Merge tag 'block-6.5-2023-07-03' of git://git.kernel.dk/linux

Pull more block updates from Jens Axboe:
 "Mostly items that came in a bit late for the initial pull request,
  wanted to make sure they had the appropriate amount of linux-next soak
  before going upstream.

  Outside of stragglers, just generic fixes for either merge window
  items, or longer standing bugs"

* tag 'block-6.5-2023-07-03' of git://git.kernel.dk/linux: (25 commits)
  md/raid0: add discard support for the 'original' layout
  nvme: disable controller on reset state failure
  nvme: sync timeout work on failed reset
  nvme: ensure unquiesce on teardown
  cdrom/gdrom: Fix build error
  nvme: improved uring polling
  block: add request polling helper
  nvme-mpath: fix I/O failure with EAGAIN when failing over I/O
  nvme: host: fix command name spelling
  blk-sysfs: add a new attr_group for blk_mq
  blk-iocost: move wbt_enable/disable_default() out of spinlock
  blk-wbt: cleanup rwb_enabled() and wbt_disabled()
  blk-wbt: remove dead code to handle wbt enable/disable with io inflight
  blk-wbt: don't create wbt sysfs entry if CONFIG_BLK_WBT is disabled
  blk-mq: fix two misuses on RQF_USE_SCHED
  blk-throttle: Fix io statistics for cgroup v1
  bcache: Fix bcache device claiming
  bcache: Alloc holder object before async registration
  raid10: avoid spin_lock from fastpath from raid10_unplug()
  md: fix 'delete_mutex' deadlock
  ...

7 files changed, 130 insertions, 100 deletions

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index e2a803683105..0ae2b3676293 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -1369,7 +1369,7 @@ static void cached_dev_free(struct closure *cl)
 		put_page(virt_to_page(dc->sb_disk));
 
 	if (!IS_ERR_OR_NULL(dc->bdev))
-		blkdev_put(dc->bdev, bcache_kobj);
+		blkdev_put(dc->bdev, dc);
 
 	wake_up(&unregister_wait);
 
@@ -1453,7 +1453,6 @@ static int register_bdev(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 
 	memcpy(&dc->sb, sb, sizeof(struct cache_sb));
 	dc->bdev = bdev;
-	dc->bdev->bd_holder = dc;
 	dc->sb_disk = sb_disk;
 
 	if (cached_dev_init(dc, sb->block_size << 9))
@@ -2218,7 +2217,7 @@ void bch_cache_release(struct kobject *kobj)
 		put_page(virt_to_page(ca->sb_disk));
 
 	if (!IS_ERR_OR_NULL(ca->bdev))
-		blkdev_put(ca->bdev, bcache_kobj);
+		blkdev_put(ca->bdev, ca);
 
 	kfree(ca);
 	module_put(THIS_MODULE);
@@ -2345,7 +2344,6 @@ static int register_cache(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 
 	memcpy(&ca->sb, sb, sizeof(struct cache_sb));
 	ca->bdev = bdev;
-	ca->bdev->bd_holder = ca;
 	ca->sb_disk = sb_disk;
 
 	if (bdev_max_discard_sectors((bdev)))
@@ -2359,7 +2357,7 @@ static int register_cache(struct cache_sb *sb, struct cache_sb_disk *sb_disk,
 		 * call blkdev_put() to bdev in bch_cache_release(). So we
 		 * explicitly call blkdev_put() here.
 		 */
-		blkdev_put(bdev, bcache_kobj);
+		blkdev_put(bdev, ca);
 		if (ret == -ENOMEM)
 			err = "cache_alloc(): -ENOMEM";
 		else if (ret == -EPERM)
@@ -2448,6 +2446,7 @@ struct async_reg_args {
 	struct cache_sb *sb;
 	struct cache_sb_disk *sb_disk;
 	struct block_device *bdev;
+	void *holder;
 };
 
 static void register_bdev_worker(struct work_struct *work)
@@ -2455,22 +2454,13 @@ static void register_bdev_worker(struct work_struct *work)
 	int fail = false;
 	struct async_reg_args *args =
 		container_of(work, struct async_reg_args, reg_work.work);
-	struct cached_dev *dc;
-
-	dc = kzalloc(sizeof(*dc), GFP_KERNEL);
-	if (!dc) {
-		fail = true;
-		put_page(virt_to_page(args->sb_disk));
-		blkdev_put(args->bdev, bcache_kobj);
-		goto out;
-	}
 
 	mutex_lock(&bch_register_lock);
-	if (register_bdev(args->sb, args->sb_disk, args->bdev, dc) < 0)
+	if (register_bdev(args->sb, args->sb_disk, args->bdev, args->holder)
+	    < 0)
 		fail = true;
 	mutex_unlock(&bch_register_lock);
 
-out:
 	if (fail)
 		pr_info("error %s: fail to register backing device\n",
 			args->path);
@@ -2485,21 +2475,11 @@ static void register_cache_worker(struct work_struct *work)
 	int fail = false;
 	struct async_reg_args *args =
 		container_of(work, struct async_reg_args, reg_work.work);
-	struct cache *ca;
-
-	ca = kzalloc(sizeof(*ca), GFP_KERNEL);
-	if (!ca) {
-		fail = true;
-		put_page(virt_to_page(args->sb_disk));
-		blkdev_put(args->bdev, bcache_kobj);
-		goto out;
-	}
 
 	/* blkdev_put() will be called in bch_cache_release() */
-	if (register_cache(args->sb, args->sb_disk, args->bdev, ca) != 0)
+	if (register_cache(args->sb, args->sb_disk, args->bdev, args->holder))
 		fail = true;
 
-out:
 	if (fail)
 		pr_info("error %s: fail to register cache device\n",
 			args->path);
@@ -2520,6 +2500,13 @@ static void register_device_async(struct async_reg_args *args)
 	queue_delayed_work(system_wq, &args->reg_work, 10);
 }
 
+static void *alloc_holder_object(struct cache_sb *sb)
+{
+	if (SB_IS_BDEV(sb))
+		return kzalloc(sizeof(struct cached_dev), GFP_KERNEL);
+	return kzalloc(sizeof(struct cache), GFP_KERNEL);
+}
+
 static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			       const char *buffer, size_t size)
 {
@@ -2527,9 +2514,11 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 	char *path = NULL;
 	struct cache_sb *sb;
 	struct cache_sb_disk *sb_disk;
-	struct block_device *bdev;
+	struct block_device *bdev, *bdev2;
+	void *holder = NULL;
 	ssize_t ret;
 	bool async_registration = false;
+	bool quiet = false;
 
 #ifdef CONFIG_BCACHE_ASYNC_REGISTRATION
 	async_registration = true;
@@ -2558,10 +2547,34 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 
 	ret = -EINVAL;
 	err = "failed to open device";
-	bdev = blkdev_get_by_path(strim(path), BLK_OPEN_READ | BLK_OPEN_WRITE,
-				  bcache_kobj, NULL);
+	bdev = blkdev_get_by_path(strim(path), BLK_OPEN_READ, NULL, NULL);
+	if (IS_ERR(bdev))
+		goto out_free_sb;
+
+	err = "failed to set blocksize";
+	if (set_blocksize(bdev, 4096))
+		goto out_blkdev_put;
+
+	err = read_super(sb, bdev, &sb_disk);
+	if (err)
+		goto out_blkdev_put;
+
+	holder = alloc_holder_object(sb);
+	if (!holder) {
+		ret = -ENOMEM;
+		err = "cannot allocate memory";
+		goto out_put_sb_page;
+	}
+
+	/* Now reopen in exclusive mode with proper holder */
+	bdev2 = blkdev_get_by_dev(bdev->bd_dev, BLK_OPEN_READ | BLK_OPEN_WRITE,
+				  holder, NULL);
+	blkdev_put(bdev, NULL);
+	bdev = bdev2;
 	if (IS_ERR(bdev)) {
-		if (bdev == ERR_PTR(-EBUSY)) {
+		ret = PTR_ERR(bdev);
+		bdev = NULL;
+		if (ret == -EBUSY) {
 			dev_t dev;
 
 			mutex_lock(&bch_register_lock);
@@ -2571,20 +2584,14 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 			else
 				err = "device busy";
 			mutex_unlock(&bch_register_lock);
-			if (attr == &ksysfs_register_quiet)
-				goto done;
+			if (attr == &ksysfs_register_quiet) {
+				quiet = true;
+				ret = size;
+			}
 		}
-		goto out_free_sb;
+		goto out_free_holder;
 	}
 
-	err = "failed to set blocksize";
-	if (set_blocksize(bdev, 4096))
-		goto out_blkdev_put;
-
-	err = read_super(sb, bdev, &sb_disk);
-	if (err)
-		goto out_blkdev_put;
-
 	err = "failed to register device";
 
 	if (async_registration) {
@@ -2595,59 +2602,46 @@ static ssize_t register_bcache(struct kobject *k, struct kobj_attribute *attr,
 		if (!args) {
 			ret = -ENOMEM;
 			err = "cannot allocate memory";
-			goto out_put_sb_page;
+			goto out_free_holder;
 		}
 
 		args->path	= path;
 		args->sb	= sb;
 		args->sb_disk	= sb_disk;
 		args->bdev	= bdev;
+		args->holder	= holder;
 		register_device_async(args);
 		/* No wait and returns to user space */
 		goto async_done;
 	}
 
 	if (SB_IS_BDEV(sb)) {
-		struct cached_dev *dc = kzalloc(sizeof(*dc), GFP_KERNEL);
-
-		if (!dc) {
-			ret = -ENOMEM;
-			err = "cannot allocate memory";
-			goto out_put_sb_page;
-		}
-
 		mutex_lock(&bch_register_lock);
-		ret = register_bdev(sb, sb_disk, bdev, dc);
+		ret = register_bdev(sb, sb_disk, bdev, holder);
 		mutex_unlock(&bch_register_lock);
 		/* blkdev_put() will be called in cached_dev_free() */
 		if (ret < 0)
 			goto out_free_sb;
 	} else {
-		struct cache *ca = kzalloc(sizeof(*ca), GFP_KERNEL);
-
-		if (!ca) {
-			ret = -ENOMEM;
-			err = "cannot allocate memory";
-			goto out_put_sb_page;
-		}
-
 		/* blkdev_put() will be called in bch_cache_release() */
-		ret = register_cache(sb, sb_disk, bdev, ca);
+		ret = register_cache(sb, sb_disk, bdev, holder);
 		if (ret)
 			goto out_free_sb;
 	}
 
-done:
 	kfree(sb);
 	kfree(path);
 	module_put(THIS_MODULE);
 async_done:
 	return size;
 
+out_free_holder:
+	kfree(holder);
 out_put_sb_page:
 	put_page(virt_to_page(sb_disk));
 out_blkdev_put:
-	blkdev_put(bdev, register_bcache);
+	if (bdev)
+		blkdev_put(bdev, holder);
 out_free_sb:
 	kfree(sb);
 out_free_path:
@@ -2656,7 +2650,8 @@ out_free_path:
 out_module_put:
 	module_put(THIS_MODULE);
 out:
-	pr_info("error %s: %s\n", path?path:"", err);
+	if (!quiet)
+		pr_info("error %s: %s\n", path?path:"", err);
 	return ret;
 }
 
diff --git a/drivers/md/md.c b/drivers/md/md.c
index cf3733c90c47..2e38ef421d69 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -643,7 +643,6 @@ void mddev_init(struct mddev *mddev)
 {
 	mutex_init(&mddev->open_mutex);
 	mutex_init(&mddev->reconfig_mutex);
-	mutex_init(&mddev->delete_mutex);
 	mutex_init(&mddev->bitmap_info.mutex);
 	INIT_LIST_HEAD(&mddev->disks);
 	INIT_LIST_HEAD(&mddev->all_mddevs);
@@ -749,26 +748,15 @@ static void mddev_free(struct mddev *mddev)
 
 static const struct attribute_group md_redundancy_group;
 
-static void md_free_rdev(struct mddev *mddev)
+void mddev_unlock(struct mddev *mddev)
 {
 	struct md_rdev *rdev;
 	struct md_rdev *tmp;
+	LIST_HEAD(delete);
 
-	mutex_lock(&mddev->delete_mutex);
-	if (list_empty(&mddev->deleting))
-		goto out;
-
-	list_for_each_entry_safe(rdev, tmp, &mddev->deleting, same_set) {
-		list_del_init(&rdev->same_set);
-		kobject_del(&rdev->kobj);
-		export_rdev(rdev, mddev);
-	}
-out:
-	mutex_unlock(&mddev->delete_mutex);
-}
+	if (!list_empty(&mddev->deleting))
+		list_splice_init(&mddev->deleting, &delete);
 
-void mddev_unlock(struct mddev *mddev)
-{
 	if (mddev->to_remove) {
 		/* These cannot be removed under reconfig_mutex as
 		 * an access to the files will try to take reconfig_mutex
@@ -808,7 +796,11 @@ void mddev_unlock(struct mddev *mddev)
 	} else
 		mutex_unlock(&mddev->reconfig_mutex);
 
-	md_free_rdev(mddev);
+	list_for_each_entry_safe(rdev, tmp, &delete, same_set) {
+		list_del_init(&rdev->same_set);
+		kobject_del(&rdev->kobj);
+		export_rdev(rdev, mddev);
+	}
 
 	md_wakeup_thread(mddev->thread);
 	wake_up(&mddev->sb_wait);
@@ -2458,7 +2450,7 @@ static void export_rdev(struct md_rdev *rdev, struct mddev *mddev)
 	if (test_bit(AutoDetected, &rdev->flags))
 		md_autodetect_dev(rdev->bdev->bd_dev);
 #endif
-	blkdev_put(rdev->bdev, mddev->major_version == -2 ? &claim_rdev : rdev);
+	blkdev_put(rdev->bdev, mddev->external ? &claim_rdev : rdev);
 	rdev->bdev = NULL;
 	kobject_put(&rdev->kobj);
 }
@@ -2488,9 +2480,7 @@ static void md_kick_rdev_from_array(struct md_rdev *rdev)
 	 * reconfig_mutex is held, hence it can't be called under
 	 * reconfig_mutex and it's delayed to mddev_unlock().
 	 */
-	mutex_lock(&mddev->delete_mutex);
 	list_add(&rdev->same_set, &mddev->deleting);
-	mutex_unlock(&mddev->delete_mutex);
 }
 
 static void export_array(struct mddev *mddev)
@@ -6140,7 +6130,7 @@ static void md_clean(struct mddev *mddev)
 	mddev->resync_min = 0;
 	mddev->resync_max = MaxSector;
 	mddev->reshape_position = MaxSector;
-	mddev->external = 0;
+	/* we still need mddev->external in export_rdev, do not clear it yet */
 	mddev->persistent = 0;
 	mddev->level = LEVEL_NONE;
 	mddev->clevel[0] = 0;
diff --git a/drivers/md/md.h b/drivers/md/md.h
index bfd2306bc750..1aef86bf3fc3 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -531,11 +531,9 @@ struct mddev {
 
 	/*
 	 * Temporarily store rdev that will be finally removed when
-	 * reconfig_mutex is unlocked.
+	 * reconfig_mutex is unlocked, protected by reconfig_mutex.
 	 */
 	struct list_head		deleting;
-	/* Protect the deleting list */
-	struct mutex			delete_mutex;
 
 	bool	has_superblocks:1;
 	bool	fail_last_dev:1;
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index f8ee9a95e25d..d1ac73fcd852 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -270,6 +270,18 @@ static int create_strip_zones(struct mddev *mddev, struct r0conf **private_conf)
 		goto abort;
 	}
 
+	if (conf->layout == RAID0_ORIG_LAYOUT) {
+		for (i = 1; i < conf->nr_strip_zones; i++) {
+			sector_t first_sector = conf->strip_zone[i-1].zone_end;
+
+			sector_div(first_sector, mddev->chunk_sectors);
+			zone = conf->strip_zone + i;
+			/* disk_shift is first disk index used in the zone */
+			zone->disk_shift = sector_div(first_sector,
+						      zone->nb_dev);
+		}
+	}
+
 	pr_debug("md/raid0:%s: done.\n", mdname(mddev));
 	*private_conf = conf;
 
@@ -431,6 +443,20 @@ exit_acct_set:
 	return ret;
 }
 
+/*
+ * Convert disk_index to the disk order in which it is read/written.
+ *  For example, if we have 4 disks, they are numbered 0,1,2,3. If we
+ *  write the disks starting at disk 3, then the read/write order would
+ *  be disk 3, then 0, then 1, and then disk 2 and we want map_disk_shift()
+ *  to map the disks as follows 0,1,2,3 => 1,2,3,0. So disk 0 would map
+ *  to 1, 1 to 2, 2 to 3, and 3 to 0. That way we can compare disks in
+ *  that 'output' space to understand the read/write disk ordering.
+ */
+static int map_disk_shift(int disk_index, int num_disks, int disk_shift)
+{
+	return ((disk_index + num_disks - disk_shift) % num_disks);
+}
+
 static void raid0_handle_discard(struct mddev *mddev, struct bio *bio)
 {
 	struct r0conf *conf = mddev->private;
@@ -444,7 +470,9 @@ static void raid0_handle_discard(struct mddev *mddev, struct bio *bio)
 	sector_t end_disk_offset;
 	unsigned int end_disk_index;
 	unsigned int disk;
+	sector_t orig_start, orig_end;
 
+	orig_start = start;
 	zone = find_zone(conf, &start);
 
 	if (bio_end_sector(bio) > zone->zone_end) {
@@ -458,6 +486,7 @@ static void raid0_handle_discard(struct mddev *mddev, struct bio *bio)
 	} else
 		end = bio_end_sector(bio);
 
+	orig_end = end;
 	if (zone != conf->strip_zone)
 		end = end - zone[-1].zone_end;
 
@@ -469,13 +498,26 @@ static void raid0_handle_discard(struct mddev *mddev, struct bio *bio)
 	last_stripe_index = end;
 	sector_div(last_stripe_index, stripe_size);
 
-	start_disk_index = (int)(start - first_stripe_index * stripe_size) /
-		mddev->chunk_sectors;
+	/* In the first zone the original and alternate layouts are the same */
+	if ((conf->layout == RAID0_ORIG_LAYOUT) && (zone != conf->strip_zone)) {
+		sector_div(orig_start, mddev->chunk_sectors);
+		start_disk_index = sector_div(orig_start, zone->nb_dev);
+		start_disk_index = map_disk_shift(start_disk_index,
+						  zone->nb_dev,
+						  zone->disk_shift);
+		sector_div(orig_end, mddev->chunk_sectors);
+		end_disk_index = sector_div(orig_end, zone->nb_dev);
+		end_disk_index = map_disk_shift(end_disk_index,
+						zone->nb_dev, zone->disk_shift);
+	} else {
+		start_disk_index = (int)(start - first_stripe_index * stripe_size) /
+			mddev->chunk_sectors;
+		end_disk_index = (int)(end - last_stripe_index * stripe_size) /
+			mddev->chunk_sectors;
+	}
 	start_disk_offset = ((int)(start - first_stripe_index * stripe_size) %
 		mddev->chunk_sectors) +
 		first_stripe_index * mddev->chunk_sectors;
-	end_disk_index = (int)(end - last_stripe_index * stripe_size) /
-		mddev->chunk_sectors;
 	end_disk_offset = ((int)(end - last_stripe_index * stripe_size) %
 		mddev->chunk_sectors) +
 		last_stripe_index * mddev->chunk_sectors;
@@ -483,18 +525,22 @@ static void raid0_handle_discard(struct mddev *mddev, struct bio *bio)
 	for (disk = 0; disk < zone->nb_dev; disk++) {
 		sector_t dev_start, dev_end;
 		struct md_rdev *rdev;
+		int compare_disk;
+
+		compare_disk = map_disk_shift(disk, zone->nb_dev,
+					      zone->disk_shift);
 
-		if (disk < start_disk_index)
+		if (compare_disk < start_disk_index)
 			dev_start = (first_stripe_index + 1) *
 				mddev->chunk_sectors;
-		else if (disk > start_disk_index)
+		else if (compare_disk > start_disk_index)
 			dev_start = first_stripe_index * mddev->chunk_sectors;
 		else
 			dev_start = start_disk_offset;
 
-		if (disk < end_disk_index)
+		if (compare_disk < end_disk_index)
 			dev_end = (last_stripe_index + 1) * mddev->chunk_sectors;
-		else if (disk > end_disk_index)
+		else if (compare_disk > end_disk_index)
 			dev_end = last_stripe_index * mddev->chunk_sectors;
 		else
 			dev_end = end_disk_offset;
diff --git a/drivers/md/raid0.h b/drivers/md/raid0.h
index 3816e5477db1..8cc761ca7423 100644
--- a/drivers/md/raid0.h
+++ b/drivers/md/raid0.h
@@ -6,6 +6,7 @@ struct strip_zone {
 	sector_t zone_end;	/* Start of the next zone (in sectors) */
 	sector_t dev_start;	/* Zone offset in real dev (in sectors) */
 	int	 nb_dev;	/* # of devices attached to the zone */
+	int	 disk_shift;	/* start disk for the original layout */
 };
 
 /* Linux 3.14 (20d0189b101) made an unintended change to
diff --git a/drivers/md/raid1-10.c b/drivers/md/raid1-10.c
index 169ebe296f2d..3f22edec70e7 100644
--- a/drivers/md/raid1-10.c
+++ b/drivers/md/raid1-10.c
@@ -116,7 +116,7 @@ static void md_bio_reset_resync_pages(struct bio *bio, struct resync_pages *rp,
 
 static inline void raid1_submit_write(struct bio *bio)
 {
-	struct md_rdev *rdev = (struct md_rdev *)bio->bi_bdev;
+	struct md_rdev *rdev = (void *)bio->bi_bdev;
 
 	bio->bi_next = NULL;
 	bio_set_dev(bio, rdev->bdev);
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index d0de8c9fb3cf..5051149e27bb 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -325,7 +325,7 @@ static void raid_end_bio_io(struct r10bio *r10_bio)
 	if (!test_bit(R10BIO_Uptodate, &r10_bio->state))
 		bio->bi_status = BLK_STS_IOERR;
 
-	if (blk_queue_io_stat(bio->bi_bdev->bd_disk->queue))
+	if (r10_bio->start_time)
 		bio_end_io_acct(bio, r10_bio->start_time);
 	bio_endio(bio);
 	/*
@@ -1118,7 +1118,7 @@ static void raid10_unplug(struct blk_plug_cb *cb, bool from_schedule)
 		spin_lock_irq(&conf->device_lock);
 		bio_list_merge(&conf->pending_bio_list, &plug->pending);
 		spin_unlock_irq(&conf->device_lock);
-		wake_up(&conf->wait_barrier);
+		wake_up_barrier(conf);
 		md_wakeup_thread(mddev->thread);
 		kfree(plug);
 		return;
@@ -1127,7 +1127,7 @@ static void raid10_unplug(struct blk_plug_cb *cb, bool from_schedule)
 	/* we aren't scheduling, so we can do the write-out directly. */
 	bio = bio_list_get(&plug->pending);
 	raid1_prepare_flush_writes(mddev->bitmap);
-	wake_up(&conf->wait_barrier);
+	wake_up_barrier(conf);
 
 	while (bio) { /* submit pending writes */
 		struct bio *next = bio->bi_next;