md: initialize 'writes_pending' while allocating mddev

Currently 'writes_pending' is initialized in pers->run for raid1/5/10,
and it's freed while deleing mddev, instead of pers->free. pers->run can
be called multiple times before mddev is deleted, and a helper
mddev_init_writes_pending() is used to prevent 'writes_pending' to be
initialized multiple times, this usage is safe but a litter weird.

On the other hand, 'writes_pending' is only initialized for raid1/5/10,
however, it's used in common layer, for example:

array_state_store
set_in_sync
if (!mddev->in_sync) -> in_sync is used for all levels
// access writes_pending

There might be some implicit dependency that I don't recognized to make
sure 'writes_pending' can only be accessed for raid1/5/10, but there are
no comments about that.

By the way, it make sense to initialize 'writes_pending' in common layer
because there are already three levels use it.

Signed-off-by: Yu Kuai <[e-mail]>
Signed-off-by: Song Liu <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]

drivers/md/md.c

···
 	wake_up(&mddev->sb_wait);
 }
 
+static void no_op(struct percpu_ref *r) {}
+
 int mddev_init(struct mddev *mddev)
 {
 
 	if (percpu_ref_init(&mddev->active_io, active_io_release,
 			    PERCPU_REF_ALLOW_REINIT, GFP_KERNEL))
 		return -ENOMEM;
+
+	if (percpu_ref_init(&mddev->writes_pending, no_op,
+			    PERCPU_REF_ALLOW_REINIT, GFP_KERNEL)) {
+		percpu_ref_exit(&mddev->active_io);
+		return -ENOMEM;
+	}
+
+	/* We want to start with the refcount at zero */
+	percpu_ref_put(&mddev->writes_pending);
 
 	mutex_init(&mddev->open_mutex);
 	mutex_init(&mddev->reconfig_mutex);
···
 void mddev_destroy(struct mddev *mddev)
 {
 	percpu_ref_exit(&mddev->active_io);
+	percpu_ref_exit(&mddev->writes_pending);
 }
 EXPORT_SYMBOL_GPL(mddev_destroy);
 
···
 	kobject_put(&mddev->kobj);
 }
 
-static void no_op(struct percpu_ref *r) {}
-
-int mddev_init_writes_pending(struct mddev *mddev)
-{
-	if (mddev->writes_pending.percpu_count_ptr)
-		return 0;
-	if (percpu_ref_init(&mddev->writes_pending, no_op,
-			    PERCPU_REF_ALLOW_REINIT, GFP_KERNEL) < 0)
-		return -ENOMEM;
-	/* We want to start with the refcount at zero */
-	percpu_ref_put(&mddev->writes_pending);
-	return 0;
-}
-EXPORT_SYMBOL_GPL(mddev_init_writes_pending);
-
 struct mddev *md_alloc(dev_t dev, char *name)
 {
 	/*
···
 	 */
 	__md_stop_writes(mddev);
 	__md_stop(mddev);
-	percpu_ref_exit(&mddev->writes_pending);
 }
 
 EXPORT_SYMBOL_GPL(md_stop);
···
 {
 	struct mddev *mddev = disk->private_data;
 
-	percpu_ref_exit(&mddev->writes_pending);
 	mddev_free(mddev);
 }
 

drivers/md/md.h

···
 extern void md_wakeup_thread(struct md_thread __rcu *thread);
 extern void md_check_recovery(struct mddev *mddev);
 extern void md_reap_sync_thread(struct mddev *mddev);
-extern int mddev_init_writes_pending(struct mddev *mddev);
 extern bool md_write_start(struct mddev *mddev, struct bio *bi);
 extern void md_write_inc(struct mddev *mddev, struct bio *bi);
 extern void md_write_end(struct mddev *mddev);

drivers/md/raid1.c

···
 			mdname(mddev));
 		return -EIO;
 	}
-	if (mddev_init_writes_pending(mddev) < 0)
-		return -ENOMEM;
+
 	/*
 	 * copy the already verified devices into our private RAID1
 	 * bookkeeping area. [whatever we allocate in run(),

drivers/md/raid10.c

···
 	sector_t min_offset_diff = 0;
 	int first = 1;
 
-	if (mddev_init_writes_pending(mddev) < 0)
-		return -ENOMEM;
-
 	if (mddev->private == NULL) {
 		conf = setup_conf(mddev);
 		if (IS_ERR(conf))

drivers/md/raid5.c

···
 	long long min_offset_diff = 0;
 	int first = 1;
 
-	if (mddev_init_writes_pending(mddev) < 0)
-		return -ENOMEM;
-
 	if (mddev->recovery_cp != MaxSector)
 		pr_notice("md/raid:%s: not clean -- starting background reconstruction\n",
 			  mdname(mddev));