= (MAX_BSETS - 1) - b->level) {
-			bch_btree_sort(b);
+		if (b->sets[i].data->keys < crit) {
+			bch_btree_sort_partial(b, i);
 			return;
 		}
 	}
 
+	/* Sort if we'd overflow */
+	if (b->nsets + 1 == MAX_BSETS) {
+		bch_btree_sort(b);
+		return;
+	}
+
+out:
 	bset_build_written_tree(b);
 }
 
diff --git a/drivers/md/bcache/bset.h b/drivers/md/bcache/bset.h
index 57a9cff4154..ae115a253d7 100644
--- a/drivers/md/bcache/bset.h
+++ b/drivers/md/bcache/bset.h
@@ -1,6 +1,8 @@
 #ifndef _BCACHE_BSET_H
 #define _BCACHE_BSET_H
 
+#include <linux/slab.h>
+
 /*
  * BKEYS:
  *
@@ -142,6 +144,8 @@
 
 /* Btree key comparison/iteration */
 
+#define MAX_BSETS		4U
+
 struct btree_iter {
 	size_t size, used;
 	struct btree_iter_set {
diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c
index 7a5658f04e6..ee372884c40 100644
--- a/drivers/md/bcache/btree.c
+++ b/drivers/md/bcache/btree.c
@@ -24,6 +24,7 @@
 #include "btree.h"
 #include "debug.h"
 #include "request.h"
+#include "writeback.h"
 
 #include <linux/slab.h>
 #include <linux/bitops.h>
@@ -134,44 +135,17 @@ static uint64_t btree_csum_set(struct btree *b, struct bset *i)
 	return crc ^ 0xffffffffffffffffULL;
 }
 
-static void btree_bio_endio(struct bio *bio, int error)
+static void bch_btree_node_read_done(struct btree *b)
 {
-	struct closure *cl = bio->bi_private;
-	struct btree *b = container_of(cl, struct btree, io.cl);
-
-	if (error)
-		set_btree_node_io_error(b);
-
-	bch_bbio_count_io_errors(b->c, bio, error, (bio->bi_rw & WRITE)
-				 ? "writing btree" : "reading btree");
-	closure_put(cl);
-}
-
-static void btree_bio_init(struct btree *b)
-{
-	BUG_ON(b->bio);
-	b->bio = bch_bbio_alloc(b->c);
-
-	b->bio->bi_end_io	= btree_bio_endio;
-	b->bio->bi_private	= &b->io.cl;
-}
-
-void bch_btree_read_done(struct closure *cl)
-{
-	struct btree *b = container_of(cl, struct btree, io.cl);
-	struct bset *i = b->sets[0].data;
-	struct btree_iter *iter = b->c->fill_iter;
 	const char *err = "bad btree header";
-	BUG_ON(b->nsets || b->written);
-
-	bch_bbio_free(b->bio, b->c);
-	b->bio = NULL;
+	struct bset *i = b->sets[0].data;
+	struct btree_iter *iter;
 
-	mutex_lock(&b->c->fill_lock);
+	iter = mempool_alloc(b->c->fill_iter, GFP_NOWAIT);
+	iter->size = b->c->sb.bucket_size / b->c->sb.block_size;
 	iter->used = 0;
 
-	if (btree_node_io_error(b) ||
-	    !i->seq)
+	if (!i->seq)
 		goto err;
 
 	for (;
@@ -228,17 +202,8 @@ void bch_btree_read_done(struct closure *cl)
 	if (b->written < btree_blocks(b))
 		bch_bset_init_next(b);
 out:
-
-	mutex_unlock(&b->c->fill_lock);
-
-	spin_lock(&b->c->btree_read_time_lock);
-	bch_time_stats_update(&b->c->btree_read_time, b->io_start_time);
-	spin_unlock(&b->c->btree_read_time_lock);
-
-	smp_wmb(); /* read_done is our write lock */
-	set_btree_node_read_done(b);
-
-	closure_return(cl);
+	mempool_free(iter, b->c->fill_iter);
+	return;
 err:
 	set_btree_node_io_error(b);
 	bch_cache_set_error(b->c, "%s at bucket %zu, block %zu, %u keys",
@@ -247,48 +212,69 @@ err:
 	goto out;
 }
 
-void bch_btree_read(struct btree *b)
+static void btree_node_read_endio(struct bio *bio, int error)
+{
+	struct closure *cl = bio->bi_private;
+	closure_put(cl);
+}
+
+void bch_btree_node_read(struct btree *b)
 {
-	BUG_ON(b->nsets || b->written);
+	uint64_t start_time = local_clock();
+	struct closure cl;
+	struct bio *bio;
+
+	trace_bcache_btree_read(b);
+
+	closure_init_stack(&cl);
+
+	bio = bch_bbio_alloc(b->c);
+	bio->bi_rw	= REQ_META|READ_SYNC;
+	bio->bi_size	= KEY_SIZE(&b->key) << 9;
+	bio->bi_end_io	= btree_node_read_endio;
+	bio->bi_private	= &cl;
+
+	bch_bio_map(bio, b->sets[0].data);
+
+	bch_submit_bbio(bio, b->c, &b->key, 0);
+	closure_sync(&cl);
 
-	if (!closure_trylock(&b->io.cl, &b->c->cl))
-		BUG();
+	if (!test_bit(BIO_UPTODATE, &bio->bi_flags))
+		set_btree_node_io_error(b);
 
-	b->io_start_time = local_clock();
+	bch_bbio_free(bio, b->c);
 
-	btree_bio_init(b);
-	b->bio->bi_rw	= REQ_META|READ_SYNC;
-	b->bio->bi_size	= KEY_SIZE(&b->key) << 9;
+	if (btree_node_io_error(b))
+		goto err;
 
-	bch_bio_map(b->bio, b->sets[0].data);
+	bch_btree_node_read_done(b);
 
-	pr_debug("%s", pbtree(b));
-	trace_bcache_btree_read(b->bio);
-	bch_submit_bbio(b->bio, b->c, &b->key, 0);
+	spin_lock(&b->c->btree_read_time_lock);
+	bch_time_stats_update(&b->c->btree_read_time, start_time);
+	spin_unlock(&b->c->btree_read_time_lock);
 
-	continue_at(&b->io.cl, bch_btree_read_done, system_wq);
+	return;
+err:
+	bch_cache_set_error(b->c, "io error reading bucket %lu",
+			    PTR_BUCKET_NR(b->c, &b->key, 0));
 }
 
 static void btree_complete_write(struct btree *b, struct btree_write *w)
 {
 	if (w->prio_blocked &&
 	    !atomic_sub_return(w->prio_blocked, &b->c->prio_blocked))
-		wake_up(&b->c->alloc_wait);
+		wake_up_allocators(b->c);
 
 	if (w->journal) {
 		atomic_dec_bug(w->journal);
 		__closure_wake_up(&b->c->journal.wait);
 	}
 
-	if (w->owner)
-		closure_put(w->owner);
-
 	w->prio_blocked	= 0;
 	w->journal	= NULL;
-	w->owner	= NULL;
 }
 
-static void __btree_write_done(struct closure *cl)
+static void __btree_node_write_done(struct closure *cl)
 {
 	struct btree *b = container_of(cl, struct btree, io.cl);
 	struct btree_write *w = btree_prev_write(b);
@@ -304,7 +290,7 @@ static void __btree_write_done(struct closure *cl)
 	closure_return(cl);
 }
 
-static void btree_write_done(struct closure *cl)
+static void btree_node_write_done(struct closure *cl)
 {
 	struct btree *b = container_of(cl, struct btree, io.cl);
 	struct bio_vec *bv;
@@ -313,10 +299,22 @@ static void btree_write_done(struct closure *cl)
 	__bio_for_each_segment(bv, b->bio, n, 0)
 		__free_page(bv->bv_page);
 
-	__btree_write_done(cl);
+	__btree_node_write_done(cl);
 }
 
-static void do_btree_write(struct btree *b)
+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);
+
+	if (error)
+		set_btree_node_io_error(b);
+
+	bch_bbio_count_io_errors(b->c, bio, error, "writing btree");
+	closure_put(cl);
+}
+
+static void do_btree_node_write(struct btree *b)
 {
 	struct closure *cl = &b->io.cl;
 	struct bset *i = b->sets[b->nsets].data;
@@ -325,15 +323,34 @@ static void do_btree_write(struct btree *b)
 	i->version	= BCACHE_BSET_VERSION;
 	i->csum		= btree_csum_set(b, i);
 
-	btree_bio_init(b);
-	b->bio->bi_rw	= REQ_META|WRITE_SYNC;
-	b->bio->bi_size	= set_blocks(i, b->c) * block_bytes(b->c);
+	BUG_ON(b->bio);
+	b->bio = bch_bbio_alloc(b->c);
+
+	b->bio->bi_end_io	= btree_node_write_endio;
+	b->bio->bi_private	= &b->io.cl;
+	b->bio->bi_rw		= REQ_META|WRITE_SYNC|REQ_FUA;
+	b->bio->bi_size		= set_blocks(i, b->c) * block_bytes(b->c);
 	bch_bio_map(b->bio, i);
 
+	/*
+	 * If we're appending to a leaf node, we don't technically need FUA -
+	 * this write just needs to be persisted before the next journal write,
+	 * which will be marked FLUSH|FUA.
+	 *
+	 * Similarly if we're writing a new btree root - the pointer is going to
+	 * be in the next journal entry.
+	 *
+	 * But if we're writing a new btree node (that isn't a root) or
+	 * appending to a non leaf btree node, we need either FUA or a flush
+	 * when we write the parent with the new pointer. FUA is cheaper than a
+	 * flush, and writes appending to leaf nodes aren't blocking anything so
+	 * just make all btree node writes FUA to keep things sane.
+	 */
+
 	bkey_copy(&k.key, &b->key);
 	SET_PTR_OFFSET(&k.key, 0, PTR_OFFSET(&k.key, 0) + bset_offset(b, i));
 
-	if (!bch_bio_alloc_pages(b->bio, GFP_NOIO)) {
+	if (!bio_alloc_pages(b->bio, GFP_NOIO)) {
 		int j;
 		struct bio_vec *bv;
 		void *base = (void *) ((unsigned long) i & ~(PAGE_SIZE - 1));
@@ -342,40 +359,41 @@ static void do_btree_write(struct btree *b)
 			memcpy(page_address(bv->bv_page),
 			       base + j * PAGE_SIZE, PAGE_SIZE);
 
-		trace_bcache_btree_write(b->bio);
 		bch_submit_bbio(b->bio, b->c, &k.key, 0);
 
-		continue_at(cl, btree_write_done, NULL);
+		continue_at(cl, btree_node_write_done, NULL);
 	} else {
 		b->bio->bi_vcnt = 0;
 		bch_bio_map(b->bio, i);
 
-		trace_bcache_btree_write(b->bio);
 		bch_submit_bbio(b->bio, b->c, &k.key, 0);
 
 		closure_sync(cl);
-		__btree_write_done(cl);
+		__btree_node_write_done(cl);
 	}
 }
 
-static void __btree_write(struct btree *b)
+void bch_btree_node_write(struct btree *b, struct closure *parent)
 {
 	struct bset *i = b->sets[b->nsets].data;
 
+	trace_bcache_btree_write(b);
+
 	BUG_ON(current->bio_list);
+	BUG_ON(b->written >= btree_blocks(b));
+	BUG_ON(b->written && !i->keys);
+	BUG_ON(b->sets->data->seq != i->seq);
+	bch_check_key_order(b, i);
 
-	closure_lock(&b->io, &b->c->cl);
 	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);
+
 	clear_bit(BTREE_NODE_dirty,	 &b->flags);
 	change_bit(BTREE_NODE_write_idx, &b->flags);
 
-	bch_check_key_order(b, i);
-	BUG_ON(b->written && !i->keys);
-
-	do_btree_write(b);
-
-	pr_debug("%s block %i keys %i", pbtree(b), b->written, i->keys);
+	do_btree_node_write(b);
 
 	b->written += set_blocks(i, b->c);
 	atomic_long_add(set_blocks(i, b->c) * b->c->sb.block_size,
@@ -387,37 +405,31 @@ static void __btree_write(struct btree *b)
 		bch_bset_init_next(b);
 }
 
-static void btree_write_work(struct work_struct *w)
+static void btree_node_write_work(struct work_struct *w)
 {
 	struct btree *b = container_of(to_delayed_work(w), struct btree, work);
 
-	down_write(&b->lock);
+	rw_lock(true, b, b->level);
 
 	if (btree_node_dirty(b))
-		__btree_write(b);
-	up_write(&b->lock);
+		bch_btree_node_write(b, NULL);
+	rw_unlock(true, b);
 }
 
-void bch_btree_write(struct btree *b, bool now, struct btree_op *op)
+static void bch_btree_leaf_dirty(struct btree *b, struct btree_op *op)
 {
 	struct bset *i = b->sets[b->nsets].data;
 	struct btree_write *w = btree_current_write(b);
 
-	BUG_ON(b->written &&
-	       (b->written >= btree_blocks(b) ||
-		i->seq != b->sets[0].data->seq ||
-		!i->keys));
+	BUG_ON(!b->written);
+	BUG_ON(!i->keys);
 
-	if (!btree_node_dirty(b)) {
-		set_btree_node_dirty(b);
-		queue_delayed_work(btree_io_wq, &b->work,
-				   msecs_to_jiffies(30000));
-	}
+	if (!btree_node_dirty(b))
+		queue_delayed_work(btree_io_wq, &b->work, 30 * HZ);
 
-	w->prio_blocked += b->prio_blocked;
-	b->prio_blocked = 0;
+	set_btree_node_dirty(b);
 
-	if (op && op->journal && !b->level) {
+	if (op && op->journal) {
 		if (w->journal &&
 		    journal_pin_cmp(b->c, w, op)) {
 			atomic_dec_bug(w->journal);
@@ -430,23 +442,10 @@ void bch_btree_write(struct btree *b, bool now, struct btree_op *op)
 		}
 	}
 
-	if (current->bio_list)
-		return;
-
 	/* Force write if set is too big */
-	if (now ||
-	    b->level ||
-	    set_bytes(i) > PAGE_SIZE - 48) {
-		if (op && now) {
-			/* Must wait on multiple writes */
-			BUG_ON(w->owner);
-			w->owner = &op->cl;
-			closure_get(&op->cl);
-		}
-
-		__btree_write(b);
-	}
-	BUG_ON(!b->written);
+	if (set_bytes(i) > PAGE_SIZE - 48 &&
+	    !current->bio_list)
+		bch_btree_node_write(b, NULL);
 }
 
 /*
@@ -559,7 +558,7 @@ static struct btree *mca_bucket_alloc(struct cache_set *c,
 	init_rwsem(&b->lock);
 	lockdep_set_novalidate_class(&b->lock);
 	INIT_LIST_HEAD(&b->list);
-	INIT_DELAYED_WORK(&b->work, btree_write_work);
+	INIT_DELAYED_WORK(&b->work, btree_node_write_work);
 	b->c = c;
 	closure_init_unlocked(&b->io);
 
@@ -582,7 +581,7 @@ static int mca_reap(struct btree *b, struct closure *cl, unsigned min_order)
 	BUG_ON(btree_node_dirty(b) && !b->sets[0].data);
 
 	if (cl && btree_node_dirty(b))
-		bch_btree_write(b, true, NULL);
+		bch_btree_node_write(b, NULL);
 
 	if (cl)
 		closure_wait_event_async(&b->io.wait, cl,
@@ -623,6 +622,13 @@ static int bch_mca_shrink(struct shrinker *shrink, struct shrink_control *sc)
 	else if (!mutex_trylock(&c->bucket_lock))
 		return -1;
 
+	/*
+	 * It's _really_ critical that we don't free too many btree nodes - we
+	 * have to always leave ourselves a reserve. The reserve is how we
+	 * guarantee that allocating memory for a new btree node can always
+	 * succeed, so that inserting keys into the btree can always succeed and
+	 * IO can always make forward progress:
+	 */
 	nr /= c->btree_pages;
 	nr = min_t(unsigned long, nr, mca_can_free(c));
 
@@ -766,6 +772,8 @@ static struct btree *mca_cannibalize(struct cache_set *c, struct bkey *k,
 	int ret = -ENOMEM;
 	struct btree *i;
 
+	trace_bcache_btree_cache_cannibalize(c);
+
 	if (!cl)
 		return ERR_PTR(-ENOMEM);
 
@@ -784,7 +792,6 @@ static struct btree *mca_cannibalize(struct cache_set *c, struct bkey *k,
 		return ERR_PTR(-EAGAIN);
 	}
 
-	/* XXX: tracepoint */
 	c->try_harder = cl;
 	c->try_harder_start = local_clock();
 retry:
@@ -905,6 +912,9 @@ retry:
 	b = mca_find(c, k);
 
 	if (!b) {
+		if (current->bio_list)
+			return ERR_PTR(-EAGAIN);
+
 		mutex_lock(&c->bucket_lock);
 		b = mca_alloc(c, k, level, &op->cl);
 		mutex_unlock(&c->bucket_lock);
@@ -914,7 +924,7 @@ retry:
 		if (IS_ERR(b))
 			return b;
 
-		bch_btree_read(b);
+		bch_btree_node_read(b);
 
 		if (!write)
 			downgrade_write(&b->lock);
@@ -937,15 +947,12 @@ retry:
 	for (; i <= b->nsets; i++)
 		prefetch(b->sets[i].data);
 
-	if (!closure_wait_event(&b->io.wait, &op->cl,
-				btree_node_read_done(b))) {
-		rw_unlock(write, b);
-		b = ERR_PTR(-EAGAIN);
-	} else if (btree_node_io_error(b)) {
+	if (btree_node_io_error(b)) {
 		rw_unlock(write, b);
-		b = ERR_PTR(-EIO);
-	} else
-		BUG_ON(!b->written);
+		return ERR_PTR(-EIO);
+	}
+
+	BUG_ON(!b->written);
 
 	return b;
 }
@@ -959,7 +966,7 @@ static void btree_node_prefetch(struct cache_set *c, struct bkey *k, int level)
 	mutex_unlock(&c->bucket_lock);
 
 	if (!IS_ERR_OR_NULL(b)) {
-		bch_btree_read(b);
+		bch_btree_node_read(b);
 		rw_unlock(true, b);
 	}
 }
@@ -970,24 +977,19 @@ static void btree_node_free(struct btree *b, struct btree_op *op)
 {
 	unsigned i;
 
+	trace_bcache_btree_node_free(b);
+
 	/*
 	 * The BUG_ON() in btree_node_get() implies that we must have a write
 	 * lock on parent to free or even invalidate a node
 	 */
 	BUG_ON(op->lock <= b->level);
 	BUG_ON(b == b->c->root);
-	pr_debug("bucket %s", pbtree(b));
 
 	if (btree_node_dirty(b))
 		btree_complete_write(b, btree_current_write(b));
 	clear_bit(BTREE_NODE_dirty, &b->flags);
 
-	if (b->prio_blocked &&
-	    !atomic_sub_return(b->prio_blocked, &b->c->prio_blocked))
-		wake_up(&b->c->alloc_wait);
-
-	b->prio_blocked = 0;
-
 	cancel_delayed_work(&b->work);
 
 	mutex_lock(&b->c->bucket_lock);
@@ -1028,17 +1030,20 @@ retry:
 		goto retry;
 	}
 
-	set_btree_node_read_done(b);
 	b->accessed = 1;
 	bch_bset_init_next(b);
 
 	mutex_unlock(&c->bucket_lock);
+
+	trace_bcache_btree_node_alloc(b);
 	return b;
 err_free:
 	bch_bucket_free(c, &k.key);
 	__bkey_put(c, &k.key);
 err:
 	mutex_unlock(&c->bucket_lock);
+
+	trace_bcache_btree_node_alloc_fail(b);
 	return b;
 }
 
@@ -1137,11 +1142,8 @@ static int btree_gc_mark_node(struct btree *b, unsigned *keys,
 		gc->nkeys++;
 
 		gc->data += KEY_SIZE(k);
-		if (KEY_DIRTY(k)) {
+		if (KEY_DIRTY(k))
 			gc->dirty += KEY_SIZE(k);
-			if (d)
-				d->sectors_dirty_gc += KEY_SIZE(k);
-		}
 	}
 
 	for (t = b->sets; t <= &b->sets[b->nsets]; t++)
@@ -1166,14 +1168,11 @@ static struct btree