bcachefs: Add a mempool for the replicas delta list

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

fs/bcachefs/bcachefs.h

@@ -546,6 +546,8 @@ struct btree_iter_buf {
 	struct btree_iter	*iter;
 };
 
+#define REPLICAS_DELTA_LIST_MAX	(1U << 16)
+
 struct bch_fs {
 	struct closure		cl;
 
@@ -573,6 +575,7 @@ struct bch_fs {
 	struct bch_replicas_cpu replicas;
 	struct bch_replicas_cpu replicas_gc;
 	struct mutex		replicas_gc_lock;
+	mempool_t		replicas_delta_pool;
 
 	struct journal_entry_res btree_root_journal_res;
 	struct journal_entry_res replicas_journal_res;

fs/bcachefs/btree_iter.c

@@ -2337,7 +2337,14 @@ int bch2_trans_exit(struct btree_trans *trans)
 
 	bch2_journal_preres_put(&trans->c->journal, &trans->journal_preres);
 
+	if (trans->fs_usage_deltas) {
+		if (trans->fs_usage_deltas->size + sizeof(trans->fs_usage_deltas) ==
+		    REPLICAS_DELTA_LIST_MAX)
+			mempool_free(trans->fs_usage_deltas,
+				     &trans->c->replicas_delta_pool);
+		else
 			kfree(trans->fs_usage_deltas);
+	}
 
 	if (trans->mem_bytes == BTREE_TRANS_MEM_MAX)
 		mempool_free(trans->mem, &trans->c->btree_trans_mem_pool);

fs/bcachefs/buckets.c

@@ -560,10 +560,26 @@ replicas_deltas_realloc(struct btree_trans *trans, unsigned more)
 {
 	struct replicas_delta_list *d = trans->fs_usage_deltas;
 	unsigned new_size = d ? (d->size + more) * 2 : 128;
+	unsigned alloc_size = sizeof(*d) + new_size;
+
+	WARN_ON_ONCE(alloc_size > REPLICAS_DELTA_LIST_MAX);
 
 	if (!d || d->used + more > d->size) {
-		d = krealloc(d, sizeof(*d) + new_size, GFP_NOIO|__GFP_ZERO);
-		BUG_ON(!d);
+		d = krealloc(d, alloc_size, GFP_NOIO|__GFP_ZERO);
+
+		BUG_ON(!d && alloc_size > REPLICAS_DELTA_LIST_MAX);
+
+		if (!d) {
+			d = mempool_alloc(&trans->c->replicas_delta_pool, GFP_NOIO);
+			memset(d, 0, REPLICAS_DELTA_LIST_MAX);
+
+			if (trans->fs_usage_deltas)
+				memcpy(d, trans->fs_usage_deltas,
+				       trans->fs_usage_deltas->size + sizeof(*d));
+
+			new_size = REPLICAS_DELTA_LIST_MAX - sizeof(*d);
+			kfree(trans->fs_usage_deltas);
+		}
 
 		d->size = new_size;
 		trans->fs_usage_deltas = d;

fs/bcachefs/replicas.c

@@ -1033,11 +1033,27 @@ unsigned bch2_dev_has_data(struct bch_fs *c, struct bch_dev *ca)
 	return ret;
 }
 
+void bch2_fs_replicas_exit(struct bch_fs *c)
+{
+	unsigned i;
+
+	kfree(c->usage_scratch);
+	for (i = 0; i < ARRAY_SIZE(c->usage); i++)
+		free_percpu(c->usage[i]);
+	kfree(c->usage_base);
+	kfree(c->replicas.entries);
+	kfree(c->replicas_gc.entries);
+
+	mempool_exit(&c->replicas_delta_pool);
+}
+
 int bch2_fs_replicas_init(struct bch_fs *c)
 {
 	bch2_journal_entry_res_resize(&c->journal,
 			&c->replicas_journal_res,
 			reserve_journal_replicas(c, &c->replicas));
 
-	return replicas_table_update(c, &c->replicas);
+	return mempool_init_kmalloc_pool(&c->replicas_delta_pool, 1,
+					 REPLICAS_DELTA_LIST_MAX) ?:
+		replicas_table_update(c, &c->replicas);
 }

fs/bcachefs/replicas.h

@@ -77,6 +77,7 @@ int bch2_sb_replicas_to_cpu_replicas(struct bch_fs *);
 extern const struct bch_sb_field_ops bch_sb_field_ops_replicas;
 extern const struct bch_sb_field_ops bch_sb_field_ops_replicas_v0;
 
+void bch2_fs_replicas_exit(struct bch_fs *);
 int bch2_fs_replicas_init(struct bch_fs *);
 
 #endif /* _BCACHEFS_REPLICAS_H */

fs/bcachefs/super.c

@@ -476,6 +476,7 @@ static void __bch2_fs_free(struct bch_fs *c)
 	bch2_fs_btree_iter_exit(c);
 	bch2_fs_btree_key_cache_exit(&c->btree_key_cache);
 	bch2_fs_btree_cache_exit(c);
+	bch2_fs_replicas_exit(c);
 	bch2_fs_journal_exit(&c->journal);
 	bch2_io_clock_exit(&c->io_clock[WRITE]);
 	bch2_io_clock_exit(&c->io_clock[READ]);
@@ -484,10 +485,6 @@ static void __bch2_fs_free(struct bch_fs *c)
 	bch2_journal_entries_free(&c->journal_entries);
 	percpu_free_rwsem(&c->mark_lock);
 	free_percpu(c->online_reserved);
-	kfree(c->usage_scratch);
-	for (i = 0; i < ARRAY_SIZE(c->usage); i++)
-		free_percpu(c->usage[i]);
-	kfree(c->usage_base);
 
 	if (c->btree_iters_bufs)
 		for_each_possible_cpu(cpu)
@@ -500,8 +497,6 @@ static void __bch2_fs_free(struct bch_fs *c)
 	bioset_exit(&c->btree_bio);
 	mempool_exit(&c->fill_iter);
 	percpu_ref_exit(&c->writes);
-	kfree(c->replicas.entries);
-	kfree(c->replicas_gc.entries);
 	kfree(rcu_dereference_protected(c->disk_groups, 1));
 	kfree(c->journal_seq_blacklist_table);
 	kfree(c->unused_inode_hints);