Update bcachefs sources to feaca6edbd24 mean and variance: Promote to lib/math

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

45 files changed, 515 insertions, 989 deletions

diff --git a/.bcachefs_revision b/.bcachefs_revision
index 97936a15..e41bb401 100644
--- a/.bcachefs_revision
+++ b/.bcachefs_revision
@@ -1 +1 @@
-8c94740b1bf8645d3398170f41c9c88b78332252
+feaca6edbd240bbd98d261097a97037c56a09eec
diff --git a/cmd_fsck.c b/cmd_fsck.c
index 0954a83c..f7dcae98 100644
--- a/cmd_fsck.c
+++ b/cmd_fsck.c
@@ -97,11 +97,11 @@ int cmd_fsck(int argc, char *argv[])
 		exit(8);
 	}
 
-	if (test_bit(BCH_FS_ERRORS_FIXED, &c->flags)) {
+	if (test_bit(BCH_FS_errors_fixed, &c->flags)) {
 		fprintf(stderr, "%s: errors fixed\n", c->name);
 		ret |= 1;
 	}
-	if (test_bit(BCH_FS_ERROR, &c->flags)) {
+	if (test_bit(BCH_FS_error, &c->flags)) {
 		fprintf(stderr, "%s: still has errors\n", c->name);
 		ret |= 4;
 	}
diff --git a/libbcachefs/backpointers.c b/libbcachefs/backpointers.c
index 4c8bcf23..8c66333b 100644
--- a/libbcachefs/backpointers.c
+++ b/libbcachefs/backpointers.c
@@ -136,15 +136,30 @@ static noinline int backpointer_mod_err(struct btree_trans *trans,
 }
 
 int bch2_bucket_backpointer_mod_nowritebuffer(struct btree_trans *trans,
-				struct bkey_i_backpointer *bp_k,
+				struct bpos bucket,
 				struct bch_backpointer bp,
 				struct bkey_s_c orig_k,
 				bool insert)
 {
 	struct btree_iter bp_iter;
 	struct bkey_s_c k;
+	struct bkey_i_backpointer *bp_k;
 	int ret;
 
+	bp_k = bch2_trans_kmalloc_nomemzero(trans, sizeof(struct bkey_i_backpointer));
+	ret = PTR_ERR_OR_ZERO(bp_k);
+	if (ret)
+		return ret;
+
+	bkey_backpointer_init(&bp_k->k_i);
+	bp_k->k.p = bucket_pos_to_bp(trans->c, bucket, bp.bucket_offset);
+	bp_k->v = bp;
+
+	if (!insert) {
+		bp_k->k.type = KEY_TYPE_deleted;
+		set_bkey_val_u64s(&bp_k->k, 0);
+	}
+
 	k = bch2_bkey_get_iter(trans, &bp_iter, BTREE_ID_backpointers,
 			       bp_k->k.p,
 			       BTREE_ITER_INTENT|
diff --git a/libbcachefs/backpointers.h b/libbcachefs/backpointers.h
index ab866fee..737e2396 100644
--- a/libbcachefs/backpointers.h
+++ b/libbcachefs/backpointers.h
@@ -63,7 +63,7 @@ static inline struct bpos bucket_pos_to_bp(const struct bch_fs *c,
 	return ret;
 }
 
-int bch2_bucket_backpointer_mod_nowritebuffer(struct btree_trans *, struct bkey_i_backpointer *,
+int bch2_bucket_backpointer_mod_nowritebuffer(struct btree_trans *, struct bpos bucket,
 				struct bch_backpointer, struct bkey_s_c, bool);
 
 static inline int bch2_bucket_backpointer_mod(struct btree_trans *trans,
@@ -72,28 +72,21 @@ static inline int bch2_bucket_backpointer_mod(struct btree_trans *trans,
 				struct bkey_s_c orig_k,
 				bool insert)
 {
-	struct bch_fs *c = trans->c;
-	struct bkey_i_backpointer *bp_k;
-	int ret;
+	if (unlikely(bch2_backpointers_no_use_write_buffer))
+		return bch2_bucket_backpointer_mod_nowritebuffer(trans, bucket, bp, orig_k, insert);
 
-	bp_k = bch2_trans_kmalloc_nomemzero(trans, sizeof(struct bkey_i_backpointer));
-	ret = PTR_ERR_OR_ZERO(bp_k);
-	if (ret)
-		return ret;
+	struct bkey_i_backpointer bp_k;
 
-	bkey_backpointer_init(&bp_k->k_i);
-	bp_k->k.p = bucket_pos_to_bp(c, bucket, bp.bucket_offset);
-	bp_k->v = bp;
+	bkey_backpointer_init(&bp_k.k_i);
+	bp_k.k.p = bucket_pos_to_bp(trans->c, bucket, bp.bucket_offset);
+	bp_k.v = bp;
 
 	if (!insert) {
-		bp_k->k.type = KEY_TYPE_deleted;
-		set_bkey_val_u64s(&bp_k->k, 0);
+		bp_k.k.type = KEY_TYPE_deleted;
+		set_bkey_val_u64s(&bp_k.k, 0);
 	}
 
-	if (unlikely(bch2_backpointers_no_use_write_buffer))
-		return bch2_bucket_backpointer_mod_nowritebuffer(trans, bp_k, bp, orig_k, insert);
-
-	return bch2_trans_update_buffered(trans, BTREE_ID_backpointers, &bp_k->k_i);
+	return bch2_trans_update_buffered(trans, BTREE_ID_backpointers, &bp_k.k_i);
 }
 
 static inline enum bch_data_type bkey_ptr_data_type(enum btree_id btree_id, unsigned level,
diff --git a/libbcachefs/bcachefs.h b/libbcachefs/bcachefs.h
index 295efeda..2e9f4af3 100644
--- a/libbcachefs/bcachefs.h
+++ b/libbcachefs/bcachefs.h
@@ -406,7 +406,6 @@ BCH_DEBUG_PARAMS_DEBUG()
 	x(blocked_journal_max_in_flight)	\
 	x(blocked_allocate)			\
 	x(blocked_allocate_open_bucket)		\
-	x(blocked_write_buffer_full)		\
 	x(nocow_lock_contended)
 
 enum bch_time_stats {
@@ -567,32 +566,38 @@ struct bch_dev {
 	struct io_count __percpu *io_done;
 };
 
-enum {
-	/* startup: */
-	BCH_FS_STARTED,
-	BCH_FS_MAY_GO_RW,
-	BCH_FS_RW,
-	BCH_FS_WAS_RW,
-
-	/* shutdown: */
-	BCH_FS_STOPPING,
-	BCH_FS_EMERGENCY_RO,
-	BCH_FS_GOING_RO,
-	BCH_FS_WRITE_DISABLE_COMPLETE,
-	BCH_FS_CLEAN_SHUTDOWN,
-
-	/* fsck passes: */
-	BCH_FS_FSCK_DONE,
-	BCH_FS_INITIAL_GC_UNFIXED,	/* kill when we enumerate fsck errors */
-	BCH_FS_NEED_ANOTHER_GC,
-
-	BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS,
-
-	/* errors: */
-	BCH_FS_ERROR,
-	BCH_FS_TOPOLOGY_ERROR,
-	BCH_FS_ERRORS_FIXED,
-	BCH_FS_ERRORS_NOT_FIXED,
+/*
+ * fsck_done - kill?
+ *
+ * replace with something more general from enumated fsck passes/errors:
+ * initial_gc_unfixed
+ * error
+ * topology error
+ */
+
+#define BCH_FS_FLAGS()			\
+	x(started)			\
+	x(may_go_rw)			\
+	x(rw)				\
+	x(was_rw)			\
+	x(stopping)			\
+	x(emergency_ro)			\
+	x(going_ro)			\
+	x(write_disable_complete)	\
+	x(clean_shutdown)		\
+	x(fsck_done)			\
+	x(initial_gc_unfixed)		\
+	x(need_another_gc)		\
+	x(need_delete_dead_snapshots)	\
+	x(error)			\
+	x(topology_error)		\
+	x(errors_fixed)			\
+	x(errors_not_fixed)
+
+enum bch_fs_flags {
+#define x(n)		BCH_FS_##n,
+	BCH_FS_FLAGS()
+#undef x
 };
 
 struct btree_debug {
@@ -1068,20 +1073,10 @@ static inline void bch2_write_ref_get(struct bch_fs *c, enum bch_write_ref ref)
 #endif
 }
 
-static inline bool __bch2_write_ref_tryget(struct bch_fs *c, enum bch_write_ref ref)
-{
-#ifdef BCH_WRITE_REF_DEBUG
-	return !test_bit(BCH_FS_GOING_RO, &c->flags) &&
-		atomic_long_inc_not_zero(&c->writes[ref]);
-#else
-	return percpu_ref_tryget(&c->writes);
-#endif
-}
-
 static inline bool bch2_write_ref_tryget(struct bch_fs *c, enum bch_write_ref ref)
 {
 #ifdef BCH_WRITE_REF_DEBUG
-	return !test_bit(BCH_FS_GOING_RO, &c->flags) &&
+	return !test_bit(BCH_FS_going_ro, &c->flags) &&
 		atomic_long_inc_not_zero(&c->writes[ref]);
 #else
 	return percpu_ref_tryget_live(&c->writes);
@@ -1100,7 +1095,7 @@ static inline void bch2_write_ref_put(struct bch_fs *c, enum bch_write_ref ref)
 		if (atomic_long_read(&c->writes[i]))
 			return;
 
-	set_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags);
+	set_bit(BCH_FS_write_disable_complete, &c->flags);
 	wake_up(&bch2_read_only_wait);
 #else
 	percpu_ref_put(&c->writes);
diff --git a/libbcachefs/bcachefs_format.h b/libbcachefs/bcachefs_format.h
index 96778007..be0367f3 100644
--- a/libbcachefs/bcachefs_format.h
+++ b/libbcachefs/bcachefs_format.h
@@ -1539,7 +1539,7 @@ struct bch_sb_field_disk_groups {
 	x(move_extent_write,				36)	\
 	x(move_extent_finish,				37)	\
 	x(move_extent_fail,				38)	\
-	x(move_extent_alloc_mem_fail,			39)	\
+	x(move_extent_start_fail,			39)	\
 	x(copygc,					40)	\
 	x(copygc_wait,					41)	\
 	x(gc_gens_end,					42)	\
@@ -1576,7 +1576,9 @@ struct bch_sb_field_disk_groups {
 	x(write_super,					73)	\
 	x(trans_restart_would_deadlock_recursion_limit,	74)	\
 	x(trans_restart_write_buffer_flush,		75)	\
-	x(trans_restart_split_race,			76)
+	x(trans_restart_split_race,			76)	\
+	x(write_buffer_flush_slowpath,			77)	\
+	x(write_buffer_flush_sync,			78)
 
 enum bch_persistent_counters {
 #define x(t, n, ...) BCH_COUNTER_##t,
@@ -2135,8 +2137,7 @@ static inline __u64 __bset_magic(struct bch_sb *sb)
 	x(clock,		7)		\
 	x(dev_usage,		8)		\
 	x(log,			9)		\
-	x(overwrite,		10)		\
-	x(write_buffer_keys,	11)
+	x(overwrite,		10)
 
 enum {
 #define x(f, nr)	BCH_JSET_ENTRY_##f	= nr,
diff --git a/libbcachefs/btree_cache.c b/libbcachefs/btree_cache.c
index 72dea90e..47e7770d 100644
--- a/libbcachefs/btree_cache.c
+++ b/libbcachefs/btree_cache.c
@@ -13,13 +13,6 @@
 
 #include <linux/prefetch.h>
 #include <linux/sched/mm.h>
-#include <linux/seq_buf.h>
-
-#define BTREE_CACHE_NOT_FREED_INCREMENT(counter) \
-do {						 \
-	if (shrinker_counter)			 \
-		bc->not_freed_##counter++;	 \
-} while (0)
 
 const char * const bch2_btree_node_flags[] = {
 #define x(f)	#f,
@@ -208,7 +201,7 @@ static inline struct btree *btree_cache_find(struct btree_cache *bc,
  * this version is for btree nodes that have already been freed (we're not
  * reaping a real btree node)
  */
-static int __btree_node_reclaim(struct bch_fs *c, struct btree *b, bool flush, bool shrinker_counter)
+static int __btree_node_reclaim(struct bch_fs *c, struct btree *b, bool flush)
 {
 	struct btree_cache *bc = &c->btree_cache;
 	int ret = 0;
@@ -218,64 +211,38 @@ wait_on_io:
 	if (b->flags & ((1U << BTREE_NODE_dirty)|
 			(1U << BTREE_NODE_read_in_flight)|
 			(1U << BTREE_NODE_write_in_flight))) {
-		if (!flush) {
-			if (btree_node_dirty(b))
-				BTREE_CACHE_NOT_FREED_INCREMENT(dirty);
-			else if (btree_node_read_in_flight(b))
-				BTREE_CACHE_NOT_FREED_INCREMENT(read_in_flight);
-			else if (btree_node_write_in_flight(b))
-				BTREE_CACHE_NOT_FREED_INCREMENT(write_in_flight);
+		if (!flush)
 			return -BCH_ERR_ENOMEM_btree_node_reclaim;
-		}
 
 		/* XXX: waiting on IO with btree cache lock held */
 		bch2_btree_node_wait_on_read(b);
 		bch2_btree_node_wait_on_write(b);
 	}
 
-	if (!six_trylock_intent(&b->c.lock)) {
-		BTREE_CACHE_NOT_FREED_INCREMENT(lock_intent);
+	if (!six_trylock_intent(&b->c.lock))
 		return -BCH_ERR_ENOMEM_btree_node_reclaim;
-	}
 
-	if (!six_trylock_write(&b->c.lock)) {
-		BTREE_CACHE_NOT_FREED_INCREMENT(lock_write);
+	if (!six_trylock_write(&b->c.lock))
 		goto out_unlock_intent;
-	}
 
 	/* recheck under lock */
 	if (b->flags & ((1U << BTREE_NODE_read_in_flight)|
 			(1U << BTREE_NODE_write_in_flight))) {
-		if (!flush) {
-			if (btree_node_read_in_flight(b))
-				BTREE_CACHE_NOT_FREED_INCREMENT(read_in_flight);
-			else if (btree_node_write_in_flight(b))
-				BTREE_CACHE_NOT_FREED_INCREMENT(write_in_flight);
+		if (!flush)
 			goto out_unlock;
-		}
 		six_unlock_write(&b->c.lock);
 		six_unlock_intent(&b->c.lock);
 		goto wait_on_io;
 	}
 
-	if (btree_node_noevict(b)) {
-		BTREE_CACHE_NOT_FREED_INCREMENT(noevict);
-		goto out_unlock;
-	}
-	if (btree_node_write_blocked(b)) {
-		BTREE_CACHE_NOT_FREED_INCREMENT(write_blocked);
-		goto out_unlock;
-	}
-	if (btree_node_will_make_reachable(b)) {
-		BTREE_CACHE_NOT_FREED_INCREMENT(will_make_reachable);
+	if (btree_node_noevict(b) ||
+	    btree_node_write_blocked(b) ||
+	    btree_node_will_make_reachable(b))
 		goto out_unlock;
-	}
 
 	if (btree_node_dirty(b)) {
-		if (!flush) {
-			BTREE_CACHE_NOT_FREED_INCREMENT(dirty);
+		if (!flush)
 			goto out_unlock;
-		}
 		/*
 		 * Using the underscore version because we don't want to compact
 		 * bsets after the write, since this node is about to be evicted
@@ -305,14 +272,14 @@ out_unlock_intent:
 	goto out;
 }
 
-static int btree_node_reclaim(struct bch_fs *c, struct btree *b, bool shrinker_counter)
+static int btree_node_reclaim(struct bch_fs *c, struct btree *b)
 {
-	return __btree_node_reclaim(c, b, false, shrinker_counter);
+	return __btree_node_reclaim(c, b, false);
 }
 
 static int btree_node_write_and_reclaim(struct bch_fs *c, struct btree *b)
 {
-	return __btree_node_reclaim(c, b, true, false);
+	return __btree_node_reclaim(c, b, true);
 }
 
 static unsigned long bch2_btree_cache_scan(struct shrinker *shrink,
@@ -360,12 +327,11 @@ static unsigned long bch2_btree_cache_scan(struct shrinker *shrink,
 		if (touched >= nr)
 			goto out;
 
-		if (!btree_node_reclaim(c, b, true)) {
+		if (!btree_node_reclaim(c, b)) {
 			btree_node_data_free(c, b);
 			six_unlock_write(&b->c.lock);
 			six_unlock_intent(&b->c.lock);
 			freed++;
-			bc->freed++;
 		}
 	}
 restart:
@@ -374,11 +340,9 @@ restart:
 
 		if (btree_node_accessed(b)) {
 			clear_btree_node_accessed(b);
-			bc->not_freed_access_bit++;
-		} else if (!btree_node_reclaim(c, b, true)) {
+		} else if (!btree_node_reclaim(c, b)) {
 			freed++;
 			btree_node_data_free(c, b);
-			bc->freed++;
 
 			bch2_btree_node_hash_remove(bc, b);
 			six_unlock_write(&b->c.lock);
@@ -428,17 +392,6 @@ static unsigned long bch2_btree_cache_count(struct shrinker *shrink,
 	return btree_cache_can_free(bc);
 }
 
-static void bch2_btree_cache_shrinker_to_text(struct seq_buf *s, struct shrinker *shrink)
-{
-	struct bch_fs *c = shrink->private_data;
-	char *cbuf;
-	size_t buflen = seq_buf_get_buf(s, &cbuf);
-	struct printbuf out = PRINTBUF_EXTERN(cbuf, buflen);
-
-	bch2_btree_cache_to_text(&out, &c->btree_cache);
-	seq_buf_commit(s, out.pos);
-}
-
 void bch2_fs_btree_cache_exit(struct bch_fs *c)
 {
 	struct btree_cache *bc = &c->btree_cache;
@@ -525,7 +478,6 @@ int bch2_fs_btree_cache_init(struct bch_fs *c)
 	bc->shrink = shrink;
 	shrink->count_objects	= bch2_btree_cache_count;
 	shrink->scan_objects	= bch2_btree_cache_scan;
-	shrink->to_text		= bch2_btree_cache_shrinker_to_text;
 	shrink->seeks		= 4;
 	shrink->private_data	= c;
 	shrinker_register(shrink);
@@ -599,7 +551,7 @@ static struct btree *btree_node_cannibalize(struct bch_fs *c)
 	struct btree *b;
 
 	list_for_each_entry_reverse(b, &bc->live, list)
-		if (!btree_node_reclaim(c, b, false))
+		if (!btree_node_reclaim(c, b))
 			return b;
 
 	while (1) {
@@ -635,7 +587,7 @@ struct btree *bch2_btree_node_mem_alloc(struct btree_trans *trans, bool pcpu_rea
 	 * disk node. Check the freed list before allocating a new one:
 	 */
 	list_for_each_entry(b, freed, list)
-		if (!btree_node_reclaim(c, b, false)) {
+		if (!btree_node_reclaim(c, b)) {
 			list_del_init(&b->list);
 			goto got_node;
 		}
@@ -661,7 +613,7 @@ got_node:
 	 * the list. Check if there's any freed nodes there:
 	 */
 	list_for_each_entry(b2, &bc->freeable, list)
-		if (!btree_node_reclaim(c, b2, false)) {
+		if (!btree_node_reclaim(c, b2)) {
 			swap(b->data, b2->data);
 			swap(b->aux_data, b2->aux_data);
 			btree_node_to_freedlist(bc, b2);
@@ -1257,21 +1209,9 @@ void bch2_btree_node_to_text(struct printbuf *out, struct bch_fs *c, const struc
 	       stats.failed);
 }
 
-void bch2_btree_cache_to_text(struct printbuf *out, const struct btree_cache *bc)
+void bch2_btree_cache_to_text(struct printbuf *out, const struct bch_fs *c)
 {
-	prt_printf(out, "nr nodes:\t\t%u\n", bc->used);
-	prt_printf(out, "nr dirty:\t\t%u\n", atomic_read(&bc->dirty));
-	prt_printf(out, "cannibalize lock:\t%p\n", bc->alloc_lock);
-
-	prt_printf(out, "freed:\t\t\t\t%u\n", bc->freed);
-	prt_printf(out, "not freed, dirty:\t\t%u\n", bc->not_freed_dirty);
-	prt_printf(out, "not freed, write in flight:\t%u\n", bc->not_freed_write_in_flight);
-	prt_printf(out, "not freed, read in flight:\t%u\n", bc->not_freed_read_in_flight);
-	prt_printf(out, "not freed, lock intent failed:\t%u\n", bc->not_freed_lock_intent);
-	prt_printf(out, "not freed, lock write failed:\t%u\n", bc->not_freed_lock_write);
-	prt_printf(out, "not freed, access bit:\t\t%u\n", bc->not_freed_access_bit);
-	prt_printf(out, "not freed, no evict failed:\t%u\n", bc->not_freed_noevict);
-	prt_printf(out, "not freed, write blocked:\t%u\n", bc->not_freed_write_blocked);
-	prt_printf(out, "not freed, will make reachable:\t%u\n", bc->not_freed_will_make_reachable);
-
+	prt_printf(out, "nr nodes:\t\t%u\n", c->btree_cache.used);
+	prt_printf(out, "nr dirty:\t\t%u\n", atomic_read(&c->btree_cache.dirty));
+	prt_printf(out, "cannibalize lock:\t%p\n", c->btree_cache.alloc_lock);
 }
diff --git a/libbcachefs/btree_cache.h b/libbcachefs/btree_cache.h
index bfe1d748..cfb80b20 100644
--- a/libbcachefs/btree_cache.h
+++ b/libbcachefs/btree_cache.h
@@ -126,6 +126,6 @@ static inline struct btree *btree_node_root(struct bch_fs *c, struct btree *b)
 const char *bch2_btree_id_str(enum btree_id);
 void bch2_btree_pos_to_text(struct printbuf *, struct bch_fs *, const struct btree *);
 void bch2_btree_node_to_text(struct printbuf *, struct bch_fs *, const struct btree *);
-void bch2_btree_cache_to_text(struct printbuf *, const struct btree_cache *);
+void bch2_btree_cache_to_text(struct printbuf *, const struct bch_fs *);
 
 #endif /* _BCACHEFS_BTREE_CACHE_H */
diff --git a/libbcachefs/btree_gc.c b/libbcachefs/btree_gc.c
index 90f5bcfa..70e47880 100644
--- a/libbcachefs/btree_gc.c
+++ b/libbcachefs/btree_gc.c
@@ -108,7 +108,7 @@ static int bch2_gc_check_topology(struct bch_fs *c,
 				ret = bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_topology);
 				goto err;
 			} else {
-				set_bit(BCH_FS_INITIAL_GC_UNFIXED, &c->flags);
+				set_bit(BCH_FS_initial_gc_unfixed, &c->flags);
 			}
 		}
 	}
@@ -134,7 +134,7 @@ static int bch2_gc_check_topology(struct bch_fs *c,
 			ret = bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_topology);
 			goto err;
 		} else {
-			set_bit(BCH_FS_INITIAL_GC_UNFIXED, &c->flags);
+			set_bit(BCH_FS_initial_gc_unfixed, &c->flags);
 		}
 	}
 
@@ -619,7 +619,7 @@ static int bch2_check_fix_ptrs(struct btree_trans *trans, enum btree_id btree_id
 				g->data_type		= 0;
 				g->dirty_sectors	= 0;
 				g->cached_sectors	= 0;
-				set_bit(BCH_FS_NEED_ANOTHER_GC, &c->flags);
+				set_bit(BCH_FS_need_another_gc, &c->flags);
 			} else {
 				do_update = true;
 			}
@@ -664,7 +664,7 @@ static int bch2_check_fix_ptrs(struct btree_trans *trans, enum btree_id btree_id
 				 bch2_bkey_val_to_text(&buf, c, *k), buf.buf))) {
 			if (data_type == BCH_DATA_btree) {
 				g->data_type	= data_type;
-				set_bit(BCH_FS_NEED_ANOTHER_GC, &c->flags);
+				set_bit(BCH_FS_need_another_gc, &c->flags);
 			} else {
 				do_update = true;
 			}
@@ -996,7 +996,7 @@ static int bch2_gc_btree_init_recurse(struct btree_trans *trans, struct btree *b
 					/* Continue marking when opted to not
 					 * fix the error: */
 					ret = 0;
-					set_bit(BCH_FS_INITIAL_GC_UNFIXED, &c->flags);
+					set_bit(BCH_FS_initial_gc_unfixed, &c->flags);
 					continue;
 				}
 			} else if (ret) {
@@ -1847,7 +1847,7 @@ again:
 #endif
 	c->gc_count++;
 
-	if (test_bit(BCH_FS_NEED_ANOTHER_GC, &c->flags) ||
+	if (test_bit(BCH_FS_need_another_gc, &c->flags) ||
 	    (!iter && bch2_test_restart_gc)) {
 		if (iter++ > 2) {
 			bch_info(c, "Unable to fix bucket gens, looping");
@@ -1859,7 +1859,7 @@ again:
 		 * XXX: make sure gens we fixed got saved
 		 */
 		bch_info(c, "Second GC pass needed, restarting:");
-		clear_bit(BCH_FS_NEED_ANOTHER_GC, &c->flags);
+		clear_bit(BCH_FS_need_another_gc, &c->flags);
 		__gc_pos_set(c, gc_phase(GC_PHASE_NOT_RUNNING));
 
 		bch2_gc_stripes_reset(c, metadata_only);
diff --git a/libbcachefs/btree_iter.c b/libbcachefs/btree_iter.c
index a52fd206..bdc80808 100644
--- a/libbcachefs/btree_iter.c
+++ b/libbcachefs/btree_iter.c
@@ -781,7 +781,7 @@ static int btree_path_prefetch(struct btree_trans *trans, struct btree_path *pat
 	struct btree_node_iter node_iter = l->iter;
 	struct bkey_packed *k;
 	struct bkey_buf tmp;
-	unsigned nr = test_bit(BCH_FS_STARTED, &c->flags)
+	unsigned nr = test_bit(BCH_FS_started, &c->flags)
 		? (path->level > 1 ? 0 :  2)
 		: (path->level > 1 ? 1 : 16);
 	bool was_locked = btree_node_locked(path, path->level);
@@ -816,7 +816,7 @@ static int btree_path_prefetch_j(struct btree_trans *trans, struct btree_path *p
 	struct bch_fs *c = trans->c;
 	struct bkey_s_c k;
 	struct bkey_buf tmp;
-	unsigned nr = test_bit(BCH_FS_STARTED, &c->flags)
+	unsigned nr = test_bit(BCH_FS_started, &c->flags)
 		? (path->level > 1 ? 0 :  2)
 		: (path->level > 1 ? 1 : 16);
 	bool was_locked = btree_node_locked(path, path->level);
diff --git a/libbcachefs/btree_journal_iter.c b/libbcachefs/btree_journal_iter.c
index 7a5e0a89..4c084ce4 100644
--- a/libbcachefs/btree_journal_iter.c
+++ b/libbcachefs/btree_journal_iter.c
@@ -177,7 +177,7 @@ int bch2_journal_key_insert_take(struct bch_fs *c, enum btree_id id,
 	struct journal_keys *keys = &c->journal_keys;
 	size_t idx = bch2_journal_key_search(keys, id, level, k->k.p);
 
-	BUG_ON(test_bit(BCH_FS_RW, &c->flags));
+	BUG_ON(test_bit(BCH_FS_rw, &c->flags));
 
 	if (idx < keys->size &&
 	    journal_key_cmp(&n, &keys->d[idx]) == 0) {
diff --git a/libbcachefs/btree_key_cache.c b/libbcachefs/btree_key_cache.c
index e14e9b4c..c64f8db0 100644
--- a/libbcachefs/btree_key_cache.c
+++ b/libbcachefs/btree_key_cache.c
@@ -13,7 +13,6 @@
 #include "trace.h"
 
 #include <linux/sched/mm.h>
-#include <linux/seq_buf.h>
 
 static inline bool btree_uses_pcpu_readers(enum btree_id id)
 {
@@ -779,7 +778,7 @@ bool bch2_btree_insert_key_cached(struct btree_trans *trans,
 	ck->valid = true;
 
 	if (!test_bit(BKEY_CACHED_DIRTY, &ck->flags)) {
-		EBUG_ON(test_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags));
+		EBUG_ON(test_bit(BCH_FS_clean_shutdown, &c->flags));
 		set_bit(BKEY_CACHED_DIRTY, &ck->flags);
 		atomic_long_inc(&c->btree_key_cache.nr_dirty);
 
@@ -1008,7 +1007,7 @@ void bch2_fs_btree_key_cache_exit(struct btree_key_cache *bc)
 
 	if (atomic_long_read(&bc->nr_dirty) &&
 	    !bch2_journal_error(&c->journal) &&
-	    test_bit(BCH_FS_WAS_RW, &c->flags))
+	    test_bit(BCH_FS_was_rw, &c->flags))
 		panic("btree key cache shutdown error: nr_dirty nonzero (%li)\n",
 		      atomic_long_read(&bc->nr_dirty));
 
@@ -1029,18 +1028,6 @@ void bch2_fs_btree_key_cache_init_early(struct btree_key_cache *c)
 	INIT_LIST_HEAD(&c->freed_nonpcpu);
 }
 
-static void bch2_btree_key_cache_shrinker_to_text(struct seq_buf *s, struct shrinker *shrink)
-{
-	struct bch_fs *c = shrink->private_data;
-	struct btree_key_cache *bc = &c->btree_key_cache;
-	char *cbuf;
-	size_t buflen = seq_buf_get_buf(s, &cbuf);
-	struct printbuf out = PRINTBUF_EXTERN(cbuf, buflen);
-
-	bch2_btree_key_cache_to_text(&out, bc);
-	seq_buf_commit(s, out.pos);
-}
-
 int bch2_fs_btree_key_cache_init(struct btree_key_cache *bc)
 {
 	struct bch_fs *c = container_of(bc, struct bch_fs, btree_key_cache);
@@ -1064,7 +1051,6 @@ int bch2_fs_btree_key_cache_init(struct btree_key_cache *bc)
 	shrink->seeks		= 0;
 	shrink->count_objects	= bch2_btree_key_cache_count;
 	shrink->scan_objects	= bch2_btree_key_cache_scan;
-	shrink->to_text		= bch2_btree_key_cache_shrinker_to_text;
 	shrink->private_data	= c;
 	shrinker_register(shrink);
 	return 0;
diff --git a/libbcachefs/btree_trans_commit.c b/libbcachefs/btree_trans_commit.c
index 7210d5c2..336350bd 100644
--- a/libbcachefs/btree_trans_commit.c
+++ b/libbcachefs/btree_trans_commit.c
@@ -287,7 +287,7 @@ inline void bch2_btree_insert_key_leaf(struct btree_trans *trans,
 	bch2_btree_add_journal_pin(c, b, journal_seq);
 
 	if (unlikely(!btree_node_dirty(b))) {
-		EBUG_ON(test_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags));
+		EBUG_ON(test_bit(BCH_FS_clean_shutdown, &c->flags));
 		set_btree_node_dirty_acct(c, b);
 	}
 
@@ -659,6 +659,10 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 		i->k->k.needs_whiteout = false;
 	}
 
+	if (trans->nr_wb_updates &&
+	    trans->nr_wb_updates + c->btree_write_buffer.state.nr > c->btree_write_buffer.size)
+		return -BCH_ERR_btree_insert_need_flush_buffer;
+
 	/*
 	 * Don't get journal reservation until after we know insert will
 	 * succeed:
@@ -693,6 +697,14 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 	    bch2_trans_fs_usage_apply(trans, trans->fs_usage_deltas))
 		return -BCH_ERR_btree_insert_need_mark_replicas;
 
+	if (trans->nr_wb_updates) {
+		EBUG_ON(flags & BCH_TRANS_COMMIT_no_journal_res);
+
+		ret = bch2_btree_insert_keys_write_buffer(trans);
+		if (ret)
+			goto revert_fs_usage;
+	}
+
 	h = trans->hooks;
 	while (h) {
 		ret = h->fn(trans, h);
@@ -754,7 +766,7 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 
 		trans_for_each_wb_update(trans, wb) {
 			entry = bch2_journal_add_entry(j, &trans->journal_res,
-					       BCH_JSET_ENTRY_write_buffer_keys,
+					       BCH_JSET_ENTRY_btree_keys,
 					       wb->btree, 0,
 					       wb->k.k.u64s);
 			bkey_copy((struct bkey_i *) entry->start, &wb->k);
@@ -938,6 +950,30 @@ int bch2_trans_commit_error(struct btree_trans *trans, unsigned flags,
 
 		ret = bch2_trans_relock(trans);
 		break;
+	case -BCH_ERR_btree_insert_need_flush_buffer: {
+		struct btree_write_buffer *wb = &c->btree_write_buffer;
+
+		ret = 0;
+
+		if (wb->state.nr > wb->size * 3 / 4) {
+			bch2_trans_unlock(trans);
+			mutex_lock(&wb->flush_lock);
+
+			if (wb->state.nr > wb->size * 3 / 4) {
+				bch2_trans_begin(trans);
+				ret = bch2_btree_write_buffer_flush_locked(trans);
+				mutex_unlock(&wb->flush_lock);
+				if (!ret) {
+					trace_and_count(c, trans_restart_write_buffer_flush, trans, _THIS_IP_);
+					ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_write_buffer_flush);
+				}
+			} else {
+				mutex_unlock(&wb->flush_lock);
+				ret = bch2_trans_relock(trans);
+			}
+		}
+		break;
+	}
 	default:
 		BUG_ON(ret >= 0);
 		break;
@@ -959,7 +995,7 @@ bch2_trans_commit_get_rw_cold(struct btree_trans *trans, unsigned flags)
 	int ret;
 
 	if (likely(!(flags & BCH_TRANS_COMMIT_lazy_rw)) ||
-	    test_bit(BCH_FS_STARTED, &c->flags))
+	    test_bit(BCH_FS_started, &c->flags))
 		return -BCH_ERR_erofs_trans_commit;
 
 	ret = drop_locks_do(trans, bch2_fs_read_write_early(c));
@@ -1024,7 +1060,7 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags)
 			return ret;
 	}
 
-	if (unlikely(!test_bit(BCH_FS_MAY_GO_RW, &c->flags))) {
+	if (unlikely(!test_bit(BCH_FS_may_go_rw, &c->flags))) {
 		ret = do_bch2_trans_commit_to_journal_replay(trans);
 		goto out_reset;
 	}
@@ -1036,7 +1072,21 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags)
 			goto out_reset;
 	}
 
-	EBUG_ON(test_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags));
+	if (c->btree_write_buffer.state.nr > c->btree_write_buffer.size / 2 &&
+	    mutex_trylock(&c->btree_write_buffer.flush_lock)) {
+		bch2_trans_begin(trans);
+		bch2_trans_unlock(trans);
+
+		ret = bch2_btree_write_buffer_flush_locked(trans);
+		mutex_unlock(&c->btree_write_buffer.flush_lock);
+		if (!ret) {
+			trace_and_count(c, trans_restart_write_buffer_flush, trans, _THIS_IP_);
+			ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_write_buffer_flush);
+		}
+		goto out;
+	}
+
+	EBUG_ON(test_bit(BCH_FS_clean_shutdown, &c->flags));
 
 	trans->journal_u64s		= trans->extra_journal_entries.nr;
 	trans->journal_transaction_names = READ_ONCE(c->opts.journal_transaction_names);
diff --git a/libbcachefs/btree_types.h b/libbcachefs/btree_types.h
index 14983e77..2326bceb 100644
--- a/libbcachefs/btree_types.h
+++ b/libbcachefs/btree_types.h
@@ -162,16 +162,6 @@ struct btree_cache {
 	/* Number of elements in live + freeable lists */
 	unsigned		used;
 	unsigned		reserve;
-	unsigned		freed;
-	unsigned		not_freed_lock_intent;
-	unsigned		not_freed_lock_write;
-	unsigned		not_freed_dirty;
-	unsigned		not_freed_read_in_flight;
-	unsigned		not_freed_write_in_flight;
-	unsigned		not_freed_noevict;
-	unsigned		not_freed_write_blocked;
-	unsigned		not_freed_will_make_reachable;
-	unsigned		not_freed_access_bit;
 	atomic_t		dirty;
 	struct shrinker		*shrink;
 
diff --git a/libbcachefs/btree_update.c b/libbcachefs/btree_update.c
index 1837f848..ba42f578 100644
--- a/libbcachefs/btree_update.c
+++ b/libbcachefs/btree_update.c
@@ -693,20 +693,6 @@ int bch2_btree_delete_at(struct btree_trans *trans,
 	return bch2_btree_delete_extent_at(trans, iter, 0, update_flags);
 }
 
-int bch2_btree_delete_at_buffered(struct btree_trans *trans,
-				  enum btree_id btree, struct bpos pos)
-{
-	struct bkey_i *k;
-
-	k = bch2_trans_kmalloc(trans, sizeof(*k));
-	if (IS_ERR(k))
-		return PTR_ERR(k);
-
-	bkey_init(&k->k);
-	k->k.p = pos;
-	return bch2_trans_update_buffered(trans, btree, k);
-}
-
 int bch2_btree_delete(struct btree_trans *trans,
 		      enum btree_id btree, struct bpos pos,
 		      unsigned update_flags)
@@ -811,19 +797,13 @@ int bch2_btree_delete_range(struct bch_fs *c, enum btree_id id,
 int bch2_btree_bit_mod(struct btree_trans *trans, enum btree_id btree,
 		       struct bpos pos, bool set)
 {
-	struct bkey_i *k;
-	int ret = 0;
+	struct bkey_i k;
 
-	k = bch2_trans_kmalloc_nomemzero(trans, sizeof(*k));
-	ret = PTR_ERR_OR_ZERO(k);
-	if (unlikely(ret))
-		return ret;
-
-	bkey_init(&k->k);
-	k->k.type = set ? KEY_TYPE_set : KEY_TYPE_deleted;
-	k->k.p = pos;
+	bkey_init(&k.k);
+	k.k.type = set ? KEY_TYPE_set : KEY_TYPE_deleted;
+	k.k.p = pos;
 
-	return bch2_trans_update_buffered(trans, btree, k);
+	return bch2_trans_update_buffered(trans, btree, &k);
 }
 
 __printf(2, 0)
diff --git a/libbcachefs/btree_update.h b/libbcachefs/btree_update.h
index 14a2315a..fa19f321 100644
--- a/libbcachefs/btree_update.h
+++ b/libbcachefs/btree_update.h
@@ -47,7 +47,6 @@ enum bch_trans_commit_flags {
 int bch2_btree_delete_extent_at(struct btree_trans *, struct btree_iter *,
 				unsigned, unsigned);
 int bch2_btree_delete_at(struct btree_trans *, struct btree_iter *, unsigned);
-int bch2_btree_delete_at_buffered(struct btree_trans *, enum btree_id, struct bpos);
 int bch2_btree_delete(struct btree_trans *, enum btree_id, struct bpos, unsigned);
 
 int bch2_btree_insert_nonextent(struct btree_trans *, enum btree_id,
@@ -65,6 +64,12 @@ int bch2_btree_delete_range(struct bch_fs *, enum btree_id,
 
 int bch2_btree_bit_mod(struct btree_trans *, enum btree_id, struct bpos, bool);
 
+static inline int bch2_btree_delete_at_buffered(struct btree_trans *trans,
+						enum btree_id btree, struct bpos pos)
+{
+	return bch2_btree_bit_mod(trans, btree, pos, false);
+}
+
 int __bch2_insert_snapshot_whiteouts(struct btree_trans *, enum btree_id,
 				     struct bpos, struct bpos);
 
diff --git a/libbcachefs/btree_update_interior.c b/libbcachefs/btree_update_interior.c
index bfe4d797..68627061 100644
--- a/libbcachefs/btree_update_interior.c
+++ b/libbcachefs/btree_update_interior.c
@@ -1082,8 +1082,12 @@ bch2_btree_update_start(struct btree_trans *trans, struct btree_path *path,
 			break;
 		}
 
+		/*
+		 * Always check for space for two keys, even if we won't have to
+		 * split at prior level - it might have been a merge instead:
+		 */
 		if (bch2_btree_node_insert_fits(c, path->l[update_level].b,
-					BKEY_BTREE_PTR_U64s_MAX * (1 + split)))
+						BKEY_BTREE_PTR_U64s_MAX * 2))
 			break;
 
 		split = path->l[update_level].b->nr.live_u64s > BTREE_SPLIT_THRESHOLD(c);
@@ -2052,7 +2056,7 @@ void bch2_btree_node_rewrite_async(struct bch_fs *c, struct btree *b)
 	a->seq		= b->data->keys.seq;
 	INIT_WORK(&a->work, async_btree_node_rewrite_work);
 
-	if (unlikely(!test_bit(BCH_FS_MAY_GO_RW, &c->flags))) {
+	if (unlikely(!test_bit(BCH_FS_may_go_rw, &c->flags))) {
 		mutex_lock(&c->pending_node_rewrites_lock);
 		list_add(&a->list, &c->pending_node_rewrites);
 		mutex_unlock(&c->pending_node_rewrites_lock);
@@ -2060,7 +2064,7 @@ void bch2_btree_node_rewrite_async(struct bch_fs *c, struct btree *b)
 	}
 
 	if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_node_rewrite)) {
-		if (test_bit(BCH_FS_STARTED, &c->flags)) {
+		if (test_bit(BCH_FS_started, &c->flags)) {
 			bch_err(c, "%s: error getting c->writes ref", __func__);
 			kfree(a);
 			return;
diff --git a/libbcachefs/btree_write_buffer.c b/libbcachefs/btree_write_buffer.c
index d3c38d2c..6a191568 100644
--- a/libbcachefs/btree_write_buffer.c
+++ b/libbcachefs/btree_write_buffer.c
@@ -7,130 +7,43 @@
 #include "btree_write_buffer.h"
 #include "error.h"
 #include "journal.h"
-#include "journal_io.h"
 #include "journal_reclaim.h"
 
-#include <linux/prefetch.h>
+#include <linux/sort.h>
 
 static int bch2_btree_write_buffer_journal_flush(struct journal *,
 				struct journal_entry_pin *, u64);
 
-static int bch2_journal_keys_to_write_buffer(struct bch_fs *, struct journal_buf *);
-
-static inline bool __wb_key_cmp(const struct wb_key_ref *l, const struct wb_key_ref *r)
-{
-	return (cmp_int(l->hi, r->hi) ?:
-		cmp_int(l->mi, r->mi) ?:
-		cmp_int(l->lo, r->lo)) >= 0;
-}
-
-static inline bool wb_key_cmp(const struct wb_key_ref *l, const struct wb_key_ref *r)
-{
-#ifdef CONFIG_X86_64
-	int cmp;
-
-	asm("mov   (%[l]), %%rax;"
-	    "sub   (%[r]), %%rax;"
-	    "mov  8(%[l]), %%rax;"
-	    "sbb  8(%[r]), %%rax;"
-	    "mov 16(%[l]), %%rax;"
-	    "sbb 16(%[r]), %%rax;"
-	    : "=@ccae" (cmp)
-	    : [l] "r" (l), [r] "r" (r)
-	    : "rax", "cc");
-
-	EBUG_ON(cmp != __wb_key_cmp(l, r));
-	return cmp;
-#else
-	return __wb_key_cmp(l, r);
-#endif
-}
-
-/* Compare excluding idx, the low 24 bits: */
-static inline bool wb_key_eq(const void *_l, const void *_r)
+static int btree_write_buffered_key_cmp(const void *_l, const void *_r)
 {
-	const struct wb_key_ref *l = _l;
-	const struct wb_key_ref *r = _r;
+	const struct btree_write_buffered_key *l = _l;
+	const struct btree_write_buffered_key *r = _r;
 
-	return !((l->hi ^ r->hi)|
-		 (l->mi ^ r->mi)|
-		 ((l->lo >> 24) ^ (r->lo >> 24)));
+	return  cmp_int(l->btree, r->btree) ?:
+		bpos_cmp(l->k.k.p, r->k.k.p) ?:
+		cmp_int(l->journal_seq, r->journal_seq) ?:
+		cmp_int(l->journal_offset, r->journal_offset);
 }
 
-static noinline void wb_sort(struct wb_key_ref *base, size_t num)
+static int btree_write_buffered_journal_cmp(const void *_l, const void *_r)
 {
-	size_t n = num, a = num / 2;
-
-	if (!a)		/* num < 2 || size == 0 */
-		return;
-
-	for (;;) {
-		size_t b, c, d;
-
-		if (a)			/* Building heap: sift down --a */
-			--a;
-		else if (--n)		/* Sorting: Extract root to --n */
-			swap(base[0], base[n]);
-		else			/* Sort complete */
-			break;
-
-		/*
-		 * Sift element at "a" down into heap.  This is the
-		 * "bottom-up" variant, which significantly reduces
-		 * calls to cmp_func(): we find the sift-down path all
-		 * the way to the leaves (one compare per level), then
-		 * backtrack to find where to insert the target element.
-		 *
-		 * Because elements tend to sift down close to the leaves,
-		 * this uses fewer compares than doing two per level
-		 * on the way down.  (A bit more than half as many on
-		 * average, 3/4 worst-case.)
-		 */
-		for (b = a; c = 2*b + 1, (d = c + 1) < n;)
-			b = wb_key_cmp(base + c, base + d) ? c : d;
-		if (d == n)		/* Special case last leaf with no sibling */
-			b = c;
-
-		/* Now backtrack from "b" to the correct location for "a" */
-		while (b != a && wb_key_cmp(base + a, base + b))
-			b = (b - 1) / 2;
-		c = b;			/* Where "a" belongs */
-		while (b != a) {	/* Shift it into place */
-			b = (b - 1) / 2;
-			swap(base[b], base[c]);
-		}
-	}
-}
-
-static noinline int wb_flush_one_slowpath(struct btree_trans *trans,
-					  struct btree_iter *iter,
-					  struct btree_write_buffered_key *wb)
-{
-	bch2_btree_node_unlock_write(trans, iter->path, iter->path->l[0].b);
-
-	trans->journal_res.seq = wb->journal_seq;
+	const struct btree_write_buffered_key *l = _l;
+	const struct btree_write_buffered_key *r = _r;
 
-	return bch2_trans_update(trans, iter, &wb->k,
-				 BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?:
-		bch2_trans_commit(trans, NULL, NULL,
-				  BCH_TRANS_COMMIT_no_enospc|
-				  BCH_TRANS_COMMIT_no_check_rw|
-				  BCH_TRANS_COMMIT_no_journal_res|
-				  BCH_TRANS_COMMIT_journal_reclaim);
+	return  cmp_int(l->journal_seq, r->journal_seq);
 }
 
-static inline int wb_flush_one(struct btree_trans *trans, struct btree_iter *iter,
-			       struct btree_write_buffered_key *wb,
-			       bool *write_locked, size_t *fast)
+static int bch2_btree_write_buffer_flush_one(struct btree_trans *trans,
+					     struct btree_iter *iter,
+					     struct btree_write_buffered_key *wb,
+					     unsigned commit_flags,
+					     bool *write_locked,
+					     size_t *fast)
 {
 	struct bch_fs *c = trans->c;
 	struct btree_path *path;
 	int ret;
 
-	EBUG_ON(!wb->journal_seq);
-	EBUG_ON(!c->btree_write_buffer.flushing.pin.seq);
-	EBUG_ON(c->btree_write_buffer.flushing.pin.seq > wb->journal_seq);
-
 	ret = bch2_btree_iter_traverse(iter);
 	if (ret)
 		return ret;
@@ -153,14 +66,46 @@ static inline int wb_flush_one(struct btree_trans *trans, struct btree_iter *ite
 		*write_locked = true;
 	}
 
-	if (unlikely(!bch2_btree_node_insert_fits(c, path->l[0].b, wb->k.k.u64s))) {
+	if (!bch2_btree_node_insert_fits(c, path->l[0].b, wb->k.k.u64s)) {
+		bch2_btree_node_unlock_write(trans, path, path->l[0].b);
 		*write_locked = false;
-		return wb_flush_one_slowpath(trans, iter, wb);
+		goto trans_commit;
 	}
 
 	bch2_btree_insert_key_leaf(trans, path, &wb->k, wb->journal_seq);
 	(*fast)++;
 	return 0;
+trans_commit:
+	trans->journal_res.seq = wb->journal_seq;
+
+	return  bch2_trans_update(trans, iter, &wb->k,
+				  BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?:
+		bch2_trans_commit(trans, NULL, NULL,
+				  commit_flags|
+				  BCH_TRANS_COMMIT_no_check_rw|
+				  BCH_TRANS_COMMIT_no_enospc|
+				  BCH_TRANS_COMMIT_no_journal_res|
+				  BCH_TRANS_COMMIT_journal_reclaim);
+}
+
+static union btree_write_buffer_state btree_write_buffer_switch(struct btree_write_buffer *wb)
+{
+	union btree_write_buffer_state old, new;
+	u64 v = READ_ONCE(wb->state.v);
+
+	do {
+		old.v = new.v = v;
+
+		new.nr = 0;
+		new.idx++;
+	} while ((v = atomic64_cmpxchg_acquire(&wb->state.counter, old.v, new.v)) != old.v);
+
+	while (old.idx == 0 ? wb->state.ref0 : wb->state.ref1)
+		cpu_relax();
+
+	smp_mb();
+
+	return old;
 }
 
 /*
@@ -192,79 +137,31 @@ btree_write_buffered_insert(struct btree_trans *trans,
 	return ret;
 }
 
-static void move_keys_from_inc_to_flushing(struct btree_write_buffer *wb)
-{
-	struct bch_fs *c = container_of(wb, struct bch_fs, btree_write_buffer);
-	struct journal *j = &c->journal;
-
-	if (!wb->inc.keys.nr)
-		return;
-
-	bch2_journal_pin_add(j, wb->inc.keys.data[0].journal_seq, &wb->flushing.pin,
-			     bch2_btree_write_buffer_journal_flush);
-
-	darray_resize(&wb->flushing.keys, min_t(size_t, 1U << 20, wb->flushing.keys.nr + wb->inc.keys.nr));
-	darray_resize(&wb->sorted, wb->flushing.keys.size);
-
-	if (!wb->flushing.keys.nr && wb->sorted.size >= wb->inc.keys.nr) {
-		swap(wb->flushing.keys, wb->inc.keys);
-		goto out;
-	}
-
-	size_t nr = min(darray_room(wb->flushing.keys),
-			wb->sorted.size - wb->flushing.keys.nr);
-	nr = min(nr, wb->inc.keys.nr);
-
-	memcpy(&darray_top(wb->flushing.keys),
-	       wb->inc.keys.data,
-	       sizeof(wb->inc.keys.data[0]) * nr);
-
-	memmove(wb->inc.keys.data,
-		wb->inc.keys.data + nr,
-	       sizeof(wb->inc.keys.data[0]) * (wb->inc.keys.nr - nr));
-
-	wb->flushing.keys.nr	+= nr;
-	wb->inc.keys.nr		-= nr;
-out:
-	if (!wb->inc.keys.nr)
-		bch2_journal_pin_drop(j, &wb->inc.pin);
-	else
-		bch2_journal_pin_update(j, wb->inc.keys.data[0].journal_seq, &wb->inc.pin,
-					bch2_btree_write_buffer_journal_flush);
-
-	if (j->watermark) {
-		spin_lock(&j->lock);
-		bch2_journal_set_watermark(j);
-		spin_unlock(&j->lock);
-	}
-
-	BUG_ON(wb->sorted.size < wb->flushing.keys.nr);
-}
-
-static int bch2_btree_write_buffer_flush_locked(struct btree_trans *trans)
+int bch2_btree_write_buffer_flush_locked(struct btree_trans *trans)
 {
 	struct bch_fs *c = trans->c;
 	struct journal *j = &c->journal;
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
-	struct wb_key_ref *i;
+	struct journal_entry_pin pin;
+	struct btree_write_buffered_key *i, *keys;
 	struct btree_iter iter = { NULL };
-	size_t skipped = 0, fast = 0, slowpath = 0;
+	size_t nr = 0, skipped = 0, fast = 0, slowpath = 0;
 	bool write_locked = false;
+	union btree_write_buffer_state s;
 	int ret = 0;
 
-	bch2_trans_unlock(trans);
-	bch2_trans_begin(trans);
+	memset(&pin, 0, sizeof(pin));
 
-	mutex_lock(&wb->inc.lock);
-	move_keys_from_inc_to_flushing(wb);
-	mutex_unlock(&wb->inc.lock);
+	bch2_journal_pin_copy(j, &pin, &wb->journal_pin,
+			      bch2_btree_write_buffer_journal_flush);
+	bch2_journal_pin_drop(j, &wb->journal_pin);
 
-	for (size_t i = 0; i < wb->flushing.keys.nr; i++) {
-		wb->sorted.data[i].idx = i;
-		wb->sorted.data[i].btree = wb->flushing.keys.data[i].btree;
-		memcpy(&wb->sorted.data[i].pos, &wb->flushing.keys.data[i].k.k.p, sizeof(struct bpos));
-	}
-	wb->sorted.nr = wb->flushing.keys.nr;
+	s = btree_write_buffer_switch(wb);
+	keys = wb->keys[s.idx];
+	nr = s.nr;
+
+	if (race_fault())
+		goto slowpath;
 
 	/*
 	 * We first sort so that we can detect and skip redundant updates, and
@@ -280,151 +177,111 @@ static int bch2_btree_write_buffer_flush_locked(struct btree_trans *trans)
 	 * If that happens, simply skip the key so we can optimistically insert
 	 * as many keys as possible in the fast path.
 	 */
-	wb_sort(wb->sorted.data, wb->sorted.nr);
-
-	darray_for_each(wb->sorted, i) {
-		struct btree_write_buffered_key *k = &wb->flushing.keys.data[i->idx];
-
-		for (struct wb_key_ref *n = i + 1; n < min(i + 4, &darray_top(wb->sorted)); n++)
-			prefetch(&wb->flushing.keys.data[n->idx]);
-
-		BUG_ON(!k->journal_seq);
-
-		if (i + 1 < &darray_top(wb->sorted) &&
-		    wb_key_eq(i, i + 1)) {
-			struct btree_write_buffered_key *n = &wb->flushing.keys.data[i[1].idx];
+	sort(keys, nr, sizeof(keys[0]),
+	     btree_write_buffered_key_cmp, NULL);
 
+	for (i = keys; i < keys + nr; i++) {
+		if (i + 1 < keys + nr &&
+		    i[0].btree == i[1].btree &&
+		    bpos_eq(i[0].k.k.p, i[1].k.k.p)) {
 			skipped++;
-			n->journal_seq = min_t(u64, n->journal_seq, k->journal_seq);
-			k->journal_seq = 0;
+			i->journal_seq = 0;
 			continue;
 		}
 
 		if (write_locked &&
-		    (iter.path->btree_id != k->btree ||
-		     bpos_gt(k->k.k.p, iter.path->l[0].b->key.k.p))) {
+		    (iter.path->btree_id != i->btree ||
+		     bpos_gt(i->k.k.p, iter.path->l[0].b->key.k.p))) {
 			bch2_btree_node_unlock_write(trans, iter.path, iter.path->l[0].b);
 			write_locked = false;
 		}
 
-		if (!iter.path || iter.path->btree_id != k->btree) {
+		if (!iter.path || iter.path->btree_id != i->btree) {
 			bch2_trans_iter_exit(trans, &iter);
-			bch2_trans_iter_init(trans, &iter, k->btree, k->k.k.p,
+			bch2_trans_iter_init(trans, &iter, i->btree, i->k.k.p,
 					     BTREE_ITER_INTENT|BTREE_ITER_ALL_SNAPSHOTS);
 		}
 
-		bch2_btree_iter_set_pos(&iter, k->k.k.p);
+		bch2_btree_iter_set_pos(&iter, i->k.k.p);
 		iter.path->preserve = false;
 
 		do {
-			if (race_fault()) {
-				ret = -BCH_ERR_journal_reclaim_would_deadlock;
-				break;
-			}
-
-			ret = wb_flush_one(trans, &iter, k, &write_locked, &fast);
+			ret = bch2_btree_write_buffer_flush_one(trans, &iter, i, 0,
+								&write_locked, &fast);
 			if (!write_locked)
 				bch2_trans_begin(trans);
 		} while (bch2_err_matches(ret, BCH_ERR_transaction_restart));
 
-		if (!ret) {
-			k->journal_seq = 0;
-		} else if (ret == -BCH_ERR_journal_reclaim_would_deadlock) {
+		if (ret == -BCH_ERR_journal_reclaim_would_deadlock) {
 			slowpath++;
-			ret = 0;
-		} else
+			continue;
+		}
+		if (ret)
 			break;
+
+		i->journal_seq = 0;
 	}
 
 	if (write_locked)
 		bch2_btree_node_unlock_write(trans, iter.path, iter.path->l[0].b);
 	bch2_trans_iter_exit(trans, &iter);
 
-	if (ret)
-		goto err;
-
-	if (slowpath) {
-		/*
-		 * Flush in the order they were present in the journal, so that
-		 * we can release journal pins:
-		 * The fastpath zapped the seq of keys that were successfully flushed so
-		 * we can skip those here.
-		 */
-		trace_write_buffer_flush_slowpath(trans, slowpath, wb->flushing.keys.nr);
-
-		struct btree_write_buffered_key *i;
-		darray_for_each(wb->flushing.keys, i) {
-			if (!i->journal_seq)
-				continue;
-
-			bch2_journal_pin_update(j, i->journal_seq, &wb->flushing.pin,
-						bch2_btree_write_buffer_journal_flush);
-
-			bch2_trans_begin(trans);
-
-			ret = commit_do(trans, NULL, NULL,
-					BCH_WATERMARK_reclaim|
-					BCH_TRANS_COMMIT_no_check_rw|
-					BCH_TRANS_COMMIT_no_enospc|
-					BCH_TRANS_COMMIT_no_journal_res|
-					BCH_TRANS_COMMIT_journal_reclaim,
-					btree_write_buffered_insert(trans, i));
-			if (ret)
-				goto err;
-		}
-	}
-err:
+	trace_write_buffer_flush(trans, nr, skipped, fast, wb->size);
+
+	if (slowpath)
+		goto slowpath;
+
 	bch2_fs_fatal_err_on(ret, c, "%s: insert error %s", __func__, bch2_err_str(ret));
-	trace_write_buffer_flush(trans, wb->flushing.keys.nr, skipped, fast, 0);
-	bch2_journal_pin_drop(j, &wb->flushing.pin);
-	wb->flushing.keys.nr = 0;
+out:
+	bch2_journal_pin_drop(j, &pin);
 	return ret;
-}
+slowpath:
+	trace_and_count(c, write_buffer_flush_slowpath, trans, slowpath, nr);
 
-static int fetch_wb_keys_from_journal(struct bch_fs *c, u64 seq)
-{
-	struct journal *j = &c->journal;
-	struct journal_buf *buf;
-	int ret = 0;
+	/*
+	 * Now sort the rest by journal seq and bump the journal pin as we go.
+	 * The slowpath zapped the seq of keys that were successfully flushed so
+	 * we can skip those here.
+	 */
+	sort(keys, nr, sizeof(keys[0]),
+	     btree_write_buffered_journal_cmp,
+	     NULL);
 
-	mutex_lock(&j->buf_lock);
-	while ((buf = bch2_next_write_buffer_flush_journal_buf(j, seq)))
-		if (bch2_journal_keys_to_write_buffer(c, buf)) {
-			ret = -ENOMEM;
+	for (i = keys; i < keys + nr; i++) {
+		if (!i->journal_seq)
+			continue;
+
+		bch2_journal_pin_update(j, i->journal_seq, &pin,
+			      bch2_btree_write_buffer_journal_flush);
+
+		ret = commit_do(trans, NULL, NULL,
+				BCH_WATERMARK_reclaim|
+				BCH_TRANS_COMMIT_no_check_rw|
+				BCH_TRANS_COMMIT_no_enospc|
+				BCH_TRANS_COMMIT_no_journal_res|
+				BCH_TRANS_COMMIT_journal_reclaim,
+				btree_write_buffered_insert(trans, i));
+		if (bch2_fs_fatal_err_on(ret, c, "%s: insert error %s", __func__, bch2_err_str(ret)))
 			break;
-		}
-	mutex_unlock(&j->buf_lock);
+	}
 
-	return ret;
+	goto out;
 }
 
 int bch2_btree_write_buffer_flush_sync(struct btree_trans *trans)
 {
 	struct bch_fs *c = trans->c;
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-	int ret = 0, fetch_from_journal_err;
-
-	trace_write_buffer_flush_sync(trans, _RET_IP_);
-retry:
-	bch2_trans_unlock(trans);
 
-	bch2_journal_block_reservations(&c->journal);
-	fetch_from_journal_err = fetch_wb_keys_from_journal(c, U64_MAX);
-	bch2_journal_unblock(&c->journal);
-
-	/*
-	 * On memory allocation failure, bch2_btree_write_buffer_flush_locked()
-	 * is not guaranteed to empty wb->inc:
-	 */
-	mutex_lock(&wb->flushing.lock);
-	while (!ret &&
-	       (wb->flushing.keys.nr || wb->inc.keys.nr))
-		ret = bch2_btree_write_buffer_flush_locked(trans);
-	mutex_unlock(&wb->flushing.lock);
+	if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_btree_write_buffer))
+		return -BCH_ERR_erofs_no_writes;
 
-	if (!ret && fetch_from_journal_err)
-		goto retry;
+	trace_and_count(c, write_buffer_flush_sync, trans, _RET_IP_);
 
+	bch2_trans_unlock(trans);
+	mutex_lock(&c->btree_write_buffer.flush_lock);
+	int ret = bch2_btree_write_buffer_flush_locked(trans);
+	mutex_unlock(&c->btree_write_buffer.flush_lock);
+	bch2_write_ref_put(c, BCH_WRITE_REF_btree_write_buffer);
 	return ret;
 }
 
@@ -434,9 +291,9 @@ int bch2_btree_write_buffer_flush_nocheck_rw(struct btree_trans *trans)
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
 	int ret = 0;
 
-	if (mutex_trylock(&wb->flushing.lock)) {
+	if (mutex_trylock(&wb->flush_lock)) {
 		ret = bch2_btree_write_buffer_flush_locked(trans);
-		mutex_unlock(&wb->flushing.lock);
+		mutex_unlock(&wb->flush_lock);
 	}
 
 	return ret;
@@ -459,195 +316,85 @@ static int bch2_btree_write_buffer_journal_flush(struct journal *j,
 {
 	struct bch_fs *c = container_of(j, struct bch_fs, journal);
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
-	int ret, fetch_from_journal_err;
 
-	do {
-		fetch_from_journal_err = fetch_wb_keys_from_journal(c, seq);
-
-		mutex_lock(&wb->flushing.lock);
-		ret = bch2_trans_run(c, bch2_btree_write_buffer_flush_locked(trans));
-		mutex_unlock(&wb->flushing.lock);
-	} while (!ret &&
-		 (fetch_from_journal_err ||
-		  (wb->flushing.pin.seq && wb->flushing.pin.seq <= seq) ||
-		  (wb->inc.pin.seq && wb->inc.pin.seq <= seq)));
+	mutex_lock(&wb->flush_lock);
+	int ret = bch2_trans_run(c, bch2_btree_write_buffer_flush_locked(trans));
+	mutex_unlock(&wb->flush_lock);
 
 	return ret;
 }
 
-static void bch2_btree_write_buffer_flush_work(struct work_struct *work)
+static inline u64 btree_write_buffer_ref(int idx)
 {
-	struct bch_fs *c = container_of(work, struct bch_fs, btree_write_buffer.flush_work);
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-	int ret;
-
-	mutex_lock(&wb->flushing.lock);
-	do {
-		ret = bch2_trans_run(c, bch2_btree_write_buffer_flush_locked(trans));
-	} while (!ret && bch2_btree_write_buffer_should_flush(c));
-	mutex_unlock(&wb->flushing.lock);
-
-	bch2_write_ref_put(c, BCH_WRITE_REF_btree_write_buffer);
+	return ((union btree_write_buffer_state) {
+		.ref0 = idx == 0,
+		.ref1 = idx == 1,
+	}).v;
 }
 
-int __bch2_journal_key_to_wb(struct bch_fs *c,
-			     struct journal_keys_to_wb *dst,
-			     enum btree_id btree, struct bkey_i *k)
+int bch2_btree_insert_keys_write_buffer(struct btree_trans *trans)
 {
+	struct bch_fs *c = trans->c;
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
-	int ret;
-retry:
-	ret = darray_make_room_gfp(&dst->wb->keys, 1, GFP_KERNEL);
-	if (!ret && dst->wb == &wb->flushing)
-		ret = darray_resize(&wb->sorted, wb->flushing.keys.size);
-
-	if (unlikely(ret)) {
-		if (dst->wb == &c->btree_write_buffer.flushing) {
-			mutex_unlock(&dst->wb->lock);
-			dst->wb = &c->btree_write_buffer.inc;
-			bch2_journal_pin_add(&c->journal, dst->seq, &dst->wb->pin,
-					     bch2_btree_write_buffer_journal_flush);
-			goto retry;
-		}
-
-		return ret;
-	}
-
-	dst->room = darray_room(dst->wb->keys);
-	if (dst->wb == &wb->flushing)
-		dst->room = min(dst->room, wb->sorted.size - wb->flushing.keys.nr);
-	BUG_ON(!dst->room);
-	BUG_ON(!dst->seq);
-
-	struct btree_write_buffered_key *wb_k = &darray_top(dst->wb->keys);
-	wb_k->journal_seq	= dst->seq;
-	wb_k->btree		= btree;
-	bkey_copy(&wb_k->k, k);
-	dst->wb->keys.nr++;
-	dst->room--;
-	return 0;
-}
+	struct btree_write_buffered_key *i;
+	union btree_write_buffer_state old, new;
+	int ret = 0;
+	u64 v;
 
-void bch2_journal_keys_to_write_buffer_start(struct bch_fs *c, struct journal_keys_to_wb *dst, u64 seq)
-{
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
+	trans_for_each_wb_update(trans, i) {
+		EBUG_ON(i->k.k.u64s > BTREE_WRITE_BUFERED_U64s_MAX);
 
-	if (mutex_trylock(&wb->flushing.lock)) {
-		mutex_lock(&wb->inc.lock);
-		move_keys_from_inc_to_flushing(wb);
+		i->journal_seq		= trans->journal_res.seq;
+		i->journal_offset	= trans->journal_res.offset;
+	}
 
-		/*
-		 * Attempt to skip wb->inc, and add keys directly to
-		 * wb->flushing, saving us a copy later:
-		 */
+	preempt_disable();
+	v = READ_ONCE(wb->state.v);
+	do {
+		old.v = new.v = v;
 
-		if (!wb->inc.keys.nr) {
-			dst->wb = &wb->flushing;
-		} else {
-			mutex_unlock(&wb->flushing.lock);
-			dst->wb = &wb->inc;
+		new.v += btree_write_buffer_ref(new.idx);
+		new.nr += trans->nr_wb_updates;
+		if (new.nr > wb->size) {
+			ret = -BCH_ERR_btree_insert_need_flush_buffer;
+			goto out;
 		}
-	} else {
-		mutex_lock(&wb->inc.lock);
-		dst->wb = &wb->inc;
-	}
+	} while ((v = atomic64_cmpxchg_acquire(&wb->state.counter, old.v, new.v)) != old.v);
 
-	dst->room = darray_room(dst->wb->keys);
-	if (dst->wb == &wb->flushing)
-		dst->room = min(dst->room, wb->sorted.size - wb->flushing.keys.nr);
-	dst->seq = seq;
+	memcpy(wb->keys[new.idx] + old.nr,
+	       trans->wb_updates,
+	       sizeof(trans->wb_updates[0]) * trans->nr_wb_updates);
 
-	bch2_journal_pin_add(&c->journal, seq, &dst->wb->pin,
+	bch2_journal_pin_add(&c->journal, trans->journal_res.seq, &wb->journal_pin,
 			     bch2_btree_write_buffer_journal_flush);
-}
 
-void bch2_journal_keys_to_write_buffer_end(struct bch_fs *c, struct journal_keys_to_wb *dst)
-{
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-
-	if (!dst->wb->keys.nr)
-		bch2_journal_pin_drop(&c->journal, &dst->wb->pin);
-
-	if (bch2_btree_write_buffer_should_flush(c) &&
-	    __bch2_write_ref_tryget(c, BCH_WRITE_REF_btree_write_buffer) &&
-	    !queue_work(system_unbound_wq, &c->btree_write_buffer.flush_work))
-		bch2_write_ref_put(c, BCH_WRITE_REF_btree_write_buffer);
-
-	if (dst->wb == &wb->flushing)
-		mutex_unlock(&wb->flushing.lock);
-	mutex_unlock(&wb->inc.lock);
-}
-
-static int bch2_journal_keys_to_write_buffer(struct bch_fs *c, struct journal_buf *buf)
-{
-	struct journal_keys_to_wb dst;
-	struct jset_entry *entry;
-	struct bkey_i *k;
-	int ret = 0;
-
-	bch2_journal_keys_to_write_buffer_start(c, &dst, le64_to_cpu(buf->data->seq));
-
-	for_each_jset_entry_type(entry, buf->data, BCH_JSET_ENTRY_write_buffer_keys) {
-		jset_entry_for_each_key(entry, k) {
-			ret = bch2_journal_key_to_wb(c, &dst, entry->btree_id, k);
-			if (ret)
-				goto out;
-		}
-
-		entry->type = BCH_JSET_ENTRY_btree_keys;
-	}
-
-	buf->need_flush_to_write_buffer = false;
+	atomic64_sub_return_release(btree_write_buffer_ref(new.idx), &wb->state.counter);
 out:
-	bch2_journal_keys_to_write_buffer_end(c, &dst);
-	return ret;
-}
-
-static int wb_keys_resize(struct btree_write_buffer_keys *wb, size_t new_size)
-{
-	if (wb->keys.size >= new_size)
-		return 0;
-
-	if (!mutex_trylock(&wb->lock))
-		return -EINTR;
-
-	int ret = darray_resize(&wb->keys, new_size);
-	mutex_unlock(&wb->lock);
+	preempt_enable();
 	return ret;
 }
 
-int bch2_btree_write_buffer_resize(struct bch_fs *c, size_t new_size)
-{
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-
-	return wb_keys_resize(&wb->flushing, new_size) ?:
-		wb_keys_resize(&wb->inc, new_size);
-}
-
 void bch2_fs_btree_write_buffer_exit(struct bch_fs *c)
 {
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
 
-	BUG_ON((wb->inc.keys.nr || wb->flushing.keys.nr) &&
-	       !bch2_journal_error(&c->journal));
+	BUG_ON(wb->state.nr && !bch2_journal_error(&c->journal));
 
-	darray_exit(&wb->sorted);
-	darray_exit(&wb->flushing.keys);
-	darray_exit(&wb->inc.keys);
+	kvfree(wb->keys[1]);
+	kvfree(wb->keys[0]);
 }
 
 int bch2_fs_btree_write_buffer_init(struct bch_fs *c)
 {
 	struct btree_write_buffer *wb = &c->btree_write_buffer;
 
-	mutex_init(&wb->inc.lock);
-	mutex_init(&wb->flushing.lock);
-	INIT_WORK(&wb->flush_work, bch2_btree_write_buffer_flush_work);
+	mutex_init(&wb->flush_lock);
+	wb->size = c->opts.btree_write_buffer_size;
 
-	/* Will be resized by journal as needed: */
-	unsigned initial_size = 1 << 16;
+	wb->keys[0] = kvmalloc_array(wb->size, sizeof(*wb->keys[0]), GFP_KERNEL);
+	wb->keys[1] = kvmalloc_array(wb->size, sizeof(*wb->keys[1]), GFP_KERNEL);
+	if (!wb->keys[0] || !wb->keys[1])
+		return -BCH_ERR_ENOMEM_fs_btree_write_buffer_init;
 
-	return  darray_make_room(&wb->inc.keys, initial_size) ?:
-		darray_make_room(&wb->flushing.keys, initial_size) ?:
-		darray_make_room(&wb->sorted, initial_size);
+	return 0;
 }
diff --git a/libbcachefs/btree_write_buffer.h b/libbcachefs/btree_write_buffer.h
index 1f645f52..dec2c9a8 100644
--- a/libbcachefs/btree_write_buffer.h
+++ b/libbcachefs/btree_write_buffer.h
@@ -2,59 +2,13 @@
 #ifndef _BCACHEFS_BTREE_WRITE_BUFFER_H
 #define _BCACHEFS_BTREE_WRITE_BUFFER_H
 
-#include "bkey.h"
-
-static inline bool bch2_btree_write_buffer_should_flush(struct bch_fs *c)
-{
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-
-	return wb->inc.keys.nr + wb->flushing.keys.nr > wb->inc.keys.size / 4;
-}
-
-static inline bool bch2_btree_write_buffer_must_wait(struct bch_fs *c)
-{
-	struct btree_write_buffer *wb = &c->btree_write_buffer;
-
-	return wb->inc.keys.nr > wb->inc.keys.size * 3 / 4;
-}
-
-struct btree_trans;
-int bch2_btree_write_buffer_flush_sync(struct btree_trans *);
+int bch2_btree_write_buffer_flush_locked(struct btree_trans *);
 int bch2_btree_write_buffer_flush_nocheck_rw(struct btree_trans *);
+int bch2_btree_write_buffer_flush_sync(struct btree_trans *);
 int bch2_btree_write_buffer_tryflush(struct btree_trans *);
 
-struct journal_keys_to_wb {
-	struct btree_write_buffer_keys	*wb;
-	size_t				room;
-	u64				seq;
-};
-
-int __bch2_journal_key_to_wb(struct bch_fs *,
-			     struct journal_keys_to_wb *,
-			     enum btree_id, struct bkey_i *);
-
-static inline int bch2_journal_key_to_wb(struct bch_fs *c,
-			     struct journal_keys_to_wb *dst,
-			     enum btree_id btree, struct bkey_i *k)
-{
-	EBUG_ON(!dst->seq);
-
-	if (unlikely(!dst->room))
-		return __bch2_journal_key_to_wb(c, dst, btree, k);
-
-	struct btree_write_buffered_key *wb_k = &darray_top(dst->wb->keys);
-	wb_k->journal_seq	= dst->seq;
-	wb_k->btree		= btree;
-	bkey_copy(&wb_k->k, k);
-	dst->wb->keys.nr++;
-	dst->room--;
-	return 0;
-}
-
-void bch2_journal_keys_to_write_buffer_start(struct bch_fs *, struct journal_keys_to_wb *, u64);
-void bch2_journal_keys_to_write_buffer_end(struct bch_fs *, struct journal_keys_to_wb *);
+int bch2_btree_insert_keys_write_buffer(struct btree_trans *);
 
-int bch2_btree_write_buffer_resize(struct bch_fs *, size_t);
 void bch2_fs_btree_write_buffer_exit(struct bch_fs *);
 int bch2_fs_btree_write_buffer_init(struct bch_fs *);
 
diff --git a/libbcachefs/btree_write_buffer_types.h b/libbcachefs/btree_write_buffer_types.h
index 9b9433de..99993ba7 100644
--- a/libbcachefs/btree_write_buffer_types.h
+++ b/libbcachefs/btree_write_buffer_types.h
@@ -2,56 +2,43 @@
 #ifndef _BCACHEFS_BTREE_WRITE_BUFFER_TYPES_H
 #define _BCACHEFS_BTREE_WRITE_BUFFER_TYPES_H
 
-#include "darray.h"
 #include "journal_types.h"
 
 #define BTREE_WRITE_BUFERED_VAL_U64s_MAX	4
 #define BTREE_WRITE_BUFERED_U64s_MAX	(BKEY_U64s + BTREE_WRITE_BUFERED_VAL_U64s_MAX)
 
-struct wb_key_ref {
-union {
-	struct {
-#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
-		unsigned			idx:24;
-		u8				pos[sizeof(struct bpos)];
-		enum btree_id			btree:8;
-#else
-		enum btree_id			btree:8;
-		u8				pos[sizeof(struct bpos)];
-		unsigned			idx:24;
-#endif
-	} __packed;
-	struct {
-#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
-		u64 lo;
-		u64 mi;
-		u64 hi;
-#else
-		u64 hi;
-		u64 mi;
-		u64 lo;
-#endif
-	};
-};
-};
-
 struct btree_write_buffered_key {
-	enum btree_id			btree:8;
-	u64				journal_seq:56;
+	u64			journal_seq;
+	unsigned		journal_offset;
+	enum btree_id		btree;
 	__BKEY_PADDED(k, BTREE_WRITE_BUFERED_VAL_U64s_MAX);
 };
 
-struct btree_write_buffer_keys {
-	DARRAY(struct btree_write_buffered_key) keys;
-	struct journal_entry_pin	pin;
-	struct mutex			lock;
+union btree_write_buffer_state {
+	struct {
+		atomic64_t	counter;
+	};
+
+	struct {
+		u64		v;
+	};
+
+	struct {
+		u64			nr:23;
+		u64			idx:1;
+		u64			ref0:20;
+		u64			ref1:20;
+	};
 };
 
 struct btree_write_buffer {
-	DARRAY(struct wb_key_ref)	sorted;
-	struct btree_write_buffer_keys	inc;
-	struct btree_write_buffer_keys	flushing;
-	struct work_struct		flush_work;
+	struct mutex			flush_lock;
+	struct journal_entry_pin	journal_pin;
+
+	union btree_write_buffer_state	state;
+	size_t				size;
+
+	struct btree_write_buffered_key	*keys[2];
 };
 
 #endif /* _BCACHEFS_BTREE_WRITE_BUFFER_TYPES_H */
diff --git a/libbcachefs/buckets.c b/libbcachefs/buckets.c
index 50eb6ba2..312bd0c8 100644
--- a/libbcachefs/buckets.c
+++ b/libbcachefs/buckets.c
@@ -334,7 +334,7 @@ static void bch2_dev_usage_update(struct bch_fs *c, struct bch_dev *ca,
 	preempt_enable();
 }
 
-struct bch_alloc_v4 bucket_m_to_alloc(struct bucket b)
+static inline struct bch_alloc_v4 bucket_m_to_alloc(struct bucket b)
 {
 	return (struct bch_alloc_v4) {
 		.gen		= b.gen,
@@ -346,13 +346,12 @@ struct bch_alloc_v4 bucket_m_to_alloc(struct bucket b)
 }
 
 static void bch2_dev_usage_update_m(struct bch_fs *c, struct bch_dev *ca,
-				    struct bucket old, struct bucket new,
-				    u64 journal_seq, bool gc)
+				    struct bucket old, struct bucket new)
 {
 	bch2_dev_usage_update(c, ca,
 			      bucket_m_to_alloc(old),
 			      bucket_m_to_alloc(new),
-			      journal_seq, gc);
+			      0, true);
 }
 
 static inline int __update_replicas(struct bch_fs *c,
@@ -658,7 +657,7 @@ int bch2_mark_metadata_bucket(struct bch_fs *c, struct bch_dev *ca,
 err:
 	bucket_unlock(g);
 	if (!ret)
-		bch2_dev_usage_update_m(c, ca, old, new, 0, true);
+		bch2_dev_usage_update_m(c, ca, old, new);
 	percpu_up_read(&c->mark_lock);
 	return ret;
 }
@@ -773,7 +772,6 @@ static int mark_stripe_bucket(struct btree_trans *trans,
 			      unsigned flags)
 {
 	struct bch_fs *c = trans->c;
-	u64 journal_seq = trans->journal_res.seq;
 	const struct bch_stripe *s = bkey_s_c_to_stripe(k).v;
 	unsigned nr_data = s->nr_blocks - s->nr_redundant;
 	bool parity = ptr_idx >= nr_data;
@@ -820,7 +818,7 @@ static int mark_stripe_bucket(struct btree_trans *trans,
 err:
 	bucket_unlock(g);
 	if (!ret)
-		bch2_dev_usage_update_m(c, ca, old, new, journal_seq, true);
+		bch2_dev_usage_update_m(c, ca, old, new);
 	percpu_up_read(&c->mark_lock);
 	printbuf_exit(&buf);
 	return ret;
@@ -843,8 +841,12 @@ static int __mark_pointer(struct btree_trans *trans,
 		return ret;
 
 	*dst_sectors += sectors;
-	*bucket_data_type = *dirty_sectors || *cached_sectors
-		? ptr_data_type : 0;
+
+	if (!*dirty_sectors && !*cached_sectors)
+		*bucket_data_type = 0;
+	else if (*bucket_data_type != BCH_DATA_stripe)
+		*bucket_data_type = ptr_data_type;
+
 	return 0;
 }
 
@@ -855,7 +857,6 @@ static int bch2_mark_pointer(struct btree_trans *trans,
 			     s64 sectors,
 			     unsigned flags)
 {
-	u64 journal_seq = trans->journal_res.seq;
 	struct bch_fs *c = trans->c;
 	struct bch_dev *ca = bch_dev_bkey_exists(c, p.ptr.dev);
 	struct bucket old, new, *g;
@@ -882,7 +883,7 @@ static int bch2_mark_pointer(struct btree_trans *trans,
 	new = *g;
 	bucket_unlock(g);
 	if (!ret)
-		bch2_dev_usage_update_m(c, ca, old, new, journal_seq, true);
+		bch2_dev_usage_update_m(c, ca, old, new);
 	percpu_up_read(&c->mark_lock);
 
 	return ret;
diff --git a/libbcachefs/chardev.c b/libbcachefs/chardev.c
index 118f0c0c..ba0436ae 100644
--- a/libbcachefs/chardev.c
+++ b/libbcachefs/chardev.c
@@ -418,7 +418,7 @@ static long bch2_ioctl_fs_usage(struct bch_fs *c,
 	unsigned i;
 	int ret = 0;
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EINVAL;
 
 	if (get_user(replica_entries_bytes, &user_arg->replica_entries_bytes))
@@ -492,7 +492,7 @@ static long bch2_ioctl_dev_usage(struct bch_fs *c,
 	struct bch_dev *ca;
 	unsigned i;
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EINVAL;
 
 	if (copy_from_user(&arg, user_arg, sizeof(arg)))
@@ -533,7 +533,7 @@ static long bch2_ioctl_dev_usage_v2(struct bch_fs *c,
 	struct bch_dev *ca;
 	int ret = 0;
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EINVAL;
 
 	if (copy_from_user(&arg, user_arg, sizeof(arg)))
@@ -725,7 +725,7 @@ long bch2_fs_ioctl(struct bch_fs *c, unsigned cmd, void __user *arg)
 		BCH_IOCTL(disk_get_idx, struct bch_ioctl_disk_get_idx);
 	}
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EINVAL;
 
 	switch (cmd) {
diff --git a/libbcachefs/ec.c b/libbcachefs/ec.c
index 2a02bf00..bc8b556f 100644
--- a/libbcachefs/ec.c
+++ b/libbcachefs/ec.c
@@ -1005,7 +1005,7 @@ static int ec_stripe_update_extents(struct bch_fs *c, struct ec_stripe_buf *s)
 	unsigned i, nr_data = v->nr_blocks - v->nr_redundant;
 	int ret = 0;
 
-	ret = bch2_btree_write_buffer_flush_sync(trans);
+	ret = bch2_btree_write_buffer_flush_nocheck_rw(trans);
 	if (ret)
 		goto err;
 
@@ -1415,7 +1415,7 @@ __bch2_ec_stripe_head_get(struct btree_trans *trans,
 	if (ret)
 		return ERR_PTR(ret);
 
-	if (test_bit(BCH_FS_GOING_RO, &c->flags)) {
+	if (test_bit(BCH_FS_going_ro, &c->flags)) {
 		h = ERR_PTR(-BCH_ERR_erofs_no_writes);
 		goto found;
 	}
diff --git a/libbcachefs/errcode.h b/libbcachefs/errcode.h
index 4d35e5c6..e3e2be79 100644
--- a/libbcachefs/errcode.h
+++ b/libbcachefs/errcode.h
@@ -150,6 +150,7 @@
 	x(BCH_ERR_btree_insert_fail,	btree_insert_need_mark_replicas)	\
 	x(BCH_ERR_btree_insert_fail,	btree_insert_need_journal_res)		\
 	x(BCH_ERR_btree_insert_fail,	btree_insert_need_journal_reclaim)	\
+	x(BCH_ERR_btree_insert_fail,	btree_insert_need_flush_buffer)		\
 	x(0,				backpointer_to_overwritten_btree_node)	\
 	x(0,				lock_fail_root_changed)			\
 	x(0,				journal_reclaim_would_deadlock)		\
diff --git a/libbcachefs/error.c b/libbcachefs/error.c
index 7b28d379..655e3ba9 100644
--- a/libbcachefs/error.c
+++ b/libbcachefs/error.c
@@ -7,7 +7,7 @@
 
 bool bch2_inconsistent_error(struct bch_fs *c)
 {
-	set_bit(BCH_FS_ERROR, &c->flags);
+	set_bit(BCH_FS_error, &c->flags);
 
 	switch (c->opts.errors) {
 	case BCH_ON_ERROR_continue:
@@ -26,8 +26,8 @@ bool bch2_inconsistent_error(struct bch_fs *c)
 
 void bch2_topology_error(struct bch_fs *c)
 {
-	set_bit(BCH_FS_TOPOLOGY_ERROR, &c->flags);
-	if (test_bit(BCH_FS_FSCK_DONE, &c->flags))
+	set_bit(BCH_FS_topology_error, &c->flags);
+	if (test_bit(BCH_FS_fsck_done, &c->flags))
 		bch2_inconsistent_error(c);
 }
 
@@ -114,7 +114,7 @@ static struct fsck_err_state *fsck_err_get(struct bch_fs *c, const char *fmt)
 {
 	struct fsck_err_state *s;
 
-	if (test_bit(BCH_FS_FSCK_DONE, &c->flags))
+	if (test_bit(BCH_FS_fsck_done, &c->flags))
 		return NULL;
 
 	list_for_each_entry(s, &c->fsck_error_msgs, list)
@@ -193,7 +193,7 @@ int bch2_fsck_err(struct bch_fs *c,
 		prt_printf(out, bch2_log_msg(c, ""));
 #endif
 
-	if (test_bit(BCH_FS_FSCK_DONE, &c->flags)) {
+	if (test_bit(BCH_FS_fsck_done, &c->flags)) {
 		if (c->opts.errors != BCH_ON_ERROR_continue ||
 		    !(flags & (FSCK_CAN_FIX|FSCK_CAN_IGNORE))) {
 			prt_str(out, ", shutting down");
@@ -253,7 +253,7 @@ int bch2_fsck_err(struct bch_fs *c,
 	if (print)
 		bch2_print_string_as_lines(KERN_ERR, out->buf);
 
-	if (!test_bit(BCH_FS_FSCK_DONE, &c->flags) &&
+	if (!test_bit(BCH_FS_fsck_done, &c->flags) &&
 	    (ret != -BCH_ERR_fsck_fix &&
 	     ret != -BCH_ERR_fsck_ignore))
 		bch_err(c, "Unable to continue, halting");
@@ -271,10 +271,10 @@ int bch2_fsck_err(struct bch_fs *c,
 		bch2_inconsistent_error(c);
 
 	if (ret == -BCH_ERR_fsck_fix) {
-		set_bit(BCH_FS_ERRORS_FIXED, &c->flags);
+		set_bit(BCH_FS_errors_fixed, &c->flags);
 	} else {
-		set_bit(BCH_FS_ERRORS_NOT_FIXED, &c->flags);
-		set_bit(BCH_FS_ERROR, &c->flags);
+		set_bit(BCH_FS_errors_not_fixed, &c->flags);
+		set_bit(BCH_FS_error, &c->flags);
 	}
 
 	return ret;
diff --git a/libbcachefs/fs-io-buffered.c b/libbcachefs/fs-io-buffered.c
index 52f0e7ac..637a83e4 100644
--- a/libbcachefs/fs-io-buffered.c
+++ b/libbcachefs/fs-io-buffered.c
@@ -638,7 +638,7 @@ do_io:
 		/* Check for writing past i_size: */
 		WARN_ONCE((bio_end_sector(&w->io->op.wbio.bio) << 9) >
 			  round_up(i_size, block_bytes(c)) &&
-			  !test_bit(BCH_FS_EMERGENCY_RO, &c->flags),
+			  !test_bit(BCH_FS_emergency_ro, &c->flags),
 			  "writing past i_size: %llu > %llu (unrounded %llu)\n",
 			  bio_end_sector(&w->io->op.wbio.bio) << 9,
 			  round_up(i_size, block_bytes(c)),
diff --git a/libbcachefs/fsck.c b/libbcachefs/fsck.c
index b1c89a48..8cf4bcf9 100644
--- a/libbcachefs/fsck.c
+++ b/libbcachefs/fsck.c
@@ -448,7 +448,7 @@ static int snapshots_seen_update(struct bch_fs *c, struct snapshots_seen *s,
 				bch2_btree_id_str(btree_id),
 				pos.inode, pos.offset,
 				i->id, n.id, n.equiv);
-			set_bit(BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS, &c->flags);
+			set_bit(BCH_FS_need_delete_dead_snapshots, &c->flags);
 			return bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_delete_dead_snapshots);
 		}
 	}
diff --git a/libbcachefs/inode.c b/libbcachefs/inode.c
index a2c96f7c..b861ab2e 100644
--- a/libbcachefs/inode.c
+++ b/libbcachefs/inode.c
@@ -1173,7 +1173,7 @@ again:
 			break;
 
 		if (ret) {
-			if (!test_bit(BCH_FS_RW, &c->flags)) {
+			if (!test_bit(BCH_FS_rw, &c->flags)) {
 				bch2_trans_unlock(trans);
 				bch2_fs_lazy_rw(c);
 			}
diff --git a/libbcachefs/journal.c b/libbcachefs/journal.c
index 86b148d9..d5540c85 100644
--- a/libbcachefs/journal.c
+++ b/libbcachefs/journal.c
@@ -10,7 +10,6 @@
 #include "bkey_methods.h"
 #include "btree_gc.h"
 #include "btree_update.h"
-#include "btree_write_buffer.h"
 #include "buckets.h"
 #include "error.h"
 #include "journal.h"
@@ -148,7 +147,6 @@ void bch2_journal_buf_put_final(struct journal *j, u64 seq, bool write)
 		bch2_journal_reclaim_fast(j);
 	if (write)
 		closure_call(&j->io, bch2_journal_write, c->io_complete_wq, NULL);
-	wake_up(&j->wait);
 }
 
 /*
@@ -332,7 +330,6 @@ static int journal_entry_open(struct journal *j)
 	buf->must_flush	= false;
 	buf->separate_flush = false;
 	buf->flush_time	= 0;
-	buf->need_flush_to_write_buffer = true;
 
 	memset(buf->data, 0, sizeof(*buf->data));
 	buf->data->seq	= cpu_to_le64(journal_cur_seq(j));
@@ -769,75 +766,6 @@ void bch2_journal_block(struct journal *j)
 	journal_quiesce(j);
 }
 
-/*
- * XXX: ideally this would not be closing the current journal entry, but
- * otherwise we do not have a way to avoid racing with res_get() - j->blocked
- * will race.
- */
-static bool journal_reservations_stopped(struct journal *j)
-{
-	union journal_res_state s;
-
-	journal_entry_close(j);
-
-	s.v = atomic64_read_acquire(&j->reservations.counter);
-
-	return  s.buf0_count == 0 &&
-		s.buf1_count == 0 &&
-		s.buf2_count == 0 &&
-		s.buf3_count == 0;
-}
-
-void bch2_journal_block_reservations(struct journal *j)
-{
-	spin_lock(&j->lock);
-	j->blocked++;
-	spin_unlock(&j->lock);
-
-	wait_event(j->wait, journal_reservations_stopped(j));
-}
-
-static struct journal_buf *__bch2_next_write_buffer_flush_journal_buf(struct journal *j, u64 max_seq)
-{
-	spin_lock(&j->lock);
-	max_seq = min(max_seq, journal_cur_seq(j));
-
-	for (u64 seq = journal_last_unwritten_seq(j);
-	     seq <= max_seq;
-	     seq++) {
-		unsigned idx = seq & JOURNAL_BUF_MASK;
-		struct journal_buf *buf = j->buf + idx;
-		union journal_res_state s;
-
-		if (!buf->need_flush_to_write_buffer)
-			continue;
-
-		if (seq == journal_cur_seq(j))
-			__journal_entry_close(j, JOURNAL_ENTRY_CLOSED_VAL);
-
-		s.v = atomic64_read_acquire(&j->reservations.counter);
-
-		if (journal_state_count(s, idx)) {
-			spin_unlock(&j->lock);
-			return ERR_PTR(-EAGAIN);
-		}
-
-		spin_unlock(&j->lock);
-		return buf;
-	}
-
-	spin_unlock(&j->lock);
-	return NULL;
-}
-
-struct journal_buf *bch2_next_write_buffer_flush_journal_buf(struct journal *j, u64 max_seq)
-{
-	struct journal_buf *ret;
-
-	wait_event(j->wait, (ret = __bch2_next_write_buffer_flush_journal_buf(j, max_seq)) != ERR_PTR(-EAGAIN));
-	return ret;
-}
-
 /* allocate journal on a device: */
 
 static int __bch2_set_nr_journal_buckets(struct bch_dev *ca, unsigned nr,
@@ -1289,7 +1217,6 @@ int bch2_fs_journal_init(struct journal *j)
 	static struct lock_class_key res_key;
 	unsigned i;
 
-	mutex_init(&j->buf_lock);
 	spin_lock_init(&j->lock);
 	spin_lock_init(&j->err_lock);
 	init_waitqueue_head(&j->wait);
diff --git a/libbcachefs/journal.h b/libbcachefs/journal.h
index b5185f97..c85d01cf 100644
--- a/libbcachefs/journal.h
+++ b/libbcachefs/journal.h
@@ -259,7 +259,7 @@ static inline union journal_res_state journal_state_buf_put(struct journal *j, u
 {
 	union journal_res_state s;
 
-	s.v = atomic64_sub_return_release(((union journal_res_state) {
+	s.v = atomic64_sub_return(((union journal_res_state) {
 				    .buf0_count = idx == 0,
 				    .buf1_count = idx == 1,
 				    .buf2_count = idx == 2,
@@ -427,8 +427,6 @@ static inline void bch2_journal_set_replay_done(struct journal *j)
 
 void bch2_journal_unblock(struct journal *);
 void bch2_journal_block(struct journal *);
-void bch2_journal_block_reservations(struct journal *);
-struct journal_buf *bch2_next_write_buffer_flush_journal_buf(struct journal *j, u64 max_seq);
 
 void __bch2_journal_debug_to_text(struct printbuf *, struct journal *);
 void bch2_journal_debug_to_text(struct printbuf *, struct journal *);
diff --git a/libbcachefs/journal_io.c b/libbcachefs/journal_io.c
index c2a65523..fe4565fc 100644
--- a/libbcachefs/journal_io.c
+++ b/libbcachefs/journal_io.c
@@ -4,7 +4,6 @@
 #include "alloc_foreground.h"
 #include "btree_io.h"
 #include "btree_update_interior.h"
-#include "btree_write_buffer.h"
 #include "buckets.h"
 #include "checksum.h"
 #include "disk_groups.h"
@@ -722,22 +721,6 @@ static void journal_entry_overwrite_to_text(struct printbuf *out, struct bch_fs
 	journal_entry_btree_keys_to_text(out, c, entry);
 }
 
-static int journal_entry_write_buffer_keys_validate(struct bch_fs *c,
-				struct jset *jset,
-				struct jset_entry *entry,
-				unsigned version, int big_endian,
-				enum bkey_invalid_flags flags)
-{
-	return journal_entry_btree_keys_validate(c, jset, entry,
-				version, big_endian, READ);
-}
-
-static void journal_entry_write_buffer_keys_to_text(struct printbuf *out, struct bch_fs *c,
-					    struct jset_entry *entry)
-{
-	journal_entry_btree_keys_to_text(out, c, entry);
-}
-
 struct jset_entry_ops {
 	int (*validate)(struct bch_fs *, struct jset *,
 			struct jset_entry *, unsigned, int,
@@ -1518,8 +1501,6 @@ done:
 
 static void journal_buf_realloc(struct journal *j, struct journal_buf *buf)
 {
-	struct bch_fs *c = container_of(j, struct bch_fs, journal);
-
 	/* we aren't holding j->lock: */
 	unsigned new_size = READ_ONCE(j->buf_size_want);
 	void *new_buf;
@@ -1527,11 +1508,6 @@ static void journal_buf_realloc(struct journal *j, struct journal_buf *buf)
 	if (buf->buf_size >= new_size)
 		return;
 
-	size_t btree_write_buffer_size = new_size / 64;
-
-	if (bch2_btree_write_buffer_resize(c, btree_write_buffer_size))
-		return;
-
 	new_buf = kvpmalloc(new_size, GFP_NOFS|__GFP_NOWARN);
 	if (!new_buf)
 		return;
@@ -1621,7 +1597,6 @@ static CLOSURE_CALLBACK(journal_write_done)
 	} while ((v = atomic64_cmpxchg(&j->reservations.counter,
 				       old.v, new.v)) != old.v);
 
-	bch2_journal_reclaim_fast(j);
 	bch2_journal_space_available(j);
 
 	track_event_change(&c->times[BCH_TIME_blocked_journal_max_in_flight],
@@ -1725,11 +1700,9 @@ static int bch2_journal_write_prep(struct journal *j, struct journal_buf *w)
 	struct bch_fs *c = container_of(j, struct bch_fs, journal);
 	struct jset_entry *start, *end, *i, *next, *prev = NULL;
 	struct jset *jset = w->data;
-	struct journal_keys_to_wb wb = { NULL };
 	unsigned sectors, bytes, u64s;
-	unsigned long btree_roots_have = 0;
 	bool validate_before_checksum = false;
-	u64 seq = le64_to_cpu(jset->seq);
+	unsigned long btree_roots_have = 0;
 	int ret;
 
 	/*
@@ -1757,28 +1730,9 @@ static int bch2_journal_write_prep(struct journal *j, struct journal_buf *w)
 		 * to c->btree_roots we have to get any missing btree roots and
 		 * add them to this journal entry:
 		 */
-		switch (i->type) {
-		case BCH_JSET_ENTRY_btree_root:
+		if (i->type == BCH_JSET_ENTRY_btree_root) {
 			bch2_journal_entry_to_btree_root(c, i);
 			__set_bit(i->btree_id, &btree_roots_have);
-			break;
-		case BCH_JSET_ENTRY_write_buffer_keys:
-			EBUG_ON(!w->need_flush_to_write_buffer);
-
-			if (!wb.wb)
-				bch2_journal_keys_to_write_buffer_start(c, &wb, seq);
-
-			struct bkey_i *k;
-			jset_entry_for_each_key(i, k) {
-				ret = bch2_journal_key_to_wb(c, &wb, i->btree_id, k);
-				if (ret) {
-					bch2_fs_fatal_error(c, "-ENOMEM flushing journal keys to btree write buffer");
-					bch2_journal_keys_to_write_buffer_end(c, &wb);
-					return ret;
-				}
-			}
-			i->type = BCH_JSET_ENTRY_btree_keys;
-			break;
 		}
 
 		/* Can we merge with previous entry? */
@@ -1801,10 +1755,6 @@ static int bch2_journal_write_prep(struct journal *j, struct journal_buf *w)
 			memmove_u64s_down(prev, i, jset_u64s(u64s));
 	}
 
-	if (wb.wb)
-		bch2_journal_keys_to_write_buffer_end(c, &wb);
-	w->need_flush_to_write_buffer = false;
-
 	prev = prev ? vstruct_next(prev) : jset->start;
 	jset->u64s = cpu_to_le32((u64 *) prev - jset->_data);
 
@@ -1812,7 +1762,8 @@ static int bch2_journal_write_prep(struct journal *j, struct journal_buf *w)
 
 	end	= bch2_btree_roots_to_journal_entries(c, end, btree_roots_have);
 
-	bch2_journal_super_entries_add_common(c, &end, seq);
+	bch2_journal_super_entries_add_common(c, &end,
+				le64_to_cpu(jset->seq));
 	u64s	= (u64 *) end - (u64 *) start;
 	BUG_ON(u64s > j->entry_u64s_reserved);
 
@@ -1835,7 +1786,7 @@ static int bch2_journal_write_prep(struct journal *j, struct journal_buf *w)
 	SET_JSET_CSUM_TYPE(jset, bch2_meta_checksum_type(c));
 
 	if (!JSET_NO_FLUSH(jset) && journal_entry_empty(jset))
-		j->last_empty_seq = seq;
+		j->last_empty_seq = le64_to_cpu(jset->seq);
 
 	if (bch2_csum_type_is_encryption(JSET_CSUM_TYPE(jset)))
 		validate_before_checksum = true;
@@ -1931,11 +1882,9 @@ CLOSURE_CALLBACK(bch2_journal_write)
 	if (ret)
 		goto err;
 
-	mutex_lock(&j->buf_lock);
 	journal_buf_realloc(j, w);
 
 	ret = bch2_journal_write_prep(j, w);
-	mutex_unlock(&j->buf_lock);
 	if (ret)
 		goto err;
 
diff --git a/libbcachefs/journal_reclaim.c b/libbcachefs/journal_reclaim.c
index 2aa4c0c6..658aaa2c 100644
--- a/libbcachefs/journal_reclaim.c
+++ b/libbcachefs/journal_reclaim.c
@@ -3,7 +3,6 @@
 #include "bcachefs.h"
 #include "btree_key_cache.h"
 #include "btree_update.h"
-#include "btree_write_buffer.h"
 #include "buckets.h"
 #include "errcode.h"
 #include "error.h"
@@ -51,23 +50,20 @@ unsigned bch2_journal_dev_buckets_available(struct journal *j,
 	return available;
 }
 
-void bch2_journal_set_watermark(struct journal *j)
+static inline void journal_set_watermark(struct journal *j)
 {
 	struct bch_fs *c = container_of(j, struct bch_fs, journal);
 	bool low_on_space = j->space[journal_space_clean].total * 4 <=
 		j->space[journal_space_total].total;
 	bool low_on_pin = fifo_free(&j->pin) < j->pin.size / 4;
-	bool low_on_wb = bch2_btree_write_buffer_must_wait(c);
-	unsigned watermark = low_on_space || low_on_pin || low_on_wb
+	unsigned watermark = low_on_space || low_on_pin
 		? BCH_WATERMARK_reclaim
 		: BCH_WATERMARK_stripe;
 
 	if (track_event_change(&c->times[BCH_TIME_blocked_journal_low_on_space],
 			       &j->low_on_space_start, low_on_space) ||
 	    track_event_change(&c->times[BCH_TIME_blocked_journal_low_on_pin],
-			       &j->low_on_pin_start, low_on_pin) ||
-	    track_event_change(&c->times[BCH_TIME_blocked_write_buffer_full],
-			       &j->write_buffer_full_start, low_on_wb))
+			       &j->low_on_pin_start, low_on_pin))
 		trace_and_count(c, journal_full, c);
 
 	swap(watermark, j->watermark);
@@ -234,7 +230,7 @@ void bch2_journal_space_available(struct journal *j)
 	else
 		clear_bit(JOURNAL_MAY_SKIP_FLUSH, &j->flags);
 
-	bch2_journal_set_watermark(j);
+	journal_set_watermark(j);
 out:
 	j->cur_entry_sectors	= !ret ? j->space[journal_space_discarded].next_entry : 0;
 	j->cur_entry_error	= ret;
@@ -307,7 +303,6 @@ void bch2_journal_reclaim_fast(struct journal *j)
 	 * all btree nodes got written out
 	 */
 	while (!fifo_empty(&j->pin) &&
-	       j->pin.front <= j->seq_ondisk &&
 	       !atomic_read(&fifo_peek_front(&j->pin).count)) {
 		j->pin.front++;
 		popped = true;
diff --git a/libbcachefs/journal_reclaim.h b/libbcachefs/journal_reclaim.h
index ec84c334..7b15d682 100644
--- a/libbcachefs/journal_reclaim.h
+++ b/libbcachefs/journal_reclaim.h
@@ -16,7 +16,6 @@ static inline void journal_reclaim_kick(struct journal *j)
 unsigned bch2_journal_dev_buckets_available(struct journal *,
 					    struct journal_device *,
 					    enum journal_space_from);
-void bch2_journal_set_watermark(struct journal *);
 void bch2_journal_space_available(struct journal *);
 
 static inline bool journal_pin_active(struct journal_entry_pin *pin)
diff --git a/libbcachefs/journal_seq_blacklist.c b/libbcachefs/journal_seq_blacklist.c
index f9d9aa95..0200e299 100644
--- a/libbcachefs/journal_seq_blacklist.c
+++ b/libbcachefs/journal_seq_blacklist.c
@@ -267,7 +267,7 @@ retry:
 
 		while (!(ret = PTR_ERR_OR_ZERO(b)) &&
 		       b &&
-		       !test_bit(BCH_FS_STOPPING, &c->flags))
+		       !test_bit(BCH_FS_stopping, &c->flags))
 			b = bch2_btree_iter_next_node(&iter);
 
 		if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
diff --git a/libbcachefs/journal_types.h b/libbcachefs/journal_types.h
index 85c543af..2427cce6 100644
--- a/libbcachefs/journal_types.h
+++ b/libbcachefs/journal_types.h
@@ -36,7 +36,6 @@ struct journal_buf {
 	bool			noflush;	/* write has already been kicked off, and was noflush */
 	bool			must_flush;	/* something wants a flush */
 	bool			separate_flush;
-	bool			need_flush_to_write_buffer;
 };
 
 /*
@@ -183,12 +182,6 @@ struct journal {
 	darray_u64		early_journal_entries;
 
 	/*
-	 * Protects journal_buf->data, when accessing without a jorunal
-	 * reservation: for synchronization between the btree write buffer code
-	 * and the journal write path:
-	 */
-	struct mutex		buf_lock;
-	/*
 	 * Two journal entries -- one is currently open for new entries, the
 	 * other is possibly being written out.
 	 */
@@ -278,7 +271,6 @@ struct journal {
 	u64			low_on_space_start;
 	u64			low_on_pin_start;
 	u64			max_in_flight_start;
-	u64			write_buffer_full_start;
 
 	struct bch2_time_stats	*flush_write_time;
 	struct bch2_time_stats	*noflush_write_time;
diff --git a/libbcachefs/lru.c b/libbcachefs/lru.c
index 5340f2d0..82c08a98 100644
--- a/libbcachefs/lru.c
+++ b/libbcachefs/lru.c
@@ -123,9 +123,11 @@ int bch2_check_lru_key(struct btree_trans *trans,
 	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)) {
-			*last_flushed_pos = lru_k.k->p;
-			ret = bch2_btree_write_buffer_flush_sync(trans) ?:
-				-BCH_ERR_transaction_restart_write_buffer_flush;
+			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;
+			}
 			goto out;
 		}
 
diff --git a/libbcachefs/move.c b/libbcachefs/move.c
index c5518a86..db14ec37 100644
--- a/libbcachefs/move.c
+++ b/libbcachefs/move.c
@@ -56,17 +56,6 @@ static void trace_move_extent_read2(struct bch_fs *c, struct bkey_s_c k)
 	}
 }
 
-static void trace_move_extent_alloc_mem_fail2(struct bch_fs *c, struct bkey_s_c k)
-{
-	if (trace_move_extent_alloc_mem_fail_enabled()) {
-		struct printbuf buf = PRINTBUF;
-
-		bch2_bkey_val_to_text(&buf, c, k);
-		trace_move_extent_alloc_mem_fail(c, buf.buf);
-		printbuf_exit(&buf);
-	}
-}
-
 struct moving_io {
 	struct list_head		read_list;
 	struct list_head		io_list;
@@ -356,8 +345,16 @@ err:
 	if (ret == -BCH_ERR_data_update_done)
 		return 0;
 
-	this_cpu_inc(c->counters[BCH_COUNTER_move_extent_alloc_mem_fail]);
-	trace_move_extent_alloc_mem_fail2(c, k);
+	this_cpu_inc(c->counters[BCH_COUNTER_move_extent_start_fail]);
+	if (trace_move_extent_start_fail_enabled()) {
+		struct printbuf buf = PRINTBUF;
+
+		bch2_bkey_val_to_text(&buf, c, k);
+		prt_str(&buf, ": ");
+		prt_str(&buf, bch2_err_str(ret));
+		trace_move_extent_start_fail(c, buf.buf);
+		printbuf_exit(&buf);
+	}
 	return ret;
 }
 
diff --git a/libbcachefs/opts.h b/libbcachefs/opts.h
index b7f9990c..8526f177 100644
--- a/libbcachefs/opts.h
+++ b/libbcachefs/opts.h
@@ -233,6 +233,11 @@ enum fsck_err_opts {
 	  OPT_BOOL(),							\
 	  BCH2_NO_SB_OPT,		true,				\
 	  NULL,		"Stash pointer to in memory btree node in btree ptr")\
+	x(btree_write_buffer_size, u32,					\
+	  OPT_FS|OPT_MOUNT,						\
+	  OPT_UINT(16, (1U << 20) - 1),					\
+	  BCH2_NO_SB_OPT,		1U << 13,			\
+	  NULL,		"Number of btree write buffer entries")		\
 	x(gc_reserve_percent,		u8,				\
 	  OPT_FS|OPT_FORMAT|OPT_MOUNT|OPT_RUNTIME,			\
 	  OPT_UINT(5, 21),						\
diff --git a/libbcachefs/recovery.c b/libbcachefs/recovery.c
index 5f4f76e6..3f8c3ba1 100644
--- a/libbcachefs/recovery.c
+++ b/libbcachefs/recovery.c
@@ -530,7 +530,7 @@ static int bch2_set_may_go_rw(struct bch_fs *c)
 	move_gap(keys->d, keys->nr, keys->size, keys->gap, keys->nr);
 	keys->gap = keys->nr;
 
-	set_bit(BCH_FS_MAY_GO_RW, &c->flags);
+	set_bit(BCH_FS_may_go_rw, &c->flags);
 	if (keys->nr)
 		return bch2_fs_read_write_early(c);
 	return 0;
@@ -876,13 +876,13 @@ use_clean:
 
 	/* If we fixed errors, verify that fs is actually clean now: */
 	if (IS_ENABLED(CONFIG_BCACHEFS_DEBUG) &&
-	    test_bit(BCH_FS_ERRORS_FIXED, &c->flags) &&
-	    !test_bit(BCH_FS_ERRORS_NOT_FIXED, &c->flags) &&
-	    !test_bit(BCH_FS_ERROR, &c->flags)) {
+	    test_bit(BCH_FS_errors_fixed, &c->flags) &&
+	    !test_bit(BCH_FS_errors_not_fixed, &c->flags) &&
+	    !test_bit(BCH_FS_error, &c->flags)) {
 		bch2_flush_fsck_errs(c);
 
 		bch_info(c, "Fixed errors, running fsck a second time to verify fs is clean");
-		clear_bit(BCH_FS_ERRORS_FIXED, &c->flags);
+		clear_bit(BCH_FS_errors_fixed, &c->flags);
 
 		c->curr_recovery_pass = BCH_RECOVERY_PASS_check_alloc_info;
 
@@ -890,13 +890,13 @@ use_clean:
 		if (ret)
 			goto err;
 
-		if (test_bit(BCH_FS_ERRORS_FIXED, &c->flags) ||
-		    test_bit(BCH_FS_ERRORS_NOT_FIXED, &c->flags)) {
+		if (test_bit(BCH_FS_errors_fixed, &c->flags) ||
+		    test_bit(BCH_FS_errors_not_fixed, &c->flags)) {
 			bch_err(c, "Second fsck run was not clean");
-			set_bit(BCH_FS_ERRORS_NOT_FIXED, &c->flags);
+			set_bit(BCH_FS_errors_not_fixed, &c->flags);
 		}
 
-		set_bit(BCH_FS_ERRORS_FIXED, &c->flags);
+		set_bit(BCH_FS_errors_fixed, &c->flags);
 	}
 
 	if (enabled_qtypes(c)) {
@@ -913,14 +913,14 @@ use_clean:
 		write_sb = true;
 	}
 
-	if (!test_bit(BCH_FS_ERROR, &c->flags)) {
+	if (!test_bit(BCH_FS_error, &c->flags)) {
 		c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_alloc_info);
 		write_sb = true;
 	}
 
 	if (c->opts.fsck &&
-	    !test_bit(BCH_FS_ERROR, &c->flags) &&
-	    !test_bit(BCH_FS_ERRORS_NOT_FIXED, &c->flags)) {
+	    !test_bit(BCH_FS_error, &c->flags) &&
+	    !test_bit(BCH_FS_errors_not_fixed, &c->flags)) {
 		SET_BCH_SB_HAS_ERRORS(c->disk_sb.sb, 0);
 		SET_BCH_SB_HAS_TOPOLOGY_ERRORS(c->disk_sb.sb, 0);
 		write_sb = true;
@@ -954,7 +954,7 @@ use_clean:
 
 	ret = 0;
 out:
-	set_bit(BCH_FS_FSCK_DONE, &c->flags);
+	set_bit(BCH_FS_fsck_done, &c->flags);
 	bch2_flush_fsck_errs(c);
 
 	if (!c->opts.keep_journal &&
@@ -962,7 +962,7 @@ out:
 		bch2_journal_keys_put_initial(c);
 	kfree(clean);
 
-	if (!ret && test_bit(BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS, &c->flags)) {
+	if (!ret && test_bit(BCH_FS_need_delete_dead_snapshots, &c->flags)) {
 		bch2_fs_read_write_early(c);
 		bch2_delete_dead_snapshots_async(c);
 	}
@@ -1001,8 +1001,8 @@ int bch2_fs_initialize(struct bch_fs *c)
 	mutex_unlock(&c->sb_lock);
 
 	c->curr_recovery_pass = ARRAY_SIZE(recovery_pass_fns);
-	set_bit(BCH_FS_MAY_GO_RW, &c->flags);
-	set_bit(BCH_FS_FSCK_DONE, &c->flags);
+	set_bit(BCH_FS_may_go_rw, &c->flags);
+	set_bit(BCH_FS_fsck_done, &c->flags);
 
 	for (i = 0; i < BTREE_ID_NR; i++)
 		bch2_btree_root_alloc(c, i);
diff --git a/libbcachefs/snapshot.c b/libbcachefs/snapshot.c
index b23550b4..e473c788 100644
--- a/libbcachefs/snapshot.c
+++ b/libbcachefs/snapshot.c
@@ -318,7 +318,7 @@ int bch2_mark_snapshot(struct btree_trans *trans,
 		__set_is_ancestor_bitmap(c, id);
 
 		if (BCH_SNAPSHOT_DELETED(s.v)) {
-			set_bit(BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS, &c->flags);
+			set_bit(BCH_FS_need_delete_dead_snapshots, &c->flags);
 			if (c->curr_recovery_pass > BCH_RECOVERY_PASS_delete_dead_snapshots)
 				bch2_delete_dead_snapshots_async(c);
 		}
@@ -1376,10 +1376,10 @@ int bch2_delete_dead_snapshots(struct bch_fs *c)
 	u32 *i, id;
 	int ret = 0;
 
-	if (!test_and_clear_bit(BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS, &c->flags))
+	if (!test_and_clear_bit(BCH_FS_need_delete_dead_snapshots, &c->flags))
 		return 0;
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags)) {
+	if (!test_bit(BCH_FS_started, &c->flags)) {
 		ret = bch2_fs_read_write_early(c);
 		if (ret) {
 			bch_err_msg(c, ret, "deleting dead snapshots: error going rw");
@@ -1680,7 +1680,7 @@ static int bch2_check_snapshot_needs_deletion(struct btree_trans *trans, struct
 	if (BCH_SNAPSHOT_DELETED(snap.v) ||
 	    bch2_snapshot_equiv(c, k.k->p.offset) != k.k->p.offset ||
 	    (ret = bch2_snapshot_needs_delete(trans, k)) > 0) {
-		set_bit(BCH_FS_NEED_DELETE_DEAD_SNAPSHOTS, &c->flags);
+		set_bit(BCH_FS_need_delete_dead_snapshots, &c->flags);
 		return 0;
 	}
 
diff --git a/libbcachefs/super-io.c b/libbcachefs/super-io.c
index 512d5665..136c0140 100644
--- a/libbcachefs/super-io.c
+++ b/libbcachefs/super-io.c
@@ -916,9 +916,9 @@ int bch2_write_super(struct bch_fs *c)
 
 	le64_add_cpu(&c->disk_sb.sb->seq, 1);
 
-	if (test_bit(BCH_FS_ERROR, &c->flags))
+	if (test_bit(BCH_FS_error, &c->flags))
 		SET_BCH_SB_HAS_ERRORS(c->disk_sb.sb, 1);
-	if (test_bit(BCH_FS_TOPOLOGY_ERROR, &c->flags))
+	if (test_bit(BCH_FS_topology_error, &c->flags))
 		SET_BCH_SB_HAS_TOPOLOGY_ERRORS(c->disk_sb.sb, 1);
 
 	SET_BCH_SB_BIG_ENDIAN(c->disk_sb.sb, CPU_BIG_ENDIAN);
diff --git a/libbcachefs/super.c b/libbcachefs/super.c
index 552d55dd..e7f186b4 100644
--- a/libbcachefs/super.c
+++ b/libbcachefs/super.c
@@ -73,6 +73,13 @@ MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Kent Overstreet <kent.overstreet@gmail.com>");
 MODULE_DESCRIPTION("bcachefs filesystem");
 
+const char * const bch2_fs_flag_strs[] = {
+#define x(n)		#n,
+	BCH_FS_FLAGS()
+#undef x
+	NULL
+};
+
 #define KTYPE(type)							\
 static const struct attribute_group type ## _group = {			\
 	.attrs = type ## _files						\
@@ -240,8 +247,8 @@ static void __bch2_fs_read_only(struct bch_fs *c)
 		    journal_cur_seq(&c->journal));
 
 	if (test_bit(JOURNAL_REPLAY_DONE, &c->journal.flags) &&
-	    !test_bit(BCH_FS_EMERGENCY_RO, &c->flags))
-		set_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags);
+	    !test_bit(BCH_FS_emergency_ro, &c->flags))
+		set_bit(BCH_FS_clean_shutdown, &c->flags);
 	bch2_fs_journal_stop(&c->journal);
 
 	/*
@@ -256,19 +263,19 @@ static void bch2_writes_disabled(struct percpu_ref *writes)
 {
 	struct bch_fs *c = container_of(writes, struct bch_fs, writes);
 
-	set_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags);
+	set_bit(BCH_FS_write_disable_complete, &c->flags);
 	wake_up(&bch2_read_only_wait);
 }
 #endif
 
 void bch2_fs_read_only(struct bch_fs *c)
 {
-	if (!test_bit(BCH_FS_RW, &c->flags)) {
+	if (!test_bit(BCH_FS_rw, &c->flags)) {
 		bch2_journal_reclaim_stop(&c->journal);
 		return;
 	}
 
-	BUG_ON(test_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags));
+	BUG_ON(test_bit(BCH_FS_write_disable_complete, &c->flags));
 
 	bch_verbose(c, "going read-only");
 
@@ -276,7 +283,7 @@ void bch2_fs_read_only(struct bch_fs *c)
 	 * Block new foreground-end write operations from starting - any new
 	 * writes will return -EROFS:
 	 */
-	set_bit(BCH_FS_GOING_RO, &c->flags);
+	set_bit(BCH_FS_going_ro, &c->flags);
 #ifndef BCH_WRITE_REF_DEBUG
 	percpu_ref_kill(&c->writes);
 #else
@@ -296,36 +303,35 @@ void bch2_fs_read_only(struct bch_fs *c)
 	 * that going RO is complete:
 	 */
 	wait_event(bch2_read_only_wait,
-		   test_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags) ||
-		   test_bit(BCH_FS_EMERGENCY_RO, &c->flags));
+		   test_bit(BCH_FS_write_disable_complete, &c->flags) ||
+		   test_bit(BCH_FS_emergency_ro, &c->flags));
 
-	bool writes_disabled = test_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags);
+	bool writes_disabled = test_bit(BCH_FS_write_disable_complete, &c->flags);
 	if (writes_disabled)
 		bch_verbose(c, "finished waiting for writes to stop");
 
 	__bch2_fs_read_only(c);
 
 	wait_event(bch2_read_only_wait,
-		   test_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags));
+		   test_bit(BCH_FS_write_disable_complete, &c->flags));
 
 	if (!writes_disabled)
 		bch_verbose(c, "finished waiting for writes to stop");
 
-	clear_bit(BCH_FS_WRITE_DISABLE_COMPLETE, &c->flags);
-	clear_bit(BCH_FS_GOING_RO, &c->flags);
-	clear_bit(BCH_FS_RW, &c->flags);
+	clear_bit(BCH_FS_write_disable_complete, &c->flags);
+	clear_bit(BCH_FS_going_ro, &c->flags);
+	clear_bit(BCH_FS_rw, &c->flags);
 
 	if (!bch2_journal_error(&c->journal) &&
-	    !test_bit(BCH_FS_ERROR, &c->flags) &&
-	    !test_bit(BCH_FS_EMERGENCY_RO, &c->flags) &&
-	    test_bit(BCH_FS_STARTED, &c->flags) &&
-	    test_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags) &&
+	    !test_bit(BCH_FS_error, &c->flags) &&
+	    !test_bit(BCH_FS_emergency_ro, &c->flags) &&
+	    test_bit(BCH_FS_started, &c->flags) &&
+	    test_bit(BCH_FS_clean_shutdown, &c->flags) &&
 	    !c->opts.norecovery) {
 		BUG_ON(c->journal.last_empty_seq != journal_cur_seq(&c->journal));
 		BUG_ON(atomic_read(&c->btree_cache.dirty));
 		BUG_ON(atomic_long_read(&c->btree_key_cache.nr_dirty));
-		BUG_ON(c->btree_write_buffer.inc.keys.nr);
-		BUG_ON(c->btree_write_buffer.flushing.keys.nr);
+		BUG_ON(c->btree_write_buffer.state.nr);
 
 		bch_verbose(c, "marking filesystem clean");
 		bch2_fs_mark_clean(c);
@@ -351,7 +357,7 @@ static void bch2_fs_read_only_async(struct bch_fs *c)
 
 bool bch2_fs_emergency_read_only(struct bch_fs *c)
 {
-	bool ret = !test_and_set_bit(BCH_FS_EMERGENCY_RO, &c->flags);
+	bool ret = !test_and_set_bit(BCH_FS_emergency_ro, &c->flags);
 
 	bch2_journal_halt(&c->journal);
 	bch2_fs_read_only_async(c);
@@ -392,12 +398,12 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early)
 	unsigned i;
 	int ret;
 
-	if (test_bit(BCH_FS_INITIAL_GC_UNFIXED, &c->flags)) {
+	if (test_bit(BCH_FS_initial_gc_unfixed, &c->flags)) {
 		bch_err(c, "cannot go rw, unfixed btree errors");
 		return -BCH_ERR_erofs_unfixed_errors;
 	}
 
-	if (test_bit(BCH_FS_RW, &c->flags))
+	if (test_bit(BCH_FS_rw, &c->flags))
 		return 0;
 
 	if (c->opts.norecovery)
@@ -420,7 +426,7 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early)
 	if (ret)
 		goto err;
 
-	clear_bit(BCH_FS_CLEAN_SHUTDOWN, &c->flags);
+	clear_bit(BCH_FS_clean_shutdown, &c->flags);
 
 	/*
 	 * First journal write must be a flush write: after a clean shutdown we
@@ -434,8 +440,8 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early)
 		bch2_dev_allocator_add(c, ca);
 	bch2_recalc_capacity(c);
 
-	set_bit(BCH_FS_RW, &c->flags);
-	set_bit(BCH_FS_WAS_RW, &c->flags);
+	set_bit(BCH_FS_rw, &c->flags);
+	set_bit(BCH_FS_was_rw, &c->flags);
 
 #ifndef BCH_WRITE_REF_DEBUG
 	percpu_ref_reinit(&c->writes);
@@ -468,7 +474,7 @@ static int __bch2_fs_read_write(struct bch_fs *c, bool early)
 	bch2_do_pending_node_rewrites(c);
 	return 0;
 err:
-	if (test_bit(BCH_FS_RW, &c->flags))
+	if (test_bit(BCH_FS_rw, &c->flags))
 		bch2_fs_read_only(c);
 	else
 		__bch2_fs_read_only(c);
@@ -568,7 +574,7 @@ void __bch2_fs_stop(struct bch_fs *c)
 
 	bch_verbose(c, "shutting down");
 
-	set_bit(BCH_FS_STOPPING, &c->flags);
+	set_bit(BCH_FS_stopping, &c->flags);
 
 	cancel_work_sync(&c->journal_seq_blacklist_gc_work);
 
@@ -960,7 +966,7 @@ int bch2_fs_start(struct bch_fs *c)
 
 	down_write(&c->state_lock);
 
-	BUG_ON(test_bit(BCH_FS_STARTED, &c->flags));
+	BUG_ON(test_bit(BCH_FS_started, &c->flags));
 
 	mutex_lock(&c->sb_lock);
 
@@ -995,12 +1001,12 @@ int bch2_fs_start(struct bch_fs *c)
 		goto err;
 	}
 
-	set_bit(BCH_FS_STARTED, &c->flags);
+	set_bit(BCH_FS_started, &c->flags);
 
 	if (c->opts.read_only || c->opts.nochanges) {
 		bch2_fs_read_only(c);
 	} else {
-		ret = !test_bit(BCH_FS_RW, &c->flags)
+		ret = !test_bit(BCH_FS_rw, &c->flags)
 			? bch2_fs_read_write(c)
 			: bch2_fs_read_write_late(c);
 		if (ret)
diff --git a/libbcachefs/super.h b/libbcachefs/super.h
index bf762df1..dada0933 100644
--- a/libbcachefs/super.h
+++ b/libbcachefs/super.h
@@ -8,6 +8,8 @@
 
 #include <linux/math64.h>
 
+extern const char * const bch2_fs_flag_strs[];
+
 struct bch_fs *bch2_dev_to_fs(dev_t);
 struct bch_fs *bch2_uuid_to_fs(__uuid_t);
 
@@ -37,8 +39,8 @@ int bch2_fs_read_write_early(struct bch_fs *);
  */
 static inline void bch2_fs_lazy_rw(struct bch_fs *c)
 {
-	if (!test_bit(BCH_FS_RW, &c->flags) &&
-	    !test_bit(BCH_FS_WAS_RW, &c->flags))
+	if (!test_bit(BCH_FS_rw, &c->flags) &&
+	    !test_bit(BCH_FS_was_rw, &c->flags))
 		bch2_fs_read_write_early(c);
 }
 
diff --git a/libbcachefs/sysfs.c b/libbcachefs/sysfs.c
index 7223418d..4a7c93bc 100644
--- a/libbcachefs/sysfs.c
+++ b/libbcachefs/sysfs.c
@@ -145,6 +145,7 @@ rw_attribute(gc_gens_pos);
 
 read_attribute(uuid);
 read_attribute(minor);
+read_attribute(flags);
 read_attribute(bucket_size);
 read_attribute(first_bucket);
 read_attribute(nbuckets);
@@ -268,7 +269,7 @@ static int bch2_compression_stats_to_text(struct printbuf *out, struct bch_fs *c
 
 	memset(s, 0, sizeof(s));
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EPERM;
 
 	trans = bch2_trans_get(c);
@@ -384,6 +385,9 @@ SHOW(bch2_fs)
 	sysfs_print(minor,			c->minor);
 	sysfs_printf(internal_uuid, "%pU",	c->sb.uuid.b);
 
+	if (attr == &sysfs_flags)
+		prt_bitflags(out, bch2_fs_flag_strs, c->flags);
+
 	sysfs_hprint(btree_cache_size,		bch2_btree_cache_size(c));
 
 	if (attr == &sysfs_btree_write_stats)
@@ -416,7 +420,7 @@ SHOW(bch2_fs)
 		bch2_btree_updates_to_text(out, c);
 
 	if (attr == &sysfs_btree_cache)
-		bch2_btree_cache_to_text(out, &c->btree_cache);
+		bch2_btree_cache_to_text(out, c);
 
 	if (attr == &sysfs_btree_key_cache)
 		bch2_btree_key_cache_to_text(out, &c->btree_key_cache);
@@ -497,12 +501,12 @@ STORE(bch2_fs)
 
 	/* Debugging: */
 
-	if (!test_bit(BCH_FS_STARTED, &c->flags))
+	if (!test_bit(BCH_FS_started, &c->flags))
 		return -EPERM;
 
 	/* Debugging: */
 
-	if (!test_bit(BCH_FS_RW, &c->flags))
+	if (!test_bit(BCH_FS_rw, &c->flags))
 		return -EROFS;
 
 	if (attr == &sysfs_prune_cache) {
@@ -634,6 +638,7 @@ STORE(bch2_fs_internal)
 SYSFS_OPS(bch2_fs_internal);
 
 struct attribute *bch2_fs_internal_files[] = {
+	&sysfs_flags,
 	&sysfs_journal_debug,
 	&sysfs_btree_updates,
 	&sysfs_btree_cache,
diff --git a/libbcachefs/trace.h b/libbcachefs/trace.h
index 4980cfdd..6eced95c 100644
--- a/libbcachefs/trace.h
+++ b/libbcachefs/trace.h
@@ -775,9 +775,9 @@ TRACE_EVENT(move_extent_fail,
 	TP_printk("%d:%d %s", MAJOR(__entry->dev), MINOR(__entry->dev), __get_str(msg))
 );
 
-DEFINE_EVENT(str, move_extent_alloc_mem_fail,
-	TP_PROTO(struct bch_fs *c, const char *k),
-	TP_ARGS(c, k)
+DEFINE_EVENT(str, move_extent_start_fail,
+	TP_PROTO(struct bch_fs *c, const char *str),
+	TP_ARGS(c, str)
 );
 
 TRACE_EVENT(move_data,