Merge tag 'for-6.1/block-2022-10-03' of git://git.kernel.dk/linux

Pull block updates from Jens Axboe:

- NVMe pull requests via Christoph:
- handle number of queue changes in the TCP and RDMA drivers
(Daniel Wagner)
- allow changing the number of queues in nvmet (Daniel Wagner)
- also consider host_iface when checking ip options (Daniel
Wagner)
- don't map pages which can't come from HIGHMEM (Fabio M. De
Francesco)
- avoid unnecessary flush bios in nvmet (Guixin Liu)
- shrink and better pack the nvme_iod structure (Keith Busch)
- add comment for unaligned "fake" nqn (Linjun Bao)
- print actual source IP address through sysfs "address" attr
(Martin Belanger)
- various cleanups (Jackie Liu, Wolfram Sang, Genjian Zhang)
- handle effects after freeing the request (Keith Busch)
- copy firmware_rev on each init (Keith Busch)
- restrict management ioctls to admin (Keith Busch)
- ensure subsystem reset is single threaded (Keith Busch)
- report the actual number of tagset maps in nvme-pci (Keith
Busch)
- small fabrics authentication fixups (Christoph Hellwig)
- add common code for tagset allocation and freeing (Christoph
Hellwig)
- stop using the request_queue in nvmet (Christoph Hellwig)
- set min_align_mask before calculating max_hw_sectors (Rishabh
Bhatnagar)
- send a rediscover uevent when a persistent discovery controller
reconnects (Sagi Grimberg)
- misc nvmet-tcp fixes (Varun Prakash, zhenwei pi)

- MD pull request via Song:
- Various raid5 fix and clean up, by Logan Gunthorpe and David
Sloan.
- Raid10 performance optimization, by Yu Kuai.

- sbitmap wakeup hang fixes (Hugh, Keith, Jan, Yu)

- IO scheduler switching quisce fix (Keith)

- s390/dasd block driver updates (Stefan)

- support for recovery for the ublk driver (ZiyangZhang)

- rnbd drivers fixes and updates (Guoqing, Santosh, ye, Christoph)

- blk-mq and null_blk map fixes (Bart)

- various bcache fixes (Coly, Jilin, Jules)

- nbd signal hang fix (Shigeru)

- block writeback throttling fix (Yu)

- optimize the passthrough mapping handling (me)

- prepare block cgroups to being gendisk based (Christoph)

- get rid of an old PSI hack in the block layer, moving it to the
callers instead where it belongs (Christoph)

- blk-throttle fixes and cleanups (Yu)

- misc fixes and cleanups (Liu Shixin, Liu Song, Miaohe, Pankaj,
Ping-Xiang, Wolfram, Saurabh, Li Jinlin, Li Lei, Lin, Li zeming,
Miaohe, Bart, Coly, Gaosheng

* tag 'for-6.1/block-2022-10-03' of git://git.kernel.dk/linux: (162 commits)
sbitmap: fix lockup while swapping
block: add rationale for not using blk_mq_plug() when applicable
block: adapt blk_mq_plug() to not plug for writes that require a zone lock
s390/dasd: use blk_mq_alloc_disk
blk-cgroup: don't update the blkg lookup hint in blkg_conf_prep
nvmet: don't look at the request_queue in nvmet_bdev_set_limits
nvmet: don't look at the request_queue in nvmet_bdev_zone_mgmt_emulate_all
blk-mq: use quiesced elevator switch when reinitializing queues
block: replace blk_queue_nowait with bdev_nowait
nvme: remove nvme_ctrl_init_connect_q
nvme-loop: use the tagset alloc/free helpers
nvme-loop: store the generic nvme_ctrl in set->driver_data
nvme-loop: initialize sqsize later
nvme-fc: use the tagset alloc/free helpers
nvme-fc: store the generic nvme_ctrl in set->driver_data
nvme-fc: keep ctrl->sqsize in sync with opts->queue_size
nvme-rdma: use the tagset alloc/free helpers
nvme-rdma: store the generic nvme_ctrl in set->driver_data
nvme-tcp: use the tagset alloc/free helpers
nvme-tcp: store the generic nvme_ctrl in set->driver_data
...

MAINTAINERS

···
 F:	include/linux/nvme*
 F:	include/uapi/linux/nvme_ioctl.h
 
+NVM EXPRESS FABRICS AUTHENTICATION
+M:	Hannes Reinecke <[e-mail]>
+L:	[e-mail]
+S:	Supported
+F:	drivers/nvme/host/auth.c
+F:	drivers/nvme/target/auth.c
+F:	drivers/nvme/target/fabrics-cmd-auth.c
+F:	include/linux/nvme-auth.h
+
 NVM EXPRESS FC TRANSPORT DRIVERS
 M:	James Smart <[e-mail]>
 L:	[e-mail]

arch/s390/include/asm/scsw.h

···
 #define SNS2_ENV_DATA_PRESENT	0x10
 #define SNS2_INPRECISE_END	0x04
 
+/*
+ * architectured values for PPRC errors
+ */
+#define SNS7_INVALID_ON_SEC	0x0e
+
 /**
  * scsw_is_tm - check for transport mode scsw
  * @scsw: pointer to scsw

arch/s390/include/uapi/asm/dasd.h

···
 } format_data_t;
 
 /*
+ * struct dasd_copypair_swap_data_t
+ * represents all data necessary to issue a swap of the copy pair relation
+ */
+struct dasd_copypair_swap_data_t {
+	char primary[20]; /* BUSID of primary */
+	char secondary[20]; /* BUSID of secondary */
+
+	/* Reserved for future updates. */
+	__u8 reserved[64];
+};
+
+/*
  * values to be used for format_data_t.intensity
  * 0/8: normal format
  * 1/9: also write record zero
···
 #define BIODASDSATTR   _IOW(DASD_IOCTL_LETTER,2,attrib_data_t)
 /* Release Allocated Space */
 #define BIODASDRAS     _IOW(DASD_IOCTL_LETTER, 3, format_data_t)
+/* Swap copy pair relation */
+#define BIODASDCOPYPAIRSWAP _IOW(DASD_IOCTL_LETTER, 4, struct dasd_copypair_swap_data_t)
 
 /* Get Sense Path Group ID (SNID) data */
 #define BIODASDSNID    _IOWR(DASD_IOCTL_LETTER, 1, struct dasd_snid_ioctl_data)

block/bfq-cgroup.c

···
 
 #else /* CONFIG_BFQ_CGROUP_DEBUG */
 
-void bfqg_stats_update_io_add(struct bfq_group *bfqg, struct bfq_queue *bfqq,
-			      blk_opf_t opf) { }
 void bfqg_stats_update_io_remove(struct bfq_group *bfqg, blk_opf_t opf) { }
 void bfqg_stats_update_io_merged(struct bfq_group *bfqg, blk_opf_t opf) { }
 void bfqg_stats_update_completion(struct bfq_group *bfqg, u64 start_time_ns,
 				  u64 io_start_time_ns, blk_opf_t opf) { }
 void bfqg_stats_update_dequeue(struct bfq_group *bfqg) { }
-void bfqg_stats_set_start_empty_time(struct bfq_group *bfqg) { }
-void bfqg_stats_update_idle_time(struct bfq_group *bfqg) { }
 void bfqg_stats_set_start_idle_time(struct bfq_group *bfqg) { }
-void bfqg_stats_update_avg_queue_size(struct bfq_group *bfqg) { }
 
 #endif /* CONFIG_BFQ_CGROUP_DEBUG */
 

block/bfq-iosched.c

···
 	bfqq->service_from_backlogged = 0;
 	bfq_clear_bfqq_softrt_update(bfqq);
 
-	bfq_add_bfqq_busy(bfqd, bfqq);
+	bfq_add_bfqq_busy(bfqq);
 
 	/*
 	 * Expire in-service queue if preemption may be needed for
···
 		bfqq->next_rq = NULL;
 
 		if (bfq_bfqq_busy(bfqq) && bfqq != bfqd->in_service_queue) {
-			bfq_del_bfqq_busy(bfqd, bfqq, false);
+			bfq_del_bfqq_busy(bfqq, false);
 			/*
 			 * bfqq emptied. In normal operation, when
 			 * bfqq is empty, bfqq->entity.service and
···
 	 */
 	if (bfq_bfqq_busy(bfqq) && RB_EMPTY_ROOT(&bfqq->sort_list) &&
 	    bfqq != bfqd->in_service_queue)
-		bfq_del_bfqq_busy(bfqd, bfqq, false);
+		bfq_del_bfqq_busy(bfqq, false);
 
 	bfq_reassign_last_bfqq(bfqq, NULL);
 
···
 			 */
 			bfqq->budget_timeout = jiffies;
 
-		bfq_del_bfqq_busy(bfqd, bfqq, true);
+		bfq_del_bfqq_busy(bfqq, true);
 	} else {
 		bfq_requeue_bfqq(bfqd, bfqq, true);
 		/*
···
 	struct hlist_node *n;
 	struct bfq_group *bfqg = bfqq_group(bfqq);
 
-	if (bfqq->bfqd)
-		bfq_log_bfqq(bfqq->bfqd, bfqq, "put_queue: %p %d",
-			     bfqq, bfqq->ref);
+	bfq_log_bfqq(bfqq->bfqd, bfqq, "put_queue: %p %d", bfqq, bfqq->ref);
 
 	bfqq->ref--;
 	if (bfqq->ref)
···
 		hlist_del_init(&item->woken_list_node);
 	}
 
-	if (bfqq->bfqd && bfqq->bfqd->last_completed_rq_bfqq == bfqq)
+	if (bfqq->bfqd->last_completed_rq_bfqq == bfqq)
 		bfqq->bfqd->last_completed_rq_bfqq = NULL;
 
 	kmem_cache_free(bfq_pool, bfqq);

block/bfq-iosched.h

···
 /* ---------------- cgroups-support interface ---------------- */
 
 void bfqg_stats_update_legacy_io(struct request_queue *q, struct request *rq);
-void bfqg_stats_update_io_add(struct bfq_group *bfqg, struct bfq_queue *bfqq,
-			      blk_opf_t opf);
 void bfqg_stats_update_io_remove(struct bfq_group *bfqg, blk_opf_t opf);
 void bfqg_stats_update_io_merged(struct bfq_group *bfqg, blk_opf_t opf);
 void bfqg_stats_update_completion(struct bfq_group *bfqg, u64 start_time_ns,
 				  u64 io_start_time_ns, blk_opf_t opf);
 void bfqg_stats_update_dequeue(struct bfq_group *bfqg);
-void bfqg_stats_set_start_empty_time(struct bfq_group *bfqg);
-void bfqg_stats_update_idle_time(struct bfq_group *bfqg);
 void bfqg_stats_set_start_idle_time(struct bfq_group *bfqg);
-void bfqg_stats_update_avg_queue_size(struct bfq_group *bfqg);
 void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 		   struct bfq_group *bfqg);
+
+#ifdef CONFIG_BFQ_CGROUP_DEBUG
+void bfqg_stats_update_io_add(struct bfq_group *bfqg, struct bfq_queue *bfqq,
+			      blk_opf_t opf);
+void bfqg_stats_set_start_empty_time(struct bfq_group *bfqg);
+void bfqg_stats_update_idle_time(struct bfq_group *bfqg);
+void bfqg_stats_update_avg_queue_size(struct bfq_group *bfqg);
+#endif
 
 void bfq_init_entity(struct bfq_entity *entity, struct bfq_group *bfqg);
 void bfq_bic_update_cgroup(struct bfq_io_cq *bic, struct bio *bio);
···
 void bfq_activate_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq);
 void bfq_requeue_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 		      bool expiration);
-void bfq_del_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq,
-		       bool expiration);
-void bfq_add_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq);
+void bfq_del_bfqq_busy(struct bfq_queue *bfqq, bool expiration);
+void bfq_add_bfqq_busy(struct bfq_queue *bfqq);
 
 /* --------------- end of interface of B-WF2Q+ ---------------- */
 

block/bfq-wf2q.c

···
  * the service tree. As a special case, it can be invoked during an
  * expiration.
  */
-void bfq_del_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq,
-		       bool expiration)
+void bfq_del_bfqq_busy(struct bfq_queue *bfqq, bool expiration)
 {
+	struct bfq_data *bfqd = bfqq->bfqd;
+
 	bfq_log_bfqq(bfqd, bfqq, "del from busy");
 
 	bfq_clear_bfqq_busy(bfqq);
···
 /*
  * Called when an inactive queue receives a new request.
  */
-void bfq_add_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq)
+void bfq_add_bfqq_busy(struct bfq_queue *bfqq)
 {
+	struct bfq_data *bfqd = bfqq->bfqd;
+
 	bfq_log_bfqq(bfqd, bfqq, "add to busy");
 
 	bfq_activate_bfqq(bfqd, bfqq);

block/bio.c

···
 static int __bio_clone(struct bio *bio, struct bio *bio_src, gfp_t gfp)
 {
 	bio_set_flag(bio, BIO_CLONED);
-	if (bio_flagged(bio_src, BIO_THROTTLED))
-		bio_set_flag(bio, BIO_THROTTLED);
 	bio->bi_ioprio = bio_src->bi_ioprio;
 	bio->bi_iter = bio_src->bi_iter;
 
···
 
 	bio->bi_iter.bi_size += len;
 	bio->bi_vcnt++;
-
-	if (!bio_flagged(bio, BIO_WORKINGSET) && unlikely(PageWorkingset(page)))
-		bio_set_flag(bio, BIO_WORKINGSET);
 }
 EXPORT_SYMBOL_GPL(__bio_add_page);
 
···
  * fit into the bio, or are requested in @iter, whatever is smaller. If
  * MM encounters an error pinning the requested pages, it stops. Error
  * is returned only if 0 pages could be pinned.
- *
- * It's intended for direct IO, so doesn't do PSI tracking, the caller is
- * responsible for setting BIO_WORKINGSET if necessary.
  */
 int bio_iov_iter_get_pages(struct bio *bio, struct iov_iter *iter)
 {
···
 		ret = __bio_iov_iter_get_pages(bio, iter);
 	} while (!ret && iov_iter_count(iter) && !bio_full(bio, 0));
 
-	/* don't account direct I/O as memory stall */
-	bio_clear_flag(bio, BIO_WORKINGSET);
 	return bio->bi_vcnt ? 0 : ret;
 }
 EXPORT_SYMBOL_GPL(bio_iov_iter_get_pages);
···
 	cpuhp_setup_state_multi(CPUHP_BIO_DEAD, "block/bio:dead", NULL,
 					bio_cpu_dead);
 
-	if (bioset_init(&fs_bio_set, BIO_POOL_SIZE, 0, BIOSET_NEED_BVECS))
+	if (bioset_init(&fs_bio_set, BIO_POOL_SIZE, 0,
+			BIOSET_NEED_BVECS | BIOSET_PERCPU_CACHE))
 		panic("bio: can't allocate bios\n");
 
 	if (bioset_integrity_create(&fs_bio_set, BIO_POOL_SIZE))

block/blk-cgroup.c

···
 /**
  * blkg_alloc - allocate a blkg
  * @blkcg: block cgroup the new blkg is associated with
- * @q: request_queue the new blkg is associated with
+ * @disk: gendisk the new blkg is associated with
  * @gfp_mask: allocation mask to use
  *
  * Allocate a new blkg assocating @blkcg and @q.
  */
-static struct blkcg_gq *blkg_alloc(struct blkcg *blkcg, struct request_queue *q,
+static struct blkcg_gq *blkg_alloc(struct blkcg *blkcg, struct gendisk *disk,
 				   gfp_t gfp_mask)
 {
 	struct blkcg_gq *blkg;
 	int i, cpu;
 
 	/* alloc and init base part */
-	blkg = kzalloc_node(sizeof(*blkg), gfp_mask, q->node);
+	blkg = kzalloc_node(sizeof(*blkg), gfp_mask, disk->queue->node);
 	if (!blkg)
 		return NULL;
 
···
 	if (!blkg->iostat_cpu)
 		goto err_free;
 
-	if (!blk_get_queue(q))
+	if (!blk_get_queue(disk->queue))
 		goto err_free;
 
-	blkg->q = q;
+	blkg->q = disk->queue;
 	INIT_LIST_HEAD(&blkg->q_node);
 	spin_lock_init(&blkg->async_bio_lock);
 	bio_list_init(&blkg->async_bios);
···
 		struct blkcg_policy *pol = blkcg_policy[i];
 		struct blkg_policy_data *pd;
 
-		if (!blkcg_policy_enabled(q, pol))
+		if (!blkcg_policy_enabled(disk->queue, pol))
 			continue;
 
 		/* alloc per-policy data and attach it to blkg */
-		pd = pol->pd_alloc_fn(gfp_mask, q, blkcg);
+		pd = pol->pd_alloc_fn(gfp_mask, disk->queue, blkcg);
 		if (!pd)
 			goto err_free;
 
···
 	return NULL;
 }
 
-struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg,
-				      struct request_queue *q, bool update_hint)
-{
-	struct blkcg_gq *blkg;
-
-	/*
-	 * Hint didn't match.  Look up from the radix tree.  Note that the
-	 * hint can only be updated under queue_lock as otherwise @blkg
-	 * could have already been removed from blkg_tree.  The caller is
-	 * responsible for grabbing queue_lock if @update_hint.
-	 */
-	blkg = radix_tree_lookup(&blkcg->blkg_tree, q->id);
-	if (blkg && blkg->q == q) {
-		if (update_hint) {
-			lockdep_assert_held(&q->queue_lock);
-			rcu_assign_pointer(blkcg->blkg_hint, blkg);
-		}
-		return blkg;
-	}
-
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(blkg_lookup_slowpath);
-
 /*
  * If @new_blkg is %NULL, this function tries to allocate a new one as
  * necessary using %GFP_NOWAIT.  @new_blkg is always consumed on return.
  */
-static struct blkcg_gq *blkg_create(struct blkcg *blkcg,
-				    struct request_queue *q,
+static struct blkcg_gq *blkg_create(struct blkcg *blkcg, struct gendisk *disk,
 				    struct blkcg_gq *new_blkg)
 {
 	struct blkcg_gq *blkg;
 	int i, ret;
 
-	lockdep_assert_held(&q->queue_lock);
+	lockdep_assert_held(&disk->queue->queue_lock);
 
 	/* request_queue is dying, do not create/recreate a blkg */
-	if (blk_queue_dying(q)) {
+	if (blk_queue_dying(disk->queue)) {
 		ret = -ENODEV;
 		goto err_free_blkg;
 	}
···
 
 	/* allocate */
 	if (!new_blkg) {
-		new_blkg = blkg_alloc(blkcg, q, GFP_NOWAIT | __GFP_NOWARN);
+		new_blkg = blkg_alloc(blkcg, disk, GFP_NOWAIT | __GFP_NOWARN);
 		if (unlikely(!new_blkg)) {
 			ret = -ENOMEM;
 			goto err_put_css;
···
 
 	/* link parent */
 	if (blkcg_parent(blkcg)) {
-		blkg->parent = __blkg_lookup(blkcg_parent(blkcg), q, false);
+		blkg->parent = blkg_lookup(blkcg_parent(blkcg), disk->queue);
 		if (WARN_ON_ONCE(!blkg->parent)) {
 			ret = -ENODEV;
 			goto err_put_css;
···
 
 	/* insert */
 	spin_lock(&blkcg->lock);
-	ret = radix_tree_insert(&blkcg->blkg_tree, q->id, blkg);
+	ret = radix_tree_insert(&blkcg->blkg_tree, disk->queue->id, blkg);
 	if (likely(!ret)) {
 		hlist_add_head_rcu(&blkg->blkcg_node, &blkcg->blkg_list);
-		list_add(&blkg->q_node, &q->blkg_list);
+		list_add(&blkg->q_node, &disk->queue->blkg_list);
 
 		for (i = 0; i < BLKCG_MAX_POLS; i++) {
 			struct blkcg_policy *pol = blkcg_policy[i];
···
 /**
  * blkg_lookup_create - lookup blkg, try to create one if not there
  * @blkcg: blkcg of interest
- * @q: request_queue of interest
+ * @disk: gendisk of interest
  *
- * Lookup blkg for the @blkcg - @q pair.  If it doesn't exist, try to
+ * Lookup blkg for the @blkcg - @disk pair.  If it doesn't exist, try to
  * create one.  blkg creation is performed recursively from blkcg_root such
  * that all non-root blkg's have access to the parent blkg.  This function
- * should be called under RCU read lock and takes @q->queue_lock.
+ * should be called under RCU read lock and takes @disk->queue->queue_lock.
  *
  * Returns the blkg or the closest blkg if blkg_create() fails as it walks
  * down from root.
  */
 static struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg,
-		struct request_queue *q)
+		struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct blkcg_gq *blkg;
 	unsigned long flags;
 
···
 		return blkg;
 
 	spin_lock_irqsave(&q->queue_lock, flags);
-	blkg = __blkg_lookup(blkcg, q, true);
-	if (blkg)
+	blkg = blkg_lookup(blkcg, q);
+	if (blkg) {
+		if (blkcg != &blkcg_root &&
+		    blkg != rcu_dereference(blkcg->blkg_hint))
+			rcu_assign_pointer(blkcg->blkg_hint, blkg);
 		goto found;
+	}
 
 	/*
 	 * Create blkgs walking down from blkcg_root to @blkcg, so that all
···
 		struct blkcg_gq *ret_blkg = q->root_blkg;
 
 		while (parent) {
-			blkg = __blkg_lookup(parent, q, false);
+			blkg = blkg_lookup(parent, q);
 			if (blkg) {
 				/* remember closest blkg */
 				ret_blkg = blkg;
···
 			parent = blkcg_parent(parent);
 		}
 
-		blkg = blkg_create(pos, q, NULL);
+		blkg = blkg_create(pos, disk, NULL);
 		if (IS_ERR(blkg)) {
 			blkg = ret_blkg;
 			break;
···
 	percpu_ref_kill(&blkg->refcnt);
 }
 
-/**
- * blkg_destroy_all - destroy all blkgs associated with a request_queue
- * @q: request_queue of interest
- *
- * Destroy all blkgs associated with @q.
- */
-static void blkg_destroy_all(struct request_queue *q)
+static void blkg_destroy_all(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct blkcg_gq *blkg, *n;
 	int count = BLKG_DESTROY_BATCH_SIZE;
 
···
 }
 EXPORT_SYMBOL_GPL(__blkg_prfill_u64);
 
-/* Performs queue bypass and policy enabled checks then looks up blkg. */
-static struct blkcg_gq *blkg_lookup_check(struct blkcg *blkcg,
-					  const struct blkcg_policy *pol,
-					  struct request_queue *q)
-{
-	WARN_ON_ONCE(!rcu_read_lock_held());
-	lockdep_assert_held(&q->queue_lock);
-
-	if (!blkcg_policy_enabled(q, pol))
-		return ERR_PTR(-EOPNOTSUPP);
-	return __blkg_lookup(blkcg, q, true /* update_hint */);
-}
-
 /**
  * blkcg_conf_open_bdev - parse and open bdev for per-blkg config update
  * @inputp: input string pointer
···
 	__acquires(rcu) __acquires(&bdev->bd_queue->queue_lock)
 {
 	struct block_device *bdev;
+	struct gendisk *disk;
 	struct request_queue *q;
 	struct blkcg_gq *blkg;
 	int ret;
···
 	bdev = blkcg_conf_open_bdev(&input);
 	if (IS_ERR(bdev))
 		return PTR_ERR(bdev);
-
-	q = bdev_get_queue(bdev);
+	disk = bdev->bd_disk;
+	q = disk->queue;
 
 	/*
 	 * blkcg_deactivate_policy() requires queue to be frozen, we can grab
···
 	rcu_read_lock();
 	spin_lock_irq(&q->queue_lock);
 
-	blkg = blkg_lookup_check(blkcg, pol, q);
-	if (IS_ERR(blkg)) {
-		ret = PTR_ERR(blkg);
+	if (!blkcg_policy_enabled(q, pol)) {
+		ret = -EOPNOTSUPP;
 		goto fail_unlock;
 	}
 
+	blkg = blkg_lookup(blkcg, q);
 	if (blkg)
 		goto success;
 
···
 		struct blkcg_gq *new_blkg;
 
 		parent = blkcg_parent(blkcg);
-		while (parent && !__blkg_lookup(parent, q, false)) {
+		while (parent && !blkg_lookup(parent, q)) {
 			pos = parent;
 			parent = blkcg_parent(parent);
 		}
···
 		spin_unlock_irq(&q->queue_lock);
 		rcu_read_unlock();
 
-		new_blkg = blkg_alloc(pos, q, GFP_KERNEL);
+		new_blkg = blkg_alloc(pos, disk, GFP_KERNEL);
 		if (unlikely(!new_blkg)) {
 			ret = -ENOMEM;
 			goto fail_exit_queue;
···
 		rcu_read_lock();
 		spin_lock_irq(&q->queue_lock);
 
-		blkg = blkg_lookup_check(pos, pol, q);
-		if (IS_ERR(blkg)) {
-			ret = PTR_ERR(blkg);
+		if (!blkcg_policy_enabled(q, pol)) {
 			blkg_free(new_blkg);
+			ret = -EOPNOTSUPP;
 			goto fail_preloaded;
 		}
 
+		blkg = blkg_lookup(pos, q);
 		if (blkg) {
 			blkg_free(new_blkg);
 		} else {
-			blkg = blkg_create(pos, q, new_blkg);
+			blkg = blkg_create(pos, disk, new_blkg);
 			if (IS_ERR(blkg)) {
 				ret = PTR_ERR(blkg);
 				goto fail_preloaded;
···
 	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
 	while ((dev = class_dev_iter_next(&iter))) {
 		struct block_device *bdev = dev_to_bdev(dev);
-		struct blkcg_gq *blkg =
-			blk_queue_root_blkg(bdev_get_queue(bdev));
+		struct blkcg_gq *blkg = bdev->bd_disk->queue->root_blkg;
 		struct blkg_iostat tmp;
 		int cpu;
 		unsigned long flags;
···
 	return 0;
 }
 
-/**
- * blkcg_init_queue - initialize blkcg part of request queue
- * @q: request_queue to initialize
- *
- * Called from blk_alloc_queue(). Responsible for initializing blkcg
- * part of new request_queue @q.
- *
- * RETURNS:
- * 0 on success, -errno on failure.
- */
-int blkcg_init_queue(struct request_queue *q)
+int blkcg_init_disk(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct blkcg_gq *new_blkg, *blkg;
 	bool preloaded;
 	int ret;
 
 	INIT_LIST_HEAD(&q->blkg_list);
 
-	new_blkg = blkg_alloc(&blkcg_root, q, GFP_KERNEL);
+	new_blkg = blkg_alloc(&blkcg_root, disk, GFP_KERNEL);
 	if (!new_blkg)
 		return -ENOMEM;
 
···
 	/* Make sure the root blkg exists. */
 	/* spin_lock_irq can serve as RCU read-side critical section. */
 	spin_lock_irq(&q->queue_lock);
-	blkg = blkg_create(&blkcg_root, q, new_blkg);
+	blkg = blkg_create(&blkcg_root, disk, new_blkg);
 	if (IS_ERR(blkg))
 		goto err_unlock;
 	q->root_blkg = blkg;
···
 	if (preloaded)
 		radix_tree_preload_end();
 
-	ret = blk_ioprio_init(q);
+	ret = blk_ioprio_init(disk);
 	if (ret)
 		goto err_destroy_all;
 
-	ret = blk_throtl_init(q);
+	ret = blk_throtl_init(disk);
 	if (ret)
-		goto err_destroy_all;
+		goto err_ioprio_exit;
 
-	ret = blk_iolatency_init(q);
-	if (ret) {
-		blk_throtl_exit(q);
-		blk_ioprio_exit(q);
-		goto err_destroy_all;
-	}
+	ret = blk_iolatency_init(disk);
+	if (ret)
+		goto err_throtl_exit;
 
 	return 0;
 
+err_throtl_exit:
+	blk_throtl_exit(disk);
+err_ioprio_exit:
+	blk_ioprio_exit(disk);
 err_destroy_all:
-	blkg_destroy_all(q);
+	blkg_destroy_all(disk);
 	return ret;
 err_unlock:
 	spin_unlock_irq(&q->queue_lock);
···
 	return PTR_ERR(blkg);
 }
 
-/**
- * blkcg_exit_queue - exit and release blkcg part of request_queue
- * @q: request_queue being released
- *
- * Called from blk_exit_queue().  Responsible for exiting blkcg part.
- */
-void blkcg_exit_queue(struct request_queue *q)
+void blkcg_exit_disk(struct gendisk *disk)
 {
-	blkg_destroy_all(q);
-	blk_throtl_exit(q);
+	blkg_destroy_all(disk);
+	blk_throtl_exit(disk);
 }
 
 static void blkcg_bind(struct cgroup_subsys_state *root_css)
···
 
 /**
  * blkcg_schedule_throttle - this task needs to check for throttling
- * @q: the request queue IO was submitted on
+ * @gendisk: disk to throttle
  * @use_memdelay: do we charge this to memory delay for PSI
  *
  * This is called by the IO controller when we know there's delay accumulated
  * for the blkg for this task.  We do not pass the blkg because there are places
  * we call this that may not have that information, the swapping code for
- * instance will only have a request_queue at that point.  This set's the
+ * instance will only have a block_device at that point.  This set's the
  * notify_resume for the task to check and see if it requires throttling before
  * returning to user space.
  *
···
  * throttle once.  If the task needs to be throttled again it'll need to be
  * re-set at the next time we see the task.
  */
-void blkcg_schedule_throttle(struct request_queue *q, bool use_memdelay)
+void blkcg_schedule_throttle(struct gendisk *disk, bool use_memdelay)
 {
+	struct request_queue *q = disk->queue;
+
 	if (unlikely(current->flags & PF_KTHREAD))
 		return;
 
···
 	struct blkcg_gq *blkg, *ret_blkg = NULL;
 
 	rcu_read_lock();
-	blkg = blkg_lookup_create(css_to_blkcg(css),
-				  bdev_get_queue(bio->bi_bdev));
+	blkg = blkg_lookup_create(css_to_blkcg(css), bio->bi_bdev->bd_disk);
 	while (blkg) {
 		if (blkg_tryget(blkg)) {
 			ret_blkg = blkg;

block/blk-cgroup.h

···
 extern struct blkcg blkcg_root;
 extern bool blkcg_debug_stats;
 
-struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg,
-				      struct request_queue *q, bool update_hint);
-int blkcg_init_queue(struct request_queue *q);
-void blkcg_exit_queue(struct request_queue *q);
+int blkcg_init_disk(struct gendisk *disk);
+void blkcg_exit_disk(struct gendisk *disk);
 
 /* Blkio controller policy registration */
 int blkcg_policy_register(struct blkcg_policy *pol);
···
 }
 
 /**
- * __blkg_lookup - internal version of blkg_lookup()
+ * blkg_lookup - lookup blkg for the specified blkcg - q pair
  * @blkcg: blkcg of interest
  * @q: request_queue of interest
- * @update_hint: whether to update lookup hint with the result or not
  *
- * This is internal version and shouldn't be used by policy
- * implementations.  Looks up blkgs for the @blkcg - @q pair regardless of
- * @q's bypass state.  If @update_hint is %true, the caller should be
- * holding @q->queue_lock and lookup hint is updated on success.
+ * Lookup blkg for the @blkcg - @q pair.
+
+ * Must be called in a RCU critical section.
  */
-static inline struct blkcg_gq *__blkg_lookup(struct blkcg *blkcg,
-					     struct request_queue *q,
-					     bool update_hint)
+static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg,
+					   struct request_queue *q)
 {
 	struct blkcg_gq *blkg;
+
+	WARN_ON_ONCE(!rcu_read_lock_held());
 
 	if (blkcg == &blkcg_root)
 		return q->root_blkg;
···
 	if (blkg && blkg->q == q)
 		return blkg;
 
-	return blkg_lookup_slowpath(blkcg, q, update_hint);
-}
-
-/**
- * blkg_lookup - lookup blkg for the specified blkcg - q pair
- * @blkcg: blkcg of interest
- * @q: request_queue of interest
- *
- * Lookup blkg for the @blkcg - @q pair.  This function should be called
- * under RCU read lock.
- */
-static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg,
-					   struct request_queue *q)
-{
-	WARN_ON_ONCE(!rcu_read_lock_held());
-	return __blkg_lookup(blkcg, q, false);
-}
-
-/**
- * blk_queue_root_blkg - return blkg for the (blkcg_root, @q) pair
- * @q: request_queue of interest
- *
- * Lookup blkg for @q at the root level. See also blkg_lookup().
- */
-static inline struct blkcg_gq *blk_queue_root_blkg(struct request_queue *q)
-{
-	return q->root_blkg;
+	blkg = radix_tree_lookup(&blkcg->blkg_tree, q->id);
+	if (blkg && blkg->q != q)
+		blkg = NULL;
+	return blkg;
 }
 
 /**
···
  */
 #define blkg_for_each_descendant_pre(d_blkg, pos_css, p_blkg)		\
 	css_for_each_descendant_pre((pos_css), &(p_blkg)->blkcg->css)	\
-		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
-					      (p_blkg)->q, false)))
+		if (((d_blkg) = blkg_lookup(css_to_blkcg(pos_css),	\
+					    (p_blkg)->q)))
 
 /**
  * blkg_for_each_descendant_post - post-order walk of a blkg's descendants
···
  */
 #define blkg_for_each_descendant_post(d_blkg, pos_css, p_blkg)		\
 	css_for_each_descendant_post((pos_css), &(p_blkg)->blkcg->css)	\
-		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
-					      (p_blkg)->q, false)))
+		if (((d_blkg) = blkg_lookup(css_to_blkcg(pos_css),	\
+					    (p_blkg)->q)))
 
 bool __blkcg_punt_bio_submit(struct bio *bio);
 
···
 };
 
 static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, void *key) { return NULL; }
-static inline struct blkcg_gq *blk_queue_root_blkg(struct request_queue *q)
-{ return NULL; }
-static inline int blkcg_init_queue(struct request_queue *q) { return 0; }
-static inline void blkcg_exit_queue(struct request_queue *q) { }
+static inline int blkcg_init_disk(struct gendisk *disk) { return 0; }
+static inline void blkcg_exit_disk(struct gendisk *disk) { }
 static inline int blkcg_policy_register(struct blkcg_policy *pol) { return 0; }
 static inline void blkcg_policy_unregister(struct blkcg_policy *pol) { }
 static inline int blkcg_activate_policy(struct request_queue *q,

block/blk-core.c

···
 #include <linux/t10-pi.h>
 #include <linux/debugfs.h>
 #include <linux/bpf.h>
-#include <linux/psi.h>
 #include <linux/part_stat.h>
 #include <linux/sched/sysctl.h>
 #include <linux/blk-crypto.h>
···
 late_initcall(fail_make_request_debugfs);
 #endif /* CONFIG_FAIL_MAKE_REQUEST */
 
-static inline bool bio_check_ro(struct bio *bio)
+static inline void bio_check_ro(struct bio *bio)
 {
 	if (op_is_write(bio_op(bio)) && bdev_read_only(bio->bi_bdev)) {
 		if (op_is_flush(bio->bi_opf) && !bio_sectors(bio))
-			return false;
+			return;
 		pr_warn("Trying to write to read-only block-device %pg\n",
 			bio->bi_bdev);
 		/* Older lvm-tools actually trigger this */
-		return false;
 	}
-
-	return false;
 }
 
 static noinline int should_fail_bio(struct bio *bio)
···
 	 * For a REQ_NOWAIT based request, return -EOPNOTSUPP
 	 * if queue does not support NOWAIT.
 	 */
-	if ((bio->bi_opf & REQ_NOWAIT) && !blk_queue_nowait(q))
+	if ((bio->bi_opf & REQ_NOWAIT) && !bdev_nowait(bdev))
 		goto not_supported;
 
 	if (should_fail_bio(bio))
 		goto end_io;
-	if (unlikely(bio_check_ro(bio)))
-		goto end_io;
+	bio_check_ro(bio);
 	if (!bio_flagged(bio, BIO_REMAPPED)) {
 		if (unlikely(bio_check_eod(bio)))
 			goto end_io;
···
  *
  * The success/failure status of the request, along with notification of
  * completion, is delivered asynchronously through the ->bi_end_io() callback
- * in @bio.  The bio must NOT be touched by thecaller until ->bi_end_io() has
+ * in @bio.  The bio must NOT be touched by the caller until ->bi_end_io() has
  * been called.
  */
 void submit_bio(struct bio *bio)
···
 		count_vm_events(PGPGIN, bio_sectors(bio));
 	} else if (bio_op(bio) == REQ_OP_WRITE) {
 		count_vm_events(PGPGOUT, bio_sectors(bio));
-	}
-
-	/*
-	 * If we're reading data that is part of the userspace workingset, count
-	 * submission time as memory stall.  When the device is congested, or
-	 * the submitting cgroup IO-throttled, submission can be a significant
-	 * part of overall IO time.
-	 */
-	if (unlikely(bio_op(bio) == REQ_OP_READ &&
-	    bio_flagged(bio, BIO_WORKINGSET))) {
-		unsigned long pflags;
-
-		psi_memstall_enter(&pflags);
-		submit_bio_noacct(bio);
-		psi_memstall_leave(&pflags);
-		return;
 	}
 
 	submit_bio_noacct(bio);
···
 	    !test_bit(QUEUE_FLAG_POLL, &q->queue_flags))
 		return 0;
 
+	/*
+	 * As the requests that require a zone lock are not plugged in the
+	 * first place, directly accessing the plug instead of using
+	 * blk_mq_plug() should not have any consequences during flushing for
+	 * zoned devices.
+	 */
 	blk_flush_plug(current->plug, false);
 
 	if (bio_queue_enter(bio))

block/blk-iocost.c

···
 	return rqos_to_ioc(rq_qos_id(q, RQ_QOS_COST));
 }
 
-static const char *q_name(struct request_queue *q)
-{
-	if (blk_queue_registered(q))
-		return kobject_name(q->kobj.parent);
-	else
-		return "<unknown>";
-}
-
 static const char __maybe_unused *ioc_name(struct ioc *ioc)
 {
-	return q_name(ioc->rqos.q);
+	struct gendisk *disk = ioc->rqos.q->disk;
+
+	if (!disk)
+		return "<unknown>";
+	return disk->disk_name;
 }
 
 static struct ioc_gq *pd_to_iocg(struct blkg_policy_data *pd)
···
 			int flags, void *key)
 {
 	struct iocg_wait *wait = container_of(wq_entry, struct iocg_wait, wait);
-	struct iocg_wake_ctx *ctx = (struct iocg_wake_ctx *)key;
+	struct iocg_wake_ctx *ctx = key;
 	u64 cost = abs_cost_to_cost(wait->abs_cost, ctx->hw_inuse);
 
 	ctx->vbudget -= cost;
···
 	if (use_debt) {
 		iocg_incur_debt(iocg, abs_cost, &now);
 		if (iocg_kick_delay(iocg, &now))
-			blkcg_schedule_throttle(rqos->q,
+			blkcg_schedule_throttle(rqos->q->disk,
 					(bio->bi_opf & REQ_SWAP) == REQ_SWAP);
 		iocg_unlock(iocg, ioc_locked, &flags);
 		return;
···
 	if (likely(!list_empty(&iocg->active_list))) {
 		iocg_incur_debt(iocg, abs_cost, &now);
 		if (iocg_kick_delay(iocg, &now))
-			blkcg_schedule_throttle(rqos->q,
+			blkcg_schedule_throttle(rqos->q->disk,
 					(bio->bi_opf & REQ_SWAP) == REQ_SWAP);
 	} else {
 		iocg_commit_bio(iocg, bio, abs_cost, cost);
···
 	.exit = ioc_rqos_exit,
 };
 
-static int blk_iocost_init(struct request_queue *q)
+static int blk_iocost_init(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct ioc *ioc;
 	struct rq_qos *rqos;
 	int i, cpu, ret;
···
 			     size_t nbytes, loff_t off)
 {
 	struct block_device *bdev;
+	struct gendisk *disk;
 	struct ioc *ioc;
 	u32 qos[NR_QOS_PARAMS];
 	bool enable, user;
···
 	if (IS_ERR(bdev))
 		return PTR_ERR(bdev);
 
-	ioc = q_to_ioc(bdev_get_queue(bdev));
+	disk = bdev->bd_disk;
+	ioc = q_to_ioc(disk->queue);
 	if (!ioc) {
-		ret = blk_iocost_init(bdev_get_queue(bdev));
+		ret = blk_iocost_init(disk);
 		if (ret)
 			goto err;
-		ioc = q_to_ioc(bdev_get_queue(bdev));
+		ioc = q_to_ioc(disk->queue);
 	}
 
 	spin_lock_irq(&ioc->lock);
···
 	spin_lock_irq(&ioc->lock);
 
 	if (enable) {
-		blk_stat_enable_accounting(ioc->rqos.q);
-		blk_queue_flag_set(QUEUE_FLAG_RQ_ALLOC_TIME, ioc->rqos.q);
+		blk_stat_enable_accounting(disk->queue);
+		blk_queue_flag_set(QUEUE_FLAG_RQ_ALLOC_TIME, disk->queue);
 		ioc->enabled = true;
 	} else {
-		blk_queue_flag_clear(QUEUE_FLAG_RQ_ALLOC_TIME, ioc->rqos.q);
+		blk_queue_flag_clear(QUEUE_FLAG_RQ_ALLOC_TIME, disk->queue);
 		ioc->enabled = false;
 	}
 
···
 
 	ioc = q_to_ioc(bdev_get_queue(bdev));
 	if (!ioc) {
-		ret = blk_iocost_init(bdev_get_queue(bdev));
+		ret = blk_iocost_init(bdev->bd_disk);
 		if (ret)
 			goto err;
 		ioc = q_to_ioc(bdev_get_queue(bdev));

block/blk-iolatency.c

···
 	unsigned use_delay = atomic_read(&lat_to_blkg(iolat)->use_delay);
 
 	if (use_delay)
-		blkcg_schedule_throttle(rqos->q, use_memdelay);
+		blkcg_schedule_throttle(rqos->q->disk, use_memdelay);
 
 	/*
 	 * To avoid priority inversions we want to just take a slot if we are
···
 	}
 }
 
-int blk_iolatency_init(struct request_queue *q)
+int blk_iolatency_init(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct blk_iolatency *blkiolat;
 	struct rq_qos *rqos;
 	int ret;

block/blk-ioprio.c

···
 		bio->bi_ioprio = prio;
 }
 
-void blk_ioprio_exit(struct request_queue *q)
+void blk_ioprio_exit(struct gendisk *disk)
 {
-	blkcg_deactivate_policy(q, &ioprio_policy);
+	blkcg_deactivate_policy(disk->queue, &ioprio_policy);
 }
 
-int blk_ioprio_init(struct request_queue *q)
+int blk_ioprio_init(struct gendisk *disk)
 {
-	return blkcg_activate_policy(q, &ioprio_policy);
+	return blkcg_activate_policy(disk->queue, &ioprio_policy);
 }
 
 static int __init ioprio_init(void)

block/blk-ioprio.h

···
 struct bio;
 
 #ifdef CONFIG_BLK_CGROUP_IOPRIO
-int blk_ioprio_init(struct request_queue *q);
-void blk_ioprio_exit(struct request_queue *q);
+int blk_ioprio_init(struct gendisk *disk);
+void blk_ioprio_exit(struct gendisk *disk);
 void blkcg_set_ioprio(struct bio *bio);
 #else
-static inline int blk_ioprio_init(struct request_queue *q)
+static inline int blk_ioprio_init(struct gendisk *disk)
 {
 	return 0;
 }
-static inline void blk_ioprio_exit(struct request_queue *q)
+static inline void blk_ioprio_exit(struct gendisk *disk)
 {
 }
 static inline void blkcg_set_ioprio(struct bio *bio)

block/blk-map.c

···
 	bio_init(bio, NULL, bio->bi_inline_vecs, nr_pages, req_op(rq));
 
 	if (map_data) {
-		nr_pages = 1 << map_data->page_order;
+		nr_pages = 1U << map_data->page_order;
 		i = map_data->offset / PAGE_SIZE;
 	}
 	while (len) {
···
 	return ret;
 }
 
+static void bio_map_put(struct bio *bio)
+{
+	if (bio->bi_opf & REQ_ALLOC_CACHE) {
+		bio_put(bio);
+	} else {
+		bio_uninit(bio);
+		kfree(bio);
+	}
+}
+
 static int bio_map_user_iov(struct request *rq, struct iov_iter *iter,
 		gfp_t gfp_mask)
 {
···
 	if (!iov_iter_count(iter))
 		return -EINVAL;
 
-	bio = bio_kmalloc(nr_vecs, gfp_mask);
-	if (!bio)
-		return -ENOMEM;
-	bio_init(bio, NULL, bio->bi_inline_vecs, nr_vecs, req_op(rq));
+	if (rq->cmd_flags & REQ_POLLED) {
+		blk_opf_t opf = rq->cmd_flags | REQ_ALLOC_CACHE;
+
+		bio = bio_alloc_bioset(NULL, nr_vecs, opf, gfp_mask,
+					&fs_bio_set);
+		if (!bio)
+			return -ENOMEM;
+	} else {
+		bio = bio_kmalloc(nr_vecs, gfp_mask);
+		if (!bio)
+			return -ENOMEM;
+		bio_init(bio, NULL, bio->bi_inline_vecs, nr_vecs, req_op(rq));
+	}
 
 	while (iov_iter_count(iter)) {
-		struct page **pages;
+		struct page **pages, *stack_pages[UIO_FASTIOV];
 		ssize_t bytes;
-		size_t offs, added = 0;
+		size_t offs;
 		int npages;
 
-		bytes = iov_iter_get_pages_alloc2(iter, &pages, LONG_MAX, &offs);
+		if (nr_vecs <= ARRAY_SIZE(stack_pages)) {
+			pages = stack_pages;
+			bytes = iov_iter_get_pages2(iter, pages, LONG_MAX,
+							nr_vecs, &offs);
+		} else {
+			bytes = iov_iter_get_pages_alloc2(iter, &pages,
+							LONG_MAX, &offs);
+		}
 		if (unlikely(bytes <= 0)) {
 			ret = bytes ? bytes : -EFAULT;
 			goto out_unmap;
···
 					break;
 				}
 
-				added += n;
 				bytes -= n;
 				offs = 0;
 			}
···
 		 */
 		while (j < npages)
 			put_page(pages[j++]);
-		kvfree(pages);
+		if (pages != stack_pages)
+			kvfree(pages);
 		/* couldn't stuff something into bio? */
 		if (bytes) {
 			iov_iter_revert(iter, bytes);
···
 
  out_unmap:
 	bio_release_pages(bio, false);
-	bio_uninit(bio);
-	kfree(bio);
+	bio_map_put(bio);
 	return ret;
 }
 
···
 
 		next_bio = bio;
 		bio = bio->bi_next;
-		bio_uninit(next_bio);
-		kfree(next_bio);
+		bio_map_put(next_bio);
 	}
 
 	return ret;

block/blk-mq-cpumap.c

···
 	return cpu;
 }
 
-int blk_mq_map_queues(struct blk_mq_queue_map *qmap)
+void blk_mq_map_queues(struct blk_mq_queue_map *qmap)
 {
 	unsigned int *map = qmap->mq_map;
 	unsigned int nr_queues = qmap->nr_queues;
···
 				map[cpu] = map[first_sibling];
 		}
 	}
-
-	return 0;
 }
 EXPORT_SYMBOL_GPL(blk_mq_map_queues);
 

block/blk-mq-debugfs.c

···
 		return "latency";
 	case RQ_QOS_COST:
 		return "cost";
-	case RQ_QOS_IOPRIO:
-		return "ioprio";
 	}
 	return "unknown";
 }

block/blk-mq-pci.c

···
  * that maps a queue to the CPUs that have irq affinity for the corresponding
  * vector.
  */
-int blk_mq_pci_map_queues(struct blk_mq_queue_map *qmap, struct pci_dev *pdev,
-			    int offset)
+void blk_mq_pci_map_queues(struct blk_mq_queue_map *qmap, struct pci_dev *pdev,
+			   int offset)
 {
 	const struct cpumask *mask;
 	unsigned int queue, cpu;
···
 			qmap->mq_map[cpu] = qmap->queue_offset + queue;
 	}
 
-	return 0;
+	return;
 
 fallback:
 	WARN_ON_ONCE(qmap->nr_queues > 1);
 	blk_mq_clear_mq_map(qmap);
-	return 0;
 }
 EXPORT_SYMBOL_GPL(blk_mq_pci_map_queues);

block/blk-mq-rdma.c

···
  * @set->nr_hw_queues, or @dev does not provide an affinity mask for a
  * vector, we fallback to the naive mapping.
  */
-int blk_mq_rdma_map_queues(struct blk_mq_queue_map *map,
+void blk_mq_rdma_map_queues(struct blk_mq_queue_map *map,
 		struct ib_device *dev, int first_vec)
 {
 	const struct cpumask *mask;
···
 			map->mq_map[cpu] = map->queue_offset + queue;
 	}
 
-	return 0;
+	return;
 
 fallback:
-	return blk_mq_map_queues(map);
+	blk_mq_map_queues(map);
 }
 EXPORT_SYMBOL_GPL(blk_mq_rdma_map_queues);

block/blk-mq-tag.c

···
 		 * other allocations on previous queue won't be starved.
 		 */
 		if (bt != bt_prev)
-			sbitmap_queue_wake_up(bt_prev);
+			sbitmap_queue_wake_up(bt_prev, 1);
 
 		ws = bt_wait_ptr(bt, data->hctx);
 	} while (1);

block/blk-mq-virtio.c

···
  * that maps a queue to the CPUs that have irq affinity for the corresponding
  * vector.
  */
-int blk_mq_virtio_map_queues(struct blk_mq_queue_map *qmap,
+void blk_mq_virtio_map_queues(struct blk_mq_queue_map *qmap,
 		struct virtio_device *vdev, int first_vec)
 {
 	const struct cpumask *mask;
···
 			qmap->mq_map[cpu] = qmap->queue_offset + queue;
 	}
 
-	return 0;
+	return;
+
 fallback:
-	return blk_mq_map_queues(qmap);
+	blk_mq_map_queues(qmap);
 }
 EXPORT_SYMBOL_GPL(blk_mq_virtio_map_queues);

block/blk-mq.c

···
 	WRITE_ONCE(rq->state, MQ_RQ_COMPLETE);
 
 	/*
-	 * For a polled request, always complete locally, it's pointless
-	 * to redirect the completion.
+	 * For request which hctx has only one ctx mapping,
+	 * or a polled request, always complete locally,
+	 * it's pointless to redirect the completion.
 	 */
-	if (rq->cmd_flags & REQ_POLLED)
+	if (rq->mq_hctx->nr_ctx == 1 ||
+		rq->cmd_flags & REQ_POLLED)
 		return false;
 
 	if (blk_mq_complete_need_ipi(rq)) {
···
 	WARN_ON(!blk_rq_is_passthrough(rq));
 
 	blk_account_io_start(rq);
+
+	/*
+	 * As plugging can be enabled for passthrough requests on a zoned
+	 * device, directly accessing the plug instead of using blk_mq_plug()
+	 * should not have any consequences.
+	 */
 	if (current->plug)
 		blk_add_rq_to_plug(current->plug, rq);
 	else
···
 		if (!needs_restart ||
 		    (no_tag && list_empty_careful(&hctx->dispatch_wait.entry)))
 			blk_mq_run_hw_queue(hctx, true);
-		else if (needs_restart && needs_resource)
+		else if (needs_resource)
 			blk_mq_delay_run_hw_queue(hctx, BLK_MQ_RESOURCE_DELAY);
 
 		blk_mq_update_dispatch_busy(hctx, true);
···
 	return 0;
 }
 
-static int blk_mq_update_queue_map(struct blk_mq_tag_set *set)
+static void blk_mq_update_queue_map(struct blk_mq_tag_set *set)
 {
 	/*
 	 * blk_mq_map_queues() and multiple .map_queues() implementations
···
 		for (i = 0; i < set->nr_maps; i++)
 			blk_mq_clear_mq_map(&set->map[i]);
 
-		return set->ops->map_queues(set);
+		set->ops->map_queues(set);
 	} else {
 		BUG_ON(set->nr_maps > 1);
-		return blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
+		blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
 	}
 }
 
···
 		set->map[i].nr_queues = is_kdump_kernel() ? 1 : set->nr_hw_queues;
 	}
 
-	ret = blk_mq_update_queue_map(set);
-	if (ret)
-		goto out_free_mq_map;
+	blk_mq_update_queue_map(set);
 
 	ret = blk_mq_alloc_set_map_and_rqs(set);
 	if (ret)
···
 	list_add(&qe->node, head);
 
 	/*
-	 * After elevator_switch_mq, the previous elevator_queue will be
+	 * After elevator_switch, the previous elevator_queue will be
 	 * released by elevator_release. The reference of the io scheduler
 	 * module get by elevator_get will also be put. So we need to get
 	 * a reference of the io scheduler module here to prevent it to be
 	 * removed.
 	 */
 	__module_get(qe->type->elevator_owner);
-	elevator_switch_mq(q, NULL);
+	elevator_switch(q, NULL);
 	mutex_unlock(&q->sysfs_lock);
 
 	return true;
···
 	kfree(qe);
 
 	mutex_lock(&q->sysfs_lock);
-	elevator_switch_mq(q, t);
+	elevator_switch(q, t);
 	mutex_unlock(&q->sysfs_lock);
 }
 

block/blk-mq.h

···
 static inline struct blk_plug *blk_mq_plug( struct bio *bio)
 {
 	/* Zoned block device write operation case: do not plug the BIO */
-	if (bdev_is_zoned(bio->bi_bdev) && op_is_write(bio_op(bio)))
+	if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
+	    bdev_op_is_zoned_write(bio->bi_bdev, bio_op(bio)))
 		return NULL;
 
 	/*

block/blk-rq-qos.h

···
 	RQ_QOS_WBT,
 	RQ_QOS_LATENCY,
 	RQ_QOS_COST,
-	RQ_QOS_IOPRIO,
 };
 
 struct rq_wait {

block/blk-sysfs.c

···
 
 	blk_queue_flag_set(QUEUE_FLAG_REGISTERED, q);
 	wbt_enable_default(q);
-	blk_throtl_register_queue(q);
+	blk_throtl_register(disk);
 
 	/* Now everything is ready and send out KOBJ_ADD uevent */
 	kobject_uevent(&q->kobj, KOBJ_ADD);

block/blk-throttle.c

···
 /* init a service_queue, assumes the caller zeroed it */
 static void throtl_service_queue_init(struct throtl_service_queue *sq)
 {
-	INIT_LIST_HEAD(&sq->queued[0]);
-	INIT_LIST_HEAD(&sq->queued[1]);
+	INIT_LIST_HEAD(&sq->queued[READ]);
+	INIT_LIST_HEAD(&sq->queued[WRITE]);
 	sq->pending_tree = RB_ROOT_CACHED;
 	timer_setup(&sq->pending_timer, throtl_pending_timer_fn, 0);
 }
···
 	struct throtl_grp *parent_tg = sq_to_tg(tg->service_queue.parent_sq);
 	struct throtl_data *td = tg->td;
 	int rw;
-	int has_iops_limit = 0;
 
 	for (rw = READ; rw <= WRITE; rw++) {
-		unsigned int iops_limit = tg_iops_limit(tg, rw);
-
-		tg->has_rules[rw] = (parent_tg && parent_tg->has_rules[rw]) ||
+		tg->has_rules_iops[rw] =
+			(parent_tg && parent_tg->has_rules_iops[rw]) ||
 			(td->limit_valid[td->limit_index] &&
-			 (tg_bps_limit(tg, rw) != U64_MAX ||
-			  iops_limit != UINT_MAX));
-
-		if (iops_limit != UINT_MAX)
-			has_iops_limit = 1;
+			  tg_iops_limit(tg, rw) != UINT_MAX);
+		tg->has_rules_bps[rw] =
+			(parent_tg && parent_tg->has_rules_bps[rw]) ||
+			(td->limit_valid[td->limit_index] &&
+			 (tg_bps_limit(tg, rw) != U64_MAX));
 	}
-
-	if (has_iops_limit)
-		tg->flags |= THROTL_TG_HAS_IOPS_LIMIT;
-	else
-		tg->flags &= ~THROTL_TG_HAS_IOPS_LIMIT;
 }
 
 static void throtl_pd_online(struct blkg_policy_data *pd)
···
 {
 	rb_erase_cached(n, &parent_sq->pending_tree);
 	RB_CLEAR_NODE(n);
-	--parent_sq->nr_pending;
 }
 
 static void update_min_dispatch_time(struct throtl_service_queue *parent_sq)
···
 static void throtl_dequeue_tg(struct throtl_grp *tg)
 {
 	if (tg->flags & THROTL_TG_PENDING) {
-		throtl_rb_erase(&tg->rb_node, tg->service_queue.parent_sq);
+		struct throtl_service_queue *parent_sq =
+			tg->service_queue.parent_sq;
+
+		throtl_rb_erase(&tg->rb_node, parent_sq);
+		--parent_sq->nr_pending;
 		tg->flags &= ~THROTL_TG_PENDING;
 	}
 }
···
 {
 	tg->bytes_disp[rw] = 0;
 	tg->io_disp[rw] = 0;
+	tg->carryover_bytes[rw] = 0;
+	tg->carryover_ios[rw] = 0;
 
 	/*
 	 * Previous slice has expired. We must have trimmed it after last
···
 		   tg->slice_end[rw], jiffies);
 }
 
-static inline void throtl_start_new_slice(struct throtl_grp *tg, bool rw)
+static inline void throtl_start_new_slice(struct throtl_grp *tg, bool rw,
+					  bool clear_carryover)
 {
 	tg->bytes_disp[rw] = 0;
 	tg->io_disp[rw] = 0;
 	tg->slice_start[rw] = jiffies;
 	tg->slice_end[rw] = jiffies + tg->td->throtl_slice;
+	if (clear_carryover) {
+		tg->carryover_bytes[rw] = 0;
+		tg->carryover_ios[rw] = 0;
+	}
 
 	throtl_log(&tg->service_queue,
 		   "[%c] new slice start=%lu end=%lu jiffies=%lu",
···
 		   tg->slice_start[rw], tg->slice_end[rw], jiffies);
 }
 
-static bool tg_with_in_iops_limit(struct throtl_grp *tg, struct bio *bio,
-				  u32 iops_limit, unsigned long *wait)
+static unsigned int calculate_io_allowed(u32 iops_limit,
+					 unsigned long jiffy_elapsed)
+{
+	unsigned int io_allowed;
+	u64 tmp;
+
+	/*
+	 * jiffy_elapsed should not be a big value as minimum iops can be
+	 * 1 then at max jiffy elapsed should be equivalent of 1 second as we
+	 * will allow dispatch after 1 second and after that slice should
+	 * have been trimmed.
+	 */
+
+	tmp = (u64)iops_limit * jiffy_elapsed;
+	do_div(tmp, HZ);
+
+	if (tmp > UINT_MAX)
+		io_allowed = UINT_MAX;
+	else
+		io_allowed = tmp;
+
+	return io_allowed;
+}
+
+static u64 calculate_bytes_allowed(u64 bps_limit, unsigned long jiffy_elapsed)
+{
+	return mul_u64_u64_div_u64(bps_limit, (u64)jiffy_elapsed, (u64)HZ);
+}
+
+static void __tg_update_carryover(struct throtl_grp *tg, bool rw)
+{
+	unsigned long jiffy_elapsed = jiffies - tg->slice_start[rw];
+	u64 bps_limit = tg_bps_limit(tg, rw);
+	u32 iops_limit = tg_iops_limit(tg, rw);
+
+	/*
+	 * If config is updated while bios are still throttled, calculate and
+	 * accumulate how many bytes/ios are waited across changes. And
+	 * carryover_bytes/ios will be used to calculate new wait time under new
+	 * configuration.
+	 */
+	if (bps_limit != U64_MAX)
+		tg->carryover_bytes[rw] +=
+			calculate_bytes_allowed(bps_limit, jiffy_elapsed) -
+			tg->bytes_disp[rw];
+	if (iops_limit != UINT_MAX)
+		tg->carryover_ios[rw] +=
+			calculate_io_allowed(iops_limit, jiffy_elapsed) -
+			tg->io_disp[rw];
+}
+
+static void tg_update_carryover(struct throtl_grp *tg)
+{
+	if (tg->service_queue.nr_queued[READ])
+		__tg_update_carryover(tg, READ);
+	if (tg->service_queue.nr_queued[WRITE])
+		__tg_update_carryover(tg, WRITE);
+
+	/* see comments in struct throtl_grp for meaning of these fields. */
+	throtl_log(&tg->service_queue, "%s: %llu %llu %u %u\n", __func__,
+		   tg->carryover_bytes[READ], tg->carryover_bytes[WRITE],
+		   tg->carryover_ios[READ], tg->carryover_ios[WRITE]);
+}
+
+static bool tg_within_iops_limit(struct throtl_grp *tg, struct bio *bio,
+				 u32 iops_limit, unsigned long *wait)
 {
 	bool rw = bio_data_dir(bio);
 	unsigned int io_allowed;
 	unsigned long jiffy_elapsed, jiffy_wait, jiffy_elapsed_rnd;
-	u64 tmp;
 
 	if (iops_limit == UINT_MAX) {
 		if (wait)
···
 
 	/* Round up to the next throttle slice, wait time must be nonzero */
 	jiffy_elapsed_rnd = roundup(jiffy_elapsed + 1, tg->td->throtl_slice);
-
-	/*
-	 * jiffy_elapsed_rnd should not be a big value as minimum iops can be
-	 * 1 then at max jiffy elapsed should be equivalent of 1 second as we
-	 * will allow dispatch after 1 second and after that slice should
-	 * have been trimmed.
-	 */
-
-	tmp = (u64)iops_limit * jiffy_elapsed_rnd;
-	do_div(tmp, HZ);
-
-	if (tmp > UINT_MAX)
-		io_allowed = UINT_MAX;
-	else
-		io_allowed = tmp;
-
+	io_allowed = calculate_io_allowed(iops_limit, jiffy_elapsed_rnd) +
+		     tg->carryover_ios[rw];
 	if (tg->io_disp[rw] + 1 <= io_allowed) {
 		if (wait)
 			*wait = 0;
···
 	return false;
 }
 
-static bool tg_with_in_bps_limit(struct throtl_grp *tg, struct bio *bio,
-				 u64 bps_limit, unsigned long *wait)
+static bool tg_within_bps_limit(struct throtl_grp *tg, struct bio *bio,
+				u64 bps_limit, unsigned long *wait)
 {
 	bool rw = bio_data_dir(bio);
-	u64 bytes_allowed, extra_bytes, tmp;
+	u64 bytes_allowed, extra_bytes;
 	unsigned long jiffy_elapsed, jiffy_wait, jiffy_elapsed_rnd;
 	unsigned int bio_size = throtl_bio_data_size(bio);
 
 	/* no need to throttle if this bio's bytes have been accounted */
-	if (bps_limit == U64_MAX || bio_flagged(bio, BIO_THROTTLED)) {
+	if (bps_limit == U64_MAX || bio_flagged(bio, BIO_BPS_THROTTLED)) {
 		if (wait)
 			*wait = 0;
 		return true;
···
 		jiffy_elapsed_rnd = tg->td->throtl_slice;
 
 	jiffy_elapsed_rnd = roundup(jiffy_elapsed_rnd, tg->td->throtl_slice);
-
-	tmp = bps_limit * jiffy_elapsed_rnd;
-	do_div(tmp, HZ);
-	bytes_allowed = tmp;
-
+	bytes_allowed = calculate_bytes_allowed(bps_limit, jiffy_elapsed_rnd) +
+			tg->carryover_bytes[rw];
 	if (tg->bytes_disp[rw] + bio_size <= bytes_allowed) {
 		if (wait)
 			*wait = 0;
···
 	 * slice and it should be extended instead.
 	 */
 	if (throtl_slice_used(tg, rw) && !(tg->service_queue.nr_queued[rw]))
-		throtl_start_new_slice(tg, rw);
+		throtl_start_new_slice(tg, rw, true);
 	else {
 		if (time_before(tg->slice_end[rw],
 		    jiffies + tg->td->throtl_slice))
···
 				jiffies + tg->td->throtl_slice);
 	}
 
-	if (tg_with_in_bps_limit(tg, bio, bps_limit, &bps_wait) &&
-	    tg_with_in_iops_limit(tg, bio, iops_limit, &iops_wait)) {
+	if (tg_within_bps_limit(tg, bio, bps_limit, &bps_wait) &&
+	    tg_within_iops_limit(tg, bio, iops_limit, &iops_wait)) {
 		if (wait)
 			*wait = 0;
 		return true;
···
 	unsigned int bio_size = throtl_bio_data_size(bio);
 
 	/* Charge the bio to the group */
-	if (!bio_flagged(bio, BIO_THROTTLED)) {
+	if (!bio_flagged(bio, BIO_BPS_THROTTLED)) {
 		tg->bytes_disp[rw] += bio_size;
 		tg->last_bytes_disp[rw] += bio_size;
 	}
 
 	tg->io_disp[rw]++;
 	tg->last_io_disp[rw]++;
-
-	/*
-	 * BIO_THROTTLED is used to prevent the same bio to be throttled
-	 * more than once as a throttled bio will go through blk-throtl the
-	 * second time when it eventually gets issued.  Set it when a bio
-	 * is being charged to a tg.
-	 */
-	if (!bio_flagged(bio, BIO_THROTTLED))
-		bio_set_flag(bio, BIO_THROTTLED);
 }
 
 /**
···
 	disptime = jiffies + min_wait;
 
 	/* Update dispatch time */
-	throtl_dequeue_tg(tg);
+	throtl_rb_erase(&tg->rb_node, tg->service_queue.parent_sq);
 	tg->disptime = disptime;
-	throtl_enqueue_tg(tg);
+	tg_service_queue_add(tg);
 
 	/* see throtl_add_bio_tg() */
 	tg->flags &= ~THROTL_TG_WAS_EMPTY;
···
 	sq->nr_queued[rw]--;
 
 	throtl_charge_bio(tg, bio);
+	bio_set_flag(bio, BIO_BPS_THROTTLED);
 
 	/*
 	 * If our parent is another tg, we just need to transfer @bio to
···
 		if (time_before(jiffies, tg->disptime))
 			break;
 
-		throtl_dequeue_tg(tg);
-
 		nr_disp += throtl_dispatch_tg(tg);
 
 		sq = &tg->service_queue;
-		if (sq->nr_queued[0] || sq->nr_queued[1])
+		if (sq->nr_queued[READ] || sq->nr_queued[WRITE])
 			tg_update_disptime(tg);
+		else
+			throtl_dequeue_tg(tg);
 
 		if (nr_disp >= THROTL_QUANTUM)
 			break;
···
 	 * that a group's limit are dropped suddenly and we don't want to
 	 * account recently dispatched IO with new low rate.
 	 */
-	throtl_start_new_slice(tg, READ);
-	throtl_start_new_slice(tg, WRITE);
+	throtl_start_new_slice(tg, READ, false);
+	throtl_start_new_slice(tg, WRITE, false);
 
 	if (tg->flags & THROTL_TG_PENDING) {
 		tg_update_disptime(tg);
···
 		v = U64_MAX;
 
 	tg = blkg_to_tg(ctx.blkg);
+	tg_update_carryover(tg);
 
 	if (is_u64)
 		*(u64 *)((void *)tg + of_cft(of)->private) = v;
···
 		return ret;
 
 	tg = blkg_to_tg(ctx.blkg);
+	tg_update_carryover(tg);
 
 	v[0] = tg->bps_conf[READ][index];
 	v[1] = tg->bps_conf[WRITE][index];
···
 	.pd_free_fn		= throtl_pd_free,
 };
 
+void blk_throtl_cancel_bios(struct gendisk *disk)
+{
+	struct request_queue *q = disk->queue;
+	struct cgroup_subsys_state *pos_css;
+	struct blkcg_gq *blkg;
+
+	spin_lock_irq(&q->queue_lock);
+	/*
+	 * queue_lock is held, rcu lock is not needed here technically.
+	 * However, rcu lock is still held to emphasize that following
+	 * path need RCU protection and to prevent warning from lockdep.
+	 */
+	rcu_read_lock();
+	blkg_for_each_descendant_post(blkg, pos_css, q->root_blkg) {
+		struct throtl_grp *tg = blkg_to_tg(blkg);
+		struct throtl_service_queue *sq = &tg->service_queue;
+
+		/*
+		 * Set the flag to make sure throtl_pending_timer_fn() won't
+		 * stop until all throttled bios are dispatched.
+		 */
+		blkg_to_tg(blkg)->flags |= THROTL_TG_CANCELING;
+		/*
+		 * Update disptime after setting the above flag to make sure
+		 * throtl_select_dispatch() won't exit without dispatching.
+		 */
+		tg_update_disptime(tg);
+
+		throtl_schedule_pending_timer(sq, jiffies + 1);
+	}
+	rcu_read_unlock();
+	spin_unlock_irq(&q->queue_lock);
+}
+
+#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
 static unsigned long __tg_last_low_overflow_time(struct throtl_grp *tg)
 {
 	unsigned long rtime = jiffies, wtime = jiffies;
···
 			return false;
 	}
 	return false;
-}
-
-void blk_throtl_cancel_bios(struct request_queue *q)
-{
-	struct cgroup_subsys_state *pos_css;
-	struct blkcg_gq *blkg;
-
-	spin_lock_irq(&q->queue_lock);
-	/*
-	 * queue_lock is held, rcu lock is not needed here technically.
-	 * However, rcu lock is still held to emphasize that following
-	 * path need RCU protection and to prevent warning from lockdep.
-	 */
-	rcu_read_lock();
-	blkg_for_each_descendant_post(blkg, pos_css, q->root_blkg) {
-		struct throtl_grp *tg = blkg_to_tg(blkg);
-		struct throtl_service_queue *sq = &tg->service_queue;
-
-		/*
-		 * Set the flag to make sure throtl_pending_timer_fn() won't
-		 * stop until all throttled bios are dispatched.
-		 */
-		blkg_to_tg(blkg)->flags |= THROTL_TG_CANCELING;
-		/*
-		 * Update disptime after setting the above flag to make sure
-		 * throtl_select_dispatch() won't exit without dispatching.
-		 */
-		tg_update_disptime(tg);
-
-		throtl_schedule_pending_timer(sq, jiffies + 1);
-	}
-	rcu_read_unlock();
-	spin_unlock_irq(&q->queue_lock);
 }
 
 static bool throtl_can_upgrade(struct throtl_data *td,
···
 	tg->checked_last_finish_time = last_finish_time;
 }
 
-#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
 static void throtl_update_latency_buckets(struct throtl_data *td)
 {
 	struct avg_latency_bucket avg_latency[2][LATENCY_BUCKET_SIZE];
···
 static inline void throtl_update_latency_buckets(struct throtl_data *td)
 {
 }
+
+static void blk_throtl_update_idletime(struct throtl_grp *tg)
+{
+}
+
+static void throtl_downgrade_check(struct throtl_grp *tg)
+{
+}
+
+static void throtl_upgrade_check(struct throtl_grp *tg)
+{
+}
+
+static bool throtl_can_upgrade(struct throtl_data *td,
+	struct throtl_grp *this_tg)
+{
+	return false;
+}
+
+static void throtl_upgrade_state(struct throtl_data *td)
+{
+}
 #endif
 
 bool __blk_throtl_bio(struct bio *bio)
···
 		qn = &tg->qnode_on_parent[rw];
 		sq = sq->parent_sq;
 		tg = sq_to_tg(sq);
-		if (!tg)
+		if (!tg) {
+			bio_set_flag(bio, BIO_BPS_THROTTLED);
 			goto out_unlock;
+		}
 	}
 
 	/* out-of-limit, queue to @tg */
···
 	}
 
 out_unlock:
-	bio_set_flag(bio, BIO_THROTTLED);
-
 #ifdef CONFIG_BLK_DEV_THROTTLING_LOW
 	if (throttled || !td->track_bio_latency)
 		bio->bi_issue.value |= BIO_ISSUE_THROTL_SKIP_LATENCY;
···
 }
 #endif
 
-int blk_throtl_init(struct request_queue *q)
+int blk_throtl_init(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct throtl_data *td;
 	int ret;
 
···
 	return ret;
 }
 
-void blk_throtl_exit(struct request_queue *q)
+void blk_throtl_exit(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
+
 	BUG_ON(!q->td);
 	del_timer_sync(&q->td->service_queue.pending_timer);
 	throtl_shutdown_wq(q);
···
 	kfree(q->td);
 }
 
-void blk_throtl_register_queue(struct request_queue *q)
+void blk_throtl_register(struct gendisk *disk)
 {
+	struct request_queue *q = disk->queue;
 	struct throtl_data *td;
 	int i;
 

block/blk-throttle.h

···
 enum tg_state_flags {
 	THROTL_TG_PENDING	= 1 << 0,	/* on parent's pending tree */
 	THROTL_TG_WAS_EMPTY	= 1 << 1,	/* bio_lists[] became non-empty */
-	THROTL_TG_HAS_IOPS_LIMIT = 1 << 2,	/* tg has iops limit */
-	THROTL_TG_CANCELING	= 1 << 3,	/* starts to cancel bio */
+	THROTL_TG_CANCELING	= 1 << 2,	/* starts to cancel bio */
 };
 
 enum {
···
 	unsigned int flags;
 
 	/* are there any throtl rules between this group and td? */
-	bool has_rules[2];
+	bool has_rules_bps[2];
+	bool has_rules_iops[2];
 
 	/* internally used bytes per second rate limits */
 	uint64_t bps[2][LIMIT_CNT];
···
 
 	uint64_t last_bytes_disp[2];
 	unsigned int last_io_disp[2];
+
+	/*
+	 * The following two fields are updated when new configuration is
+	 * submitted while some bios are still throttled, they record how many
+	 * bytes/ios are waited already in previous configuration, and they will
+	 * be used to calculate wait time under new configuration.
+	 */
+	uint64_t carryover_bytes[2];
+	unsigned int carryover_ios[2];
 
 	unsigned long last_check_time;
 
···
  * Internal throttling interface
  */
 #ifndef CONFIG_BLK_DEV_THROTTLING
-static inline int blk_throtl_init(struct request_queue *q) { return 0; }
-static inline void blk_throtl_exit(struct request_queue *q) { }
-static inline void blk_throtl_register_queue(struct request_queue *q) { }
+static inline int blk_throtl_init(struct gendisk *disk) { return 0; }
+static inline void blk_throtl_exit(struct gendisk *disk) { }
+static inline void blk_throtl_register(struct gendisk *disk) { }
 static inline bool blk_throtl_bio(struct bio *bio) { return false; }
-static inline void blk_throtl_cancel_bios(struct request_queue *q) { }
+static inline void blk_throtl_cancel_bios(struct gendisk *disk) { }
 #else /* CONFIG_BLK_DEV_THROTTLING */
-int blk_throtl_init(struct request_queue *q);
-void blk_throtl_exit(struct request_queue *q);
-void blk_throtl_register_queue(struct request_queue *q);
+int blk_throtl_init(struct gendisk *disk);
+void blk_throtl_exit(struct gendisk *disk);
+void blk_throtl_register(struct gendisk *disk);
 bool __blk_throtl_bio(struct bio *bio);
-void blk_throtl_cancel_bios(struct request_queue *q);
-static inline bool blk_throtl_bio(struct bio *bio)
+void blk_throtl_cancel_bios(struct gendisk *disk);
+
+static inline bool blk_should_throtl(struct bio *bio)
 {
 	struct throtl_grp *tg = blkg_to_tg(bio->bi_blkg);
+	int rw = bio_data_dir(bio);
 
-	/* no need to throttle bps any more if the bio has been throttled */
-	if (bio_flagged(bio, BIO_THROTTLED) &&
-	    !(tg->flags & THROTL_TG_HAS_IOPS_LIMIT))
-		return false;
+	/* iops limit is always counted */
+	if (tg->has_rules_iops[rw])
+		return true;
 
-	if (!tg->has_rules[bio_data_dir(bio)])
+	if (tg->has_rules_bps[rw] && !bio_flagged(bio, BIO_BPS_THROTTLED))
+		return true;
+
+	return false;
+}
+
+static inline bool blk_throtl_bio(struct bio *bio)
+{
+
+	if (!blk_should_throtl(bio))
 		return false;
 
 	return __blk_throtl_bio(bio);

block/blk-wbt.c

···
 	rwb->enable_state = WBT_STATE_ON_DEFAULT;
 	rwb->wc = 1;
 	rwb->rq_depth.default_depth = RWB_DEF_DEPTH;
+	rwb->min_lat_nsec = wbt_default_latency_nsec(q);
+
+	wbt_queue_depth_changed(&rwb->rqos);
+	wbt_set_write_cache(q, test_bit(QUEUE_FLAG_WC, &q->queue_flags));
 
 	/*
 	 * Assign rwb and add the stats callback.
···
 		goto err_free;
 
 	blk_stat_add_callback(q, rwb->cb);
-
-	rwb->min_lat_nsec = wbt_default_latency_nsec(q);
-
-	wbt_queue_depth_changed(&rwb->rqos);
-	wbt_set_write_cache(q, test_bit(QUEUE_FLAG_WC, &q->queue_flags));
 
 	return 0;
 

block/blk-zoned.c

···
 	if (!rq->q->disk->seq_zones_wlock)
 		return false;
 
-	switch (req_op(rq)) {
-	case REQ_OP_WRITE_ZEROES:
-	case REQ_OP_WRITE:
+	if (bdev_op_is_zoned_write(rq->q->disk->part0, req_op(rq)))
 		return blk_rq_zone_is_seq(rq);
-	default:
-		return false;
-	}
+
+	return false;
 }
 EXPORT_SYMBOL_GPL(blk_req_needs_zone_write_lock);
 

block/blk.h

···
 
 void blk_insert_flush(struct request *rq);
 
-int elevator_switch_mq(struct request_queue *q,
-			      struct elevator_type *new_e);
+int elevator_switch(struct request_queue *q, struct elevator_type *new_e);
 void elevator_exit(struct request_queue *q);
 int elv_register_queue(struct request_queue *q, bool uevent);
 void elv_unregister_queue(struct request_queue *q);
···
 }
 
 #ifdef CONFIG_BLK_CGROUP_IOLATENCY
-extern int blk_iolatency_init(struct request_queue *q);
+int blk_iolatency_init(struct gendisk *disk);
 #else
-static inline int blk_iolatency_init(struct request_queue *q) { return 0; }
+static inline int blk_iolatency_init(struct gendisk *disk) { return 0; };
 #endif
 
 #ifdef CONFIG_BLK_DEV_ZONED

block/elevator.c

···
 }
 EXPORT_SYMBOL_GPL(elv_unregister);
 
-int elevator_switch_mq(struct request_queue *q,
+static int elevator_switch_mq(struct request_queue *q,
 			      struct elevator_type *new_e)
 {
 	int ret;
···
  * need for the new one. this way we have a chance of going back to the old
  * one, if the new one fails init for some reason.
  */
-static int elevator_switch(struct request_queue *q, struct elevator_type *new_e)
+int elevator_switch(struct request_queue *q, struct elevator_type *new_e)
 {
 	int err;
 

block/genhd.c

···
 
 	blk_mq_freeze_queue_wait(q);
 
-	blk_throtl_cancel_bios(disk->queue);
+	blk_throtl_cancel_bios(disk);
 
 	blk_sync_queue(q);
 	blk_flush_integrity();
···
 	    !test_bit(GD_ADDED, &disk->state))
 		blk_mq_exit_queue(disk->queue);
 
-	blkcg_exit_queue(disk->queue);
+	blkcg_exit_disk(disk);
+
 	bioset_exit(&disk->bio_split);
 
 	disk_release_events(disk);
···
 	if (xa_insert(&disk->part_tbl, 0, disk->part0, GFP_KERNEL))
 		goto out_destroy_part_tbl;
 
-	if (blkcg_init_queue(q))
+	if (blkcg_init_disk(disk))
 		goto out_erase_part0;
 
 	rand_initialize_disk(disk);

block/opal_proto.h

···
 #define FIRST_TPER_SESSION_NUM	4096
 
 #define TPER_SYNC_SUPPORTED 0x01
+/* FC_LOCKING features */
+#define LOCKING_SUPPORTED_MASK 0x01
+#define LOCKING_ENABLED_MASK 0x02
+#define LOCKED_MASK 0x04
 #define MBR_ENABLED_MASK 0x10
+#define MBR_DONE_MASK 0x20
 
 #define TINY_ATOM_DATA_MASK 0x3F
 #define TINY_ATOM_SIGNED 0x40

block/sed-opal.c

···
 };
 
 struct opal_dev {
-	bool supported;
-	bool mbr_enabled;
+	u32 flags;
 
 	void *data;
 	sec_send_recv *send_recv;
···
 	return true;
 }
 
+static bool check_lcksuppt(const void *data)
+{
+	const struct d0_locking_features *lfeat = data;
+	u8 sup_feat = lfeat->supported_features;
+
+	return !!(sup_feat & LOCKING_SUPPORTED_MASK);
+}
+
+static bool check_lckenabled(const void *data)
+{
+	const struct d0_locking_features *lfeat = data;
+	u8 sup_feat = lfeat->supported_features;
+
+	return !!(sup_feat & LOCKING_ENABLED_MASK);
+}
+
+static bool check_locked(const void *data)
+{
+	const struct d0_locking_features *lfeat = data;
+	u8 sup_feat = lfeat->supported_features;
+
+	return !!(sup_feat & LOCKED_MASK);
+}
+
 static bool check_mbrenabled(const void *data)
 {
 	const struct d0_locking_features *lfeat = data;
 	u8 sup_feat = lfeat->supported_features;
 
 	return !!(sup_feat & MBR_ENABLED_MASK);
+}
+
+static bool check_mbrdone(const void *data)
+{
+	const struct d0_locking_features *lfeat = data;
+	u8 sup_feat = lfeat->supported_features;
+
+	return !!(sup_feat & MBR_DONE_MASK);
 }
 
 static bool check_sum(const void *data)
···
 	u32 hlen = be32_to_cpu(hdr->length);
 
 	print_buffer(dev->resp, hlen);
-	dev->mbr_enabled = false;
+	dev->flags &= OPAL_FL_SUPPORTED;
 
 	if (hlen > IO_BUFFER_LENGTH - sizeof(*hdr)) {
 		pr_debug("Discovery length overflows buffer (%zu+%u)/%u\n",
···
 			check_geometry(dev, body);
 			break;
 		case FC_LOCKING:
-			dev->mbr_enabled = check_mbrenabled(body->features);
+			if (check_lcksuppt(body->features))
+				dev->flags |= OPAL_FL_LOCKING_SUPPORTED;
+			if (check_lckenabled(body->features))
+				dev->flags |= OPAL_FL_LOCKING_ENABLED;
+			if (check_locked(body->features))
+				dev->flags |= OPAL_FL_LOCKED;
+			if (check_mbrenabled(body->features))
+				dev->flags |= OPAL_FL_MBR_ENABLED;
+			if (check_mbrdone(body->features))
+				dev->flags |= OPAL_FL_MBR_DONE;
 			break;
 		case FC_ENTERPRISE:
 		case FC_DATASTORE:
···
 	mutex_lock(&dev->dev_lock);
 	setup_opal_dev(dev);
 	ret = opal_discovery0_step(dev);
-	dev->supported = !ret;
+	if (!ret)
+		dev->flags |= OPAL_FL_SUPPORTED;
 	mutex_unlock(&dev->dev_lock);
 
 	return ret;
···
 
 	INIT_LIST_HEAD(&dev->unlk_lst);
 	mutex_init(&dev->dev_lock);
+	dev->flags = 0;
 	dev->data = data;
 	dev->send_recv = send_recv;
 	if (check_opal_support(dev) != 0) {
···
 	if (!dev)
 		return false;
 
-	if (!dev->supported)
+	if (!(dev->flags & OPAL_FL_SUPPORTED))
 		return false;
 
 	mutex_lock(&dev->dev_lock);
···
 			was_failure = true;
 		}
 
-		if (dev->mbr_enabled) {
+		if (dev->flags & OPAL_FL_MBR_ENABLED) {
 			ret = __opal_set_mbr_done(dev, &suspend->unlk.session.opal_key);
 			if (ret)
 				pr_debug("Failed to set MBR Done in S3 resume\n");
···
 	return ret;
 }
 
+static int opal_get_status(struct opal_dev *dev, void __user *data)
+{
+	struct opal_status sts = {0};
+
+	/*
+	 * check_opal_support() error is not fatal,
+	 * !dev->supported is a valid condition
+	 */
+	if (!check_opal_support(dev))
+		sts.flags = dev->flags;
+	if (copy_to_user(data, &sts, sizeof(sts))) {
+		pr_debug("Error copying status to userspace\n");
+		return -EFAULT;
+	}
+	return 0;
+}
+
 int sed_ioctl(struct opal_dev *dev, unsigned int cmd, void __user *arg)
 {
 	void *p;
···
 		return -EACCES;
 	if (!dev)
 		return -ENOTSUPP;
-	if (!dev->supported)
+	if (!(dev->flags & OPAL_FL_SUPPORTED))
 		return -ENOTSUPP;
 
-	p = memdup_user(arg, _IOC_SIZE(cmd));
-	if (IS_ERR(p))
-		return PTR_ERR(p);
+	if (cmd & IOC_IN) {
+		p = memdup_user(arg, _IOC_SIZE(cmd));
+		if (IS_ERR(p))
+			return PTR_ERR(p);
+	}
 
 	switch (cmd) {
 	case IOC_OPAL_SAVE:
···
 	case IOC_OPAL_GENERIC_TABLE_RW:
 		ret = opal_generic_read_write_table(dev, p);
 		break;
+	case IOC_OPAL_GET_STATUS:
+		ret = opal_get_status(dev, arg);
+		break;
 	default:
 		break;
 	}
 
-	kfree(p);
+	if (cmd & IOC_IN)
+		kfree(p);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(sed_ioctl);

drivers/block/aoe/aoeblk.c

···
 	return sysfs_emit(page, "%lu\n", d->maxbcnt);
 }
 
-static int aoedisk_debugfs_show(struct seq_file *s, void *ignored)
+static int aoe_debugfs_show(struct seq_file *s, void *ignored)
 {
 	struct aoedev *d;
 	struct aoetgt **t, **te;
···
 
 	return 0;
 }
-
-static int aoe_debugfs_open(struct inode *inode, struct file *file)
-{
-	return single_open(file, aoedisk_debugfs_show, inode->i_private);
-}
+DEFINE_SHOW_ATTRIBUTE(aoe_debugfs);
 
 static DEVICE_ATTR(state, 0444, aoedisk_show_state, NULL);
 static DEVICE_ATTR(mac, 0444, aoedisk_show_mac, NULL);
···
 static const struct attribute_group *aoe_attr_groups[] = {
 	&aoe_attr_group,
 	NULL,
-};
-
-static const struct file_operations aoe_debugfs_fops = {
-	.open = aoe_debugfs_open,
-	.read = seq_read,
-	.llseek = seq_lseek,
-	.release = single_release,
 };
 
 static void

drivers/block/brd.c

···
 	disk->minors		= max_part;
 	disk->fops		= &brd_fops;
 	disk->private_data	= brd;
-	strlcpy(disk->disk_name, buf, DISK_NAME_LEN);
+	strscpy(disk->disk_name, buf, DISK_NAME_LEN);
 	set_capacity(disk, rd_size * 2);
 	
 	/*

drivers/block/drbd/drbd_int.h

···
 extern int w_e_reissue(struct drbd_work *, int);
 extern int w_restart_disk_io(struct drbd_work *, int);
 extern int w_send_out_of_sync(struct drbd_work *, int);
-extern int w_start_resync(struct drbd_work *, int);
 
 extern void resync_timer_fn(struct timer_list *t);
 extern void start_resync_timer_fn(struct timer_list *t);

drivers/block/drbd/drbd_nl.c

···
 	struct drbd_genlmsghdr *dh;
 	int err;
 
-	strlcpy(helper_info.helper_name, name, sizeof(helper_info.helper_name));
+	strscpy(helper_info.helper_name, name, sizeof(helper_info.helper_name));
 	helper_info.helper_name_len = min(strlen(name), sizeof(helper_info.helper_name));
 	helper_info.helper_status = status;
 

drivers/block/drbd/drbd_receiver.c

···
 	if (unlikely(!req))
 		return -EIO;
 
-	/* hlist_del(&req->collision) is done in _req_may_be_done, to avoid
-	 * special casing it there for the various failure cases.
-	 * still no race with drbd_fail_pending_reads */
 	err = recv_dless_read(peer_device, req, sector, pi->size);
 	if (!err)
 		req_mod(req, DATA_RECEIVED);

drivers/block/drbd/drbd_req.h

···
 
 extern void start_new_tl_epoch(struct drbd_connection *connection);
 extern void drbd_req_destroy(struct kref *kref);
-extern void _req_may_be_done(struct drbd_request *req,
-		struct bio_and_error *m);
 extern int __req_mod(struct drbd_request *req, enum drbd_req_event what,
 		struct bio_and_error *m);
 extern void complete_master_bio(struct drbd_device *device,

drivers/block/mtip32xx/mtip32xx.c

···
 	if (!port->identify_valid)
 		return;
 
-	strlcpy(cbuf, (char *)(port->identify+10), 21);
+	strscpy(cbuf, (char *)(port->identify + 10), 21);
 	dev_info(&port->dd->pdev->dev,
 		"Serial No.: %s\n", cbuf);
 
-	strlcpy(cbuf, (char *)(port->identify+23), 9);
+	strscpy(cbuf, (char *)(port->identify + 23), 9);
 	dev_info(&port->dd->pdev->dev,
 		"Firmware Ver.: %s\n", cbuf);
 
-	strlcpy(cbuf, (char *)(port->identify+27), 41);
+	strscpy(cbuf, (char *)(port->identify + 27), 41);
 	dev_info(&port->dd->pdev->dev, "Model: %s\n", cbuf);
 
 	dev_info(&port->dd->pdev->dev, "Security: %04x %s\n",
···
 	pci_read_config_word(port->dd->pdev, PCI_REVISION_ID, &revid);
 	switch (revid & 0xFF) {
 	case 0x1:
-		strlcpy(cbuf, "A0", 3);
+		strscpy(cbuf, "A0", 3);
 		break;
 	case 0x3:
-		strlcpy(cbuf, "A2", 3);
+		strscpy(cbuf, "A2", 3);
 		break;
 	default:
-		strlcpy(cbuf, "?", 2);
+		strscpy(cbuf, "?", 2);
 		break;
 	}
 	dev_info(&port->dd->pdev->dev,

drivers/block/nbd.c

···
 	mutex_unlock(&nbd->config_lock);
 	ret = wait_event_interruptible(config->recv_wq,
 					 atomic_read(&config->recv_threads) == 0);
-	if (ret)
+	if (ret) {
 		sock_shutdown(nbd);
-	flush_workqueue(nbd->recv_workq);
+		nbd_clear_que(nbd);
+	}
 
+	flush_workqueue(nbd->recv_workq);
 	mutex_lock(&nbd->config_lock);
 	nbd_bdev_reset(nbd);
 	/* user requested, ignore socket errors */

drivers/block/null_blk/main.c

···
 	return false;
 }
 
-static int null_map_queues(struct blk_mq_tag_set *set)
+static void null_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nullb *nullb = set->driver_data;
 	int i, qoff;
···
 		} else {
 			pr_warn("tag set has unexpected nr_hw_queues: %d\n",
 				set->nr_hw_queues);
-			return -EINVAL;
+			WARN_ON_ONCE(true);
+			submit_queues = 1;
+			poll_queues = 0;
 		}
 	}
 
···
 		qoff += map->nr_queues;
 		blk_mq_map_queues(map);
 	}
-
-	return 0;
 }
 
 static int null_poll(struct blk_mq_hw_ctx *hctx, struct io_comp_batch *iob)

drivers/block/ps3vram.c

···
 	gendisk->flags |= GENHD_FL_NO_PART;
 	gendisk->fops = &ps3vram_fops;
 	gendisk->private_data = dev;
-	strlcpy(gendisk->disk_name, DEVICE_NAME, sizeof(gendisk->disk_name));
+	strscpy(gendisk->disk_name, DEVICE_NAME, sizeof(gendisk->disk_name));
 	set_capacity(gendisk, priv->size >> 9);
 	blk_queue_max_segments(gendisk->queue, BLK_MAX_SEGMENTS);
 	blk_queue_max_segment_size(gendisk->queue, BLK_MAX_SEGMENT_SIZE);

drivers/block/rnbd/Makefile

···
 		  rnbd-clt-sysfs.o \
 		  rnbd-common.o
 
+CFLAGS_rnbd-srv-trace.o = -I$(src)
+
 rnbd-server-y := rnbd-common.o \
 		  rnbd-srv.o \
-		  rnbd-srv-dev.o \
-		  rnbd-srv-sysfs.o
+		  rnbd-srv-sysfs.o \
+		  rnbd-srv-trace.o
 
 obj-$(CONFIG_BLK_DEV_RNBD_CLIENT) += rnbd-client.o
 obj-$(CONFIG_BLK_DEV_RNBD_SERVER) += rnbd-server.o

drivers/block/rnbd/rnbd-clt.c

···
 {
 	struct rnbd_queue *q = hctx->driver_data;
 	struct rnbd_clt_dev *dev = q->dev;
-	int cnt;
 
-	cnt = rtrs_clt_rdma_cq_direct(dev->sess->rtrs, hctx->queue_num);
-	return cnt;
+	return rtrs_clt_rdma_cq_direct(dev->sess->rtrs, hctx->queue_num);
 }
 
-static int rnbd_rdma_map_queues(struct blk_mq_tag_set *set)
+static void rnbd_rdma_map_queues(struct blk_mq_tag_set *set)
 {
 	struct rnbd_clt_session *sess = set->driver_data;
 
···
 			set->map[HCTX_TYPE_DEFAULT].nr_queues,
 			set->map[HCTX_TYPE_READ].nr_queues);
 	}
-
-	return 0;
 }
 
 static struct blk_mq_ops rnbd_mq_ops = {

drivers/block/rnbd/rnbd-srv-dev.c

···
-// SPDX-License-Identifier: GPL-2.0-or-later
-/*
- * RDMA Network Block Driver
- *
- * Copyright (c) 2014 - 2018 ProfitBricks GmbH. All rights reserved.
- * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
- * Copyright (c) 2019 - 2020 1&1 IONOS SE. All rights reserved.
- */
-#undef pr_fmt
-#define pr_fmt(fmt) KBUILD_MODNAME " L" __stringify(__LINE__) ": " fmt
-
-#include "rnbd-srv-dev.h"
-#include "rnbd-log.h"
-
-struct rnbd_dev *rnbd_dev_open(const char *path, fmode_t flags)
-{
-	struct rnbd_dev *dev;
-	int ret;
-
-	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
-	if (!dev)
-		return ERR_PTR(-ENOMEM);
-
-	dev->blk_open_flags = flags;
-	dev->bdev = blkdev_get_by_path(path, flags, THIS_MODULE);
-	ret = PTR_ERR_OR_ZERO(dev->bdev);
-	if (ret)
-		goto err;
-
-	dev->blk_open_flags = flags;
-
-	return dev;
-
-err:
-	kfree(dev);
-	return ERR_PTR(ret);
-}
-
-void rnbd_dev_close(struct rnbd_dev *dev)
-{
-	blkdev_put(dev->bdev, dev->blk_open_flags);
-	kfree(dev);
-}

drivers/block/rnbd/rnbd-srv-dev.h

···
-/* SPDX-License-Identifier: GPL-2.0-or-later */
-/*
- * RDMA Network Block Driver
- *
- * Copyright (c) 2014 - 2018 ProfitBricks GmbH. All rights reserved.
- * Copyright (c) 2018 - 2019 1&1 IONOS Cloud GmbH. All rights reserved.
- * Copyright (c) 2019 - 2020 1&1 IONOS SE. All rights reserved.
- */
-#ifndef RNBD_SRV_DEV_H
-#define RNBD_SRV_DEV_H
-
-#include <linux/fs.h>
-#include "rnbd-proto.h"
-
-struct rnbd_dev {
-	struct block_device	*bdev;
-	fmode_t			blk_open_flags;
-};
-
-/**
- * rnbd_dev_open() - Open a device
- * @path:	path to open
- * @flags:	open flags
- */
-struct rnbd_dev *rnbd_dev_open(const char *path, fmode_t flags);
-
-/**
- * rnbd_dev_close() - Close a device
- */
-void rnbd_dev_close(struct rnbd_dev *dev);
-
-void rnbd_endio(void *priv, int error);
-
-static inline int rnbd_dev_get_max_segs(const struct rnbd_dev *dev)
-{
-	return queue_max_segments(bdev_get_queue(dev->bdev));
-}
-
-static inline int rnbd_dev_get_max_hw_sects(const struct rnbd_dev *dev)
-{
-	return queue_max_hw_sectors(bdev_get_queue(dev->bdev));
-}
-
-static inline int rnbd_dev_get_secure_discard(const struct rnbd_dev *dev)
-{
-	return bdev_max_secure_erase_sectors(dev->bdev);
-}
-
-static inline int rnbd_dev_get_max_discard_sects(const struct rnbd_dev *dev)
-{
-	return bdev_max_discard_sectors(dev->bdev);
-}
-
-static inline int rnbd_dev_get_discard_granularity(const struct rnbd_dev *dev)
-{
-	return bdev_get_queue(dev->bdev)->limits.discard_granularity;
-}
-
-static inline int rnbd_dev_get_discard_alignment(const struct rnbd_dev *dev)
-{
-	return bdev_discard_alignment(dev->bdev);
-}
-
-#endif /* RNBD_SRV_DEV_H */

drivers/block/rnbd/rnbd-srv-trace.c

···
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * RDMA Network Block Driver
+ *
+ * Copyright (c) 2022 1&1 IONOS SE. All rights reserved.
+ */
+#include "rtrs.h"
+#include "rtrs-srv.h"
+#include "rnbd-srv.h"
+#include "rnbd-proto.h"
+
+/*
+ * We include this last to have the helpers above available for the trace
+ * event implementations.
+ */
+#define CREATE_TRACE_POINTS
+#include "rnbd-srv-trace.h"

drivers/block/rnbd/rnbd-srv-trace.h

···
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * RDMA Network Block Driver
+ *
+ * Copyright (c) 2022 1&1 IONOS SE. All rights reserved.
+ */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM rnbd_srv
+
+#if !defined(_TRACE_RNBD_SRV_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_RNBD_SRV_H
+
+#include <linux/tracepoint.h>
+
+struct rnbd_srv_session;
+struct rtrs_srv_op;
+
+DECLARE_EVENT_CLASS(rnbd_srv_link_class,
+	TP_PROTO(struct rnbd_srv_session *srv),
+
+	TP_ARGS(srv),
+
+	TP_STRUCT__entry(
+		__field(int, qdepth)
+		__string(sessname, srv->sessname)
+	),
+
+	TP_fast_assign(
+		__entry->qdepth = srv->queue_depth;
+		__assign_str(sessname, srv->sessname);
+	),
+
+	TP_printk("sessname: %s qdepth: %d",
+		   __get_str(sessname),
+		   __entry->qdepth
+	)
+);
+
+#define DEFINE_LINK_EVENT(name) \
+DEFINE_EVENT(rnbd_srv_link_class, name, \
+	TP_PROTO(struct rnbd_srv_session *srv), \
+	TP_ARGS(srv))
+
+DEFINE_LINK_EVENT(create_sess);
+DEFINE_LINK_EVENT(destroy_sess);
+
+TRACE_DEFINE_ENUM(RNBD_OP_READ);
+TRACE_DEFINE_ENUM(RNBD_OP_WRITE);
+TRACE_DEFINE_ENUM(RNBD_OP_FLUSH);
+TRACE_DEFINE_ENUM(RNBD_OP_DISCARD);
+TRACE_DEFINE_ENUM(RNBD_OP_SECURE_ERASE);
+TRACE_DEFINE_ENUM(RNBD_F_SYNC);
+TRACE_DEFINE_ENUM(RNBD_F_FUA);
+
+#define show_rnbd_rw_flags(x) \
+	__print_flags(x, "|", \
+		{ RNBD_OP_READ,		"READ" }, \
+		{ RNBD_OP_WRITE,	"WRITE" }, \
+		{ RNBD_OP_FLUSH,	"FLUSH" }, \
+		{ RNBD_OP_DISCARD,	"DISCARD" }, \
+		{ RNBD_OP_SECURE_ERASE,	"SECURE_ERASE" }, \
+		{ RNBD_F_SYNC,		"SYNC" }, \
+		{ RNBD_F_FUA,		"FUA" })
+
+TRACE_EVENT(process_rdma,
+	TP_PROTO(struct rnbd_srv_session *srv,
+		 const struct rnbd_msg_io *msg,
+		 struct rtrs_srv_op *id,
+		 u32 datalen,
+		 size_t usrlen),
+
+	TP_ARGS(srv, msg, id, datalen, usrlen),
+
+	TP_STRUCT__entry(
+		__string(sessname, srv->sessname)
+		__field(u8, dir)
+		__field(u8, ver)
+		__field(u32, device_id)
+		__field(u64, sector)
+		__field(u32, flags)
+		__field(u32, bi_size)
+		__field(u16, ioprio)
+		__field(u32, datalen)
+		__field(size_t, usrlen)
+	),
+
+	TP_fast_assign(
+		__assign_str(sessname, srv->sessname);
+		__entry->dir = id->dir;
+		__entry->ver = srv->ver;
+		__entry->device_id = le32_to_cpu(msg->device_id);
+		__entry->sector = le64_to_cpu(msg->sector);
+		__entry->bi_size = le32_to_cpu(msg->bi_size);
+		__entry->flags = le32_to_cpu(msg->rw);
+		__entry->ioprio = le16_to_cpu(msg->prio);
+		__entry->datalen = datalen;
+		__entry->usrlen = usrlen;
+	),
+
+	TP_printk("I/O req: sess: %s, type: %s, ver: %d, devid: %u, sector: %llu, bsize: %u, flags: %s, ioprio: %d, datalen: %u, usrlen: %zu",
+		   __get_str(sessname),
+		   __print_symbolic(__entry->dir,
+			 { READ,  "READ" },
+			 { WRITE, "WRITE" }),
+		   __entry->ver,
+		   __entry->device_id,
+		   __entry->sector,
+		   __entry->bi_size,
+		   show_rnbd_rw_flags(__entry->flags),
+		   __entry->ioprio,
+		   __entry->datalen,
+		   __entry->usrlen
+	)
+);
+
+TRACE_EVENT(process_msg_sess_info,
+	TP_PROTO(struct rnbd_srv_session *srv,
+		 const struct rnbd_msg_sess_info *msg),
+
+	TP_ARGS(srv, msg),
+
+	TP_STRUCT__entry(
+		__field(u8, proto_ver)
+		__field(u8, clt_ver)
+		__field(u8, srv_ver)
+		__string(sessname, srv->sessname)
+	),
+
+	TP_fast_assign(
+		__entry->proto_ver = srv->ver;
+		__entry->clt_ver = msg->ver;
+		__entry->srv_ver = RNBD_PROTO_VER_MAJOR;
+		__assign_str(sessname, srv->sessname);
+	),
+
+	TP_printk("Session %s using proto-ver %d (clt-ver: %d, srv-ver: %d)",
+		   __get_str(sessname),
+		   __entry->proto_ver,
+		   __entry->clt_ver,
+		   __entry->srv_ver
+	)
+);
+
+TRACE_DEFINE_ENUM(RNBD_ACCESS_RO);
+TRACE_DEFINE_ENUM(RNBD_ACCESS_RW);
+TRACE_DEFINE_ENUM(RNBD_ACCESS_MIGRATION);
+
+#define show_rnbd_access_mode(x) \
+	__print_symbolic(x, \
+		{ RNBD_ACCESS_RO,		"RO" }, \
+		{ RNBD_ACCESS_RW,		"RW" }, \
+		{ RNBD_ACCESS_MIGRATION,	"MIGRATION" })
+
+TRACE_EVENT(process_msg_open,
+	TP_PROTO(struct rnbd_srv_session *srv,
+		 const struct rnbd_msg_open *msg),
+
+	TP_ARGS(srv, msg),
+
+	TP_STRUCT__entry(
+		__field(u8, access_mode)
+		__string(sessname, srv->sessname)
+		__string(dev_name, msg->dev_name)
+	),
+
+	TP_fast_assign(
+		__entry->access_mode = msg->access_mode;
+		__assign_str(sessname, srv->sessname);
+		__assign_str(dev_name, msg->dev_name);
+	),
+
+	TP_printk("Open message received: session='%s' path='%s' access_mode=%s",
+		   __get_str(sessname),
+		   __get_str(dev_name),
+		   show_rnbd_access_mode(__entry->access_mode)
+	)
+);
+
+TRACE_EVENT(process_msg_close,
+	TP_PROTO(struct rnbd_srv_session *srv,
+		 const struct rnbd_msg_close *msg),
+
+	TP_ARGS(srv, msg),
+
+	TP_STRUCT__entry(
+		__field(u32, device_id)
+		__string(sessname, srv->sessname)
+	),
+
+	TP_fast_assign(
+		__entry->device_id = le32_to_cpu(msg->device_id);
+		__assign_str(sessname, srv->sessname);
+	),
+
+	TP_printk("Close message received: session='%s' device id='%d'",
+		   __get_str(sessname),
+		   __entry->device_id
+	)
+);
+
+#endif /* _TRACE_RNBD_SRV_H */
+
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE rnbd-srv-trace
+#include <trace/define_trace.h>
+

drivers/block/rnbd/rnbd-srv.c

···
 #include <linux/blkdev.h>
 
 #include "rnbd-srv.h"
-#include "rnbd-srv-dev.h"
+#include "rnbd-srv-trace.h"
 
 MODULE_DESCRIPTION("RDMA Network Block Device Server");
 MODULE_LICENSE("GPL");
···
 	kref_put(&sess_dev->kref, rnbd_sess_dev_release);
 }
 
-void rnbd_endio(void *priv, int error)
-{
-	struct rnbd_io_private *rnbd_priv = priv;
-	struct rnbd_srv_sess_dev *sess_dev = rnbd_priv->sess_dev;
-
-	rnbd_put_sess_dev(sess_dev);
-
-	rtrs_srv_resp_rdma(rnbd_priv->id, error);
-
-	kfree(priv);
-}
-
 static struct rnbd_srv_sess_dev *
 rnbd_get_sess_dev(int dev_id, struct rnbd_srv_session *srv_sess)
 {
···
 
 static void rnbd_dev_bi_end_io(struct bio *bio)
 {
-	rnbd_endio(bio->bi_private, blk_status_to_errno(bio->bi_status));
+	struct rnbd_io_private *rnbd_priv = bio->bi_private;
+	struct rnbd_srv_sess_dev *sess_dev = rnbd_priv->sess_dev;
+
+	rnbd_put_sess_dev(sess_dev);
+	rtrs_srv_resp_rdma(rnbd_priv->id, blk_status_to_errno(bio->bi_status));
+
+	kfree(rnbd_priv);
 	bio_put(bio);
 }
 
···
 	int err;
 	struct bio *bio;
 	short prio;
+
+	trace_process_rdma(srv_sess, msg, id, datalen, usrlen);
 
 	priv = kmalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv)
···
 	priv->sess_dev = sess_dev;
 	priv->id = id;
 
-	bio = bio_alloc(sess_dev->rnbd_dev->bdev, 1,
+	bio = bio_alloc(sess_dev->bdev, 1,
 			rnbd_to_bio_flags(le32_to_cpu(msg->rw)), GFP_KERNEL);
 	if (bio_add_page(bio, virt_to_page(data), datalen,
 			offset_in_page(data)) != datalen) {
···
 	rnbd_put_sess_dev(sess_dev);
 	wait_for_completion(&dc); /* wait for inflights to drop to zero */
 
-	rnbd_dev_close(sess_dev->rnbd_dev);
+	blkdev_put(sess_dev->bdev, sess_dev->open_flags);
 	mutex_lock(&sess_dev->dev->lock);
 	list_del(&sess_dev->dev_list);
 	if (sess_dev->open_flags & FMODE_WRITE)
···
 
 	if (xa_empty(&srv_sess->index_idr))
 		goto out;
+
+	trace_destroy_sess(srv_sess);
 
 	mutex_lock(&srv_sess->lock);
 	xa_for_each(&srv_sess->index_idr, index, sess_dev)
···
 
 	rtrs_srv_set_sess_priv(rtrs, srv_sess);
 
+	trace_create_sess(srv_sess);
+
 	return 0;
 }
 
···
 	mutex_unlock(&sess->lock);
 }
 
-static int process_msg_close(struct rnbd_srv_session *srv_sess,
+static void process_msg_close(struct rnbd_srv_session *srv_sess,
 			     void *data, size_t datalen, const void *usr,
 			     size_t usrlen)
 {
 	const struct rnbd_msg_close *close_msg = usr;
 	struct rnbd_srv_sess_dev *sess_dev;
 
+	trace_process_msg_close(srv_sess, close_msg);
+
 	sess_dev = rnbd_get_sess_dev(le32_to_cpu(close_msg->device_id),
 				      srv_sess);
 	if (IS_ERR(sess_dev))
-		return 0;
+		return;
 
 	rnbd_put_sess_dev(sess_dev);
 	mutex_lock(&srv_sess->lock);
 	rnbd_srv_destroy_dev_session_sysfs(sess_dev);
 	mutex_unlock(&srv_sess->lock);
-	return 0;
 }
 
 static int process_msg_open(struct rnbd_srv_session *srv_sess,
···
 	case RNBD_MSG_IO:
 		return process_rdma(srv_sess, id, data, datalen, usr, usrlen);
 	case RNBD_MSG_CLOSE:
-		ret = process_msg_close(srv_sess, data, datalen, usr, usrlen);
+		process_msg_close(srv_sess, data, datalen, usr, usrlen);
 		break;
 	case RNBD_MSG_OPEN:
 		ret = process_msg_open(srv_sess, usr, usrlen, data, datalen);
···
 		return -EINVAL;
 	}
 
+	/*
+	 * Since ret is passed to rtrs to handle the failure case, we
+	 * just return 0 at the end otherwise callers in rtrs would call
+	 * send_io_resp_imm again to print redundant err message.
+	 */
 	rtrs_srv_resp_rdma(id, ret);
 	return 0;
 }
···
 }
 
 static struct rnbd_srv_dev *
-rnbd_srv_get_or_create_srv_dev(struct rnbd_dev *rnbd_dev,
+rnbd_srv_get_or_create_srv_dev(struct block_device *bdev,
 				struct rnbd_srv_session *srv_sess,
 				enum rnbd_access_mode access_mode)
 {
 	int ret;
 	struct rnbd_srv_dev *new_dev, *dev;
 
-	new_dev = rnbd_srv_init_srv_dev(rnbd_dev->bdev);
+	new_dev = rnbd_srv_init_srv_dev(bdev);
 	if (IS_ERR(new_dev))
 		return new_dev;
 
···
 static void rnbd_srv_fill_msg_open_rsp(struct rnbd_msg_open_rsp *rsp,
 					struct rnbd_srv_sess_dev *sess_dev)
 {
-	struct rnbd_dev *rnbd_dev = sess_dev->rnbd_dev;
+	struct block_device *bdev = sess_dev->bdev;
 
 	rsp->hdr.type = cpu_to_le16(RNBD_MSG_OPEN_RSP);
-	rsp->device_id =
-		cpu_to_le32(sess_dev->device_id);
-	rsp->nsectors =
-		cpu_to_le64(get_capacity(rnbd_dev->bdev->bd_disk));
-	rsp->logical_block_size	=
-		cpu_to_le16(bdev_logical_block_size(rnbd_dev->bdev));
-	rsp->physical_block_size =
-		cpu_to_le16(bdev_physical_block_size(rnbd_dev->bdev));
-	rsp->max_segments =
-		cpu_to_le16(rnbd_dev_get_max_segs(rnbd_dev));
+	rsp->device_id = cpu_to_le32(sess_dev->device_id);
+	rsp->nsectors = cpu_to_le64(bdev_nr_sectors(bdev));
+	rsp->logical_block_size	= cpu_to_le16(bdev_logical_block_size(bdev));
+	rsp->physical_block_size = cpu_to_le16(bdev_physical_block_size(bdev));
+	rsp->max_segments = cpu_to_le16(bdev_max_segments(bdev));
 	rsp->max_hw_sectors =
-		cpu_to_le32(rnbd_dev_get_max_hw_sects(rnbd_dev));
+		cpu_to_le32(queue_max_hw_sectors(bdev_get_queue(bdev)));
 	rsp->max_write_same_sectors = 0;
-	rsp->max_discard_sectors =
-		cpu_to_le32(rnbd_dev_get_max_discard_sects(rnbd_dev));
-	rsp->discard_granularity =
-		cpu_to_le32(rnbd_dev_get_discard_granularity(rnbd_dev));
-	rsp->discard_alignment =
-		cpu_to_le32(rnbd_dev_get_discard_alignment(rnbd_dev));
-	rsp->secure_discard =
-		cpu_to_le16(rnbd_dev_get_secure_discard(rnbd_dev));
+	rsp->max_discard_sectors = cpu_to_le32(bdev_max_discard_sectors(bdev));
+	rsp->discard_granularity = cpu_to_le32(bdev_discard_granularity(bdev));
+	rsp->discard_alignment = cpu_to_le32(bdev_discard_alignment(bdev));
+	rsp->secure_discard = cpu_to_le16(bdev_max_secure_erase_sectors(bdev));
 	rsp->cache_policy = 0;
-	if (bdev_write_cache(rnbd_dev->bdev))
+	if (bdev_write_cache(bdev))
 		rsp->cache_policy |= RNBD_WRITEBACK;
-	if (bdev_fua(rnbd_dev->bdev))
+	if (bdev_fua(bdev))
 		rsp->cache_policy |= RNBD_FUA;
 }
 
 static struct rnbd_srv_sess_dev *
 rnbd_srv_create_set_sess_dev(struct rnbd_srv_session *srv_sess,
 			      const struct rnbd_msg_open *open_msg,
-			      struct rnbd_dev *rnbd_dev, fmode_t open_flags,
+			      struct block_device *bdev, fmode_t open_flags,
 			      struct rnbd_srv_dev *srv_dev)
 {
 	struct rnbd_srv_sess_dev *sdev = rnbd_sess_dev_alloc(srv_sess);
···
 
 	strscpy(sdev->pathname, open_msg->dev_name, sizeof(sdev->pathname));
 
-	sdev->rnbd_dev		= rnbd_dev;
+	sdev->bdev		= bdev;
 	sdev->sess		= srv_sess;
 	sdev->dev		= srv_dev;
 	sdev->open_flags	= open_flags;
···
 	struct rnbd_msg_sess_info_rsp *rsp = data;
 
 	srv_sess->ver = min_t(u8, sess_info_msg->ver, RNBD_PROTO_VER_MAJOR);
-	pr_debug("Session %s using protocol version %d (client version: %d, server version: %d)\n",
-		 srv_sess->sessname, srv_sess->ver,
-		 sess_info_msg->ver, RNBD_PROTO_VER_MAJOR);
+
+	trace_process_msg_sess_info(srv_sess, sess_info_msg);
 
 	rsp->hdr.type = cpu_to_le16(RNBD_MSG_SESS_INFO_RSP);
 	rsp->ver = srv_sess->ver;
···
 	struct rnbd_srv_dev *srv_dev;
 	struct rnbd_srv_sess_dev *srv_sess_dev;
 	const struct rnbd_msg_open *open_msg = msg;
+	struct block_device *bdev;
 	fmode_t open_flags;
 	char *full_path;
-	struct rnbd_dev *rnbd_dev;
 	struct rnbd_msg_open_rsp *rsp = data;
 
-	pr_debug("Open message received: session='%s' path='%s' access_mode=%d\n",
-		 srv_sess->sessname, open_msg->dev_name,
-		 open_msg->access_mode);
+	trace_process_msg_open(srv_sess, open_msg);
+
 	open_flags = FMODE_READ;
 	if (open_msg->access_mode != RNBD_ACCESS_RO)
 		open_flags |= FMODE_WRITE;
···
 		goto reject;
 	}
 
-	rnbd_dev = rnbd_dev_open(full_path, open_flags);
-	if (IS_ERR(rnbd_dev)) {
-		pr_err("Opening device '%s' on session %s failed, failed to open the block device, err: %ld\n",
-		       full_path, srv_sess->sessname, PTR_ERR(rnbd_dev));
-		ret = PTR_ERR(rnbd_dev);
+	bdev = blkdev_get_by_path(full_path, open_flags, THIS_MODULE);
+	if (IS_ERR(bdev)) {
+		ret = PTR_ERR(bdev);
+		pr_err("Opening device '%s' on session %s failed, failed to open the block device, err: %d\n",
+		       full_path, srv_sess->sessname, ret);
 		goto free_path;
 	}
 
-	srv_dev = rnbd_srv_get_or_create_srv_dev(rnbd_dev, srv_sess,
+	srv_dev = rnbd_srv_get_or_create_srv_dev(bdev, srv_sess,
 						  open_msg->access_mode);
 	if (IS_ERR(srv_dev)) {
 		pr_err("Opening device '%s' on session %s failed, creating srv_dev failed, err: %ld\n",
 		       full_path, srv_sess->sessname, PTR_ERR(srv_dev));
 		ret = PTR_ERR(srv_dev);
-		goto rnbd_dev_close;
+		goto blkdev_put;
 	}
 
 	srv_sess_dev = rnbd_srv_create_set_sess_dev(srv_sess, open_msg,
-						     rnbd_dev, open_flags,
+						     bdev, open_flags,
 						     srv_dev);
 	if (IS_ERR(srv_sess_dev)) {
 		pr_err("Opening device '%s' on session %s failed, creating sess_dev failed, err: %ld\n",
···
 	 */
 	mutex_lock(&srv_dev->lock);
 	if (!srv_dev->dev_kobj.state_in_sysfs) {
-		ret = rnbd_srv_create_dev_sysfs(srv_dev, rnbd_dev->bdev);
+		ret = rnbd_srv_create_dev_sysfs(srv_dev, bdev);
 		if (ret) {
 			mutex_unlock(&srv_dev->lock);
 			rnbd_srv_err(srv_sess_dev,
···
 		mutex_unlock(&srv_dev->lock);
 	}
 	rnbd_put_srv_dev(srv_dev);
-rnbd_dev_close:
-	rnbd_dev_close(rnbd_dev);
+blkdev_put:
+	blkdev_put(bdev, open_flags);
 free_path:
 	kfree(full_path);
 reject:

drivers/block/rnbd/rnbd-srv.h

···
 struct rnbd_srv_sess_dev {
 	/* Entry inside rnbd_srv_dev struct */
 	struct list_head		dev_list;
-	struct rnbd_dev			*rnbd_dev;
+	struct block_device		*bdev;
 	struct rnbd_srv_session		*sess;
 	struct rnbd_srv_dev		*dev;
 	struct kobject                  kobj;

drivers/block/ublk_drv.c

···
 /* All UBLK_F_* have to be included into UBLK_F_ALL */
 #define UBLK_F_ALL (UBLK_F_SUPPORT_ZERO_COPY \
 		| UBLK_F_URING_CMD_COMP_IN_TASK \
-		| UBLK_F_NEED_GET_DATA)
+		| UBLK_F_NEED_GET_DATA \
+		| UBLK_F_USER_RECOVERY \
+		| UBLK_F_USER_RECOVERY_REISSUE)
 
 /* All UBLK_PARAM_TYPE_* should be included here */
 #define UBLK_PARAM_TYPE_ALL (UBLK_PARAM_TYPE_BASIC | UBLK_PARAM_TYPE_DISCARD)
···
 
 	unsigned long io_addr;	/* mapped vm address */
 	unsigned int max_io_sz;
-	bool abort_work_pending;
+	bool force_abort;
 	unsigned short nr_io_ready;	/* how many ios setup */
 	struct ublk_device *dev;
 	struct ublk_io ios[0];
···
 	 * monitor each queue's daemon periodically
 	 */
 	struct delayed_work	monitor_work;
+	struct work_struct	quiesce_work;
 	struct work_struct	stop_work;
 };
 
···
 
 	return round_up(ubq->q_depth * sizeof(struct ublksrv_io_desc),
 			PAGE_SIZE);
+}
+
+static inline bool ublk_queue_can_use_recovery_reissue(
+		struct ublk_queue *ubq)
+{
+	if ((ubq->flags & UBLK_F_USER_RECOVERY) &&
+			(ubq->flags & UBLK_F_USER_RECOVERY_REISSUE))
+		return true;
+	return false;
+}
+
+static inline bool ublk_queue_can_use_recovery(
+		struct ublk_queue *ubq)
+{
+	if (ubq->flags & UBLK_F_USER_RECOVERY)
+		return true;
+	return false;
+}
+
+static inline bool ublk_can_use_recovery(struct ublk_device *ub)
+{
+	if (ub->dev_info.flags & UBLK_F_USER_RECOVERY)
+		return true;
+	return false;
 }
 
 static void ublk_free_disk(struct gendisk *disk)
···
  * Also aborting may not be started yet, keep in mind that one failed
  * request may be issued by block layer again.
  */
-static void __ublk_fail_req(struct ublk_io *io, struct request *req)
+static void __ublk_fail_req(struct ublk_queue *ubq, struct ublk_io *io,
+		struct request *req)
 {
 	WARN_ON_ONCE(io->flags & UBLK_IO_FLAG_ACTIVE);
 
 	if (!(io->flags & UBLK_IO_FLAG_ABORTED)) {
 		io->flags |= UBLK_IO_FLAG_ABORTED;
-		blk_mq_end_request(req, BLK_STS_IOERR);
+		if (ublk_queue_can_use_recovery_reissue(ubq))
+			blk_mq_requeue_request(req, false);
+		else
+			blk_mq_end_request(req, BLK_STS_IOERR);
 	}
 }
 
···
 
 #define UBLK_REQUEUE_DELAY_MS	3
 
+static inline void __ublk_abort_rq(struct ublk_queue *ubq,
+		struct request *rq)
+{
+	/* We cannot process this rq so just requeue it. */
+	if (ublk_queue_can_use_recovery(ubq))
+		blk_mq_requeue_request(rq, false);
+	else
+		blk_mq_end_request(rq, BLK_STS_IOERR);
+
+	mod_delayed_work(system_wq, &ubq->dev->monitor_work, 0);
+}
+
 static inline void __ublk_rq_task_work(struct request *req)
 {
 	struct ublk_queue *ubq = req->mq_hctx->driver_data;
-	struct ublk_device *ub = ubq->dev;
 	int tag = req->tag;
 	struct ublk_io *io = &ubq->ios[tag];
-	bool task_exiting = current != ubq->ubq_daemon || ubq_daemon_is_dying(ubq);
 	unsigned int mapped_bytes;
 
 	pr_devel("%s: complete: op %d, qid %d tag %d io_flags %x addr %llx\n",
 			__func__, io->cmd->cmd_op, ubq->q_id, req->tag, io->flags,
 			ublk_get_iod(ubq, req->tag)->addr);
 
-	if (unlikely(task_exiting)) {
-		blk_mq_end_request(req, BLK_STS_IOERR);
-		mod_delayed_work(system_wq, &ub->monitor_work, 0);
+	/*
+	 * Task is exiting if either:
+	 *
+	 * (1) current != ubq_daemon.
+	 * io_uring_cmd_complete_in_task() tries to run task_work
+	 * in a workqueue if ubq_daemon(cmd's task) is PF_EXITING.
+	 *
+	 * (2) current->flags & PF_EXITING.
+	 */
+	if (unlikely(current != ubq->ubq_daemon || current->flags & PF_EXITING)) {
+		__ublk_abort_rq(ubq, req);
 		return;
 	}
 
···
 	res = ublk_setup_iod(ubq, rq);
 	if (unlikely(res != BLK_STS_OK))
 		return BLK_STS_IOERR;
+	/* With recovery feature enabled, force_abort is set in
+	 * ublk_stop_dev() before calling del_gendisk(). We have to
+	 * abort all requeued and new rqs here to let del_gendisk()
+	 * move on. Besides, we cannot not call io_uring_cmd_complete_in_task()
+	 * to avoid UAF on io_uring ctx.
+	 *
+	 * Note: force_abort is guaranteed to be seen because it is set
+	 * before request queue is unqiuesced.
+	 */
+	if (ublk_queue_can_use_recovery(ubq) && unlikely(ubq->force_abort))
+		return BLK_STS_IOERR;
 
 	blk_mq_start_request(bd->rq);
 
 	if (unlikely(ubq_daemon_is_dying(ubq))) {
  fail:
-		mod_delayed_work(system_wq, &ubq->dev->monitor_work, 0);
-		return BLK_STS_IOERR;
+		__ublk_abort_rq(ubq, rq);
+		return BLK_STS_OK;
 	}
 
 	if (ublk_can_use_task_work(ubq)) {
···
 			 */
 			rq = blk_mq_tag_to_rq(ub->tag_set.tags[ubq->q_id], i);
 			if (rq)
-				__ublk_fail_req(io, rq);
+				__ublk_fail_req(ubq, io, rq);
 		}
 	}
 	ublk_put_device(ub);
···
 		struct ublk_queue *ubq = ublk_get_queue(ub, i);
 
 		if (ubq_daemon_is_dying(ubq)) {
-			schedule_work(&ub->stop_work);
+			if (ublk_queue_can_use_recovery(ubq))
+				schedule_work(&ub->quiesce_work);
+			else
+				schedule_work(&ub->stop_work);
 
 			/* abort queue is for making forward progress */
 			ublk_abort_queue(ub, ubq);
···
 	}
 
 	/*
-	 * We can't schedule monitor work after ublk_remove() is started.
+	 * We can't schedule monitor work after ub's state is not UBLK_S_DEV_LIVE.
+	 * after ublk_remove() or __ublk_quiesce_dev() is started.
 	 *
 	 * No need ub->mutex, monitor work are canceled after state is marked
-	 * as DEAD, so DEAD state is observed reliably.
+	 * as not LIVE, so new state is observed reliably.
 	 */
-	if (ub->dev_info.state != UBLK_S_DEV_DEAD)
+	if (ub->dev_info.state == UBLK_S_DEV_LIVE)
 		schedule_delayed_work(&ub->monitor_work,
 				UBLK_DAEMON_MONITOR_PERIOD);
 }
···
 		ublk_cancel_queue(ublk_get_queue(ub, i));
 }
 
-static void ublk_stop_dev(struct ublk_device *ub)
+static bool ublk_check_inflight_rq(struct request *rq, void *data)
 {
+	bool *idle = data;
+
+	if (blk_mq_request_started(rq)) {
+		*idle = false;
+		return false;
+	}
+	return true;
+}
+
+static void ublk_wait_tagset_rqs_idle(struct ublk_device *ub)
+{
+	bool idle;
+
+	WARN_ON_ONCE(!blk_queue_quiesced(ub->ub_disk->queue));
+	while (true) {
+		idle = true;
+		blk_mq_tagset_busy_iter(&ub->tag_set,
+				ublk_check_inflight_rq, &idle);
+		if (idle)
+			break;
+		msleep(UBLK_REQUEUE_DELAY_MS);
+	}
+}
+
+static void __ublk_quiesce_dev(struct ublk_device *ub)
+{
+	pr_devel("%s: quiesce ub: dev_id %d state %s\n",
+			__func__, ub->dev_info.dev_id,
+			ub->dev_info.state == UBLK_S_DEV_LIVE ?
+			"LIVE" : "QUIESCED");
+	blk_mq_quiesce_queue(ub->ub_disk->queue);
+	ublk_wait_tagset_rqs_idle(ub);
+	ub->dev_info.state = UBLK_S_DEV_QUIESCED;
+	ublk_cancel_dev(ub);
+	/* we are going to release task_struct of ubq_daemon and resets
+	 * ->ubq_daemon to NULL. So in monitor_work, check on ubq_daemon causes UAF.
+	 * Besides, monitor_work is not necessary in QUIESCED state since we have
+	 * already scheduled quiesce_work and quiesced all ubqs.
+	 *
+	 * Do not let monitor_work schedule itself if state it QUIESCED. And we cancel
+	 * it here and re-schedule it in END_USER_RECOVERY to avoid UAF.
+	 */
+	cancel_delayed_work_sync(&ub->monitor_work);
+}
+
+static void ublk_quiesce_work_fn(struct work_struct *work)
+{
+	struct ublk_device *ub =
+		container_of(work, struct ublk_device, quiesce_work);
+
 	mutex_lock(&ub->mutex);
 	if (ub->dev_info.state != UBLK_S_DEV_LIVE)
 		goto unlock;
+	__ublk_quiesce_dev(ub);
+ unlock:
+	mutex_unlock(&ub->mutex);
+}
 
+static void ublk_unquiesce_dev(struct ublk_device *ub)
+{
+	int i;
+
+	pr_devel("%s: unquiesce ub: dev_id %d state %s\n",
+			__func__, ub->dev_info.dev_id,
+			ub->dev_info.state == UBLK_S_DEV_LIVE ?
+			"LIVE" : "QUIESCED");
+	/* quiesce_work has run. We let requeued rqs be aborted
+	 * before running fallback_wq. "force_abort" must be seen
+	 * after request queue is unqiuesced. Then del_gendisk()
+	 * can move on.
+	 */
+	for (i = 0; i < ub->dev_info.nr_hw_queues; i++)
+		ublk_get_queue(ub, i)->force_abort = true;
+
+	blk_mq_unquiesce_queue(ub->ub_disk->queue);
+	/* We may have requeued some rqs in ublk_quiesce_queue() */
+	blk_mq_kick_requeue_list(ub->ub_disk->queue);
+}
+
+static void ublk_stop_dev(struct ublk_device *ub)
+{
+	mutex_lock(&ub->mutex);
+	if (ub->dev_info.state == UBLK_S_DEV_DEAD)
+		goto unlock;
+	if (ublk_can_use_recovery(ub)) {
+		if (ub->dev_info.state == UBLK_S_DEV_LIVE)
+			__ublk_quiesce_dev(ub);
+		ublk_unquiesce_dev(ub);
+	}
 	del_gendisk(ub->ub_disk);
 	ub->dev_info.state = UBLK_S_DEV_DEAD;
 	ub->dev_info.ublksrv_pid = -1;
···
 {
 	ublk_stop_dev(ub);
 	cancel_work_sync(&ub->stop_work);
+	cancel_work_sync(&ub->quiesce_work);
 	cdev_device_del(&ub->cdev, &ub->cdev_dev);
 	put_device(&ub->cdev_dev);
 }
···
 		goto out_unlock;
 	mutex_init(&ub->mutex);
 	spin_lock_init(&ub->mm_lock);
+	INIT_WORK(&ub->quiesce_work, ublk_quiesce_work_fn);
 	INIT_WORK(&ub->stop_work, ublk_stop_work_fn);
 	INIT_DELAYED_WORK(&ub->monitor_work, ublk_daemon_monitor_work);
 
···
 
 	ublk_stop_dev(ub);
 	cancel_work_sync(&ub->stop_work);
+	cancel_work_sync(&ub->quiesce_work);
 
 	ublk_put_device(ub);
 	return 0;
···
 	return ret;
 }
 
+static void ublk_queue_reinit(struct ublk_device *ub, struct ublk_queue *ubq)
+{
+	int i;
+
+	WARN_ON_ONCE(!(ubq->ubq_daemon && ubq_daemon_is_dying(ubq)));
+	/* All old ioucmds have to be completed */
+	WARN_ON_ONCE(ubq->nr_io_ready);
+	/* old daemon is PF_EXITING, put it now */
+	put_task_struct(ubq->ubq_daemon);
+	/* We have to reset it to NULL, otherwise ub won't accept new FETCH_REQ */
+	ubq->ubq_daemon = NULL;
+
+	for (i = 0; i < ubq->q_depth; i++) {
+		struct ublk_io *io = &ubq->ios[i];
+
+		/* forget everything now and be ready for new FETCH_REQ */
+		io->flags = 0;
+		io->cmd = NULL;
+		io->addr = 0;
+	}
+}
+
+static int ublk_ctrl_start_recovery(struct io_uring_cmd *cmd)
+{
+	struct ublksrv_ctrl_cmd *header = (struct ublksrv_ctrl_cmd *)cmd->cmd;
+	struct ublk_device *ub;
+	int ret = -EINVAL;
+	int i;
+
+	ub = ublk_get_device_from_id(header->dev_id);
+	if (!ub)
+		return ret;
+
+	mutex_lock(&ub->mutex);
+	if (!ublk_can_use_recovery(ub))
+		goto out_unlock;
+	/*
+	 * START_RECOVERY is only allowd after:
+	 *
+	 * (1) UB_STATE_OPEN is not set, which means the dying process is exited
+	 *     and related io_uring ctx is freed so file struct of /dev/ublkcX is
+	 *     released.
+	 *
+	 * (2) UBLK_S_DEV_QUIESCED is set, which means the quiesce_work:
+	 *     (a)has quiesced request queue
+	 *     (b)has requeued every inflight rqs whose io_flags is ACTIVE
+	 *     (c)has requeued/aborted every inflight rqs whose io_flags is NOT ACTIVE
+	 *     (d)has completed/camceled all ioucmds owned by ther dying process
+	 */
+	if (test_bit(UB_STATE_OPEN, &ub->state) ||
+			ub->dev_info.state != UBLK_S_DEV_QUIESCED) {
+		ret = -EBUSY;
+		goto out_unlock;
+	}
+	pr_devel("%s: start recovery for dev id %d.\n", __func__, header->dev_id);
+	for (i = 0; i < ub->dev_info.nr_hw_queues; i++)
+		ublk_queue_reinit(ub, ublk_get_queue(ub, i));
+	/* set to NULL, otherwise new ubq_daemon cannot mmap the io_cmd_buf */
+	ub->mm = NULL;
+	ub->nr_queues_ready = 0;
+	init_completion(&ub->completion);
+	ret = 0;
+ out_unlock:
+	mutex_unlock(&ub->mutex);
+	ublk_put_device(ub);
+	return ret;
+}
+
+static int ublk_ctrl_end_recovery(struct io_uring_cmd *cmd)
+{
+	struct ublksrv_ctrl_cmd *header = (struct ublksrv_ctrl_cmd *)cmd->cmd;
+	int ublksrv_pid = (int)header->data[0];
+	struct ublk_device *ub;
+	int ret = -EINVAL;
+
+	ub = ublk_get_device_from_id(header->dev_id);
+	if (!ub)
+		return ret;
+
+	pr_devel("%s: Waiting for new ubq_daemons(nr: %d) are ready, dev id %d...\n",
+			__func__, ub->dev_info.nr_hw_queues, header->dev_id);
+	/* wait until new ubq_daemon sending all FETCH_REQ */
+	wait_for_completion_interruptible(&ub->completion);
+	pr_devel("%s: All new ubq_daemons(nr: %d) are ready, dev id %d\n",
+			__func__, ub->dev_info.nr_hw_queues, header->dev_id);
+
+	mutex_lock(&ub->mutex);
+	if (!ublk_can_use_recovery(ub))
+		goto out_unlock;
+
+	if (ub->dev_info.state != UBLK_S_DEV_QUIESCED) {
+		ret = -EBUSY;
+		goto out_unlock;
+	}
+	ub->dev_info.ublksrv_pid = ublksrv_pid;
+	pr_devel("%s: new ublksrv_pid %d, dev id %d\n",
+			__func__, ublksrv_pid, header->dev_id);
+	blk_mq_unquiesce_queue(ub->ub_disk->queue);
+	pr_devel("%s: queue unquiesced, dev id %d.\n",
+			__func__, header->dev_id);
+	blk_mq_kick_requeue_list(ub->ub_disk->queue);
+	ub->dev_info.state = UBLK_S_DEV_LIVE;
+	schedule_delayed_work(&ub->monitor_work, UBLK_DAEMON_MONITOR_PERIOD);
+	ret = 0;
+ out_unlock:
+	mutex_unlock(&ub->mutex);
+	ublk_put_device(ub);
+	return ret;
+}
+
 static int ublk_ctrl_uring_cmd(struct io_uring_cmd *cmd,
 		unsigned int issue_flags)
 {
···
 		break;
 	case UBLK_CMD_SET_PARAMS:
 		ret = ublk_ctrl_set_params(cmd);
+		break;
+	case UBLK_CMD_START_USER_RECOVERY:
+		ret = ublk_ctrl_start_recovery(cmd);
+		break;
+	case UBLK_CMD_END_USER_RECOVERY:
+		ret = ublk_ctrl_end_recovery(cmd);
 		break;
 	default:
 		break;

drivers/block/virtio_blk.c

···
 	NULL,
 };
 
-static int virtblk_map_queues(struct blk_mq_tag_set *set)
+static void virtblk_map_queues(struct blk_mq_tag_set *set)
 {
 	struct virtio_blk *vblk = set->driver_data;
 	int i, qoff;
···
 		else
 			blk_mq_virtio_map_queues(&set->map[i], vblk->vdev, 0);
 	}
-
-	return 0;
 }
 
 static void virtblk_complete_batch(struct io_comp_batch *iob)

drivers/block/zram/zram_drv.c

···
 		goto out;
 	}
 
-	strlcpy(file_name, buf, PATH_MAX);
+	strscpy(file_name, buf, PATH_MAX);
 	/* ignore trailing newline */
 	sz = strlen(file_name);
 	if (sz > 0 && file_name[sz - 1] == '\n')
···
 	char compressor[ARRAY_SIZE(zram->compressor)];
 	size_t sz;
 
-	strlcpy(compressor, buf, sizeof(compressor));
+	strscpy(compressor, buf, sizeof(compressor));
 	/* ignore trailing newline */
 	sz = strlen(compressor);
 	if (sz > 0 && compressor[sz - 1] == '\n')
···
 	if (ret)
 		goto out_cleanup_disk;
 
-	strlcpy(zram->compressor, default_compressor, sizeof(zram->compressor));
+	strscpy(zram->compressor, default_compressor, sizeof(zram->compressor));
 
 	zram_debugfs_register(zram);
 	pr_info("Added device: %s\n", zram->disk->disk_name);

drivers/md/bcache/bcache.h

···
  *
  * BTREE NODES:
  *
- * Our unit of allocation is a bucket, and we we can't arbitrarily allocate and
+ * Our unit of allocation is a bucket, and we can't arbitrarily allocate and
  * free smaller than a bucket - so, that's how big our btree nodes are.
  *
  * (If buckets are really big we'll only use part of the bucket for a btree node

drivers/md/bcache/bset.c

···
 		 *
 		 * Don't worry event 'out' is allocated from mempool, it can
 		 * still be swapped here. Because state->pool is a page mempool
-		 * creaated by by mempool_init_page_pool(), which allocates
+		 * created by mempool_init_page_pool(), which allocates
 		 * pages by alloc_pages() indeed.
 		 */
 

drivers/md/bcache/stats.h

···
 
 void bch_mark_cache_accounting(struct cache_set *c, struct bcache_device *d,
 			       bool hit, bool bypass);
-void bch_mark_cache_readahead(struct cache_set *c, struct bcache_device *d);
 void bch_mark_cache_miss_collision(struct cache_set *c,
 				   struct bcache_device *d);
 void bch_mark_sectors_bypassed(struct cache_set *c,

drivers/md/bcache/writeback.c

···
 	dc->writeback_rate_target = target;
 }
 
+static bool idle_counter_exceeded(struct cache_set *c)
+{
+	int counter, dev_nr;
+
+	/*
+	 * If c->idle_counter is overflow (idel for really long time),
+	 * reset as 0 and not set maximum rate this time for code
+	 * simplicity.
+	 */
+	counter = atomic_inc_return(&c->idle_counter);
+	if (counter <= 0) {
+		atomic_set(&c->idle_counter, 0);
+		return false;
+	}
+
+	dev_nr = atomic_read(&c->attached_dev_nr);
+	if (dev_nr == 0)
+		return false;
+
+	/*
+	 * c->idle_counter is increased by writeback thread of all
+	 * attached backing devices, in order to represent a rough
+	 * time period, counter should be divided by dev_nr.
+	 * Otherwise the idle time cannot be larger with more backing
+	 * device attached.
+	 * The following calculation equals to checking
+	 *	(counter / dev_nr) < (dev_nr * 6)
+	 */
+	if (counter < (dev_nr * dev_nr * 6))
+		return false;
+
+	return true;
+}
+
+/*
+ * Idle_counter is increased every time when update_writeback_rate() is
+ * called. If all backing devices attached to the same cache set have
+ * identical dc->writeback_rate_update_seconds values, it is about 6
+ * rounds of update_writeback_rate() on each backing device before
+ * c->at_max_writeback_rate is set to 1, and then max wrteback rate set
+ * to each dc->writeback_rate.rate.
+ * In order to avoid extra locking cost for counting exact dirty cached
+ * devices number, c->attached_dev_nr is used to calculate the idle
+ * throushold. It might be bigger if not all cached device are in write-
+ * back mode, but it still works well with limited extra rounds of
+ * update_writeback_rate().
+ */
 static bool set_at_max_writeback_rate(struct cache_set *c,
 				       struct cached_dev *dc)
 {
···
 	/* Don't set max writeback rate if gc is running */
 	if (!c->gc_mark_valid)
 		return false;
-	/*
-	 * Idle_counter is increased everytime when update_writeback_rate() is
-	 * called. If all backing devices attached to the same cache set have
-	 * identical dc->writeback_rate_update_seconds values, it is about 6
-	 * rounds of update_writeback_rate() on each backing device before
-	 * c->at_max_writeback_rate is set to 1, and then max wrteback rate set
-	 * to each dc->writeback_rate.rate.
-	 * In order to avoid extra locking cost for counting exact dirty cached
-	 * devices number, c->attached_dev_nr is used to calculate the idle
-	 * throushold. It might be bigger if not all cached device are in write-
-	 * back mode, but it still works well with limited extra rounds of
-	 * update_writeback_rate().
-	 */
-	if (atomic_inc_return(&c->idle_counter) <
-	    atomic_read(&c->attached_dev_nr) * 6)
+
+	if (!idle_counter_exceeded(c))
 		return false;
 
 	if (atomic_read(&c->at_max_writeback_rate) != 1)
···
 	dc->writeback_rate_change = 0;
 
 	/*
-	 * Check c->idle_counter and c->at_max_writeback_rate agagain in case
-	 * new I/O arrives during before set_at_max_writeback_rate() returns.
-	 * Then the writeback rate is set to 1, and its new value should be
-	 * decided via __update_writeback_rate().
+	 * In case new I/O arrives during before
+	 * set_at_max_writeback_rate() returns.
 	 */
-	if ((atomic_read(&c->idle_counter) <
-	     atomic_read(&c->attached_dev_nr) * 6) ||
+	if (!idle_counter_exceeded(c) ||
 	    !atomic_read(&c->at_max_writeback_rate))
 		return false;
 
···
 		}
 	}
 
-	if (dc->writeback_write_wq) {
-		flush_workqueue(dc->writeback_write_wq);
+	if (dc->writeback_write_wq)
 		destroy_workqueue(dc->writeback_write_wq);
-	}
+
 	cached_dev_put(dc);
 	wait_for_kthread_stop();
 

drivers/md/dm-table.c

···
 static int device_not_nowait_capable(struct dm_target *ti, struct dm_dev *dev,
 				     sector_t start, sector_t len, void *data)
 {
-	struct request_queue *q = bdev_get_queue(dev->bdev);
-
-	return !blk_queue_nowait(q);
+	return !bdev_nowait(dev->bdev);
 }
 
 static bool dm_table_supports_nowait(struct dm_table *t)

drivers/md/md.c

···
 			}
 		}
 		sysfs_notify_dirent_safe(rdev->sysfs_state);
-		nowait = nowait && blk_queue_nowait(bdev_get_queue(rdev->bdev));
+		nowait = nowait && bdev_nowait(rdev->bdev);
 	}
 
 	if (!bioset_initialized(&mddev->bio_set)) {
···
 	 * If the new disk does not support REQ_NOWAIT,
 	 * disable on the whole MD.
 	 */
-	if (!blk_queue_nowait(bdev_get_queue(rdev->bdev))) {
+	if (!bdev_nowait(rdev->bdev)) {
 		pr_info("%s: Disabling nowait because %pg does not support nowait\n",
 			mdname(mddev), rdev->bdev);
 		blk_queue_flag_clear(QUEUE_FLAG_NOWAIT, mddev->queue);
···
 	list_for_each(tmp,&all_mddevs)
 		if (!l--) {
 			mddev = list_entry(tmp, struct mddev, all_mddevs);
-			mddev_get(mddev);
 			if (!mddev_get(mddev))
 				continue;
 			spin_unlock(&all_mddevs_lock);

drivers/md/raid0.c

···
 		int len = 0;
 
 		for (k = 0; k < conf->strip_zone[j].nb_dev; k++)
-			len += snprintf(line+len, 200-len, "%s%pg", k?"/":"",
+			len += scnprintf(line+len, 200-len, "%s%pg", k?"/":"",
 				conf->devlist[j * raid_disks + k]->bdev);
 		pr_debug("md: zone%d=[%s]\n", j, line);
 

drivers/md/raid10.c

···
 
 #include "raid1-10.c"
 
+#define NULL_CMD
+#define cmd_before(conf, cmd) \
+	do { \
+		write_sequnlock_irq(&(conf)->resync_lock); \
+		cmd; \
+	} while (0)
+#define cmd_after(conf) write_seqlock_irq(&(conf)->resync_lock)
+
+#define wait_event_barrier_cmd(conf, cond, cmd) \
+	wait_event_cmd((conf)->wait_barrier, cond, cmd_before(conf, cmd), \
+		       cmd_after(conf))
+
+#define wait_event_barrier(conf, cond) \
+	wait_event_barrier_cmd(conf, cond, NULL_CMD)
+
 /*
  * for resync bio, r10bio pointer can be retrieved from the per-bio
  * 'struct resync_pages'.
···
 	mempool_free(r10_bio, &conf->r10buf_pool);
 
 	lower_barrier(conf);
+}
+
+static void wake_up_barrier(struct r10conf *conf)
+{
+	if (wq_has_sleeper(&conf->wait_barrier))
+		wake_up(&conf->wait_barrier);
 }
 
 static void reschedule_retry(struct r10bio *r10_bio)
···
 
 static void raise_barrier(struct r10conf *conf, int force)
 {
+	write_seqlock_irq(&conf->resync_lock);
 	BUG_ON(force && !conf->barrier);
-	spin_lock_irq(&conf->resync_lock);
 
 	/* Wait until no block IO is waiting (unless 'force') */
-	wait_event_lock_irq(conf->wait_barrier, force || !conf->nr_waiting,
-			    conf->resync_lock);
+	wait_event_barrier(conf, force || !conf->nr_waiting);
 
 	/* block any new IO from starting */
-	conf->barrier++;
+	WRITE_ONCE(conf->barrier, conf->barrier + 1);
 
 	/* Now wait for all pending IO to complete */
-	wait_event_lock_irq(conf->wait_barrier,
-			    !atomic_read(&conf->nr_pending) && conf->barrier < RESYNC_DEPTH,
-			    conf->resync_lock);
+	wait_event_barrier(conf, !atomic_read(&conf->nr_pending) &&
+				 conf->barrier < RESYNC_DEPTH);
 
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static void lower_barrier(struct r10conf *conf)
 {
 	unsigned long flags;
-	spin_lock_irqsave(&conf->resync_lock, flags);
-	conf->barrier--;
-	spin_unlock_irqrestore(&conf->resync_lock, flags);
+
+	write_seqlock_irqsave(&conf->resync_lock, flags);
+	WRITE_ONCE(conf->barrier, conf->barrier - 1);
+	write_sequnlock_irqrestore(&conf->resync_lock, flags);
 	wake_up(&conf->wait_barrier);
+}
+
+static bool stop_waiting_barrier(struct r10conf *conf)
+{
+	struct bio_list *bio_list = current->bio_list;
+
+	/* barrier is dropped */
+	if (!conf->barrier)
+		return true;
+
+	/*
+	 * If there are already pending requests (preventing the barrier from
+	 * rising completely), and the pre-process bio queue isn't empty, then
+	 * don't wait, as we need to empty that queue to get the nr_pending
+	 * count down.
+	 */
+	if (atomic_read(&conf->nr_pending) && bio_list &&
+	    (!bio_list_empty(&bio_list[0]) || !bio_list_empty(&bio_list[1])))
+		return true;
+
+	/* move on if recovery thread is blocked by us */
+	if (conf->mddev->thread->tsk == current &&
+	    test_bit(MD_RECOVERY_RUNNING, &conf->mddev->recovery) &&
+	    conf->nr_queued > 0)
+		return true;
+
+	return false;
+}
+
+static bool wait_barrier_nolock(struct r10conf *conf)
+{
+	unsigned int seq = read_seqbegin(&conf->resync_lock);
+
+	if (READ_ONCE(conf->barrier))
+		return false;
+
+	atomic_inc(&conf->nr_pending);
+	if (!read_seqretry(&conf->resync_lock, seq))
+		return true;
+
+	if (atomic_dec_and_test(&conf->nr_pending))
+		wake_up_barrier(conf);
+
+	return false;
 }
 
 static bool wait_barrier(struct r10conf *conf, bool nowait)
 {
 	bool ret = true;
 
-	spin_lock_irq(&conf->resync_lock);
+	if (wait_barrier_nolock(conf))
+		return true;
+
+	write_seqlock_irq(&conf->resync_lock);
 	if (conf->barrier) {
-		struct bio_list *bio_list = current->bio_list;
-		conf->nr_waiting++;
-		/* Wait for the barrier to drop.
-		 * However if there are already pending
-		 * requests (preventing the barrier from
-		 * rising completely), and the
-		 * pre-process bio queue isn't empty,
-		 * then don't wait, as we need to empty
-		 * that queue to get the nr_pending
-		 * count down.
-		 */
 		/* Return false when nowait flag is set */
 		if (nowait) {
 			ret = false;
 		} else {
+			conf->nr_waiting++;
 			raid10_log(conf->mddev, "wait barrier");
-			wait_event_lock_irq(conf->wait_barrier,
-					    !conf->barrier ||
-					    (atomic_read(&conf->nr_pending) &&
-					     bio_list &&
-					     (!bio_list_empty(&bio_list[0]) ||
-					      !bio_list_empty(&bio_list[1]))) ||
-					     /* move on if recovery thread is
-					      * blocked by us
-					      */
-					     (conf->mddev->thread->tsk == current &&
-					      test_bit(MD_RECOVERY_RUNNING,
-						       &conf->mddev->recovery) &&
-					      conf->nr_queued > 0),
-					    conf->resync_lock);
+			wait_event_barrier(conf, stop_waiting_barrier(conf));
+			conf->nr_waiting--;
 		}
-		conf->nr_waiting--;
 		if (!conf->nr_waiting)
 			wake_up(&conf->wait_barrier);
 	}
 	/* Only increment nr_pending when we wait */
 	if (ret)
 		atomic_inc(&conf->nr_pending);
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 	return ret;
 }
 
···
 {
 	if ((atomic_dec_and_test(&conf->nr_pending)) ||
 			(conf->array_freeze_pending))
-		wake_up(&conf->wait_barrier);
+		wake_up_barrier(conf);
 }
 
 static void freeze_array(struct r10conf *conf, int extra)
···
 	 * must match the number of pending IOs (nr_pending) before
 	 * we continue.
 	 */
-	spin_lock_irq(&conf->resync_lock);
+	write_seqlock_irq(&conf->resync_lock);
 	conf->array_freeze_pending++;
-	conf->barrier++;
+	WRITE_ONCE(conf->barrier, conf->barrier + 1);
 	conf->nr_waiting++;
-	wait_event_lock_irq_cmd(conf->wait_barrier,
-				atomic_read(&conf->nr_pending) == conf->nr_queued+extra,
-				conf->resync_lock,
-				flush_pending_writes(conf));
-
+	wait_event_barrier_cmd(conf, atomic_read(&conf->nr_pending) ==
+			conf->nr_queued + extra, flush_pending_writes(conf));
 	conf->array_freeze_pending--;
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static void unfreeze_array(struct r10conf *conf)
 {
 	/* reverse the effect of the freeze */
-	spin_lock_irq(&conf->resync_lock);
-	conf->barrier--;
+	write_seqlock_irq(&conf->resync_lock);
+	WRITE_ONCE(conf->barrier, conf->barrier - 1);
 	conf->nr_waiting--;
 	wake_up(&conf->wait_barrier);
-	spin_unlock_irq(&conf->resync_lock);
+	write_sequnlock_irq(&conf->resync_lock);
 }
 
 static sector_t choose_data_offset(struct r10bio *r10_bio,
···
 	__make_request(mddev, bio, sectors);
 
 	/* In case raid10d snuck in to freeze_array */
-	wake_up(&conf->wait_barrier);
+	wake_up_barrier(conf);
 	return true;
 }
 
···
  * Otherwise, it must be degraded:
  *	- recovery is interrupted.
  *	- &mddev->degraded is bumped.
-
+ *
  * @rdev is marked as &Faulty excluding case when array is failed and
  * &mddev->fail_last_dev is off.
  */
···
 	INIT_LIST_HEAD(&conf->retry_list);
 	INIT_LIST_HEAD(&conf->bio_end_io_list);
 
-	spin_lock_init(&conf->resync_lock);
+	seqlock_init(&conf->resync_lock);
 	init_waitqueue_head(&conf->wait_barrier);
 	atomic_set(&conf->nr_pending, 0);
 
···
 				rdev->new_raid_disk = rdev->raid_disk * 2;
 				rdev->sectors = size;
 			}
-		conf->barrier = 1;
+		WRITE_ONCE(conf->barrier, 1);
 	}
 
 	return conf;

drivers/md/raid10.h

···
 	/* queue pending writes and submit them on unplug */
 	struct bio_list		pending_bio_list;
 
-	spinlock_t		resync_lock;
+	seqlock_t		resync_lock;
 	atomic_t		nr_pending;
 	int			nr_waiting;
 	int			nr_queued;

drivers/md/raid5-cache.c

···
 					 * reclaimed.  if it's 0, reclaim spaces
 					 * used by io_units which are in
 					 * IO_UNIT_STRIPE_END state (eg, reclaim
-					 * dones't wait for specific io_unit
+					 * doesn't wait for specific io_unit
 					 * switching to IO_UNIT_STRIPE_END
 					 * state) */
 	wait_queue_head_t iounit_wait;
···
 	 * superblock is updated to new log tail. Updating superblock (either
 	 * directly call md_update_sb() or depend on md thread) must hold
 	 * reconfig mutex. On the other hand, raid5_quiesce is called with
-	 * reconfig_mutex hold. The first step of raid5_quiesce() is waitting
-	 * for all IO finish, hence waitting for reclaim thread, while reclaim
-	 * thread is calling this function and waitting for reconfig mutex. So
+	 * reconfig_mutex hold. The first step of raid5_quiesce() is waiting
+	 * for all IO finish, hence waiting for reclaim thread, while reclaim
+	 * thread is calling this function and waiting for reconfig mutex. So
 	 * there is a deadlock. We workaround this issue with a trylock.
 	 * FIXME: we could miss discard if we can't take reconfig mutex
 	 */
···
 {
 	struct stripe_head *sh;
 
-	sh = raid5_get_active_stripe(conf, stripe_sect, 0, noblock, 0);
+	sh = raid5_get_active_stripe(conf, NULL, stripe_sect,
+				     noblock ? R5_GAS_NOBLOCK : 0);
 	if (!sh)
 		return NULL;  /* no more stripe available */
 

drivers/md/raid5.c

···
  */
 
 #include <linux/blkdev.h>
+#include <linux/delay.h>
 #include <linux/kthread.h>
 #include <linux/raid/pq.h>
 #include <linux/async_tx.h>
···
  */
 static bool is_inactive_blocked(struct r5conf *conf, int hash)
 {
-	int active = atomic_read(&conf->active_stripes);
-
 	if (list_empty(conf->inactive_list + hash))
 		return false;
 
 	if (!test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state))
 		return true;
 
-	return active < (conf->max_nr_stripes * 3 / 4);
+	return (atomic_read(&conf->active_stripes) <
+		(conf->max_nr_stripes * 3 / 4));
 }
 
-static struct stripe_head *__raid5_get_active_stripe(struct r5conf *conf,
+struct stripe_head *raid5_get_active_stripe(struct r5conf *conf,
 		struct stripe_request_ctx *ctx, sector_t sector,
-		bool previous, bool noblock, bool noquiesce)
+		unsigned int flags)
 {
 	struct stripe_head *sh;
 	int hash = stripe_hash_locks_hash(conf, sector);
+	int previous = !!(flags & R5_GAS_PREVIOUS);
 
 	pr_debug("get_stripe, sector %llu\n", (unsigned long long)sector);
 
 	spin_lock_irq(conf->hash_locks + hash);
 
-retry:
-	if (!noquiesce && conf->quiesce) {
-		/*
-		 * Must release the reference to batch_last before waiting,
-		 * on quiesce, otherwise the batch_last will hold a reference
-		 * to a stripe and raid5_quiesce() will deadlock waiting for
-		 * active_stripes to go to zero.
-		 */
-		if (ctx && ctx->batch_last) {
-			raid5_release_stripe(ctx->batch_last);
-			ctx->batch_last = NULL;
+	for (;;) {
+		if (!(flags & R5_GAS_NOQUIESCE) && conf->quiesce) {
+			/*
+			 * Must release the reference to batch_last before
+			 * waiting, on quiesce, otherwise the batch_last will
+			 * hold a reference to a stripe and raid5_quiesce()
+			 * will deadlock waiting for active_stripes to go to
+			 * zero.
+			 */
+			if (ctx && ctx->batch_last) {
+				raid5_release_stripe(ctx->batch_last);
+				ctx->batch_last = NULL;
+			}
+
+			wait_event_lock_irq(conf->wait_for_quiescent,
+					    !conf->quiesce,
+					    *(conf->hash_locks + hash));
 		}
 
-		wait_event_lock_irq(conf->wait_for_quiescent, !conf->quiesce,
+		sh = find_get_stripe(conf, sector, conf->generation - previous,
+				     hash);
+		if (sh)
+			break;
+
+		if (!test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state)) {
+			sh = get_free_stripe(conf, hash);
+			if (sh) {
+				r5c_check_stripe_cache_usage(conf);
+				init_stripe(sh, sector, previous);
+				atomic_inc(&sh->count);
+				break;
+			}
+
+			if (!test_bit(R5_DID_ALLOC, &conf->cache_state))
+				set_bit(R5_ALLOC_MORE, &conf->cache_state);
+		}
+
+		if (flags & R5_GAS_NOBLOCK)
+			break;
+
+		set_bit(R5_INACTIVE_BLOCKED, &conf->cache_state);
+		r5l_wake_reclaim(conf->log, 0);
+		wait_event_lock_irq(conf->wait_for_stripe,
+				    is_inactive_blocked(conf, hash),
 				    *(conf->hash_locks + hash));
+		clear_bit(R5_INACTIVE_BLOCKED, &conf->cache_state);
 	}
 
-	sh = find_get_stripe(conf, sector, conf->generation - previous, hash);
-	if (sh)
-		goto out;
-
-	if (test_bit(R5_INACTIVE_BLOCKED, &conf->cache_state))
-		goto wait_for_stripe;
-
-	sh = get_free_stripe(conf, hash);
-	if (sh) {
-		r5c_check_stripe_cache_usage(conf);
-		init_stripe(sh, sector, previous);
-		atomic_inc(&sh->count);
-		goto out;
-	}
-
-	if (!test_bit(R5_DID_ALLOC, &conf->cache_state))
-		set_bit(R5_ALLOC_MORE, &conf->cache_state);
-
-wait_for_stripe:
-	if (noblock)
-		goto out;
-
-	set_bit(R5_INACTIVE_BLOCKED, &conf->cache_state);
-	r5l_wake_reclaim(conf->log, 0);
-	wait_event_lock_irq(conf->wait_for_stripe,
-			    is_inactive_blocked(conf, hash),
-			    *(conf->hash_locks + hash));
-	clear_bit(R5_INACTIVE_BLOCKED, &conf->cache_state);
-	goto retry;
-
-out:
 	spin_unlock_irq(conf->hash_locks + hash);
 	return sh;
-}
-
-struct stripe_head *raid5_get_active_stripe(struct r5conf *conf,
-		sector_t sector, bool previous, bool noblock, bool noquiesce)
-{
-	return __raid5_get_active_stripe(conf, NULL, sector, previous, noblock,
-					 noquiesce);
 }
 
 static bool is_full_stripe_write(struct stripe_head *sh)
···
 		 * back cache (prexor with orig_page, and then xor with
 		 * page) in the read path
 		 */
-		if (s->injournal && s->failed) {
+		if (s->to_read && s->injournal && s->failed) {
 			if (test_bit(STRIPE_R5C_CACHING, &sh->state))
 				r5c_make_stripe_write_out(sh);
 			goto out;
···
 			sector_t bn = raid5_compute_blocknr(sh, i, 1);
 			sector_t s = raid5_compute_sector(conf, bn, 0,
 							  &dd_idx, NULL);
-			sh2 = raid5_get_active_stripe(conf, s, 0, 1, 1);
+			sh2 = raid5_get_active_stripe(conf, NULL, s,
+				R5_GAS_NOBLOCK | R5_GAS_NOQUIESCE);
 			if (sh2 == NULL)
 				/* so far only the early blocks of this stripe
 				 * have been requested.  When later blocks
···
 	/* Finish reconstruct operations initiated by the expansion process */
 	if (sh->reconstruct_state == reconstruct_state_result) {
 		struct stripe_head *sh_src
-			= raid5_get_active_stripe(conf, sh->sector, 1, 1, 1);
+			= raid5_get_active_stripe(conf, NULL, sh->sector,
+					R5_GAS_PREVIOUS | R5_GAS_NOBLOCK |
+					R5_GAS_NOQUIESCE);
 		if (sh_src && test_bit(STRIPE_EXPAND_SOURCE, &sh_src->state)) {
 			/* sh cannot be written until sh_src has been read.
 			 * so arrange for sh to be delayed a little
···
 
 	if (is_badblock(rdev, sector, bio_sectors(raid_bio), &first_bad,
 			&bad_sectors)) {
-		bio_put(raid_bio);
 		rdev_dec_pending(rdev, mddev);
 		return 0;
 	}
···
 		DEFINE_WAIT(w);
 		int d;
 	again:
-		sh = raid5_get_active_stripe(conf, logical_sector, 0, 0, 0);
+		sh = raid5_get_active_stripe(conf, NULL, logical_sector, 0);
 		prepare_to_wait(&conf->wait_for_overlap, &w,
 				TASK_UNINTERRUPTIBLE);
 		set_bit(R5_Overlap, &sh->dev[sh->pd_idx].flags);
···
 	enum stripe_result ret;
 	struct stripe_head *sh;
 	sector_t new_sector;
-	int previous = 0;
+	int previous = 0, flags = 0;
 	int seq, dd_idx;
 
 	seq = read_seqcount_begin(&conf->gen_lock);
···
 	pr_debug("raid456: %s, sector %llu logical %llu\n", __func__,
 		 new_sector, logical_sector);
 
-	sh = __raid5_get_active_stripe(conf, ctx, new_sector, previous,
-				       (bi->bi_opf & REQ_RAHEAD), 0);
+	if (previous)
+		flags |= R5_GAS_PREVIOUS;
+	if (bi->bi_opf & REQ_RAHEAD)
+		flags |= R5_GAS_NOBLOCK;
+	sh = raid5_get_active_stripe(conf, ctx, new_sector, flags);
 	if (unlikely(!sh)) {
 		/* cannot get stripe, just give-up */
 		bi->bi_status = BLK_STS_IOERR;
···
 	for (i = 0; i < reshape_sectors; i += RAID5_STRIPE_SECTORS(conf)) {
 		int j;
 		int skipped_disk = 0;
-		sh = raid5_get_active_stripe(conf, stripe_addr+i, 0, 0, 1);
+		sh = raid5_get_active_stripe(conf, NULL, stripe_addr+i,
+					     R5_GAS_NOQUIESCE);
 		set_bit(STRIPE_EXPANDING, &sh->state);
 		atomic_inc(&conf->reshape_stripes);
 		/* If any of this stripe is beyond the end of the old
···
 	if (last_sector >= mddev->dev_sectors)
 		last_sector = mddev->dev_sectors - 1;
 	while (first_sector <= last_sector) {
-		sh = raid5_get_active_stripe(conf, first_sector, 1, 0, 1);
+		sh = raid5_get_active_stripe(conf, NULL, first_sector,
+				R5_GAS_PREVIOUS | R5_GAS_NOQUIESCE);
 		set_bit(STRIPE_EXPAND_SOURCE, &sh->state);
 		set_bit(STRIPE_HANDLE, &sh->state);
 		raid5_release_stripe(sh);
···
 
 	md_bitmap_cond_end_sync(mddev->bitmap, sector_nr, false);
 
-	sh = raid5_get_active_stripe(conf, sector_nr, 0, 1, 0);
+	sh = raid5_get_active_stripe(conf, NULL, sector_nr,
+				     R5_GAS_NOBLOCK);
 	if (sh == NULL) {
-		sh = raid5_get_active_stripe(conf, sector_nr, 0, 0, 0);
+		sh = raid5_get_active_stripe(conf, NULL, sector_nr, 0);
 		/* make sure we don't swamp the stripe cache if someone else
 		 * is trying to get access
 		 */
···
 			/* already done this stripe */
 			continue;
 
-		sh = raid5_get_active_stripe(conf, sector, 0, 1, 1);
-
+		sh = raid5_get_active_stripe(conf, NULL, sector,
+				R5_GAS_NOBLOCK | R5_GAS_NOQUIESCE);
 		if (!sh) {
 			/* failed to get a stripe - must wait */
 			conf->retry_read_aligned = raid_bio;
···
 			spin_unlock_irq(&conf->device_lock);
 			md_check_recovery(mddev);
 			spin_lock_irq(&conf->device_lock);
+
+			/*
+			 * Waiting on MD_SB_CHANGE_PENDING below may deadlock
+			 * seeing md_check_recovery() is needed to clear
+			 * the flag when using mdmon.
+			 */
+			continue;
 		}
+
+		wait_event_lock_irq(mddev->sb_wait,
+			!test_bit(MD_SB_CHANGE_PENDING, &mddev->sb_flags),
+			conf->device_lock);
 	}
 	pr_debug("%d stripes handled\n", handled);
 

drivers/md/raid5.h

···
 }
 #endif
 
-extern void md_raid5_kick_device(struct r5conf *conf);
-extern int raid5_set_cache_size(struct mddev *mddev, int size);
-extern sector_t raid5_compute_blocknr(struct stripe_head *sh, int i, int previous);
-extern void raid5_release_stripe(struct stripe_head *sh);
-extern sector_t raid5_compute_sector(struct r5conf *conf, sector_t r_sector,
-				     int previous, int *dd_idx,
-				     struct stripe_head *sh);
-extern struct stripe_head *
-raid5_get_active_stripe(struct r5conf *conf, sector_t sector,
-			bool previous, bool noblock, bool noquiesce);
-extern int raid5_calc_degraded(struct r5conf *conf);
-extern int r5c_journal_mode_set(struct mddev *mddev, int journal_mode);
+void md_raid5_kick_device(struct r5conf *conf);
+int raid5_set_cache_size(struct mddev *mddev, int size);
+sector_t raid5_compute_blocknr(struct stripe_head *sh, int i, int previous);
+void raid5_release_stripe(struct stripe_head *sh);
+sector_t raid5_compute_sector(struct r5conf *conf, sector_t r_sector,
+		int previous, int *dd_idx, struct stripe_head *sh);
+
+struct stripe_request_ctx;
+/* get stripe from previous generation (when reshaping) */
+#define R5_GAS_PREVIOUS		(1 << 0)
+/* do not block waiting for a free stripe */
+#define R5_GAS_NOBLOCK		(1 << 1)
+/* do not block waiting for quiesce to be released */
+#define R5_GAS_NOQUIESCE	(1 << 2)
+struct stripe_head *raid5_get_active_stripe(struct r5conf *conf,
+		struct stripe_request_ctx *ctx, sector_t sector,
+		unsigned int flags);
+
+int raid5_calc_degraded(struct r5conf *conf);
+int r5c_journal_mode_set(struct mddev *mddev, int journal_mode);
 #endif

drivers/nvme/host/core.c

···
 	return effects;
 }
 
-static void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects,
-			      struct nvme_command *cmd, int status)
+void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects,
+		       struct nvme_command *cmd, int status)
 {
 	if (effects & NVME_CMD_EFFECTS_CSE_MASK) {
 		nvme_unfreeze(ctrl);
···
 		break;
 	}
 }
+EXPORT_SYMBOL_NS_GPL(nvme_passthru_end, NVME_TARGET_PASSTHRU);
 
-int nvme_execute_passthru_rq(struct request *rq)
+int nvme_execute_passthru_rq(struct request *rq, u32 *effects)
 {
 	struct nvme_command *cmd = nvme_req(rq)->cmd;
 	struct nvme_ctrl *ctrl = nvme_req(rq)->ctrl;
 	struct nvme_ns *ns = rq->q->queuedata;
-	u32 effects;
-	int  ret;
 
-	effects = nvme_passthru_start(ctrl, ns, cmd->common.opcode);
-	ret = nvme_execute_rq(rq, false);
-	if (effects) /* nothing to be done for zero cmd effects */
-		nvme_passthru_end(ctrl, effects, cmd, ret);
-
-	return ret;
+	*effects = nvme_passthru_start(ctrl, ns, cmd->common.opcode);
+	return nvme_execute_rq(rq, false);
 }
 EXPORT_SYMBOL_NS_GPL(nvme_execute_passthru_rq, NVME_TARGET_PASSTHRU);
 
···
 	if(!(ctrl->quirks & NVME_QUIRK_IGNORE_DEV_SUBNQN)) {
 		nqnlen = strnlen(id->subnqn, NVMF_NQN_SIZE);
 		if (nqnlen > 0 && nqnlen < NVMF_NQN_SIZE) {
-			strlcpy(subsys->subnqn, id->subnqn, NVMF_NQN_SIZE);
+			strscpy(subsys->subnqn, id->subnqn, NVMF_NQN_SIZE);
 			return;
 		}
 
···
 			dev_warn(ctrl->device, "missing or invalid SUBNQN field.\n");
 	}
 
-	/* Generate a "fake" NQN per Figure 254 in NVMe 1.3 + ECN 001 */
+	/*
+	 * Generate a "fake" NQN similar to the one in Section 4.5 of the NVMe
+	 * Base Specification 2.0.  It is slightly different from the format
+	 * specified there due to historic reasons, and we can't change it now.
+	 */
 	off = snprintf(subsys->subnqn, NVMF_NQN_SIZE,
 			"nqn.2014.08.org.nvmexpress:%04x%04x",
 			le16_to_cpu(id->vid), le16_to_cpu(id->ssvid));
···
 	nvme_init_subnqn(subsys, ctrl, id);
 	memcpy(subsys->serial, id->sn, sizeof(subsys->serial));
 	memcpy(subsys->model, id->mn, sizeof(subsys->model));
-	memcpy(subsys->firmware_rev, id->fr, sizeof(subsys->firmware_rev));
 	subsys->vendor_id = le16_to_cpu(id->vid);
 	subsys->cmic = id->cmic;
 
···
 				ctrl->quirks |= core_quirks[i].quirks;
 		}
 	}
+	memcpy(ctrl->subsys->firmware_rev, id->fr,
+	       sizeof(ctrl->subsys->firmware_rev));
 
 	if (force_apst && (ctrl->quirks & NVME_QUIRK_NO_DEEPEST_PS)) {
 		dev_warn(ctrl->device, "forcibly allowing all power states due to nvme_core.force_apst -- use at your own risk\n");
···
 }
 EXPORT_SYMBOL_GPL(nvme_complete_async_event);
 
+int nvme_alloc_admin_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set,
+		const struct blk_mq_ops *ops, unsigned int flags,
+		unsigned int cmd_size)
+{
+	int ret;
+
+	memset(set, 0, sizeof(*set));
+	set->ops = ops;
+	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
+	if (ctrl->ops->flags & NVME_F_FABRICS)
+		set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = ctrl->numa_node;
+	set->flags = flags;
+	set->cmd_size = cmd_size;
+	set->driver_data = ctrl;
+	set->nr_hw_queues = 1;
+	set->timeout = NVME_ADMIN_TIMEOUT;
+	ret = blk_mq_alloc_tag_set(set);
+	if (ret)
+		return ret;
+
+	ctrl->admin_q = blk_mq_init_queue(set);
+	if (IS_ERR(ctrl->admin_q)) {
+		ret = PTR_ERR(ctrl->admin_q);
+		goto out_free_tagset;
+	}
+
+	if (ctrl->ops->flags & NVME_F_FABRICS) {
+		ctrl->fabrics_q = blk_mq_init_queue(set);
+		if (IS_ERR(ctrl->fabrics_q)) {
+			ret = PTR_ERR(ctrl->fabrics_q);
+			goto out_cleanup_admin_q;
+		}
+	}
+
+	ctrl->admin_tagset = set;
+	return 0;
+
+out_cleanup_admin_q:
+	blk_mq_destroy_queue(ctrl->fabrics_q);
+out_free_tagset:
+	blk_mq_free_tag_set(ctrl->admin_tagset);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(nvme_alloc_admin_tag_set);
+
+void nvme_remove_admin_tag_set(struct nvme_ctrl *ctrl)
+{
+	blk_mq_destroy_queue(ctrl->admin_q);
+	if (ctrl->ops->flags & NVME_F_FABRICS)
+		blk_mq_destroy_queue(ctrl->fabrics_q);
+	blk_mq_free_tag_set(ctrl->admin_tagset);
+}
+EXPORT_SYMBOL_GPL(nvme_remove_admin_tag_set);
+
+int nvme_alloc_io_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set,
+		const struct blk_mq_ops *ops, unsigned int flags,
+		unsigned int cmd_size)
+{
+	int ret;
+
+	memset(set, 0, sizeof(*set));
+	set->ops = ops;
+	set->queue_depth = ctrl->sqsize + 1;
+	set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = ctrl->numa_node;
+	set->flags = flags;
+	set->cmd_size = cmd_size,
+	set->driver_data = ctrl;
+	set->nr_hw_queues = ctrl->queue_count - 1;
+	set->timeout = NVME_IO_TIMEOUT;
+	if (ops->map_queues)
+		set->nr_maps = ctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
+	ret = blk_mq_alloc_tag_set(set);
+	if (ret)
+		return ret;
+
+	if (ctrl->ops->flags & NVME_F_FABRICS) {
+		ctrl->connect_q = blk_mq_init_queue(set);
+        	if (IS_ERR(ctrl->connect_q)) {
+			ret = PTR_ERR(ctrl->connect_q);
+			goto out_free_tag_set;
+		}
+	}
+
+	ctrl->tagset = set;
+	return 0;
+
+out_free_tag_set:
+	blk_mq_free_tag_set(set);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(nvme_alloc_io_tag_set);
+
+void nvme_remove_io_tag_set(struct nvme_ctrl *ctrl)
+{
+	if (ctrl->ops->flags & NVME_F_FABRICS)
+		blk_mq_destroy_queue(ctrl->connect_q);
+	blk_mq_free_tag_set(ctrl->tagset);
+}
+EXPORT_SYMBOL_GPL(nvme_remove_io_tag_set);
+
 void nvme_stop_ctrl(struct nvme_ctrl *ctrl)
 {
 	nvme_mpath_stop(ctrl);
···
 	nvme_start_keep_alive(ctrl);
 
 	nvme_enable_aen(ctrl);
+
+	/*
+	 * persistent discovery controllers need to send indication to userspace
+	 * to re-read the discovery log page to learn about possible changes
+	 * that were missed. We identify persistent discovery controllers by
+	 * checking that they started once before, hence are reconnecting back.
+	 */
+	if (test_and_set_bit(NVME_CTRL_STARTED_ONCE, &ctrl->flags) &&
+	    nvme_discovery_ctrl(ctrl))
+		nvme_change_uevent(ctrl, "NVME_EVENT=rediscover");
 
 	if (ctrl->queue_count > 1) {
 		nvme_queue_scan(ctrl);

drivers/nvme/host/fabrics.c

···
 		goto out_unlock;
 
 	kref_init(&host->ref);
-	strlcpy(host->nqn, hostnqn, NVMF_NQN_SIZE);
+	strscpy(host->nqn, hostnqn, NVMF_NQN_SIZE);
 
 	list_add_tail(&host->list, &nvmf_hosts);
 out_unlock:
···
 		return false;
 
 	/*
-	 * Checking the local address is rough. In most cases, none is specified
-	 * and the host port is selected by the stack.
+	 * Checking the local address or host interfaces is rough.
+	 *
+	 * In most cases, none is specified and the host port or
+	 * host interface is selected by the stack.
 	 *
 	 * Assume no match if:
-	 * -  local address is specified and address is not the same
-	 * -  local address is not specified but remote is, or vice versa
-	 *    (admin using specific host_traddr when it matters).
+	 * -  local address or host interface is specified and address
+	 *    or host interface is not the same
+	 * -  local address or host interface is not specified but
+	 *    remote is, or vice versa (admin using specific
+	 *    host_traddr/host_iface when it matters).
 	 */
 	if ((opts->mask & NVMF_OPT_HOST_TRADDR) &&
 	    (ctrl->opts->mask & NVMF_OPT_HOST_TRADDR)) {
···
 			return false;
 	} else if ((opts->mask & NVMF_OPT_HOST_TRADDR) ||
 		   (ctrl->opts->mask & NVMF_OPT_HOST_TRADDR)) {
+		return false;
+	}
+
+	if ((opts->mask & NVMF_OPT_HOST_IFACE) &&
+	    (ctrl->opts->mask & NVMF_OPT_HOST_IFACE)) {
+		if (strcmp(opts->host_iface, ctrl->opts->host_iface))
+			return false;
+	} else if ((opts->mask & NVMF_OPT_HOST_IFACE) ||
+		   (ctrl->opts->mask & NVMF_OPT_HOST_IFACE)) {
 		return false;
 	}
 

drivers/nvme/host/fc.c

···
 {
 	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(rq);
 
-	return __nvme_fc_exit_request(set->driver_data, op);
+	return __nvme_fc_exit_request(to_fc_ctrl(set->driver_data), op);
 }
 
 static int
···
 nvme_fc_init_request(struct blk_mq_tag_set *set, struct request *rq,
 		unsigned int hctx_idx, unsigned int numa_node)
 {
-	struct nvme_fc_ctrl *ctrl = set->driver_data;
+	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(set->driver_data);
 	struct nvme_fcp_op_w_sgl *op = blk_mq_rq_to_pdu(rq);
 	int queue_idx = (set == &ctrl->tag_set) ? hctx_idx + 1 : 0;
 	struct nvme_fc_queue *queue = &ctrl->queues[queue_idx];
···
 	}
 }
 
-static inline void
-__nvme_fc_init_hctx(struct blk_mq_hw_ctx *hctx, struct nvme_fc_ctrl *ctrl,
-		unsigned int qidx)
+static inline int
+__nvme_fc_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int qidx)
 {
+	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(data);
 	struct nvme_fc_queue *queue = &ctrl->queues[qidx];
 
 	hctx->driver_data = queue;
 	queue->hctx = hctx;
+	return 0;
 }
 
 static int
-nvme_fc_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
-		unsigned int hctx_idx)
+nvme_fc_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int hctx_idx)
 {
-	struct nvme_fc_ctrl *ctrl = data;
-
-	__nvme_fc_init_hctx(hctx, ctrl, hctx_idx + 1);
-
-	return 0;
+	return __nvme_fc_init_hctx(hctx, data, hctx_idx + 1);
 }
 
 static int
 nvme_fc_init_admin_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_fc_ctrl *ctrl = data;
-
-	__nvme_fc_init_hctx(hctx, ctrl, hctx_idx);
-
-	return 0;
+	return __nvme_fc_init_hctx(hctx, data, hctx_idx);
 }
 
 static void
···
 		container_of(ref, struct nvme_fc_ctrl, ref);
 	unsigned long flags;
 
-	if (ctrl->ctrl.tagset) {
-		blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-		blk_mq_free_tag_set(&ctrl->tag_set);
-	}
+	if (ctrl->ctrl.tagset)
+		nvme_remove_io_tag_set(&ctrl->ctrl);
 
 	/* remove from rport list */
 	spin_lock_irqsave(&ctrl->rport->lock, flags);
···
 	spin_unlock_irqrestore(&ctrl->rport->lock, flags);
 
 	nvme_start_admin_queue(&ctrl->ctrl);
-	blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-	blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-	blk_mq_free_tag_set(&ctrl->admin_tag_set);
+	nvme_remove_admin_tag_set(&ctrl->ctrl);
 
 	kfree(ctrl->queues);
 
···
 	nvme_fc_ctrl_put(ctrl);
 }
 
-static int nvme_fc_map_queues(struct blk_mq_tag_set *set)
+static void nvme_fc_map_queues(struct blk_mq_tag_set *set)
 {
-	struct nvme_fc_ctrl *ctrl = set->driver_data;
+	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(set->driver_data);
 	int i;
 
 	for (i = 0; i < set->nr_maps; i++) {
···
 		else
 			blk_mq_map_queues(map);
 	}
-	return 0;
 }
 
 static const struct blk_mq_ops nvme_fc_mq_ops = {
···
 
 	nvme_fc_init_io_queues(ctrl);
 
-	memset(&ctrl->tag_set, 0, sizeof(ctrl->tag_set));
-	ctrl->tag_set.ops = &nvme_fc_mq_ops;
-	ctrl->tag_set.queue_depth = ctrl->ctrl.opts->queue_size;
-	ctrl->tag_set.reserved_tags = NVMF_RESERVED_TAGS;
-	ctrl->tag_set.numa_node = ctrl->ctrl.numa_node;
-	ctrl->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
-	ctrl->tag_set.cmd_size =
-		struct_size((struct nvme_fcp_op_w_sgl *)NULL, priv,
-			    ctrl->lport->ops->fcprqst_priv_sz);
-	ctrl->tag_set.driver_data = ctrl;
-	ctrl->tag_set.nr_hw_queues = ctrl->ctrl.queue_count - 1;
-	ctrl->tag_set.timeout = NVME_IO_TIMEOUT;
-
-	ret = blk_mq_alloc_tag_set(&ctrl->tag_set);
+	ret = nvme_alloc_io_tag_set(&ctrl->ctrl, &ctrl->tag_set,
+			&nvme_fc_mq_ops, BLK_MQ_F_SHOULD_MERGE,
+			struct_size((struct nvme_fcp_op_w_sgl *)NULL, priv,
+				    ctrl->lport->ops->fcprqst_priv_sz));
 	if (ret)
 		return ret;
 
-	ctrl->ctrl.tagset = &ctrl->tag_set;
-
-	ret = nvme_ctrl_init_connect_q(&(ctrl->ctrl));
-	if (ret)
-		goto out_free_tag_set;
-
 	ret = nvme_fc_create_hw_io_queues(ctrl, ctrl->ctrl.sqsize + 1);
 	if (ret)
-		goto out_cleanup_blk_queue;
+		goto out_cleanup_tagset;
 
 	ret = nvme_fc_connect_io_queues(ctrl, ctrl->ctrl.sqsize + 1);
 	if (ret)
···
 
 out_delete_hw_queues:
 	nvme_fc_delete_hw_io_queues(ctrl);
-out_cleanup_blk_queue:
-	blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-out_free_tag_set:
-	blk_mq_free_tag_set(&ctrl->tag_set);
+out_cleanup_tagset:
+	nvme_remove_io_tag_set(&ctrl->ctrl);
 	nvme_fc_free_io_queues(ctrl);
 
 	/* force put free routine to ignore io queues */
···
 			"to maxcmd\n",
 			opts->queue_size, ctrl->ctrl.maxcmd);
 		opts->queue_size = ctrl->ctrl.maxcmd;
-	}
-
-	if (opts->queue_size > ctrl->ctrl.sqsize + 1) {
-		/* warn if sqsize is lower than queue_size */
-		dev_warn(ctrl->ctrl.device,
-			"queue_size %zu > ctrl sqsize %u, reducing "
-			"to sqsize\n",
-			opts->queue_size, ctrl->ctrl.sqsize + 1);
-		opts->queue_size = ctrl->ctrl.sqsize + 1;
+		ctrl->ctrl.sqsize = opts->queue_size - 1;
 	}
 
 	ret = nvme_fc_init_aen_ops(ctrl);
···
 
 	nvme_fc_init_queue(ctrl, 0);
 
-	memset(&ctrl->admin_tag_set, 0, sizeof(ctrl->admin_tag_set));
-	ctrl->admin_tag_set.ops = &nvme_fc_admin_mq_ops;
-	ctrl->admin_tag_set.queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-	ctrl->admin_tag_set.reserved_tags = NVMF_RESERVED_TAGS;
-	ctrl->admin_tag_set.numa_node = ctrl->ctrl.numa_node;
-	ctrl->admin_tag_set.cmd_size =
-		struct_size((struct nvme_fcp_op_w_sgl *)NULL, priv,
-			    ctrl->lport->ops->fcprqst_priv_sz);
-	ctrl->admin_tag_set.driver_data = ctrl;
-	ctrl->admin_tag_set.nr_hw_queues = 1;
-	ctrl->admin_tag_set.timeout = NVME_ADMIN_TIMEOUT;
-	ctrl->admin_tag_set.flags = BLK_MQ_F_NO_SCHED;
-
-	ret = blk_mq_alloc_tag_set(&ctrl->admin_tag_set);
+	ret = nvme_alloc_admin_tag_set(&ctrl->ctrl, &ctrl->admin_tag_set,
+			&nvme_fc_admin_mq_ops, BLK_MQ_F_NO_SCHED,
+			struct_size((struct nvme_fcp_op_w_sgl *)NULL, priv,
+				    ctrl->lport->ops->fcprqst_priv_sz));
 	if (ret)
 		goto out_free_queues;
-	ctrl->ctrl.admin_tagset = &ctrl->admin_tag_set;
-
-	ctrl->ctrl.fabrics_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-	if (IS_ERR(ctrl->ctrl.fabrics_q)) {
-		ret = PTR_ERR(ctrl->ctrl.fabrics_q);
-		goto out_free_admin_tag_set;
-	}
-
-	ctrl->ctrl.admin_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-	if (IS_ERR(ctrl->ctrl.admin_q)) {
-		ret = PTR_ERR(ctrl->ctrl.admin_q);
-		goto out_cleanup_fabrics_q;
-	}
 
 	/*
 	 * Would have been nice to init io queues tag set as well.
···
 
 	ret = nvme_init_ctrl(&ctrl->ctrl, dev, &nvme_fc_ctrl_ops, 0);
 	if (ret)
-		goto out_cleanup_admin_q;
+		goto out_cleanup_tagset;
 
 	/* at this point, teardown path changes to ref counting on nvme ctrl */
 
···
 
 	return ERR_PTR(-EIO);
 
-out_cleanup_admin_q:
-	blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-out_cleanup_fabrics_q:
-	blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-out_free_admin_tag_set:
-	blk_mq_free_tag_set(&ctrl->admin_tag_set);
+out_cleanup_tagset:
+	nvme_remove_admin_tag_set(&ctrl->ctrl);
 out_free_queues:
 	kfree(ctrl->queues);
 out_free_ida:

drivers/nvme/host/ioctl.c

···
 		unsigned bufflen, void __user *meta_buffer, unsigned meta_len,
 		u32 meta_seed, u64 *result, unsigned timeout, bool vec)
 {
+	struct nvme_ctrl *ctrl;
 	struct request *req;
 	void *meta = NULL;
 	struct bio *bio;
+	u32 effects;
 	int ret;
 
 	req = nvme_alloc_user_request(q, cmd, ubuffer, bufflen, meta_buffer,
···
 		return PTR_ERR(req);
 
 	bio = req->bio;
+	ctrl = nvme_req(req)->ctrl;
 
-	ret = nvme_execute_passthru_rq(req);
+	ret = nvme_execute_passthru_rq(req, &effects);
 
 	if (result)
 		*result = le64_to_cpu(nvme_req(req)->result.u64);
···
 	if (bio)
 		blk_rq_unmap_user(bio);
 	blk_mq_free_request(req);
+
+	if (effects)
+		nvme_passthru_end(ctrl, effects, cmd, ret);
+
 	return ret;
 }
 
···
 	case NVME_IOCTL_IO_CMD:
 		return nvme_dev_user_cmd(ctrl, argp);
 	case NVME_IOCTL_RESET:
+		if (!capable(CAP_SYS_ADMIN))
+			return -EACCES;
 		dev_warn(ctrl->device, "resetting controller\n");
 		return nvme_reset_ctrl_sync(ctrl);
 	case NVME_IOCTL_SUBSYS_RESET:
+		if (!capable(CAP_SYS_ADMIN))
+			return -EACCES;
 		return nvme_reset_subsystem(ctrl);
 	case NVME_IOCTL_RESCAN:
+		if (!capable(CAP_SYS_ADMIN))
+			return -EACCES;
 		nvme_queue_scan(ctrl);
 		return 0;
 	default:

drivers/nvme/host/nvme.h

···
 #endif
 };
 
+enum nvme_ctrl_flags {
+	NVME_CTRL_FAILFAST_EXPIRED	= 0,
+	NVME_CTRL_ADMIN_Q_STOPPED	= 1,
+	NVME_CTRL_STARTED_ONCE		= 2,
+};
+
 struct nvme_ctrl {
 	bool comp_seen;
 	enum nvme_ctrl_state state;
···
 	u16 maxcmd;
 	int nr_reconnects;
 	unsigned long flags;
-#define NVME_CTRL_FAILFAST_EXPIRED	0
-#define NVME_CTRL_ADMIN_Q_STOPPED	1
 	struct nvmf_ctrl_options *opts;
 
 	struct page *discard_page;
···
 static inline void nvme_should_fail(struct request *req) {}
 #endif
 
+bool nvme_wait_reset(struct nvme_ctrl *ctrl);
+int nvme_try_sched_reset(struct nvme_ctrl *ctrl);
+
 static inline int nvme_reset_subsystem(struct nvme_ctrl *ctrl)
 {
+	int ret;
+
 	if (!ctrl->subsystem)
 		return -ENOTTY;
-	return ctrl->ops->reg_write32(ctrl, NVME_REG_NSSR, 0x4E564D65);
+	if (!nvme_wait_reset(ctrl))
+		return -EBUSY;
+
+	ret = ctrl->ops->reg_write32(ctrl, NVME_REG_NSSR, 0x4E564D65);
+	if (ret)
+		return ret;
+
+	return nvme_try_sched_reset(ctrl);
 }
 
 /*
···
 void nvme_cancel_admin_tagset(struct nvme_ctrl *ctrl);
 bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
 		enum nvme_ctrl_state new_state);
-bool nvme_wait_reset(struct nvme_ctrl *ctrl);
 int nvme_disable_ctrl(struct nvme_ctrl *ctrl);
 int nvme_enable_ctrl(struct nvme_ctrl *ctrl);
 int nvme_shutdown_ctrl(struct nvme_ctrl *ctrl);
···
 void nvme_start_ctrl(struct nvme_ctrl *ctrl);
 void nvme_stop_ctrl(struct nvme_ctrl *ctrl);
 int nvme_init_ctrl_finish(struct nvme_ctrl *ctrl);
+int nvme_alloc_admin_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set,
+		const struct blk_mq_ops *ops, unsigned int flags,
+		unsigned int cmd_size);
+void nvme_remove_admin_tag_set(struct nvme_ctrl *ctrl);
+int nvme_alloc_io_tag_set(struct nvme_ctrl *ctrl, struct blk_mq_tag_set *set,
+		const struct blk_mq_ops *ops, unsigned int flags,
+		unsigned int cmd_size);
+void nvme_remove_io_tag_set(struct nvme_ctrl *ctrl);
 
 void nvme_remove_namespaces(struct nvme_ctrl *ctrl);
 
···
 void nvme_stop_keep_alive(struct nvme_ctrl *ctrl);
 int nvme_reset_ctrl(struct nvme_ctrl *ctrl);
 int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl);
-int nvme_try_sched_reset(struct nvme_ctrl *ctrl);
 int nvme_delete_ctrl(struct nvme_ctrl *ctrl);
 void nvme_queue_scan(struct nvme_ctrl *ctrl);
 int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp, u8 csi,
···
 }
 #endif
 
-static inline int nvme_ctrl_init_connect_q(struct nvme_ctrl *ctrl)
-{
-	ctrl->connect_q = blk_mq_init_queue(ctrl->tagset);
-	if (IS_ERR(ctrl->connect_q))
-		return PTR_ERR(ctrl->connect_q);
-	return 0;
-}
-
 static inline struct nvme_ns *nvme_get_ns_from_dev(struct device *dev)
 {
 	return dev_to_disk(dev)->private_data;
···
 
 u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 			 u8 opcode);
-int nvme_execute_passthru_rq(struct request *rq);
+int nvme_execute_passthru_rq(struct request *rq, u32 *effects);
+void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects,
+		       struct nvme_command *cmd, int status);
 struct nvme_ctrl *nvme_ctrl_from_file(struct file *file);
 struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid);
 void nvme_put_ns(struct nvme_ns *ns);

drivers/nvme/host/pci.c

···
 struct nvme_iod {
 	struct nvme_request req;
 	struct nvme_command cmd;
-	struct nvme_queue *nvmeq;
 	bool use_sgl;
-	int aborted;
-	int npages;		/* In the PRP list. 0 means small pool in use */
-	dma_addr_t first_dma;
+	bool aborted;
+	s8 nr_allocations;	/* PRP list pool allocations. 0 means small
+				   pool in use */
 	unsigned int dma_len;	/* length of single DMA segment mapping */
+	dma_addr_t first_dma;
 	dma_addr_t meta_dma;
 	struct sg_table sgt;
 };
···
 {
 	struct nvme_dev *dev = set->driver_data;
 	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	int queue_idx = (set == &dev->tagset) ? hctx_idx + 1 : 0;
-	struct nvme_queue *nvmeq = &dev->queues[queue_idx];
-
-	BUG_ON(!nvmeq);
-	iod->nvmeq = nvmeq;
 
 	nvme_req(req)->ctrl = &dev->ctrl;
 	nvme_req(req)->cmd = &iod->cmd;
···
 	return 0;
 }
 
-static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
+static void nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 	int i, qoff, offset;
···
 		qoff += map->nr_queues;
 		offset += map->nr_queues;
 	}
-
-	return 0;
 }
 
 /*
···
 
 static inline bool nvme_pci_use_sgls(struct nvme_dev *dev, struct request *req)
 {
-	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
+	struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
 	int nseg = blk_rq_nr_phys_segments(req);
 	unsigned int avg_seg_size;
 
···
 
 	if (!nvme_ctrl_sgl_supported(&dev->ctrl))
 		return false;
-	if (!iod->nvmeq->qid)
+	if (!nvmeq->qid)
 		return false;
 	if (!sgl_threshold || avg_seg_size < sgl_threshold)
 		return false;
···
 	dma_addr_t dma_addr = iod->first_dma;
 	int i;
 
-	for (i = 0; i < iod->npages; i++) {
+	for (i = 0; i < iod->nr_allocations; i++) {
 		__le64 *prp_list = nvme_pci_iod_list(req)[i];
 		dma_addr_t next_dma_addr = le64_to_cpu(prp_list[last_prp]);
 
···
 	dma_addr_t dma_addr = iod->first_dma;
 	int i;
 
-	for (i = 0; i < iod->npages; i++) {
+	for (i = 0; i < iod->nr_allocations; i++) {
 		struct nvme_sgl_desc *sg_list = nvme_pci_iod_list(req)[i];
 		dma_addr_t next_dma_addr = le64_to_cpu((sg_list[last_sg]).addr);
 
···
 
 	dma_unmap_sgtable(dev->dev, &iod->sgt, rq_dma_dir(req), 0);
 
-	if (iod->npages == 0)
+	if (iod->nr_allocations == 0)
 		dma_pool_free(dev->prp_small_pool, nvme_pci_iod_list(req)[0],
 			      iod->first_dma);
 	else if (iod->use_sgl)
···
 	nprps = DIV_ROUND_UP(length, NVME_CTRL_PAGE_SIZE);
 	if (nprps <= (256 / 8)) {
 		pool = dev->prp_small_pool;
-		iod->npages = 0;
+		iod->nr_allocations = 0;
 	} else {
 		pool = dev->prp_page_pool;
-		iod->npages = 1;
+		iod->nr_allocations = 1;
 	}
 
 	prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma);
 	if (!prp_list) {
-		iod->npages = -1;
+		iod->nr_allocations = -1;
 		return BLK_STS_RESOURCE;
 	}
 	list[0] = prp_list;
···
 			prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma);
 			if (!prp_list)
 				goto free_prps;
-			list[iod->npages++] = prp_list;
+			list[iod->nr_allocations++] = prp_list;
 			prp_list[0] = old_prp_list[i - 1];
 			old_prp_list[i - 1] = cpu_to_le64(prp_dma);
 			i = 1;
···
 
 	if (entries <= (256 / sizeof(struct nvme_sgl_desc))) {
 		pool = dev->prp_small_pool;
-		iod->npages = 0;
+		iod->nr_allocations = 0;
 	} else {
 		pool = dev->prp_page_pool;
-		iod->npages = 1;
+		iod->nr_allocations = 1;
 	}
 
 	sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma);
 	if (!sg_list) {
-		iod->npages = -1;
+		iod->nr_allocations = -1;
 		return BLK_STS_RESOURCE;
 	}
 
···
 				goto free_sgls;
 
 			i = 0;
-			nvme_pci_iod_list(req)[iod->npages++] = sg_list;
+			nvme_pci_iod_list(req)[iod->nr_allocations++] = sg_list;
 			sg_list[i++] = *link;
 			nvme_pci_sgl_set_seg(link, sgl_dma, entries);
 		}
···
 	int rc;
 
 	if (blk_rq_nr_phys_segments(req) == 1) {
+		struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
 		struct bio_vec bv = req_bvec(req);
 
 		if (!is_pci_p2pdma_page(bv.bv_page)) {
···
 				return nvme_setup_prp_simple(dev, req,
 							     &cmnd->rw, &bv);
 
-			if (iod->nvmeq->qid && sgl_threshold &&
+			if (nvmeq->qid && sgl_threshold &&
 			    nvme_ctrl_sgl_supported(&dev->ctrl))
 				return nvme_setup_sgl_simple(dev, req,
 							     &cmnd->rw, &bv);
···
 	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
 	blk_status_t ret;
 
-	iod->aborted = 0;
-	iod->npages = -1;
+	iod->aborted = false;
+	iod->nr_allocations = -1;
 	iod->sgt.nents = 0;
 
 	ret = nvme_setup_cmd(req->q->queuedata, req);
···
 
 static __always_inline void nvme_pci_unmap_rq(struct request *req)
 {
-	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	struct nvme_dev *dev = iod->nvmeq->dev;
+	struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
+	struct nvme_dev *dev = nvmeq->dev;
 
-	if (blk_integrity_rq(req))
+	if (blk_integrity_rq(req)) {
+	        struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
+
 		dma_unmap_page(dev->dev, iod->meta_dma,
 			       rq_integrity_vec(req)->bv_len, rq_data_dir(req));
+	}
+
 	if (blk_rq_nr_phys_segments(req))
 		nvme_unmap_data(dev, req);
 }
···
 
 static void abort_endio(struct request *req, blk_status_t error)
 {
-	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	struct nvme_queue *nvmeq = iod->nvmeq;
+	struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
 
 	dev_warn(nvmeq->dev->ctrl.device,
 		 "Abort status: 0x%x", nvme_req(req)->status);
···
 static enum blk_eh_timer_return nvme_timeout(struct request *req)
 {
 	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	struct nvme_queue *nvmeq = iod->nvmeq;
+	struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
 	struct nvme_dev *dev = nvmeq->dev;
 	struct request *abort_req;
 	struct nvme_command cmd = { };
···
 		atomic_inc(&dev->ctrl.abort_limit);
 		return BLK_EH_RESET_TIMER;
 	}
-	iod->aborted = 1;
+	iod->aborted = true;
 
 	cmd.abort.opcode = nvme_admin_abort_cmd;
 	cmd.abort.cid = nvme_cid(req);
···
 
 	set->ops = &nvme_mq_ops;
 	set->nr_hw_queues = dev->online_queues - 1;
-	set->nr_maps = 2; /* default + read */
+	set->nr_maps = 1;
+	if (dev->io_queues[HCTX_TYPE_READ])
+		set->nr_maps = 2;
 	if (dev->io_queues[HCTX_TYPE_POLL])
-		set->nr_maps++;
+		set->nr_maps = 3;
 	set->timeout = NVME_IO_TIMEOUT;
 	set->numa_node = dev->ctrl.numa_node;
 	set->queue_depth = min_t(unsigned, dev->q_depth, BLK_MQ_MAX_DEPTH) - 1;
···
 		nvme_start_admin_queue(&dev->ctrl);
 	}
 
+	dma_set_min_align_mask(dev->dev, NVME_CTRL_PAGE_SIZE - 1);
+
 	/*
 	 * Limit the max command size to prevent iod->sg allocations going
 	 * over a single page.
···
 	 * Don't limit the IOMMU merged segment size.
 	 */
 	dma_set_max_seg_size(dev->dev, 0xffffffff);
-	dma_set_min_align_mask(dev->dev, NVME_CTRL_PAGE_SIZE - 1);
 
 	mutex_unlock(&dev->shutdown_lock);
 
···
 	BUILD_BUG_ON(sizeof(struct nvme_create_sq) != 64);
 	BUILD_BUG_ON(sizeof(struct nvme_delete_queue) != 64);
 	BUILD_BUG_ON(IRQ_AFFINITY_MAX_SETS < 2);
+	BUILD_BUG_ON(DIV_ROUND_UP(nvme_pci_npages_prp(), NVME_CTRL_PAGE_SIZE) >
+		     S8_MAX);
 
 	return pci_register_driver(&nvme_driver);
 }

drivers/nvme/host/rdma.c

···
 		struct request *rq, unsigned int hctx_idx,
 		unsigned int numa_node)
 {
-	struct nvme_rdma_ctrl *ctrl = set->driver_data;
+	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(set->driver_data);
 	struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq);
 	int queue_idx = (set == &ctrl->tag_set) ? hctx_idx + 1 : 0;
 	struct nvme_rdma_queue *queue = &ctrl->queues[queue_idx];
···
 static int nvme_rdma_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_rdma_ctrl *ctrl = data;
+	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(data);
 	struct nvme_rdma_queue *queue = &ctrl->queues[hctx_idx + 1];
 
 	BUG_ON(hctx_idx >= ctrl->ctrl.queue_count);
···
 static int nvme_rdma_init_admin_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_rdma_ctrl *ctrl = data;
+	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(data);
 	struct nvme_rdma_queue *queue = &ctrl->queues[0];
 
 	BUG_ON(hctx_idx != 0);
···
 	return ret;
 }
 
-static int nvme_rdma_start_io_queues(struct nvme_rdma_ctrl *ctrl)
+static int nvme_rdma_start_io_queues(struct nvme_rdma_ctrl *ctrl,
+				     int first, int last)
 {
 	int i, ret = 0;
 
-	for (i = 1; i < ctrl->ctrl.queue_count; i++) {
+	for (i = first; i < last; i++) {
 		ret = nvme_rdma_start_queue(ctrl, i);
 		if (ret)
 			goto out_stop_queues;
···
 	return 0;
 
 out_stop_queues:
-	for (i--; i >= 1; i--)
+	for (i--; i >= first; i--)
 		nvme_rdma_stop_queue(&ctrl->queues[i]);
 	return ret;
 }
···
 	return ret;
 }
 
-static int nvme_rdma_alloc_admin_tag_set(struct nvme_ctrl *nctrl)
+static int nvme_rdma_alloc_tag_set(struct nvme_ctrl *ctrl)
 {
-	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);
-	struct blk_mq_tag_set *set = &ctrl->admin_tag_set;
-	int ret;
+	unsigned int cmd_size = sizeof(struct nvme_rdma_request) +
+				NVME_RDMA_DATA_SGL_SIZE;
 
-	memset(set, 0, sizeof(*set));
-	set->ops = &nvme_rdma_admin_mq_ops;
-	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-	set->reserved_tags = NVMF_RESERVED_TAGS;
-	set->numa_node = nctrl->numa_node;
-	set->cmd_size = sizeof(struct nvme_rdma_request) +
-			NVME_RDMA_DATA_SGL_SIZE;
-	set->driver_data = ctrl;
-	set->nr_hw_queues = 1;
-	set->timeout = NVME_ADMIN_TIMEOUT;
-	set->flags = BLK_MQ_F_NO_SCHED;
-	ret = blk_mq_alloc_tag_set(set);
-	if (!ret)
-		ctrl->ctrl.admin_tagset = set;
-	return ret;
+	if (ctrl->max_integrity_segments)
+		cmd_size += sizeof(struct nvme_rdma_sgl) +
+			    NVME_RDMA_METADATA_SGL_SIZE;
+
+	return nvme_alloc_io_tag_set(ctrl, &to_rdma_ctrl(ctrl)->tag_set,
+			&nvme_rdma_mq_ops, BLK_MQ_F_SHOULD_MERGE, cmd_size);
 }
 
-static int nvme_rdma_alloc_tag_set(struct nvme_ctrl *nctrl)
+static void nvme_rdma_destroy_admin_queue(struct nvme_rdma_ctrl *ctrl)
 {
-	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);
-	struct blk_mq_tag_set *set = &ctrl->tag_set;
-	int ret;
-
-	memset(set, 0, sizeof(*set));
-	set->ops = &nvme_rdma_mq_ops;
-	set->queue_depth = nctrl->sqsize + 1;
-	set->reserved_tags = NVMF_RESERVED_TAGS;
-	set->numa_node = nctrl->numa_node;
-	set->flags = BLK_MQ_F_SHOULD_MERGE;
-	set->cmd_size = sizeof(struct nvme_rdma_request) +
-			NVME_RDMA_DATA_SGL_SIZE;
-	if (nctrl->max_integrity_segments)
-		set->cmd_size += sizeof(struct nvme_rdma_sgl) +
-				 NVME_RDMA_METADATA_SGL_SIZE;
-	set->driver_data = ctrl;
-	set->nr_hw_queues = nctrl->queue_count - 1;
-	set->timeout = NVME_IO_TIMEOUT;
-	set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
-	ret = blk_mq_alloc_tag_set(set);
-	if (!ret)
-		ctrl->ctrl.tagset = set;
-	return ret;
-}
-
-static void nvme_rdma_destroy_admin_queue(struct nvme_rdma_ctrl *ctrl,
-		bool remove)
-{
-	if (remove) {
-		blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-		blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-		blk_mq_free_tag_set(ctrl->ctrl.admin_tagset);
-	}
 	if (ctrl->async_event_sqe.data) {
 		cancel_work_sync(&ctrl->ctrl.async_event_work);
 		nvme_rdma_free_qe(ctrl->device->dev, &ctrl->async_event_sqe,
···
 		goto out_free_queue;
 
 	if (new) {
-		error = nvme_rdma_alloc_admin_tag_set(&ctrl->ctrl);
+		error = nvme_alloc_admin_tag_set(&ctrl->ctrl,
+				&ctrl->admin_tag_set, &nvme_rdma_admin_mq_ops,
+				BLK_MQ_F_NO_SCHED,
+				sizeof(struct nvme_rdma_request) +
+				NVME_RDMA_DATA_SGL_SIZE);
 		if (error)
 			goto out_free_async_qe;
 
-		ctrl->ctrl.fabrics_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-		if (IS_ERR(ctrl->ctrl.fabrics_q)) {
-			error = PTR_ERR(ctrl->ctrl.fabrics_q);
-			goto out_free_tagset;
-		}
-
-		ctrl->ctrl.admin_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-		if (IS_ERR(ctrl->ctrl.admin_q)) {
-			error = PTR_ERR(ctrl->ctrl.admin_q);
-			goto out_cleanup_fabrics_q;
-		}
 	}
 
 	error = nvme_rdma_start_queue(ctrl, 0);
 	if (error)
-		goto out_cleanup_queue;
+		goto out_remove_admin_tag_set;
 
 	error = nvme_enable_ctrl(&ctrl->ctrl);
 	if (error)
···
 out_stop_queue:
 	nvme_rdma_stop_queue(&ctrl->queues[0]);
 	nvme_cancel_admin_tagset(&ctrl->ctrl);
-out_cleanup_queue:
+out_remove_admin_tag_set:
 	if (new)
-		blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-out_cleanup_fabrics_q:
-	if (new)
-		blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-out_free_tagset:
-	if (new)
-		blk_mq_free_tag_set(ctrl->ctrl.admin_tagset);
+		nvme_remove_admin_tag_set(&ctrl->ctrl);
 out_free_async_qe:
 	if (ctrl->async_event_sqe.data) {
 		nvme_rdma_free_qe(ctrl->device->dev, &ctrl->async_event_sqe,
···
 	return error;
 }
 
-static void nvme_rdma_destroy_io_queues(struct nvme_rdma_ctrl *ctrl,
-		bool remove)
-{
-	if (remove) {
-		blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-		blk_mq_free_tag_set(ctrl->ctrl.tagset);
-	}
-	nvme_rdma_free_io_queues(ctrl);
-}
-
 static int nvme_rdma_configure_io_queues(struct nvme_rdma_ctrl *ctrl, bool new)
 {
-	int ret;
+	int ret, nr_queues;
 
 	ret = nvme_rdma_alloc_io_queues(ctrl);
 	if (ret)
···
 		ret = nvme_rdma_alloc_tag_set(&ctrl->ctrl);
 		if (ret)
 			goto out_free_io_queues;
-
-		ret = nvme_ctrl_init_connect_q(&(ctrl->ctrl));
-		if (ret)
-			goto out_free_tag_set;
 	}
 
-	ret = nvme_rdma_start_io_queues(ctrl);
+	/*
+	 * Only start IO queues for which we have allocated the tagset
+	 * and limitted it to the available queues. On reconnects, the
+	 * queue number might have changed.
+	 */
+	nr_queues = min(ctrl->tag_set.nr_hw_queues + 1, ctrl->ctrl.queue_count);
+	ret = nvme_rdma_start_io_queues(ctrl, 1, nr_queues);
 	if (ret)
-		goto out_cleanup_connect_q;
+		goto out_cleanup_tagset;
 
 	if (!new) {
 		nvme_start_queues(&ctrl->ctrl);
···
 		nvme_unfreeze(&ctrl->ctrl);
 	}
 
+	/*
+	 * If the number of queues has increased (reconnect case)
+	 * start all new queues now.
+	 */
+	ret = nvme_rdma_start_io_queues(ctrl, nr_queues,
+					ctrl->tag_set.nr_hw_queues + 1);
+	if (ret)
+		goto out_wait_freeze_timed_out;
+
 	return 0;
 
 out_wait_freeze_timed_out:
 	nvme_stop_queues(&ctrl->ctrl);
 	nvme_sync_io_queues(&ctrl->ctrl);
 	nvme_rdma_stop_io_queues(ctrl);
-out_cleanup_connect_q:
+out_cleanup_tagset:
 	nvme_cancel_tagset(&ctrl->ctrl);
 	if (new)
-		blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-out_free_tag_set:
-	if (new)
-		blk_mq_free_tag_set(ctrl->ctrl.tagset);
+		nvme_remove_io_tag_set(&ctrl->ctrl);
 out_free_io_queues:
 	nvme_rdma_free_io_queues(ctrl);
 	return ret;
···
 	blk_sync_queue(ctrl->ctrl.admin_q);
 	nvme_rdma_stop_queue(&ctrl->queues[0]);
 	nvme_cancel_admin_tagset(&ctrl->ctrl);
-	if (remove)
+	if (remove) {
 		nvme_start_admin_queue(&ctrl->ctrl);
-	nvme_rdma_destroy_admin_queue(ctrl, remove);
+		nvme_remove_admin_tag_set(&ctrl->ctrl);
+	}
+	nvme_rdma_destroy_admin_queue(ctrl);
 }
 
 static void nvme_rdma_teardown_io_queues(struct nvme_rdma_ctrl *ctrl,
···
 		nvme_sync_io_queues(&ctrl->ctrl);
 		nvme_rdma_stop_io_queues(ctrl);
 		nvme_cancel_tagset(&ctrl->ctrl);
-		if (remove)
+		if (remove) {
 			nvme_start_queues(&ctrl->ctrl);
-		nvme_rdma_destroy_io_queues(ctrl, remove);
+			nvme_remove_io_tag_set(&ctrl->ctrl);
+		}
+		nvme_rdma_free_io_queues(ctrl);
 	}
 }
 
···
 		nvme_sync_io_queues(&ctrl->ctrl);
 		nvme_rdma_stop_io_queues(ctrl);
 		nvme_cancel_tagset(&ctrl->ctrl);
-		nvme_rdma_destroy_io_queues(ctrl, new);
+		if (new)
+			nvme_remove_io_tag_set(&ctrl->ctrl);
+		nvme_rdma_free_io_queues(ctrl);
 	}
 destroy_admin:
 	nvme_stop_admin_queue(&ctrl->ctrl);
 	blk_sync_queue(ctrl->ctrl.admin_q);
 	nvme_rdma_stop_queue(&ctrl->queues[0]);
 	nvme_cancel_admin_tagset(&ctrl->ctrl);
-	nvme_rdma_destroy_admin_queue(ctrl, new);
+	if (new)
+		nvme_remove_admin_tag_set(&ctrl->ctrl);
+	nvme_rdma_destroy_admin_queue(ctrl);
 	return ret;
 }
 
···
 	nvme_complete_rq(rq);
 }
 
-static int nvme_rdma_map_queues(struct blk_mq_tag_set *set)
+static void nvme_rdma_map_queues(struct blk_mq_tag_set *set)
 {
-	struct nvme_rdma_ctrl *ctrl = set->driver_data;
+	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(set->driver_data);
 	struct nvmf_ctrl_options *opts = ctrl->ctrl.opts;
 
 	if (opts->nr_write_queues && ctrl->io_queues[HCTX_TYPE_READ]) {
···
 		ctrl->io_queues[HCTX_TYPE_DEFAULT],
 		ctrl->io_queues[HCTX_TYPE_READ],
 		ctrl->io_queues[HCTX_TYPE_POLL]);
-
-	return 0;
 }
 
 static const struct blk_mq_ops nvme_rdma_mq_ops = {

drivers/nvme/host/tcp.c

···
 	/* send state */
 	struct nvme_tcp_request *request;
 
-	int			queue_size;
 	u32			maxh2cdata;
 	size_t			cmnd_capsule_len;
 	struct nvme_tcp_ctrl	*ctrl;
···
 		struct request *rq, unsigned int hctx_idx,
 		unsigned int numa_node)
 {
-	struct nvme_tcp_ctrl *ctrl = set->driver_data;
+	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(set->driver_data);
 	struct nvme_tcp_request *req = blk_mq_rq_to_pdu(rq);
 	struct nvme_tcp_cmd_pdu *pdu;
 	int queue_idx = (set == &ctrl->tag_set) ? hctx_idx + 1 : 0;
···
 static int nvme_tcp_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_tcp_ctrl *ctrl = data;
+	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(data);
 	struct nvme_tcp_queue *queue = &ctrl->queues[hctx_idx + 1];
 
 	hctx->driver_data = queue;
···
 static int nvme_tcp_init_admin_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_tcp_ctrl *ctrl = data;
+	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(data);
 	struct nvme_tcp_queue *queue = &ctrl->queues[0];
 
 	hctx->driver_data = queue;
···
 	queue->io_cpu = cpumask_next_wrap(n - 1, cpu_online_mask, -1, false);
 }
 
-static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl,
-		int qid, size_t queue_size)
+static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl, int qid)
 {
 	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
 	struct nvme_tcp_queue *queue = &ctrl->queues[qid];
···
 	INIT_LIST_HEAD(&queue->send_list);
 	mutex_init(&queue->send_mutex);
 	INIT_WORK(&queue->io_work, nvme_tcp_io_work);
-	queue->queue_size = queue_size;
 
 	if (qid > 0)
 		queue->cmnd_capsule_len = nctrl->ioccsz * 16;
···
 	return ret;
 }
 
-static int nvme_tcp_alloc_admin_tag_set(struct nvme_ctrl *nctrl)
-{
-	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
-	struct blk_mq_tag_set *set = &ctrl->admin_tag_set;
-	int ret;
-
-	memset(set, 0, sizeof(*set));
-	set->ops = &nvme_tcp_admin_mq_ops;
-	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-	set->reserved_tags = NVMF_RESERVED_TAGS;
-	set->numa_node = nctrl->numa_node;
-	set->flags = BLK_MQ_F_BLOCKING;
-	set->cmd_size = sizeof(struct nvme_tcp_request);
-	set->driver_data = ctrl;
-	set->nr_hw_queues = 1;
-	set->timeout = NVME_ADMIN_TIMEOUT;
-	ret = blk_mq_alloc_tag_set(set);
-	if (!ret)
-		nctrl->admin_tagset = set;
-	return ret;
-}
-
-static int nvme_tcp_alloc_tag_set(struct nvme_ctrl *nctrl)
-{
-	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
-	struct blk_mq_tag_set *set = &ctrl->tag_set;
-	int ret;
-
-	memset(set, 0, sizeof(*set));
-	set->ops = &nvme_tcp_mq_ops;
-	set->queue_depth = nctrl->sqsize + 1;
-	set->reserved_tags = NVMF_RESERVED_TAGS;
-	set->numa_node = nctrl->numa_node;
-	set->flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING;
-	set->cmd_size = sizeof(struct nvme_tcp_request);
-	set->driver_data = ctrl;
-	set->nr_hw_queues = nctrl->queue_count - 1;
-	set->timeout = NVME_IO_TIMEOUT;
-	set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
-	ret = blk_mq_alloc_tag_set(set);
-	if (!ret)
-		nctrl->tagset = set;
-	return ret;
-}
-
 static void nvme_tcp_free_admin_queue(struct nvme_ctrl *ctrl)
 {
 	if (to_tcp_ctrl(ctrl)->async_req.pdu) {
···
 		nvme_tcp_stop_queue(ctrl, i);
 }
 
-static int nvme_tcp_start_io_queues(struct nvme_ctrl *ctrl)
+static int nvme_tcp_start_io_queues(struct nvme_ctrl *ctrl,
+				    int first, int last)
 {
 	int i, ret;
 
-	for (i = 1; i < ctrl->queue_count; i++) {
+	for (i = first; i < last; i++) {
 		ret = nvme_tcp_start_queue(ctrl, i);
 		if (ret)
 			goto out_stop_queues;
···
 	return 0;
 
 out_stop_queues:
-	for (i--; i >= 1; i--)
+	for (i--; i >= first; i--)
 		nvme_tcp_stop_queue(ctrl, i);
 	return ret;
 }
···
 {
 	int ret;
 
-	ret = nvme_tcp_alloc_queue(ctrl, 0, NVME_AQ_DEPTH);
+	ret = nvme_tcp_alloc_queue(ctrl, 0);
 	if (ret)
 		return ret;
 
···
 	int i, ret;
 
 	for (i = 1; i < ctrl->queue_count; i++) {
-		ret = nvme_tcp_alloc_queue(ctrl, i, ctrl->sqsize + 1);
+		ret = nvme_tcp_alloc_queue(ctrl, i);
 		if (ret)
 			goto out_free_queues;
 	}
···
 static void nvme_tcp_destroy_io_queues(struct nvme_ctrl *ctrl, bool remove)
 {
 	nvme_tcp_stop_io_queues(ctrl);
-	if (remove) {
-		blk_mq_destroy_queue(ctrl->connect_q);
-		blk_mq_free_tag_set(ctrl->tagset);
-	}
+	if (remove)
+		nvme_remove_io_tag_set(ctrl);
 	nvme_tcp_free_io_queues(ctrl);
 }
 
 static int nvme_tcp_configure_io_queues(struct nvme_ctrl *ctrl, bool new)
 {
-	int ret;
+	int ret, nr_queues;
 
 	ret = nvme_tcp_alloc_io_queues(ctrl);
 	if (ret)
 		return ret;
 
 	if (new) {
-		ret = nvme_tcp_alloc_tag_set(ctrl);
+		ret = nvme_alloc_io_tag_set(ctrl, &to_tcp_ctrl(ctrl)->tag_set,
+				&nvme_tcp_mq_ops,
+				BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING,
+				sizeof(struct nvme_tcp_request));
 		if (ret)
 			goto out_free_io_queues;
-
-		ret = nvme_ctrl_init_connect_q(ctrl);
-		if (ret)
-			goto out_free_tag_set;
 	}
 
-	ret = nvme_tcp_start_io_queues(ctrl);
+	/*
+	 * Only start IO queues for which we have allocated the tagset
+	 * and limitted it to the available queues. On reconnects, the
+	 * queue number might have changed.
+	 */
+	nr_queues = min(ctrl->tagset->nr_hw_queues + 1, ctrl->queue_count);
+	ret = nvme_tcp_start_io_queues(ctrl, 1, nr_queues);
 	if (ret)
 		goto out_cleanup_connect_q;
 
···
 		nvme_unfreeze(ctrl);
 	}
 
+	/*
+	 * If the number of queues has increased (reconnect case)
+	 * start all new queues now.
+	 */
+	ret = nvme_tcp_start_io_queues(ctrl, nr_queues,
+				       ctrl->tagset->nr_hw_queues + 1);
+	if (ret)
+		goto out_wait_freeze_timed_out;
+
 	return 0;
 
 out_wait_freeze_timed_out:
···
 out_cleanup_connect_q:
 	nvme_cancel_tagset(ctrl);
 	if (new)
-		blk_mq_destroy_queue(ctrl->connect_q);
-out_free_tag_set:
-	if (new)
-		blk_mq_free_tag_set(ctrl->tagset);
+		nvme_remove_io_tag_set(ctrl);
 out_free_io_queues:
 	nvme_tcp_free_io_queues(ctrl);
 	return ret;
···
 static void nvme_tcp_destroy_admin_queue(struct nvme_ctrl *ctrl, bool remove)
 {
 	nvme_tcp_stop_queue(ctrl, 0);
-	if (remove) {
-		blk_mq_destroy_queue(ctrl->admin_q);
-		blk_mq_destroy_queue(ctrl->fabrics_q);
-		blk_mq_free_tag_set(ctrl->admin_tagset);
-	}
+	if (remove)
+		nvme_remove_admin_tag_set(ctrl);
 	nvme_tcp_free_admin_queue(ctrl);
 }
 
···
 		return error;
 
 	if (new) {
-		error = nvme_tcp_alloc_admin_tag_set(ctrl);
+		error = nvme_alloc_admin_tag_set(ctrl,
+				&to_tcp_ctrl(ctrl)->admin_tag_set,
+				&nvme_tcp_admin_mq_ops, BLK_MQ_F_BLOCKING,
+				sizeof(struct nvme_tcp_request));
 		if (error)
 			goto out_free_queue;
-
-		ctrl->fabrics_q = blk_mq_init_queue(ctrl->admin_tagset);
-		if (IS_ERR(ctrl->fabrics_q)) {
-			error = PTR_ERR(ctrl->fabrics_q);
-			goto out_free_tagset;
-		}
-
-		ctrl->admin_q = blk_mq_init_queue(ctrl->admin_tagset);
-		if (IS_ERR(ctrl->admin_q)) {
-			error = PTR_ERR(ctrl->admin_q);
-			goto out_cleanup_fabrics_q;
-		}
 	}
 
 	error = nvme_tcp_start_queue(ctrl, 0);
 	if (error)
-		goto out_cleanup_queue;
+		goto out_cleanup_tagset;
 
 	error = nvme_enable_ctrl(ctrl);
 	if (error)
···
 out_stop_queue:
 	nvme_tcp_stop_queue(ctrl, 0);
 	nvme_cancel_admin_tagset(ctrl);
-out_cleanup_queue:
+out_cleanup_tagset:
 	if (new)
-		blk_mq_destroy_queue(ctrl->admin_q);
-out_cleanup_fabrics_q:
-	if (new)
-		blk_mq_destroy_queue(ctrl->fabrics_q);
-out_free_tagset:
-	if (new)
-		blk_mq_free_tag_set(ctrl->admin_tagset);
+		nvme_remove_admin_tag_set(ctrl);
 out_free_queue:
 	nvme_tcp_free_admin_queue(ctrl);
 	return error;
···
 	return BLK_STS_OK;
 }
 
-static int nvme_tcp_map_queues(struct blk_mq_tag_set *set)
+static void nvme_tcp_map_queues(struct blk_mq_tag_set *set)
 {
-	struct nvme_tcp_ctrl *ctrl = set->driver_data;
+	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(set->driver_data);
 	struct nvmf_ctrl_options *opts = ctrl->ctrl.opts;
 
 	if (opts->nr_write_queues && ctrl->io_queues[HCTX_TYPE_READ]) {
···
 		ctrl->io_queues[HCTX_TYPE_DEFAULT],
 		ctrl->io_queues[HCTX_TYPE_READ],
 		ctrl->io_queues[HCTX_TYPE_POLL]);
-
-	return 0;
 }
 
 static int nvme_tcp_poll(struct blk_mq_hw_ctx *hctx, struct io_comp_batch *iob)
···
 	nvme_tcp_try_recv(queue);
 	clear_bit(NVME_TCP_Q_POLLING, &queue->flags);
 	return queue->nr_cqe;
+}
+
+static int nvme_tcp_get_address(struct nvme_ctrl *ctrl, char *buf, int size)
+{
+	struct nvme_tcp_queue *queue = &to_tcp_ctrl(ctrl)->queues[0];
+	struct sockaddr_storage src_addr;
+	int ret, len;
+
+	len = nvmf_get_address(ctrl, buf, size);
+
+	ret = kernel_getsockname(queue->sock, (struct sockaddr *)&src_addr);
+	if (ret > 0) {
+		if (len > 0)
+			len--; /* strip trailing newline */
+		len += scnprintf(buf + len, size - len, "%ssrc_addr=%pISc\n",
+				(len) ? "," : "", &src_addr);
+	}
+
+	return len;
 }
 
 static const struct blk_mq_ops nvme_tcp_mq_ops = {
···
 	.free_ctrl		= nvme_tcp_free_ctrl,
 	.submit_async_event	= nvme_tcp_submit_async_event,
 	.delete_ctrl		= nvme_tcp_delete_ctrl,
-	.get_address		= nvmf_get_address,
+	.get_address		= nvme_tcp_get_address,
 	.stop_ctrl		= nvme_tcp_stop_ctrl,
 };
 

drivers/nvme/target/admin-cmd.c

···
 	if (req->port->inline_data_size)
 		id->sgls |= cpu_to_le32(1 << 20);
 
-	strlcpy(id->subnqn, ctrl->subsys->subsysnqn, sizeof(id->subnqn));
+	strscpy(id->subnqn, ctrl->subsys->subsysnqn, sizeof(id->subnqn));
 
 	/*
 	 * Max command capsule size is sqe + in-capsule data size.

drivers/nvme/target/configfs.c

···
 CONFIGFS_ATTR(nvmet_subsys_, attr_pi_enable);
 #endif
 
+static ssize_t nvmet_subsys_attr_qid_max_show(struct config_item *item,
+					      char *page)
+{
+	return snprintf(page, PAGE_SIZE, "%u\n", to_subsys(item)->max_qid);
+}
+
+static ssize_t nvmet_subsys_attr_qid_max_store(struct config_item *item,
+					       const char *page, size_t cnt)
+{
+	struct nvmet_port *port = to_nvmet_port(item);
+	u16 qid_max;
+
+	if (nvmet_is_port_enabled(port, __func__))
+		return -EACCES;
+
+	if (sscanf(page, "%hu\n", &qid_max) != 1)
+		return -EINVAL;
+
+	if (qid_max < 1 || qid_max > NVMET_NR_QUEUES)
+		return -EINVAL;
+
+	down_write(&nvmet_config_sem);
+	to_subsys(item)->max_qid = qid_max;
+	up_write(&nvmet_config_sem);
+	return cnt;
+}
+CONFIGFS_ATTR(nvmet_subsys_, attr_qid_max);
+
 static struct configfs_attribute *nvmet_subsys_attrs[] = {
 	&nvmet_subsys_attr_attr_allow_any_host,
 	&nvmet_subsys_attr_attr_version,
···
 	&nvmet_subsys_attr_attr_cntlid_min,
 	&nvmet_subsys_attr_attr_cntlid_max,
 	&nvmet_subsys_attr_attr_model,
+	&nvmet_subsys_attr_attr_qid_max,
 #ifdef CONFIG_BLK_DEV_INTEGRITY
 	&nvmet_subsys_attr_attr_pi_enable,
 #endif

drivers/nvme/target/core.c

···
 	}
 	init_completion(&sq->free_done);
 	init_completion(&sq->confirm_done);
+	nvmet_auth_sq_init(sq);
 
 	return 0;
 }

drivers/nvme/target/discovery.c

···
 
 	id->oaes = cpu_to_le32(NVMET_DISC_AEN_CFG_OPTIONAL);
 
-	strlcpy(id->subnqn, ctrl->subsys->subsysnqn, sizeof(id->subnqn));
+	strscpy(id->subnqn, ctrl->subsys->subsysnqn, sizeof(id->subnqn));
 
 	status = nvmet_copy_to_sgl(req, 0, id, sizeof(*id));
 

drivers/nvme/target/fabrics-cmd-auth.c

···
 	sq->dhchap_tid = -1;
 }
 
-void nvmet_init_auth(struct nvmet_ctrl *ctrl, struct nvmet_req *req)
+void nvmet_auth_sq_init(struct nvmet_sq *sq)
 {
-	u32 result = le32_to_cpu(req->cqe->result.u32);
-
 	/* Initialize in-band authentication */
-	INIT_DELAYED_WORK(&req->sq->auth_expired_work,
-			  nvmet_auth_expired_work);
-	req->sq->authenticated = false;
-	req->sq->dhchap_step = NVME_AUTH_DHCHAP_MESSAGE_NEGOTIATE;
-	result |= (u32)NVME_CONNECT_AUTHREQ_ATR << 16;
-	req->cqe->result.u32 = cpu_to_le32(result);
+	INIT_DELAYED_WORK(&sq->auth_expired_work, nvmet_auth_expired_work);
+	sq->authenticated = false;
+	sq->dhchap_step = NVME_AUTH_DHCHAP_MESSAGE_NEGOTIATE;
 }
 
 static u16 nvmet_auth_negotiate(struct nvmet_req *req, void *d)
···
 	return 0;
 }
 
-static u16 nvmet_auth_failure2(struct nvmet_req *req, void *d)
+static u16 nvmet_auth_failure2(void *d)
 {
 	struct nvmf_auth_dhchap_failure_data *data = d;
 
···
 	}
 
 	status = nvmet_copy_from_sgl(req, 0, d, tl);
-	if (status) {
-		kfree(d);
-		goto done;
-	}
+	if (status)
+		goto done_kfree;
 
 	data = d;
 	pr_debug("%s: ctrl %d qid %d type %d id %d step %x\n", __func__,
···
 		goto done_kfree;
 		break;
 	case NVME_AUTH_DHCHAP_MESSAGE_FAILURE2:
-		status = nvmet_auth_failure2(req, d);
+		status = nvmet_auth_failure2(d);
 		if (status) {
 			pr_warn("ctrl %d qid %d: authentication failed (%d)\n",
 				ctrl->cntlid, req->sq->qid, status);

drivers/nvme/target/fabrics-cmd.c

···
 	return ret;
 }
 
+static u32 nvmet_connect_result(struct nvmet_ctrl *ctrl)
+{
+	return (u32)ctrl->cntlid |
+		(nvmet_has_auth(ctrl) ? NVME_CONNECT_AUTHREQ_ATR : 0);
+}
+
 static void nvmet_execute_admin_connect(struct nvmet_req *req)
 {
 	struct nvmf_connect_command *c = &req->cmd->connect;
···
 		ctrl->cntlid, ctrl->subsys->subsysnqn, ctrl->hostnqn,
 		ctrl->pi_support ? " T10-PI is enabled" : "",
 		nvmet_has_auth(ctrl) ? " with DH-HMAC-CHAP" : "");
-	req->cqe->result.u16 = cpu_to_le16(ctrl->cntlid);
-
-	if (nvmet_has_auth(ctrl))
-		nvmet_init_auth(ctrl, req);
+	req->cqe->result.u32 = cpu_to_le32(nvmet_connect_result(ctrl));
 out:
 	kfree(d);
 complete:
···
 	if (status)
 		goto out_ctrl_put;
 
-	/* pass back cntlid for successful completion */
-	req->cqe->result.u16 = cpu_to_le16(ctrl->cntlid);
-
 	pr_debug("adding queue %d to ctrl %d.\n", qid, ctrl->cntlid);
-	req->cqe->result.u16 = cpu_to_le16(ctrl->cntlid);
-	if (nvmet_has_auth(ctrl))
-		nvmet_init_auth(ctrl, req);
-
+	req->cqe->result.u32 = cpu_to_le32(nvmet_connect_result(ctrl));
 out:
 	kfree(d);
 complete:

drivers/nvme/target/io-cmd-bdev.c

···
 
 void nvmet_bdev_set_limits(struct block_device *bdev, struct nvme_id_ns *id)
 {
-	const struct queue_limits *ql = &bdev_get_queue(bdev)->limits;
-	/* Number of logical blocks per physical block. */
-	const u32 lpp = ql->physical_block_size / ql->logical_block_size;
 	/* Logical blocks per physical block, 0's based. */
-	const __le16 lpp0b = to0based(lpp);
+	const __le16 lpp0b = to0based(bdev_physical_block_size(bdev) /
+				      bdev_logical_block_size(bdev));
 
 	/*
 	 * For NVMe 1.2 and later, bit 1 indicates that the fields NAWUN,
···
 	/* NPWA = Namespace Preferred Write Alignment. 0's based */
 	id->npwa = id->npwg;
 	/* NPDG = Namespace Preferred Deallocate Granularity. 0's based */
-	id->npdg = to0based(ql->discard_granularity / ql->logical_block_size);
+	id->npdg = to0based(bdev_discard_granularity(bdev) /
+			    bdev_logical_block_size(bdev));
 	/* NPDG = Namespace Preferred Deallocate Alignment */
 	id->npda = id->npdg;
 	/* NOWS = Namespace Optimal Write Size */
-	id->nows = to0based(ql->io_opt / ql->logical_block_size);
+	id->nows = to0based(bdev_io_opt(bdev) / bdev_logical_block_size(bdev));
 }
 
 void nvmet_bdev_ns_disable(struct nvmet_ns *ns)
···
 {
 	struct bio *bio = &req->b.inline_bio;
 
+	if (!bdev_write_cache(req->ns->bdev)) {
+		nvmet_req_complete(req, NVME_SC_SUCCESS);
+		return;
+	}
+
 	if (!nvmet_check_transfer_len(req, 0))
 		return;
 
···
 
 u16 nvmet_bdev_flush(struct nvmet_req *req)
 {
+	if (!bdev_write_cache(req->ns->bdev))
+		return 0;
+
 	if (blkdev_issue_flush(req->ns->bdev))
 		return NVME_SC_INTERNAL | NVME_SC_DNR;
 	return 0;

drivers/nvme/target/loop.c

···
 		struct request *req, unsigned int hctx_idx,
 		unsigned int numa_node)
 {
-	struct nvme_loop_ctrl *ctrl = set->driver_data;
+	struct nvme_loop_ctrl *ctrl = to_loop_ctrl(set->driver_data);
 	struct nvme_loop_iod *iod = blk_mq_rq_to_pdu(req);
 
 	nvme_req(req)->ctrl = &ctrl->ctrl;
···
 static int nvme_loop_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_loop_ctrl *ctrl = data;
+	struct nvme_loop_ctrl *ctrl = to_loop_ctrl(data);
 	struct nvme_loop_queue *queue = &ctrl->queues[hctx_idx + 1];
 
 	BUG_ON(hctx_idx >= ctrl->ctrl.queue_count);
···
 static int nvme_loop_init_admin_hctx(struct blk_mq_hw_ctx *hctx, void *data,
 		unsigned int hctx_idx)
 {
-	struct nvme_loop_ctrl *ctrl = data;
+	struct nvme_loop_ctrl *ctrl = to_loop_ctrl(data);
 	struct nvme_loop_queue *queue = &ctrl->queues[0];
 
 	BUG_ON(hctx_idx != 0);
···
 	if (!test_and_clear_bit(NVME_LOOP_Q_LIVE, &ctrl->queues[0].flags))
 		return;
 	nvmet_sq_destroy(&ctrl->queues[0].nvme_sq);
-	blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-	blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-	blk_mq_free_tag_set(&ctrl->admin_tag_set);
+	nvme_remove_admin_tag_set(&ctrl->ctrl);
 }
 
 static void nvme_loop_free_ctrl(struct nvme_ctrl *nctrl)
···
 	list_del(&ctrl->list);
 	mutex_unlock(&nvme_loop_ctrl_mutex);
 
-	if (nctrl->tagset) {
-		blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-		blk_mq_free_tag_set(&ctrl->tag_set);
-	}
+	if (nctrl->tagset)
+		nvme_remove_io_tag_set(nctrl);
 	kfree(ctrl->queues);
 	nvmf_free_options(nctrl->opts);
 free_ctrl:
···
 {
 	int error;
 
-	memset(&ctrl->admin_tag_set, 0, sizeof(ctrl->admin_tag_set));
-	ctrl->admin_tag_set.ops = &nvme_loop_admin_mq_ops;
-	ctrl->admin_tag_set.queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-	ctrl->admin_tag_set.reserved_tags = NVMF_RESERVED_TAGS;
-	ctrl->admin_tag_set.numa_node = ctrl->ctrl.numa_node;
-	ctrl->admin_tag_set.cmd_size = sizeof(struct nvme_loop_iod) +
-		NVME_INLINE_SG_CNT * sizeof(struct scatterlist);
-	ctrl->admin_tag_set.driver_data = ctrl;
-	ctrl->admin_tag_set.nr_hw_queues = 1;
-	ctrl->admin_tag_set.timeout = NVME_ADMIN_TIMEOUT;
-	ctrl->admin_tag_set.flags = BLK_MQ_F_NO_SCHED;
-
 	ctrl->queues[0].ctrl = ctrl;
 	error = nvmet_sq_init(&ctrl->queues[0].nvme_sq);
 	if (error)
 		return error;
 	ctrl->ctrl.queue_count = 1;
 
-	error = blk_mq_alloc_tag_set(&ctrl->admin_tag_set);
+	error = nvme_alloc_admin_tag_set(&ctrl->ctrl, &ctrl->admin_tag_set,
+			&nvme_loop_admin_mq_ops, BLK_MQ_F_NO_SCHED,
+			sizeof(struct nvme_loop_iod) +
+			NVME_INLINE_SG_CNT * sizeof(struct scatterlist));
 	if (error)
 		goto out_free_sq;
-	ctrl->ctrl.admin_tagset = &ctrl->admin_tag_set;
 
-	ctrl->ctrl.fabrics_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-	if (IS_ERR(ctrl->ctrl.fabrics_q)) {
-		error = PTR_ERR(ctrl->ctrl.fabrics_q);
-		goto out_free_tagset;
-	}
-
-	ctrl->ctrl.admin_q = blk_mq_init_queue(&ctrl->admin_tag_set);
-	if (IS_ERR(ctrl->ctrl.admin_q)) {
-		error = PTR_ERR(ctrl->ctrl.admin_q);
-		goto out_cleanup_fabrics_q;
-	}
 	/* reset stopped state for the fresh admin queue */
 	clear_bit(NVME_CTRL_ADMIN_Q_STOPPED, &ctrl->ctrl.flags);
 
 	error = nvmf_connect_admin_queue(&ctrl->ctrl);
 	if (error)
-		goto out_cleanup_queue;
+		goto out_cleanup_tagset;
 
 	set_bit(NVME_LOOP_Q_LIVE, &ctrl->queues[0].flags);
 
 	error = nvme_enable_ctrl(&ctrl->ctrl);
 	if (error)
-		goto out_cleanup_queue;
+		goto out_cleanup_tagset;
 
 	ctrl->ctrl.max_hw_sectors =
 		(NVME_LOOP_MAX_SEGMENTS - 1) << (PAGE_SHIFT - 9);
···
 
 	error = nvme_init_ctrl_finish(&ctrl->ctrl);
 	if (error)
-		goto out_cleanup_queue;
+		goto out_cleanup_tagset;
 
 	return 0;
 
-out_cleanup_queue:
+out_cleanup_tagset:
 	clear_bit(NVME_LOOP_Q_LIVE, &ctrl->queues[0].flags);
-	blk_mq_destroy_queue(ctrl->ctrl.admin_q);
-out_cleanup_fabrics_q:
-	blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
-out_free_tagset:
-	blk_mq_free_tag_set(&ctrl->admin_tag_set);
+	nvme_remove_admin_tag_set(&ctrl->ctrl);
 out_free_sq:
 	nvmet_sq_destroy(&ctrl->queues[0].nvme_sq);
 	return error;
···
 	if (ret)
 		return ret;
 
-	memset(&ctrl->tag_set, 0, sizeof(ctrl->tag_set));
-	ctrl->tag_set.ops = &nvme_loop_mq_ops;
-	ctrl->tag_set.queue_depth = ctrl->ctrl.opts->queue_size;
-	ctrl->tag_set.reserved_tags = NVMF_RESERVED_TAGS;
-	ctrl->tag_set.numa_node = ctrl->ctrl.numa_node;
-	ctrl->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
-	ctrl->tag_set.cmd_size = sizeof(struct nvme_loop_iod) +
-		NVME_INLINE_SG_CNT * sizeof(struct scatterlist);
-	ctrl->tag_set.driver_data = ctrl;
-	ctrl->tag_set.nr_hw_queues = ctrl->ctrl.queue_count - 1;
-	ctrl->tag_set.timeout = NVME_IO_TIMEOUT;
-	ctrl->ctrl.tagset = &ctrl->tag_set;
-
-	ret = blk_mq_alloc_tag_set(&ctrl->tag_set);
+	ret = nvme_alloc_io_tag_set(&ctrl->ctrl, &ctrl->tag_set,
+			&nvme_loop_mq_ops, BLK_MQ_F_SHOULD_MERGE,
+			sizeof(struct nvme_loop_iod) +
+			NVME_INLINE_SG_CNT * sizeof(struct scatterlist));
 	if (ret)
 		goto out_destroy_queues;
 
-	ret = nvme_ctrl_init_connect_q(&(ctrl->ctrl));
-	if (ret)
-		goto out_free_tagset;
-
 	ret = nvme_loop_connect_io_queues(ctrl);
 	if (ret)
-		goto out_cleanup_connect_q;
+		goto out_cleanup_tagset;
 
 	return 0;
 
-out_cleanup_connect_q:
-	blk_mq_destroy_queue(ctrl->ctrl.connect_q);
-out_free_tagset:
-	blk_mq_free_tag_set(&ctrl->tag_set);
+out_cleanup_tagset:
+	nvme_remove_io_tag_set(&ctrl->ctrl);
 out_destroy_queues:
 	nvme_loop_destroy_io_queues(ctrl);
 	return ret;
···
 
 	ret = -ENOMEM;
 
-	ctrl->ctrl.sqsize = opts->queue_size - 1;
 	ctrl->ctrl.kato = opts->kato;
 	ctrl->port = nvme_loop_find_port(&ctrl->ctrl);
 
···
 			opts->queue_size, ctrl->ctrl.maxcmd);
 		opts->queue_size = ctrl->ctrl.maxcmd;
 	}
+	ctrl->ctrl.sqsize = opts->queue_size - 1;
 
 	if (opts->nr_io_queues) {
 		ret = nvme_loop_create_io_queues(ctrl);

drivers/nvme/target/nvmet.h

···
 		       bool set_ctrl);
 int nvmet_auth_set_host_hash(struct nvmet_host *host, const char *hash);
 int nvmet_setup_auth(struct nvmet_ctrl *ctrl);
-void nvmet_init_auth(struct nvmet_ctrl *ctrl, struct nvmet_req *req);
+void nvmet_auth_sq_init(struct nvmet_sq *sq);
 void nvmet_destroy_auth(struct nvmet_ctrl *ctrl);
 void nvmet_auth_sq_free(struct nvmet_sq *sq);
 int nvmet_setup_dhgroup(struct nvmet_ctrl *ctrl, u8 dhgroup_id);
···
 {
 	return 0;
 }
-static inline void nvmet_init_auth(struct nvmet_ctrl *ctrl,
-				   struct nvmet_req *req) {};
+static inline void nvmet_auth_sq_init(struct nvmet_sq *sq)
+{
+}
 static inline void nvmet_destroy_auth(struct nvmet_ctrl *ctrl) {};
 static inline void nvmet_auth_sq_free(struct nvmet_sq *sq) {};
 static inline bool nvmet_check_auth_status(struct nvmet_req *req)

drivers/nvme/target/passthru.c

···
 {
 	struct nvmet_req *req = container_of(w, struct nvmet_req, p.work);
 	struct request *rq = req->p.rq;
+	struct nvme_ctrl *ctrl = nvme_req(rq)->ctrl;
+	u32 effects;
 	int status;
 
-	status = nvme_execute_passthru_rq(rq);
+	status = nvme_execute_passthru_rq(rq, &effects);
 
 	if (status == NVME_SC_SUCCESS &&
 	    req->cmd->common.opcode == nvme_admin_identify) {
···
 	req->cqe->result = nvme_req(rq)->result;
 	nvmet_req_complete(req, status);
 	blk_mq_free_request(rq);
+
+	if (effects)
+		nvme_passthru_end(ctrl, effects, req->cmd, status);
 }
 
 static void nvmet_passthru_req_done(struct request *rq,

drivers/nvme/target/tcp.c

···
 	u32				pdu_len;
 	u32				pdu_recv;
 	int				sg_idx;
-	int				nr_mapped;
 	struct msghdr			recv_msg;
-	struct kvec			*iov;
+	struct bio_vec			*iov;
 	u32				flags;
 
 	struct list_head		entry;
···
 static struct workqueue_struct *nvmet_tcp_wq;
 static const struct nvmet_fabrics_ops nvmet_tcp_ops;
 static void nvmet_tcp_free_cmd(struct nvmet_tcp_cmd *c);
-static void nvmet_tcp_finish_cmd(struct nvmet_tcp_cmd *cmd);
 static void nvmet_tcp_free_cmd_buffers(struct nvmet_tcp_cmd *cmd);
-static void nvmet_tcp_unmap_pdu_iovec(struct nvmet_tcp_cmd *cmd);
 
 static inline u16 nvmet_tcp_cmd_tag(struct nvmet_tcp_queue *queue,
 		struct nvmet_tcp_cmd *cmd)
···
 
 static void nvmet_tcp_free_cmd_buffers(struct nvmet_tcp_cmd *cmd)
 {
-	WARN_ON(unlikely(cmd->nr_mapped > 0));
-
 	kfree(cmd->iov);
 	sgl_free(cmd->req.sg);
 	cmd->iov = NULL;
 	cmd->req.sg = NULL;
 }
 
-static void nvmet_tcp_unmap_pdu_iovec(struct nvmet_tcp_cmd *cmd)
+static void nvmet_tcp_build_pdu_iovec(struct nvmet_tcp_cmd *cmd)
 {
-	struct scatterlist *sg;
-	int i;
-
-	sg = &cmd->req.sg[cmd->sg_idx];
-
-	for (i = 0; i < cmd->nr_mapped; i++)
-		kunmap(sg_page(&sg[i]));
-
-	cmd->nr_mapped = 0;
-}
-
-static void nvmet_tcp_map_pdu_iovec(struct nvmet_tcp_cmd *cmd)
-{
-	struct kvec *iov = cmd->iov;
+	struct bio_vec *iov = cmd->iov;
 	struct scatterlist *sg;
 	u32 length, offset, sg_offset;
+	int nr_pages;
 
 	length = cmd->pdu_len;
-	cmd->nr_mapped = DIV_ROUND_UP(length, PAGE_SIZE);
+	nr_pages = DIV_ROUND_UP(length, PAGE_SIZE);
 	offset = cmd->rbytes_done;
 	cmd->sg_idx = offset / PAGE_SIZE;
 	sg_offset = offset % PAGE_SIZE;
···
 	while (length) {
 		u32 iov_len = min_t(u32, length, sg->length - sg_offset);
 
-		iov->iov_base = kmap(sg_page(sg)) + sg->offset + sg_offset;
-		iov->iov_len = iov_len;
+		iov->bv_page = sg_page(sg);
+		iov->bv_len = sg->length;
+		iov->bv_offset = sg->offset + sg_offset;
 
 		length -= iov_len;
 		sg = sg_next(sg);
···
 		sg_offset = 0;
 	}
 
-	iov_iter_kvec(&cmd->recv_msg.msg_iter, READ, cmd->iov,
-		cmd->nr_mapped, cmd->pdu_len);
+	iov_iter_bvec(&cmd->recv_msg.msg_iter, READ, cmd->iov,
+		      nr_pages, cmd->pdu_len);
 }
 
 static void nvmet_tcp_fatal_error(struct nvmet_tcp_queue *queue)
···
 	}
 
 	queue->rcv_state = NVMET_TCP_RECV_DATA;
-	nvmet_tcp_map_pdu_iovec(cmd);
+	nvmet_tcp_build_pdu_iovec(cmd);
 	cmd->flags |= NVMET_TCP_F_INIT_FAILED;
 }
 
···
 	struct nvme_tcp_data_pdu *data = &queue->pdu.data;
 	struct nvmet_tcp_cmd *cmd;
 
-	if (likely(queue->nr_cmds))
+	if (likely(queue->nr_cmds)) {
+		if (unlikely(data->ttag >= queue->nr_cmds)) {
+			pr_err("queue %d: received out of bound ttag %u, nr_cmds %u\n",
+				queue->idx, data->ttag, queue->nr_cmds);
+			nvmet_tcp_fatal_error(queue);
+			return -EPROTO;
+		}
 		cmd = &queue->cmds[data->ttag];
-	else
+	} else {
 		cmd = &queue->connect;
+	}
 
 	if (le32_to_cpu(data->data_offset) != cmd->rbytes_done) {
 		pr_err("ttag %u unexpected data offset %u (expected %u)\n",
···
 
 	cmd->pdu_len = le32_to_cpu(data->data_length);
 	cmd->pdu_recv = 0;
-	nvmet_tcp_map_pdu_iovec(cmd);
+	nvmet_tcp_build_pdu_iovec(cmd);
 	queue->cmd = cmd;
 	queue->rcv_state = NVMET_TCP_RECV_DATA;
 
···
 			return -EPROTO;
 		}
 		return nvmet_tcp_handle_icreq(queue);
+	}
+
+	if (unlikely(hdr->type == nvme_tcp_icreq)) {
+		pr_err("queue %d: received icreq pdu in state %d\n",
+			queue->idx, queue->state);
+		nvmet_tcp_fatal_error(queue);
+		return -EPROTO;
 	}
 
 	if (hdr->type == nvme_tcp_h2c_data) {
···
 	if (nvmet_tcp_need_data_in(queue->cmd)) {
 		if (nvmet_tcp_has_inline_data(queue->cmd)) {
 			queue->rcv_state = NVMET_TCP_RECV_DATA;
-			nvmet_tcp_map_pdu_iovec(queue->cmd);
+			nvmet_tcp_build_pdu_iovec(queue->cmd);
 			return 0;
 		}
 		/* send back R2T */
···
 		cmd->rbytes_done += ret;
 	}
 
-	nvmet_tcp_unmap_pdu_iovec(cmd);
 	if (queue->data_digest) {
 		nvmet_tcp_prep_recv_ddgst(cmd);
 		return 0;
···
 			queue->idx, cmd->req.cmd->common.command_id,
 			queue->pdu.cmd.hdr.type, le32_to_cpu(cmd->recv_ddgst),
 			le32_to_cpu(cmd->exp_ddgst));
-		nvmet_tcp_finish_cmd(cmd);
+		nvmet_req_uninit(&cmd->req);
+		nvmet_tcp_free_cmd_buffers(cmd);
 		nvmet_tcp_fatal_error(queue);
 		ret = -EPROTO;
 		goto out;
···
 	write_unlock_bh(&sock->sk->sk_callback_lock);
 }
 
-static void nvmet_tcp_finish_cmd(struct nvmet_tcp_cmd *cmd)
-{
-	nvmet_req_uninit(&cmd->req);
-	nvmet_tcp_unmap_pdu_iovec(cmd);
-	nvmet_tcp_free_cmd_buffers(cmd);
-}
-
 static void nvmet_tcp_uninit_data_in_cmds(struct nvmet_tcp_queue *queue)
 {
 	struct nvmet_tcp_cmd *cmd = queue->cmds;
···
 	for (i = 0; i < queue->nr_cmds; i++, cmd++) {
 		if (nvmet_tcp_need_data_in(cmd))
 			nvmet_req_uninit(&cmd->req);
-
-		nvmet_tcp_unmap_pdu_iovec(cmd);
-		nvmet_tcp_free_cmd_buffers(cmd);
 	}
 
 	if (!queue->nr_cmds && nvmet_tcp_need_data_in(&queue->connect)) {
 		/* failed in connect */
-		nvmet_tcp_finish_cmd(&queue->connect);
+		nvmet_req_uninit(&queue->connect.req);
 	}
+}
+
+static void nvmet_tcp_free_cmd_data_in_buffers(struct nvmet_tcp_queue *queue)
+{
+	struct nvmet_tcp_cmd *cmd = queue->cmds;
+	int i;
+
+	for (i = 0; i < queue->nr_cmds; i++, cmd++) {
+		if (nvmet_tcp_need_data_in(cmd))
+			nvmet_tcp_free_cmd_buffers(cmd);
+	}
+
+	if (!queue->nr_cmds && nvmet_tcp_need_data_in(&queue->connect))
+		nvmet_tcp_free_cmd_buffers(&queue->connect);
 }
 
 static void nvmet_tcp_release_queue_work(struct work_struct *w)
···
 	nvmet_tcp_uninit_data_in_cmds(queue);
 	nvmet_sq_destroy(&queue->nvme_sq);
 	cancel_work_sync(&queue->io_work);
+	nvmet_tcp_free_cmd_data_in_buffers(queue);
 	sock_release(queue->sock);
 	nvmet_tcp_free_cmds(queue);
 	if (queue->hdr_digest || queue->data_digest)

drivers/nvme/target/zns.c

···
 {
 	struct block_device *bdev = req->ns->bdev;
 	unsigned int nr_zones = bdev_nr_zones(bdev);
-	struct request_queue *q = bdev_get_queue(bdev);
 	struct bio *bio = NULL;
 	sector_t sector = 0;
 	int ret;
···
 	};
 
 	d.zbitmap = kcalloc_node(BITS_TO_LONGS(nr_zones), sizeof(*(d.zbitmap)),
-				 GFP_NOIO, q->node);
+				 GFP_NOIO, bdev->bd_disk->node_id);
 	if (!d.zbitmap) {
 		ret = -ENOMEM;
 		goto out;

drivers/s390/block/dasd.c

···
 
 #define DASD_DIAG_MOD		"dasd_diag_mod"
 
-static unsigned int queue_depth = 32;
-static unsigned int nr_hw_queues = 4;
-
-module_param(queue_depth, uint, 0444);
-MODULE_PARM_DESC(queue_depth, "Default queue depth for new DASD devices");
-
-module_param(nr_hw_queues, uint, 0444);
-MODULE_PARM_DESC(nr_hw_queues, "Default number of hardware queues for new DASD devices");
-
 /*
  * SECTION: exported variables of dasd.c
  */
···
 /*
  * SECTION: prototypes for static functions of dasd.c
  */
-static int  dasd_alloc_queue(struct dasd_block *);
-static void dasd_free_queue(struct dasd_block *);
 static int dasd_flush_block_queue(struct dasd_block *);
 static void dasd_device_tasklet(unsigned long);
 static void dasd_block_tasklet(unsigned long);
···
  */
 static int dasd_state_new_to_known(struct dasd_device *device)
 {
-	int rc;
-
 	/*
 	 * As long as the device is not in state DASD_STATE_NEW we want to
 	 * keep the reference count > 0.
 	 */
 	dasd_get_device(device);
-
-	if (device->block) {
-		rc = dasd_alloc_queue(device->block);
-		if (rc) {
-			dasd_put_device(device);
-			return rc;
-		}
-	}
 	device->state = DASD_STATE_KNOWN;
 	return 0;
 }
···
 	/* Disable extended error reporting for this device. */
 	dasd_eer_disable(device);
 	device->state = DASD_STATE_NEW;
-
-	if (device->block)
-		dasd_free_queue(device->block);
 
 	/* Give up reference we took in dasd_state_new_to_known. */
 	dasd_put_device(device);
···
 	dasd_schedule_device_bh(device);
 	if (device->block) {
 		dasd_schedule_block_bh(device->block);
-		if (device->block->request_queue)
-			blk_mq_run_hw_queues(device->block->request_queue,
-					     true);
+		if (device->block->gdp)
+			blk_mq_run_hw_queues(device->block->gdp->queue, true);
 	}
 }
 EXPORT_SYMBOL_GPL(dasd_generic_handle_state_change);
···
 	dasd_device_remove_stop_bits(block->base, DASD_STOPPED_PENDING);
 	spin_unlock_irqrestore(get_ccwdev_lock(block->base->cdev), flags);
 	dasd_schedule_block_bh(block);
-	blk_mq_run_hw_queues(block->request_queue, true);
+	blk_mq_run_hw_queues(block->gdp->queue, true);
 }
 
 /*
···
 	blk_mq_run_hw_queues(req->q, true);
 }
 
-static struct blk_mq_ops dasd_mq_ops = {
+struct blk_mq_ops dasd_mq_ops = {
 	.queue_rq = do_dasd_request,
 	.complete = dasd_request_done,
 	.timeout = dasd_times_out,
 	.init_hctx = dasd_init_hctx,
 	.exit_hctx = dasd_exit_hctx,
 };
-
-/*
- * Allocate and initialize request queue and default I/O scheduler.
- */
-static int dasd_alloc_queue(struct dasd_block *block)
-{
-	int rc;
-
-	block->tag_set.ops = &dasd_mq_ops;
-	block->tag_set.cmd_size = sizeof(struct dasd_ccw_req);
-	block->tag_set.nr_hw_queues = nr_hw_queues;
-	block->tag_set.queue_depth = queue_depth;
-	block->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
-	block->tag_set.numa_node = NUMA_NO_NODE;
-
-	rc = blk_mq_alloc_tag_set(&block->tag_set);
-	if (rc)
-		return rc;
-
-	block->request_queue = blk_mq_init_queue(&block->tag_set);
-	if (IS_ERR(block->request_queue))
-		return PTR_ERR(block->request_queue);
-
-	block->request_queue->queuedata = block;
-
-	return 0;
-}
-
-/*
- * Deactivate and free request queue.
- */
-static void dasd_free_queue(struct dasd_block *block)
-{
-	if (block->request_queue) {
-		blk_mq_destroy_queue(block->request_queue);
-		blk_mq_free_tag_set(&block->tag_set);
-		block->request_queue = NULL;
-	}
-}
 
 static int dasd_open(struct block_device *bdev, fmode_t mode)
 {
···
 	dasd_schedule_device_bh(device);
 	if (device->block) {
 		dasd_schedule_block_bh(device->block);
-		if (device->block->request_queue)
-			blk_mq_run_hw_queues(device->block->request_queue,
-					     true);
-		}
+		if (device->block->gdp)
+			blk_mq_run_hw_queues(device->block->gdp->queue, true);
+	}
 
 	if (!device->stopped)
 		wake_up(&generic_waitq);
···
 
 	if (device->block) {
 		dasd_schedule_block_bh(device->block);
-		if (device->block->request_queue)
-			blk_mq_run_hw_queues(device->block->request_queue, true);
+		if (device->block->gdp)
+			blk_mq_run_hw_queues(device->block->gdp->queue, true);
 	}
 	if (!device->stopped)
 		wake_up(&generic_waitq);
···
 /*
  * clear active requests and requeue them to block layer if possible
  */
-static int dasd_generic_requeue_all_requests(struct dasd_device *device)
+int dasd_generic_requeue_all_requests(struct dasd_device *device)
 {
 	struct list_head requeue_queue;
 	struct dasd_ccw_req *cqr, *n;
···
 	dasd_schedule_device_bh(device);
 	return rc;
 }
+EXPORT_SYMBOL_GPL(dasd_generic_requeue_all_requests);
 
 static void do_requeue_requests(struct work_struct *work)
 {

drivers/s390/block/dasd_3990_erp.c

···
 		dev_err(&device->cdev->dev, "An I/O request was rejected"
 			" because writing is inhibited\n");
 		erp = dasd_3990_erp_cleanup(erp, DASD_CQR_FAILED);
+	} else if (sense[7] & SNS7_INVALID_ON_SEC) {
+		dev_err(&device->cdev->dev, "An I/O request was rejected on a copy pair secondary device\n");
+		/* suppress dump of sense data for this error */
+		set_bit(DASD_CQR_SUPPRESS_CR, &erp->refers->flags);
+		erp = dasd_3990_erp_cleanup(erp, DASD_CQR_FAILED);
 	} else {
 		/* fatal error -  set status to FAILED
 		   internal error 09 - Command Reject */

drivers/s390/block/dasd_devmap.c

···
 
 /* This is ugly... */
 #define PRINTK_HEADER "dasd_devmap:"
-#define DASD_BUS_ID_SIZE 20
 #define DASD_MAX_PARAMS 256
 
 #include "dasd_int.h"
···
         unsigned int devindex;
         unsigned short features;
 	struct dasd_device *device;
+	struct dasd_copy_relation *copy;
 };
 
 /*
···
 /*
  * Read a device busid/devno from a string.
  */
-static int __init dasd_busid(char *str, int *id0, int *id1, int *devno)
+static int dasd_busid(char *str, int *id0, int *id1, int *devno)
 {
 	unsigned int val;
 	char *tok;
···
 	return devmap;
 }
 
-/*
- * Find devmap for device with given bus_id.
- */
 static struct dasd_devmap *
-dasd_find_busid(const char *bus_id)
+dasd_find_busid_locked(const char *bus_id)
 {
 	struct dasd_devmap *devmap, *tmp;
 	int hash;
 
-	spin_lock(&dasd_devmap_lock);
 	devmap = ERR_PTR(-ENODEV);
 	hash = dasd_hash_busid(bus_id);
 	list_for_each_entry(tmp, &dasd_hashlists[hash], list) {
···
 			break;
 		}
 	}
+	return devmap;
+}
+
+/*
+ * Find devmap for device with given bus_id.
+ */
+static struct dasd_devmap *
+dasd_find_busid(const char *bus_id)
+{
+	struct dasd_devmap *devmap;
+
+	spin_lock(&dasd_devmap_lock);
+	devmap = dasd_find_busid_locked(bus_id);
 	spin_unlock(&dasd_devmap_lock);
 	return devmap;
 }
···
 }
 
 /*
+ * allocate a PPRC data structure and call the discipline function to fill
+ */
+static int dasd_devmap_get_pprc_status(struct dasd_device *device,
+				       struct dasd_pprc_data_sc4 **data)
+{
+	struct dasd_pprc_data_sc4 *temp;
+
+	if (!device->discipline || !device->discipline->pprc_status) {
+		dev_warn(&device->cdev->dev, "Unable to query copy relation status\n");
+		return -EOPNOTSUPP;
+	}
+	temp = kzalloc(sizeof(*temp), GFP_KERNEL);
+	if (!temp)
+		return -ENOMEM;
+
+	/* get PPRC information from storage */
+	if (device->discipline->pprc_status(device, temp)) {
+		dev_warn(&device->cdev->dev, "Error during copy relation status query\n");
+		kfree(temp);
+		return -EINVAL;
+	}
+	*data = temp;
+
+	return 0;
+}
+
+/*
+ * find an entry in a PPRC device_info array by a given UID
+ * depending on the primary/secondary state of the device it has to be
+ * matched with the respective fields
+ */
+static int dasd_devmap_entry_from_pprc_data(struct dasd_pprc_data_sc4 *data,
+					    struct dasd_uid uid,
+					    bool primary)
+{
+	int i;
+
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (primary) {
+			if (data->dev_info[i].prim_cu_ssid == uid.ssid &&
+			    data->dev_info[i].primary == uid.real_unit_addr)
+				return i;
+		} else {
+			if (data->dev_info[i].sec_cu_ssid == uid.ssid &&
+			    data->dev_info[i].secondary == uid.real_unit_addr)
+				return i;
+		}
+	}
+	return -1;
+}
+
+/*
+ * check the consistency of a specified copy relation by checking
+ * the following things:
+ *
+ *   - is the given device part of a copy pair setup
+ *   - does the state of the device match the state in the PPRC status data
+ *   - does the device UID match with the UID in the PPRC status data
+ *   - to prevent misrouted IO check if the given device is present in all
+ *     related PPRC status data
+ */
+static int dasd_devmap_check_copy_relation(struct dasd_device *device,
+					   struct dasd_copy_entry *entry,
+					   struct dasd_pprc_data_sc4 *data,
+					   struct dasd_copy_relation *copy)
+{
+	struct dasd_pprc_data_sc4 *tmp_dat;
+	struct dasd_device *tmp_dev;
+	struct dasd_uid uid;
+	int i, j;
+
+	if (!device->discipline || !device->discipline->get_uid ||
+	    device->discipline->get_uid(device, &uid))
+		return 1;
+
+	i = dasd_devmap_entry_from_pprc_data(data, uid, entry->primary);
+	if (i < 0) {
+		dev_warn(&device->cdev->dev, "Device not part of a copy relation\n");
+		return 1;
+	}
+
+	/* double check which role the current device has */
+	if (entry->primary) {
+		if (data->dev_info[i].flags & 0x80) {
+			dev_warn(&device->cdev->dev, "Copy pair secondary is setup as primary\n");
+			return 1;
+		}
+		if (data->dev_info[i].prim_cu_ssid != uid.ssid ||
+		    data->dev_info[i].primary != uid.real_unit_addr) {
+			dev_warn(&device->cdev->dev,
+				 "Primary device %s does not match copy pair status primary device %04x\n",
+				 dev_name(&device->cdev->dev),
+				 data->dev_info[i].prim_cu_ssid |
+				 data->dev_info[i].primary);
+			return 1;
+		}
+	} else {
+		if (!(data->dev_info[i].flags & 0x80)) {
+			dev_warn(&device->cdev->dev, "Copy pair primary is setup as secondary\n");
+			return 1;
+		}
+		if (data->dev_info[i].sec_cu_ssid != uid.ssid ||
+		    data->dev_info[i].secondary != uid.real_unit_addr) {
+			dev_warn(&device->cdev->dev,
+				 "Secondary device %s does not match copy pair status secondary device %04x\n",
+				 dev_name(&device->cdev->dev),
+				 data->dev_info[i].sec_cu_ssid |
+				 data->dev_info[i].secondary);
+			return 1;
+		}
+	}
+
+	/*
+	 * the current device has to be part of the copy relation of all
+	 * entries to prevent misrouted IO to another copy pair
+	 */
+	for (j = 0; j < DASD_CP_ENTRIES; j++) {
+		if (entry == &copy->entry[j])
+			tmp_dev = device;
+		else
+			tmp_dev = copy->entry[j].device;
+
+		if (!tmp_dev)
+			continue;
+
+		if (dasd_devmap_get_pprc_status(tmp_dev, &tmp_dat))
+			return 1;
+
+		if (dasd_devmap_entry_from_pprc_data(tmp_dat, uid, entry->primary) < 0) {
+			dev_warn(&tmp_dev->cdev->dev,
+				 "Copy pair relation does not contain device: %s\n",
+				 dev_name(&device->cdev->dev));
+			kfree(tmp_dat);
+			return 1;
+		}
+		kfree(tmp_dat);
+	}
+	return 0;
+}
+
+/* delete device from copy relation entry */
+static void dasd_devmap_delete_copy_relation_device(struct dasd_device *device)
+{
+	struct dasd_copy_relation *copy;
+	int i;
+
+	if (!device->copy)
+		return;
+
+	copy = device->copy;
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].device == device)
+			copy->entry[i].device = NULL;
+	}
+	dasd_put_device(device);
+	device->copy = NULL;
+}
+
+/*
+ * read all required information for a copy relation setup and setup the device
+ * accordingly
+ */
+int dasd_devmap_set_device_copy_relation(struct ccw_device *cdev,
+					 bool pprc_enabled)
+{
+	struct dasd_pprc_data_sc4 *data = NULL;
+	struct dasd_copy_entry *entry = NULL;
+	struct dasd_copy_relation *copy;
+	struct dasd_devmap *devmap;
+	struct dasd_device *device;
+	int i, rc = 0;
+
+	devmap = dasd_devmap_from_cdev(cdev);
+	if (IS_ERR(devmap))
+		return PTR_ERR(devmap);
+
+	device = devmap->device;
+	if (!device)
+		return -ENODEV;
+
+	copy = devmap->copy;
+	/* no copy pair setup for this device */
+	if (!copy)
+		goto out;
+
+	rc = dasd_devmap_get_pprc_status(device, &data);
+	if (rc)
+		return rc;
+
+	/* print error if PPRC is requested but not enabled on storage server */
+	if (!pprc_enabled) {
+		dev_err(&cdev->dev, "Copy relation not enabled on storage server\n");
+		rc = -EINVAL;
+		goto out;
+	}
+
+	if (!data->dev_info[0].state) {
+		dev_warn(&device->cdev->dev, "Copy pair setup requested for device not in copy relation\n");
+		rc = -EINVAL;
+		goto out;
+	}
+	/* find entry */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured &&
+		    strncmp(dev_name(&cdev->dev),
+			    copy->entry[i].busid, DASD_BUS_ID_SIZE) == 0) {
+			entry = &copy->entry[i];
+			break;
+		}
+	}
+	if (!entry) {
+		dev_warn(&device->cdev->dev, "Copy relation entry not found\n");
+		rc = -EINVAL;
+		goto out;
+	}
+	/* check if the copy relation is valid */
+	if (dasd_devmap_check_copy_relation(device, entry, data, copy)) {
+		dev_warn(&device->cdev->dev, "Copy relation faulty\n");
+		rc = -EINVAL;
+		goto out;
+	}
+
+	dasd_get_device(device);
+	copy->entry[i].device = device;
+	device->copy = copy;
+out:
+	kfree(data);
+	return rc;
+}
+EXPORT_SYMBOL_GPL(dasd_devmap_set_device_copy_relation);
+
+/*
  * Wait queue for dasd_delete_device waits.
  */
 static DECLARE_WAIT_QUEUE_HEAD(dasd_delete_wq);
···
 	dev_set_drvdata(&device->cdev->dev, NULL);
 	spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
 
+	/* Removve copy relation */
+	dasd_devmap_delete_copy_relation_device(device);
 	/*
 	 * Drop ref_count by 3, one for the devmap reference, one for
 	 * the cdev reference and one for the passed reference.
···
 	gdp->private_data = devmap;
 	spin_unlock(&dasd_devmap_lock);
 }
+EXPORT_SYMBOL(dasd_add_link_to_gendisk);
 
 struct dasd_device *dasd_device_from_gendisk(struct gendisk *gdp)
 {
···
 		   const char *buf, size_t count)
 {
 	struct dasd_device *device;
-	struct request_queue *q;
 	unsigned long val;
 
 	device = dasd_device_from_cdev(to_ccwdev(dev));
···
 		dasd_put_device(device);
 		return -EINVAL;
 	}
-	q = device->block->request_queue;
-	if (!q) {
+	if (!device->block->gdp) {
 		dasd_put_device(device);
 		return -ENODEV;
 	}
 
 	device->blk_timeout = val;
-
-	blk_queue_rq_timeout(q, device->blk_timeout * HZ);
+	blk_queue_rq_timeout(device->block->gdp->queue, val * HZ);
 
 	dasd_put_device(device);
 	return count;
···
 static struct kobj_attribute path_fcs_attribute =
 	__ATTR(fc_security, 0444, dasd_path_fcs_show, NULL);
 
+/*
+ * print copy relation in the form
+ * primary,secondary[1] primary,secondary[2], ...
+ */
+static ssize_t
+dasd_copy_pair_show(struct device *dev,
+		    struct device_attribute *attr, char *buf)
+{
+	char prim_busid[DASD_BUS_ID_SIZE];
+	struct dasd_copy_relation *copy;
+	struct dasd_devmap *devmap;
+	int len = 0;
+	int i;
+
+	devmap = dasd_find_busid(dev_name(dev));
+	if (IS_ERR(devmap))
+		return -ENODEV;
+
+	if (!devmap->copy)
+		return -ENODEV;
+
+	copy = devmap->copy;
+	/* find primary */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured && copy->entry[i].primary) {
+			strscpy(prim_busid, copy->entry[i].busid,
+				DASD_BUS_ID_SIZE);
+			break;
+		}
+	}
+	if (!copy->entry[i].primary)
+		goto out;
+
+	/* print all secondary */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured && !copy->entry[i].primary)
+			len += sysfs_emit_at(buf, len, "%s,%s ", prim_busid,
+					     copy->entry[i].busid);
+	}
+
+	len += sysfs_emit_at(buf, len, "\n");
+out:
+	return len;
+}
+
+static int dasd_devmap_set_copy_relation(struct dasd_devmap *devmap,
+					 struct dasd_copy_relation *copy,
+					 char *busid, bool primary)
+{
+	int i;
+
+	/* find free entry */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		/* current bus_id already included, nothing to do */
+		if (copy->entry[i].configured &&
+		    strncmp(copy->entry[i].busid, busid, DASD_BUS_ID_SIZE) == 0)
+			return 0;
+
+		if (!copy->entry[i].configured)
+			break;
+	}
+	if (i == DASD_CP_ENTRIES)
+		return -EINVAL;
+
+	copy->entry[i].configured = true;
+	strscpy(copy->entry[i].busid, busid, DASD_BUS_ID_SIZE);
+	if (primary) {
+		copy->active = &copy->entry[i];
+		copy->entry[i].primary = true;
+	}
+	if (!devmap->copy)
+		devmap->copy = copy;
+
+	return 0;
+}
+
+static void dasd_devmap_del_copy_relation(struct dasd_copy_relation *copy,
+					  char *busid)
+{
+	int i;
+
+	spin_lock(&dasd_devmap_lock);
+	/* find entry */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured &&
+		    strncmp(copy->entry[i].busid, busid, DASD_BUS_ID_SIZE) == 0)
+			break;
+	}
+	if (i == DASD_CP_ENTRIES || !copy->entry[i].configured) {
+		spin_unlock(&dasd_devmap_lock);
+		return;
+	}
+
+	copy->entry[i].configured = false;
+	memset(copy->entry[i].busid, 0, DASD_BUS_ID_SIZE);
+	if (copy->active == &copy->entry[i]) {
+		copy->active = NULL;
+		copy->entry[i].primary = false;
+	}
+	spin_unlock(&dasd_devmap_lock);
+}
+
+static int dasd_devmap_clear_copy_relation(struct device *dev)
+{
+	struct dasd_copy_relation *copy;
+	struct dasd_devmap *devmap;
+	int i, rc = 1;
+
+	devmap = dasd_devmap_from_cdev(to_ccwdev(dev));
+	if (IS_ERR(devmap))
+		return 1;
+
+	spin_lock(&dasd_devmap_lock);
+	if (!devmap->copy)
+		goto out;
+
+	copy = devmap->copy;
+	/* first check if all secondary devices are offline*/
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (!copy->entry[i].configured)
+			continue;
+
+		if (copy->entry[i].device == copy->active->device)
+			continue;
+
+		if (copy->entry[i].device)
+			goto out;
+	}
+	/* clear all devmap entries */
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (strlen(copy->entry[i].busid) == 0)
+			continue;
+		if (copy->entry[i].device) {
+			dasd_put_device(copy->entry[i].device);
+			copy->entry[i].device->copy = NULL;
+			copy->entry[i].device = NULL;
+		}
+		devmap = dasd_find_busid_locked(copy->entry[i].busid);
+		devmap->copy = NULL;
+		memset(copy->entry[i].busid, 0, DASD_BUS_ID_SIZE);
+	}
+	kfree(copy);
+	rc = 0;
+out:
+	spin_unlock(&dasd_devmap_lock);
+	return rc;
+}
+
+/*
+ * parse BUSIDs from a copy pair
+ */
+static int dasd_devmap_parse_busid(const char *buf, char *prim_busid,
+				   char *sec_busid)
+{
+	char *primary, *secondary, *tmp, *pt;
+	int id0, id1, id2;
+
+	pt =  kstrdup(buf, GFP_KERNEL);
+	tmp = pt;
+	if (!tmp)
+		return -ENOMEM;
+
+	primary = strsep(&tmp, ",");
+	if (!primary) {
+		kfree(pt);
+		return -EINVAL;
+	}
+	secondary = strsep(&tmp, ",");
+	if (!secondary) {
+		kfree(pt);
+		return -EINVAL;
+	}
+	if (dasd_busid(primary, &id0, &id1, &id2)) {
+		kfree(pt);
+		return -EINVAL;
+	}
+	sprintf(prim_busid, "%01x.%01x.%04x", id0, id1, id2);
+	if (dasd_busid(secondary, &id0, &id1, &id2)) {
+		kfree(pt);
+		return -EINVAL;
+	}
+	sprintf(sec_busid, "%01x.%01x.%04x", id0, id1, id2);
+	kfree(pt);
+
+	return 0;
+}
+
+static ssize_t dasd_copy_pair_store(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t count)
+{
+	struct dasd_devmap *prim_devmap, *sec_devmap;
+	char prim_busid[DASD_BUS_ID_SIZE];
+	char sec_busid[DASD_BUS_ID_SIZE];
+	struct dasd_copy_relation *copy;
+	struct dasd_device *device;
+	bool pprc_enabled;
+	int rc;
+
+	if (strncmp(buf, "clear", strlen("clear")) == 0) {
+		if (dasd_devmap_clear_copy_relation(dev))
+			return -EINVAL;
+		return count;
+	}
+
+	rc = dasd_devmap_parse_busid(buf, prim_busid, sec_busid);
+	if (rc)
+		return rc;
+
+	if (strncmp(dev_name(dev), prim_busid, DASD_BUS_ID_SIZE) != 0 &&
+	    strncmp(dev_name(dev), sec_busid, DASD_BUS_ID_SIZE) != 0)
+		return -EINVAL;
+
+	/* allocate primary devmap if needed */
+	prim_devmap = dasd_find_busid(prim_busid);
+	if (IS_ERR(prim_devmap))
+		prim_devmap = dasd_add_busid(prim_busid, DASD_FEATURE_DEFAULT);
+
+	/* allocate secondary devmap if needed */
+	sec_devmap = dasd_find_busid(sec_busid);
+	if (IS_ERR(sec_devmap))
+		sec_devmap = dasd_add_busid(sec_busid, DASD_FEATURE_DEFAULT);
+
+	/* setting copy relation is only allowed for offline secondary */
+	if (sec_devmap->device)
+		return -EINVAL;
+
+	if (prim_devmap->copy) {
+		copy = prim_devmap->copy;
+	} else if (sec_devmap->copy) {
+		copy = sec_devmap->copy;
+	} else {
+		copy = kzalloc(sizeof(*copy), GFP_KERNEL);
+		if (!copy)
+			return -ENOMEM;
+	}
+	spin_lock(&dasd_devmap_lock);
+	rc = dasd_devmap_set_copy_relation(prim_devmap, copy, prim_busid, true);
+	if (rc) {
+		spin_unlock(&dasd_devmap_lock);
+		return rc;
+	}
+	rc = dasd_devmap_set_copy_relation(sec_devmap, copy, sec_busid, false);
+	if (rc) {
+		spin_unlock(&dasd_devmap_lock);
+		return rc;
+	}
+	spin_unlock(&dasd_devmap_lock);
+
+	/* if primary device is already online call device setup directly */
+	if (prim_devmap->device && !prim_devmap->device->copy) {
+		device = prim_devmap->device;
+		if (device->discipline->pprc_enabled) {
+			pprc_enabled = device->discipline->pprc_enabled(device);
+			rc = dasd_devmap_set_device_copy_relation(device->cdev,
+								  pprc_enabled);
+		} else {
+			rc = -EOPNOTSUPP;
+		}
+	}
+	if (rc) {
+		dasd_devmap_del_copy_relation(copy, prim_busid);
+		dasd_devmap_del_copy_relation(copy, sec_busid);
+		count = rc;
+	}
+
+	return count;
+}
+static DEVICE_ATTR(copy_pair, 0644, dasd_copy_pair_show,
+		   dasd_copy_pair_store);
+
+static ssize_t
+dasd_copy_role_show(struct device *dev,
+		    struct device_attribute *attr, char *buf)
+{
+	struct dasd_copy_relation *copy;
+	struct dasd_device *device;
+	int len, i;
+
+	device = dasd_device_from_cdev(to_ccwdev(dev));
+	if (IS_ERR(device))
+		return -ENODEV;
+
+	if (!device->copy) {
+		len = sysfs_emit(buf, "none\n");
+		goto out;
+	}
+	copy = device->copy;
+	/* only the active device is primary */
+	if (copy->active->device == device) {
+		len = sysfs_emit(buf, "primary\n");
+		goto out;
+	}
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].device == device) {
+			len = sysfs_emit(buf, "secondary\n");
+			goto out;
+		}
+	}
+	/* not in the list, no COPY role */
+	len = sysfs_emit(buf, "none\n");
+out:
+	dasd_put_device(device);
+	return len;
+}
+static DEVICE_ATTR(copy_role, 0444, dasd_copy_role_show, NULL);
+
+static ssize_t dasd_device_ping(struct device *dev,
+				struct device_attribute *attr,
+				const char *buf, size_t count)
+{
+	struct dasd_device *device;
+	size_t rc;
+
+	device = dasd_device_from_cdev(to_ccwdev(dev));
+	if (IS_ERR(device))
+		return -ENODEV;
+
+	/*
+	 * do not try during offline processing
+	 * early check only
+	 * the sleep_on function itself checks for offline
+	 * processing again
+	 */
+	if (test_bit(DASD_FLAG_OFFLINE, &device->flags)) {
+		rc = -EBUSY;
+		goto out;
+	}
+	if (!device->discipline || !device->discipline->device_ping) {
+		rc = -EOPNOTSUPP;
+		goto out;
+	}
+	rc = device->discipline->device_ping(device);
+	if (!rc)
+		rc = count;
+out:
+	dasd_put_device(device);
+	return rc;
+}
+static DEVICE_ATTR(ping, 0200, NULL, dasd_device_ping);
+
 #define DASD_DEFINE_ATTR(_name, _func)					\
 static ssize_t dasd_##_name##_show(struct device *dev,			\
 				   struct device_attribute *attr,	\
···
 	&dev_attr_hpf.attr,
 	&dev_attr_ese.attr,
 	&dev_attr_fc_security.attr,
+	&dev_attr_copy_pair.attr,
+	&dev_attr_copy_role.attr,
+	&dev_attr_ping.attr,
 	NULL,
 };
 

drivers/s390/block/dasd_diag.c

···
 static void dasd_diag_setup_blk_queue(struct dasd_block *block)
 {
 	unsigned int logical_block_size = block->bp_block;
-	struct request_queue *q = block->request_queue;
+	struct request_queue *q = block->gdp->queue;
 	int max;
 
 	max = DIAG_MAX_BLOCKS << block->s2b_shift;

drivers/s390/block/dasd_eckd.c

···
 }
 
 /*
+ * return if the device is the copy relation primary if a copy relation is active
+ */
+static int dasd_device_is_primary(struct dasd_device *device)
+{
+	if (!device->copy)
+		return 1;
+
+	if (device->copy->active->device == device)
+		return 1;
+
+	return 0;
+}
+
+static int dasd_eckd_alloc_block(struct dasd_device *device)
+{
+	struct dasd_block *block;
+	struct dasd_uid temp_uid;
+
+	if (!dasd_device_is_primary(device))
+		return 0;
+
+	dasd_eckd_get_uid(device, &temp_uid);
+	if (temp_uid.type == UA_BASE_DEVICE) {
+		block = dasd_alloc_block();
+		if (IS_ERR(block)) {
+			DBF_EVENT_DEVID(DBF_WARNING, device->cdev, "%s",
+					"could not allocate dasd block structure");
+			return PTR_ERR(block);
+		}
+		device->block = block;
+		block->base = device;
+	}
+	return 0;
+}
+
+static bool dasd_eckd_pprc_enabled(struct dasd_device *device)
+{
+	struct dasd_eckd_private *private = device->private;
+
+	return private->rdc_data.facilities.PPRC_enabled;
+}
+
+/*
  * Check device characteristics.
  * If the device is accessible using ECKD discipline, the device is enabled.
  */
···
 dasd_eckd_check_characteristics(struct dasd_device *device)
 {
 	struct dasd_eckd_private *private = device->private;
-	struct dasd_block *block;
-	struct dasd_uid temp_uid;
 	int rc, i;
 	int readonly;
 	unsigned long value;
···
 			device->default_expires = value;
 	}
 
-	dasd_eckd_get_uid(device, &temp_uid);
-	if (temp_uid.type == UA_BASE_DEVICE) {
-		block = dasd_alloc_block();
-		if (IS_ERR(block)) {
-			DBF_EVENT_DEVID(DBF_WARNING, device->cdev, "%s",
-					"could not allocate dasd "
-					"block structure");
-			rc = PTR_ERR(block);
-			goto out_err1;
-		}
-		device->block = block;
-		block->base = device;
+	/* Read Device Characteristics */
+	rc = dasd_generic_read_dev_chars(device, DASD_ECKD_MAGIC,
+					 &private->rdc_data, 64);
+	if (rc) {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev,
+				"Read device characteristic failed, rc=%d", rc);
+		goto out_err1;
 	}
+
+	/* setup PPRC for device from devmap */
+	rc = dasd_devmap_set_device_copy_relation(device->cdev,
+						  dasd_eckd_pprc_enabled(device));
+	if (rc) {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev,
+				"copy relation setup failed, rc=%d", rc);
+		goto out_err1;
+	}
+
+	/* check if block device is needed and allocate in case */
+	rc = dasd_eckd_alloc_block(device);
+	if (rc)
+		goto out_err1;
 
 	/* register lcu with alias handling, enable PAV */
 	rc = dasd_alias_make_device_known_to_lcu(device);
···
 
 	/* Read Extent Pool Information */
 	dasd_eckd_read_ext_pool_info(device);
-
-	/* Read Device Characteristics */
-	rc = dasd_generic_read_dev_chars(device, DASD_ECKD_MAGIC,
-					 &private->rdc_data, 64);
-	if (rc) {
-		DBF_EVENT_DEVID(DBF_WARNING, device->cdev,
-				"Read device characteristic failed, rc=%d", rc);
-		goto out_err3;
-	}
 
 	if ((device->features & DASD_FEATURE_USERAW) &&
 	    !(private->rdc_data.facilities.RT_in_LR)) {
···
 	return 0;
 }
 
+static struct dasd_device
+*copy_relation_find_device(struct dasd_copy_relation *copy,
+			   char *busid)
+{
+	int i;
+
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured &&
+		    strncmp(copy->entry[i].busid, busid, DASD_BUS_ID_SIZE) == 0)
+			return copy->entry[i].device;
+	}
+	return NULL;
+}
+
+/*
+ * set the new active/primary device
+ */
+static void copy_pair_set_active(struct dasd_copy_relation *copy, char *new_busid,
+				 char *old_busid)
+{
+	int i;
+
+	for (i = 0; i < DASD_CP_ENTRIES; i++) {
+		if (copy->entry[i].configured &&
+		    strncmp(copy->entry[i].busid, new_busid,
+			    DASD_BUS_ID_SIZE) == 0) {
+			copy->active = &copy->entry[i];
+			copy->entry[i].primary = true;
+		} else if (copy->entry[i].configured &&
+			   strncmp(copy->entry[i].busid, old_busid,
+				   DASD_BUS_ID_SIZE) == 0) {
+			copy->entry[i].primary = false;
+		}
+	}
+}
+
+/*
+ * The function will swap the role of a given copy pair.
+ * During the swap operation the relation of the blockdevice is disconnected
+ * from the old primary and connected to the new.
+ *
+ * IO is paused on the block queue before swap and may be resumed afterwards.
+ */
+static int dasd_eckd_copy_pair_swap(struct dasd_device *device, char *prim_busid,
+				    char *sec_busid)
+{
+	struct dasd_device *primary, *secondary;
+	struct dasd_copy_relation *copy;
+	struct dasd_block *block;
+	struct gendisk *gdp;
+
+	copy = device->copy;
+	if (!copy)
+		return DASD_COPYPAIRSWAP_INVALID;
+	primary = copy->active->device;
+	if (!primary)
+		return DASD_COPYPAIRSWAP_INVALID;
+	/* double check if swap has correct primary */
+	if (strncmp(dev_name(&primary->cdev->dev), prim_busid, DASD_BUS_ID_SIZE) != 0)
+		return DASD_COPYPAIRSWAP_PRIMARY;
+
+	secondary = copy_relation_find_device(copy, sec_busid);
+	if (!secondary)
+		return DASD_COPYPAIRSWAP_SECONDARY;
+
+	/*
+	 * usually the device should be quiesced for swap
+	 * for paranoia stop device and requeue requests again
+	 */
+	dasd_device_set_stop_bits(primary, DASD_STOPPED_PPRC);
+	dasd_device_set_stop_bits(secondary, DASD_STOPPED_PPRC);
+	dasd_generic_requeue_all_requests(primary);
+
+	/* swap DASD internal device <> block assignment */
+	block = primary->block;
+	primary->block = NULL;
+	secondary->block = block;
+	block->base = secondary;
+	/* set new primary device in COPY relation */
+	copy_pair_set_active(copy, sec_busid, prim_busid);
+
+	/* swap blocklayer device link */
+	gdp = block->gdp;
+	dasd_add_link_to_gendisk(gdp, secondary);
+
+	/* re-enable device */
+	dasd_device_remove_stop_bits(primary, DASD_STOPPED_PPRC);
+	dasd_device_remove_stop_bits(secondary, DASD_STOPPED_PPRC);
+	dasd_schedule_device_bh(secondary);
+
+	return DASD_COPYPAIRSWAP_SUCCESS;
+}
+
+/*
+ * Perform Subsystem Function - Peer-to-Peer Remote Copy Extended Query
+ */
+static int dasd_eckd_query_pprc_status(struct dasd_device *device,
+				       struct dasd_pprc_data_sc4 *data)
+{
+	struct dasd_pprc_data_sc4 *pprc_data;
+	struct dasd_psf_prssd_data *prssdp;
+	struct dasd_ccw_req *cqr;
+	struct ccw1 *ccw;
+	int rc;
+
+	cqr = dasd_smalloc_request(DASD_ECKD_MAGIC, 1 /* PSF */	+ 1 /* RSSD */,
+				   sizeof(*prssdp) + sizeof(*pprc_data) + 1,
+				   device, NULL);
+	if (IS_ERR(cqr)) {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev, "%s",
+				"Could not allocate query PPRC status request");
+		return PTR_ERR(cqr);
+	}
+	cqr->startdev = device;
+	cqr->memdev = device;
+	cqr->block = NULL;
+	cqr->retries = 256;
+	cqr->expires = 10 * HZ;
+
+	/* Prepare for Read Subsystem Data */
+	prssdp = (struct dasd_psf_prssd_data *)cqr->data;
+	memset(prssdp, 0, sizeof(struct dasd_psf_prssd_data));
+	prssdp->order = PSF_ORDER_PRSSD;
+	prssdp->suborder = PSF_SUBORDER_PPRCEQ;
+	prssdp->varies[0] = PPRCEQ_SCOPE_4;
+	pprc_data = (struct dasd_pprc_data_sc4 *)(prssdp + 1);
+
+	ccw = cqr->cpaddr;
+	ccw->cmd_code = DASD_ECKD_CCW_PSF;
+	ccw->count = sizeof(struct dasd_psf_prssd_data);
+	ccw->flags |= CCW_FLAG_CC;
+	ccw->flags |= CCW_FLAG_SLI;
+	ccw->cda = (__u32)(addr_t)prssdp;
+
+	/* Read Subsystem Data - query host access */
+	ccw++;
+	ccw->cmd_code = DASD_ECKD_CCW_RSSD;
+	ccw->count = sizeof(*pprc_data);
+	ccw->flags |= CCW_FLAG_SLI;
+	ccw->cda = (__u32)(addr_t)pprc_data;
+
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+
+	rc = dasd_sleep_on_interruptible(cqr);
+	if (rc == 0) {
+		*data = *pprc_data;
+	} else {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev,
+				"PPRC Extended Query failed with rc=%d\n",
+				rc);
+		rc = -EOPNOTSUPP;
+	}
+
+	dasd_sfree_request(cqr, cqr->memdev);
+	return rc;
+}
+
+/*
+ * ECKD NOP - no operation
+ */
+static int dasd_eckd_nop(struct dasd_device *device)
+{
+	struct dasd_ccw_req *cqr;
+	struct ccw1 *ccw;
+	int rc;
+
+	cqr = dasd_smalloc_request(DASD_ECKD_MAGIC, 1, 1, device, NULL);
+	if (IS_ERR(cqr)) {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev, "%s",
+				"Could not allocate NOP request");
+		return PTR_ERR(cqr);
+	}
+	cqr->startdev = device;
+	cqr->memdev = device;
+	cqr->block = NULL;
+	cqr->retries = 1;
+	cqr->expires = 10 * HZ;
+
+	ccw = cqr->cpaddr;
+	ccw->cmd_code = DASD_ECKD_CCW_NOP;
+	ccw->flags |= CCW_FLAG_SLI;
+
+	cqr->buildclk = get_tod_clock();
+	cqr->status = DASD_CQR_FILLED;
+
+	rc = dasd_sleep_on_interruptible(cqr);
+	if (rc != 0) {
+		DBF_EVENT_DEVID(DBF_WARNING, device->cdev,
+				"NOP failed with rc=%d\n", rc);
+		rc = -EOPNOTSUPP;
+	}
+	dasd_sfree_request(cqr, cqr->memdev);
+	return rc;
+}
+
+static int dasd_eckd_device_ping(struct dasd_device *device)
+{
+	return dasd_eckd_nop(device);
+}
+
 /*
  * Perform Subsystem Function - CUIR response
  */
···
 static void dasd_eckd_setup_blk_queue(struct dasd_block *block)
 {
 	unsigned int logical_block_size = block->bp_block;
-	struct request_queue *q = block->request_queue;
+	struct request_queue *q = block->gdp->queue;
 	struct dasd_device *device = block->base;
 	int max;
 
···
 	.ext_pool_exhaust = dasd_eckd_ext_pool_exhaust,
 	.ese_format = dasd_eckd_ese_format,
 	.ese_read = dasd_eckd_ese_read,
+	.pprc_status = dasd_eckd_query_pprc_status,
+	.pprc_enabled = dasd_eckd_pprc_enabled,
+	.copy_pair_swap = dasd_eckd_copy_pair_swap,
+	.device_ping = dasd_eckd_device_ping,
 };
 
 static int __init

drivers/s390/block/dasd_eckd.h

···
 /*****************************************************************************
  * SECTION: CCW Definitions
  ****************************************************************************/
+#define DASD_ECKD_CCW_NOP		 0x03
 #define DASD_ECKD_CCW_WRITE		 0x05
 #define DASD_ECKD_CCW_READ		 0x06
 #define DASD_ECKD_CCW_WRITE_HOME_ADDRESS 0x09
···
  * Perform Subsystem Function / Sub-Orders
  */
 #define PSF_SUBORDER_QHA		 0x1C /* Query Host Access */
+#define PSF_SUBORDER_PPRCEQ		 0x50 /* PPRC Extended Query */
 #define PSF_SUBORDER_VSQ		 0x52 /* Volume Storage Query */
 #define PSF_SUBORDER_LCQ		 0x53 /* Logical Configuration Query */
+
+/*
+ * PPRC Extended Query Scopes
+ */
+#define PPRCEQ_SCOPE_4			 0x04 /* Scope 4 for PPRC Extended Query */
 
 /*
  * CUIR response condition codes
···
 		unsigned char reserved3:8;
 		unsigned char defect_wr:1;
 		unsigned char XRC_supported:1;
-		unsigned char reserved4:1;
+		unsigned char PPRC_enabled:1;
 		unsigned char striping:1;
 		unsigned char reserved5:4;
 		unsigned char cfw:1;

drivers/s390/block/dasd_fba.c

···
 static void dasd_fba_setup_blk_queue(struct dasd_block *block)
 {
 	unsigned int logical_block_size = block->bp_block;
-	struct request_queue *q = block->request_queue;
+	struct request_queue *q = block->gdp->queue;
 	unsigned int max_bytes, max_discard_sectors;
 	int max;
 

drivers/s390/block/dasd_genhd.c

···
 
 #include "dasd_int.h"
 
-static struct lock_class_key dasd_bio_compl_lkclass;
+static unsigned int queue_depth = 32;
+static unsigned int nr_hw_queues = 4;
+
+module_param(queue_depth, uint, 0444);
+MODULE_PARM_DESC(queue_depth, "Default queue depth for new DASD devices");
+
+module_param(nr_hw_queues, uint, 0444);
+MODULE_PARM_DESC(nr_hw_queues, "Default number of hardware queues for new DASD devices");
 
 /*
  * Allocate and register gendisk structure for device.
···
 	if (base->devindex >= DASD_PER_MAJOR)
 		return -EBUSY;
 
-	gdp = blk_mq_alloc_disk_for_queue(block->request_queue,
-					  &dasd_bio_compl_lkclass);
-	if (!gdp)
-		return -ENOMEM;
+	block->tag_set.ops = &dasd_mq_ops;
+	block->tag_set.cmd_size = sizeof(struct dasd_ccw_req);
+	block->tag_set.nr_hw_queues = nr_hw_queues;
+	block->tag_set.queue_depth = queue_depth;
+	block->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
+	block->tag_set.numa_node = NUMA_NO_NODE;
+	rc = blk_mq_alloc_tag_set(&block->tag_set);
+	if (rc)
+		return rc;
+
+	gdp = blk_mq_alloc_disk(&block->tag_set, block);
+	if (IS_ERR(gdp)) {
+		blk_mq_free_tag_set(&block->tag_set);
+		return PTR_ERR(gdp);
+	}
 
 	/* Initialize gendisk structure. */
 	gdp->major = DASD_MAJOR;
···
 		block->gdp->private_data = NULL;
 		put_disk(block->gdp);
 		block->gdp = NULL;
+		blk_mq_free_tag_set(&block->tag_set);
 	}
 }
 

drivers/s390/block/dasd_int.h

···
 };
 
 /*
+ * PPRC Status data
+ */
+struct dasd_pprc_header {
+	__u8 entries;		/* 0     Number of device entries */
+	__u8 unused;		/* 1     unused */
+	__u16 entry_length;	/* 2-3   Length of device entry */
+	__u32 unused2;		/* 4-7   unused */
+} __packed;
+
+struct dasd_pprc_dev_info {
+	__u8 state;		/* 0       Copy State */
+	__u8 flags;		/* 1       Flags */
+	__u8 reserved1[2];	/* 2-3     reserved */
+	__u8 prim_lss;		/* 4       Primary device LSS */
+	__u8 primary;		/* 5       Primary device address */
+	__u8 sec_lss;		/* 6       Secondary device LSS */
+	__u8 secondary;		/* 7       Secondary device address */
+	__u16 pprc_id;		/* 8-9     Peer-to-Peer Remote Copy ID */
+	__u8 reserved2[12];	/* 10-21   reserved */
+	__u16 prim_cu_ssid;	/* 22-23   Pimary Control Unit SSID */
+	__u8 reserved3[12];	/* 24-35   reserved */
+	__u16 sec_cu_ssid;	/* 36-37   Secondary Control Unit SSID */
+	__u8 reserved4[90];	/* 38-127  reserved */
+} __packed;
+
+struct dasd_pprc_data_sc4 {
+	struct dasd_pprc_header header;
+	struct dasd_pprc_dev_info dev_info[5];
+} __packed;
+
+#define DASD_BUS_ID_SIZE 20
+#define DASD_CP_ENTRIES 5
+
+struct dasd_copy_entry {
+	char busid[DASD_BUS_ID_SIZE];
+	struct dasd_device *device;
+	bool primary;
+	bool configured;
+};
+
+struct dasd_copy_relation {
+	struct dasd_copy_entry entry[DASD_CP_ENTRIES];
+	struct dasd_copy_entry *active;
+};
+
+int dasd_devmap_set_device_copy_relation(struct ccw_device *,
+					 bool pprc_enabled);
+
+/*
  * the struct dasd_discipline is
  * sth like a table of virtual functions, if you think of dasd_eckd
  * inheriting dasd...
···
 	struct dasd_ccw_req *(*ese_format)(struct dasd_device *,
 					   struct dasd_ccw_req *, struct irb *);
 	int (*ese_read)(struct dasd_ccw_req *, struct irb *);
+	int (*pprc_status)(struct dasd_device *, struct	dasd_pprc_data_sc4 *);
+	bool (*pprc_enabled)(struct dasd_device *);
+	int (*copy_pair_swap)(struct dasd_device *, char *, char *);
+	int (*device_ping)(struct dasd_device *);
 };
 
 extern struct dasd_discipline *dasd_diag_discipline_pointer;
···
 	struct dasd_profile profile;
 	struct dasd_format_entry format_entry;
 	struct kset *paths_info;
+	struct dasd_copy_relation *copy;
 };
 
 struct dasd_block {
 	/* Block device stuff. */
 	struct gendisk *gdp;
-	struct request_queue *request_queue;
 	spinlock_t request_queue_lock;
 	struct blk_mq_tag_set tag_set;
 	struct block_device *bdev;
···
 #define DASD_STOPPED_PENDING 4         /* long busy */
 #define DASD_STOPPED_DC_WAIT 8         /* disconnected, wait */
 #define DASD_STOPPED_SU      16        /* summary unit check handling */
+#define DASD_STOPPED_PPRC    32        /* PPRC swap */
 #define DASD_STOPPED_NOSPC   128       /* no space left */
 
 /* per device flags */
···
 #define DASD_SLEEPON_END_TAG	((void *) 2)
 
 void dasd_put_device_wake(struct dasd_device *);
+
+/*
+ * return values to be returned from the copy pair swap function
+ * 0x00: swap successful
+ * 0x01: swap data invalid
+ * 0x02: no active device found
+ * 0x03: wrong primary specified
+ * 0x04: secondary device not found
+ * 0x05: swap already running
+ */
+#define DASD_COPYPAIRSWAP_SUCCESS	0
+#define DASD_COPYPAIRSWAP_INVALID	1
+#define DASD_COPYPAIRSWAP_NOACTIVE	2
+#define DASD_COPYPAIRSWAP_PRIMARY	3
+#define DASD_COPYPAIRSWAP_SECONDARY	4
+#define DASD_COPYPAIRSWAP_MULTIPLE	5
 
 /*
  * Reference count inliners
···
 extern struct dasd_profile dasd_global_profile;
 extern unsigned int dasd_global_profile_level;
 extern const struct block_device_operations dasd_device_operations;
+extern struct blk_mq_ops dasd_mq_ops;
 
 extern struct kmem_cache *dasd_page_cache;
 
···
 int dasd_generic_verify_path(struct dasd_device *, __u8);
 void dasd_generic_space_exhaust(struct dasd_device *, struct dasd_ccw_req *);
 void dasd_generic_space_avail(struct dasd_device *);
+
+int dasd_generic_requeue_all_requests(struct dasd_device *);
 
 int dasd_generic_read_dev_chars(struct dasd_device *, int, void *, int);
 char *dasd_get_sense(struct irb *);

drivers/s390/block/dasd_ioctl.c

···
 	return rc;
 }
 
+/*
+ * Swap driver iternal copy relation.
+ */
+static int
+dasd_ioctl_copy_pair_swap(struct block_device *bdev, void __user *argp)
+{
+	struct dasd_copypair_swap_data_t data;
+	struct dasd_device *device;
+	int rc;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EACCES;
+
+	device = dasd_device_from_gendisk(bdev->bd_disk);
+	if (!device)
+		return -ENODEV;
+
+	if (copy_from_user(&data, argp, sizeof(struct dasd_copypair_swap_data_t))) {
+		dasd_put_device(device);
+		return -EFAULT;
+	}
+	if (memchr_inv(data.reserved, 0, sizeof(data.reserved))) {
+		pr_warn("%s: Ivalid swap data specified.\n",
+			dev_name(&device->cdev->dev));
+		dasd_put_device(device);
+		return DASD_COPYPAIRSWAP_INVALID;
+	}
+	if (bdev_is_partition(bdev)) {
+		pr_warn("%s: The specified DASD is a partition and cannot be swapped\n",
+			dev_name(&device->cdev->dev));
+		dasd_put_device(device);
+		return DASD_COPYPAIRSWAP_INVALID;
+	}
+	if (!device->copy) {
+		pr_warn("%s: The specified DASD has no copy pair set up\n",
+			dev_name(&device->cdev->dev));
+		dasd_put_device(device);
+		return -ENODEV;
+	}
+	if (!device->discipline->copy_pair_swap) {
+		dasd_put_device(device);
+		return -EOPNOTSUPP;
+	}
+	rc = device->discipline->copy_pair_swap(device, data.primary,
+						data.secondary);
+	dasd_put_device(device);
+
+	return rc;
+}
+
 #ifdef CONFIG_DASD_PROFILE
 /*
  * Reset device profile information
···
 		break;
 	case BIODASDRAS:
 		rc = dasd_ioctl_release_space(bdev, argp);
+		break;
+	case BIODASDCOPYPAIRSWAP:
+		rc = dasd_ioctl_copy_pair_swap(bdev, argp);
 		break;
 	default:
 		/* if the discipline has an ioctl method try it. */

drivers/scsi/hisi_sas/hisi_sas_v2_hw.c

···
 
 ATTRIBUTE_GROUPS(host_v2_hw);
 
-static int map_queues_v2_hw(struct Scsi_Host *shost)
+static void map_queues_v2_hw(struct Scsi_Host *shost)
 {
 	struct hisi_hba *hisi_hba = shost_priv(shost);
 	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
···
 		for_each_cpu(cpu, mask)
 			qmap->mq_map[cpu] = qmap->queue_offset + queue;
 	}
-
-	return 0;
-
 }
 
 static struct scsi_host_template sht_v2_hw = {

drivers/scsi/hisi_sas/hisi_sas_v3_hw.c

···
 	return 0;
 }
 
-static int hisi_sas_map_queues(struct Scsi_Host *shost)
+static void hisi_sas_map_queues(struct Scsi_Host *shost)
 {
 	struct hisi_hba *hisi_hba = shost_priv(shost);
 	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
 
-	return blk_mq_pci_map_queues(qmap, hisi_hba->pci_dev,
-				     BASE_VECTORS_V3_HW);
+	blk_mq_pci_map_queues(qmap, hisi_hba->pci_dev, BASE_VECTORS_V3_HW);
 }
 
 static struct scsi_host_template sht_v3_hw = {

drivers/scsi/megaraid/megaraid_sas_base.c

···
 	return 0;
 }
 
-static int megasas_map_queues(struct Scsi_Host *shost)
+static void megasas_map_queues(struct Scsi_Host *shost)
 {
 	struct megasas_instance *instance;
 	int qoff = 0, offset;
···
 	instance = (struct megasas_instance *)shost->hostdata;
 
 	if (shost->nr_hw_queues == 1)
-		return 0;
+		return;
 
 	offset = instance->low_latency_index_start;
 
···
 		map->queue_offset = qoff;
 		blk_mq_map_queues(map);
 	}
-
-	return 0;
 }
 
 static void megasas_aen_polling(struct work_struct *work);

drivers/scsi/mpi3mr/mpi3mr_os.c

···
  *
  * Return: return zero.
  */
-static int mpi3mr_map_queues(struct Scsi_Host *shost)
+static void mpi3mr_map_queues(struct Scsi_Host *shost)
 {
 	struct mpi3mr_ioc *mrioc = shost_priv(shost);
 	int i, qoff, offset;
···
 		qoff += map->nr_queues;
 		offset += map->nr_queues;
 	}
-
-	return 0;
-
 }
 
 /**

drivers/scsi/mpt3sas/mpt3sas_scsih.c

···
  * scsih_map_queues - map reply queues with request queues
  * @shost: SCSI host pointer
  */
-static int scsih_map_queues(struct Scsi_Host *shost)
+static void scsih_map_queues(struct Scsi_Host *shost)
 {
 	struct MPT3SAS_ADAPTER *ioc =
 	    (struct MPT3SAS_ADAPTER *)shost->hostdata;
···
 	int iopoll_q_count = ioc->reply_queue_count - nr_msix_vectors;
 
 	if (shost->nr_hw_queues == 1)
-		return 0;
+		return;
 
 	for (i = 0, qoff = 0; i < shost->nr_maps; i++) {
 		map = &shost->tag_set.map[i];
···
 
 		qoff += map->nr_queues;
 	}
-	return 0;
 }
 
 /* shost template for SAS 2.0 HBA devices */

drivers/scsi/pm8001/pm8001_init.c

···
 
 struct workqueue_struct *pm8001_wq;
 
-static int pm8001_map_queues(struct Scsi_Host *shost)
+static void pm8001_map_queues(struct Scsi_Host *shost)
 {
 	struct sas_ha_struct *sha = SHOST_TO_SAS_HA(shost);
 	struct pm8001_hba_info *pm8001_ha = sha->lldd_ha;

drivers/scsi/qla2xxx/qla_nvme.c

···
 		struct blk_mq_queue_map *map)
 {
 	struct scsi_qla_host *vha = lport->private;
-	int rc;
 
-	rc = blk_mq_pci_map_queues(map, vha->hw->pdev, vha->irq_offset);
-	if (rc)
-		ql_log(ql_log_warn, vha, 0x21de,
-		       "pci map queue failed 0x%x", rc);
+	blk_mq_pci_map_queues(map, vha->hw->pdev, vha->irq_offset);
 }
 
 static void qla_nvme_localport_delete(struct nvme_fc_local_port *lport)

drivers/scsi/qla2xxx/qla_os.c

···
 
 static void qla2x00_clear_drv_active(struct qla_hw_data *);
 static void qla2x00_free_device(scsi_qla_host_t *);
-static int qla2xxx_map_queues(struct Scsi_Host *shost);
+static void qla2xxx_map_queues(struct Scsi_Host *shost);
 static void qla2x00_destroy_deferred_work(struct qla_hw_data *);
 
 u32 ql2xnvme_queues = DEF_NVME_HW_QUEUES;
···
 	clear_bit(ABORT_ISP_ACTIVE, &base_vha->dpc_flags);
 }
 
-static int qla2xxx_map_queues(struct Scsi_Host *shost)
+static void qla2xxx_map_queues(struct Scsi_Host *shost)
 {
-	int rc;
 	scsi_qla_host_t *vha = (scsi_qla_host_t *)shost->hostdata;
 	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
 
 	if (USER_CTRL_IRQ(vha->hw) || !vha->hw->mqiobase)
-		rc = blk_mq_map_queues(qmap);
+		blk_mq_map_queues(qmap);
 	else
-		rc = blk_mq_pci_map_queues(qmap, vha->hw->pdev, vha->irq_offset);
-	return rc;
+		blk_mq_pci_map_queues(qmap, vha->hw->pdev, vha->irq_offset);
 }
 
 struct scsi_host_template qla2xxx_driver_template = {

drivers/scsi/scsi_debug.c

···
 	return check_condition_result;
 }
 
-static int sdebug_map_queues(struct Scsi_Host *shost)
+static void sdebug_map_queues(struct Scsi_Host *shost)
 {
 	int i, qoff;
 
 	if (shost->nr_hw_queues == 1)
-		return 0;
+		return;
 
 	for (i = 0, qoff = 0; i < HCTX_MAX_TYPES; i++) {
 		struct blk_mq_queue_map *map = &shost->tag_set.map[i];
···
 
 		qoff += map->nr_queues;
 	}
-
-	return 0;
-
 }
 
 static int sdebug_blk_mq_poll(struct Scsi_Host *shost, unsigned int queue_num)

drivers/scsi/scsi_lib.c

···
 	return 0;
 }
 
-static int scsi_map_queues(struct blk_mq_tag_set *set)
+static void scsi_map_queues(struct blk_mq_tag_set *set)
 {
 	struct Scsi_Host *shost = container_of(set, struct Scsi_Host, tag_set);
 
 	if (shost->hostt->map_queues)
 		return shost->hostt->map_queues(shost);
-	return blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
+	blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
 }
 
 void __scsi_init_queue(struct Scsi_Host *shost, struct request_queue *q)

drivers/scsi/smartpqi/smartpqi_init.c

···
 	return 0;
 }
 
-static int pqi_map_queues(struct Scsi_Host *shost)
+static void pqi_map_queues(struct Scsi_Host *shost)
 {
 	struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
 
-	return blk_mq_pci_map_queues(&shost->tag_set.map[HCTX_TYPE_DEFAULT],
-					ctrl_info->pci_dev, 0);
+	blk_mq_pci_map_queues(&shost->tag_set.map[HCTX_TYPE_DEFAULT],
+			      ctrl_info->pci_dev, 0);
 }
 
 static inline bool pqi_is_tape_changer_device(struct pqi_scsi_dev *device)

drivers/scsi/virtio_scsi.c

···
 	return virtscsi_tmf(vscsi, cmd);
 }
 
-static int virtscsi_map_queues(struct Scsi_Host *shost)
+static void virtscsi_map_queues(struct Scsi_Host *shost)
 {
 	struct virtio_scsi *vscsi = shost_priv(shost);
 	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
 
-	return blk_mq_virtio_map_queues(qmap, vscsi->vdev, 2);
+	blk_mq_virtio_map_queues(qmap, vscsi->vdev, 2);
 }
 
 static void virtscsi_commit_rqs(struct Scsi_Host *shost, u16 hwq)

drivers/ufs/core/ufshcd.c

···
  * Associate the UFS controller queue with the default and poll HCTX types.
  * Initialize the mq_map[] arrays.
  */
-static int ufshcd_map_queues(struct Scsi_Host *shost)
+static void ufshcd_map_queues(struct Scsi_Host *shost)
 {
-	int i, ret;
+	int i;
 
 	for (i = 0; i < shost->nr_maps; i++) {
 		struct blk_mq_queue_map *map = &shost->tag_set.map[i];
···
 			WARN_ON_ONCE(true);
 		}
 		map->queue_offset = 0;
-		ret = blk_mq_map_queues(map);
-		WARN_ON_ONCE(ret);
+		blk_mq_map_queues(map);
 	}
-
-	return 0;
 }
 
 static void ufshcd_init_lrb(struct ufs_hba *hba, struct ufshcd_lrb *lrb, int i)

fs/btrfs/compression.c

···
 #include <linux/string.h>
 #include <linux/backing-dev.h>
 #include <linux/writeback.h>
+#include <linux/psi.h>
 #include <linux/slab.h>
 #include <linux/sched/mm.h>
 #include <linux/log2.h>
···
  */
 static noinline int add_ra_bio_pages(struct inode *inode,
 				     u64 compressed_end,
-				     struct compressed_bio *cb)
+				     struct compressed_bio *cb,
+				     unsigned long *pflags)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	unsigned long end_index;
···
 			cur = (pg_index << PAGE_SHIFT) + PAGE_SIZE;
 			continue;
 		}
+
+		if (PageWorkingset(page))
+			psi_memstall_enter(pflags);
 
 		ret = set_page_extent_mapped(page);
 		if (ret < 0) {
···
 	u64 em_len;
 	u64 em_start;
 	struct extent_map *em;
+	/* Initialize to 1 to make skip psi_memstall_leave unless needed */
+	unsigned long pflags = 1;
 	blk_status_t ret;
 	int ret2;
 	int i;
···
 		goto fail;
 	}
 
-	add_ra_bio_pages(inode, em_start + em_len, cb);
+	add_ra_bio_pages(inode, em_start + em_len, cb, &pflags);
 
 	/* include any pages we added in add_ra-bio_pages */
 	cb->len = bio->bi_iter.bi_size;
···
 			comp_bio = NULL;
 		}
 	}
+
+	if (!pflags)
+		psi_memstall_leave(&pflags);
 
 	if (refcount_dec_and_test(&cb->pending_ios))
 		finish_compressed_bio_read(cb);

fs/direct-io.c

···
 	unsigned long flags;
 
 	bio->bi_private = dio;
-	/* don't account direct I/O as memory stall */
-	bio_clear_flag(bio, BIO_WORKINGSET);
 
 	spin_lock_irqsave(&dio->bio_lock, flags);
 	dio->refcount++;

fs/erofs/zdata.c

···
 #include "zdata.h"
 #include "compress.h"
 #include <linux/prefetch.h>
+#include <linux/psi.h>
 
 #include <trace/events/erofs.h>
 
···
 	struct block_device *last_bdev;
 	unsigned int nr_bios = 0;
 	struct bio *bio = NULL;
+	/* initialize to 1 to make skip psi_memstall_leave unless needed */
+	unsigned long pflags = 1;
 
 	bi_private = jobqueueset_init(sb, q, fgq, force_fg);
 	qtail[JQ_BYPASS] = &q[JQ_BYPASS]->head;
···
 			if (bio && (cur != last_index + 1 ||
 				    last_bdev != mdev.m_bdev)) {
 submit_bio_retry:
+				if (!pflags)
+					psi_memstall_leave(&pflags);
 				submit_bio(bio);
 				bio = NULL;
 			}
+
+			if (unlikely(PageWorkingset(page)))
+				psi_memstall_enter(&pflags);
 
 			if (!bio) {
 				bio = bio_alloc(mdev.m_bdev, BIO_MAX_VECS,
···
 			move_to_bypass_jobqueue(pcl, qtail, owned_head);
 	} while (owned_head != Z_EROFS_PCLUSTER_TAIL);
 
-	if (bio)
+	if (bio) {
+		if (!pflags)
+			psi_memstall_leave(&pflags);
 		submit_bio(bio);
+	}
 
 	/*
 	 * although background is preferred, no one is pending for submission.

include/linux/bio.h

···
 {
 	bio_clear_flag(bio, BIO_REMAPPED);
 	if (bio->bi_bdev != bdev)
-		bio_clear_flag(bio, BIO_THROTTLED);
+		bio_clear_flag(bio, BIO_BPS_THROTTLED);
 	bio->bi_bdev = bdev;
 	bio_associate_blkg(bio);
 }

include/linux/blk-cgroup.h

···
 
 struct bio;
 struct cgroup_subsys_state;
-struct request_queue;
+struct gendisk;
 
 #define FC_APPID_LEN              129
 
 #ifdef CONFIG_BLK_CGROUP
 extern struct cgroup_subsys_state * const blkcg_root_css;
 
-void blkcg_schedule_throttle(struct request_queue *q, bool use_memdelay);
+void blkcg_schedule_throttle(struct gendisk *disk, bool use_memdelay);
 void blkcg_maybe_throttle_current(void);
 bool blk_cgroup_congested(void);
 void blkcg_pin_online(struct cgroup_subsys_state *blkcg_css);
···
 
 static inline void blkcg_maybe_throttle_current(void) { }
 static inline bool blk_cgroup_congested(void) { return false; }
-static inline void blkcg_schedule_throttle(struct request_queue *q, bool use_memdelay) { }
 static inline struct cgroup_subsys_state *bio_blkcg_css(struct bio *bio)
 {
 	return NULL;

include/linux/blk-mq-pci.h

···
 struct blk_mq_queue_map;
 struct pci_dev;
 
-int blk_mq_pci_map_queues(struct blk_mq_queue_map *qmap, struct pci_dev *pdev,
-			  int offset);
+void blk_mq_pci_map_queues(struct blk_mq_queue_map *qmap, struct pci_dev *pdev,
+			   int offset);
 
 #endif /* _LINUX_BLK_MQ_PCI_H */

include/linux/blk-mq-rdma.h

···
 struct blk_mq_tag_set;
 struct ib_device;
 
-int blk_mq_rdma_map_queues(struct blk_mq_queue_map *map,
+void blk_mq_rdma_map_queues(struct blk_mq_queue_map *map,
 		struct ib_device *dev, int first_vec);
 
 #endif /* _LINUX_BLK_MQ_RDMA_H */

include/linux/blk-mq-virtio.h

···
 struct blk_mq_queue_map;
 struct virtio_device;
 
-int blk_mq_virtio_map_queues(struct blk_mq_queue_map *qmap,
+void blk_mq_virtio_map_queues(struct blk_mq_queue_map *qmap,
 		struct virtio_device *vdev, int first_vec);
 
 #endif /* _LINUX_BLK_MQ_VIRTIO_H */

include/linux/blk-mq.h

···
 	rq_list_add(dst, rq);
 }
 
+/**
+ * enum blk_eh_timer_return - How the timeout handler should proceed
+ * @BLK_EH_DONE: The block driver completed the command or will complete it at
+ *	a later time.
+ * @BLK_EH_RESET_TIMER: Reset the request timer and continue waiting for the
+ *	request to complete.
+ */
 enum blk_eh_timer_return {
-	BLK_EH_DONE,		/* drivers has completed the command */
-	BLK_EH_RESET_TIMER,	/* reset timer and try again */
+	BLK_EH_DONE,
+	BLK_EH_RESET_TIMER,
 };
 
 #define BLK_TAG_ALLOC_FIFO 0 /* allocate starting from 0 */
···
 	 * @map_queues: This allows drivers specify their own queue mapping by
 	 * overriding the setup-time function that builds the mq_map.
 	 */
-	int (*map_queues)(struct blk_mq_tag_set *set);
+	void (*map_queues)(struct blk_mq_tag_set *set);
 
 #ifdef CONFIG_BLK_DEBUG_FS
 	/**
···
 int blk_mq_freeze_queue_wait_timeout(struct request_queue *q,
 				     unsigned long timeout);
 
-int blk_mq_map_queues(struct blk_mq_queue_map *qmap);
+void blk_mq_map_queues(struct blk_mq_queue_map *qmap);
 void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues);
 
 void blk_mq_quiesce_queue_nowait(struct request_queue *q);
···
 
 struct rq_map_data {
 	struct page **pages;
-	int page_order;
-	int nr_entries;
 	unsigned long offset;
-	int null_mapped;
-	int from_user;
+	unsigned short page_order;
+	unsigned short nr_entries;
+	bool null_mapped;
+	bool from_user;
 };
 
 int blk_rq_map_user(struct request_queue *, struct request *,

include/linux/blk_types.h

···
 	BIO_NO_PAGE_REF,	/* don't put release vec pages */
 	BIO_CLONED,		/* doesn't own data */
 	BIO_BOUNCED,		/* bio is a bounce bio */
-	BIO_WORKINGSET,		/* contains userspace workingset pages */
 	BIO_QUIET,		/* Make BIO Quiet */
 	BIO_CHAIN,		/* chained bio, ->bi_remaining in effect */
 	BIO_REFFED,		/* bio has elevated ->bi_cnt */
-	BIO_THROTTLED,		/* This bio has already been subjected to
+	BIO_BPS_THROTTLED,	/* This bio has already been subjected to
 				 * throttling rules. Don't do it again. */
 	BIO_TRACE_COMPLETION,	/* bio_endio() should trace the final completion
 				 * of this bio. */

include/linux/blkdev.h

···
 #define blk_queue_quiesced(q)	test_bit(QUEUE_FLAG_QUIESCED, &(q)->queue_flags)
 #define blk_queue_pm_only(q)	atomic_read(&(q)->pm_only)
 #define blk_queue_registered(q)	test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags)
-#define blk_queue_nowait(q)	test_bit(QUEUE_FLAG_NOWAIT, &(q)->queue_flags)
 #define blk_queue_sq_sched(q)	test_bit(QUEUE_FLAG_SQ_SCHED, &(q)->queue_flags)
 
 extern void blk_set_pm_only(struct request_queue *q);
···
 	return test_bit(QUEUE_FLAG_FUA, &bdev_get_queue(bdev)->queue_flags);
 }
 
+static inline bool bdev_nowait(struct block_device *bdev)
+{
+	return test_bit(QUEUE_FLAG_NOWAIT, &bdev_get_queue(bdev)->queue_flags);
+}
+
 static inline enum blk_zoned_model bdev_zoned_model(struct block_device *bdev)
 {
 	struct request_queue *q = bdev_get_queue(bdev);
···
 		return blk_queue_is_zoned(q);
 
 	return false;
+}
+
+static inline bool bdev_op_is_zoned_write(struct block_device *bdev,
+					  blk_opf_t op)
+{
+	if (!bdev_is_zoned(bdev))
+		return false;
+
+	return op == REQ_OP_WRITE || op == REQ_OP_WRITE_ZEROES;
 }
 
 static inline sector_t bdev_zone_sectors(struct block_device *bdev)

include/linux/nvme.h

···
 };
 
 enum {
-	NVME_CONNECT_AUTHREQ_ASCR	= (1 << 2),
-	NVME_CONNECT_AUTHREQ_ATR	= (1 << 1),
+	NVME_CONNECT_AUTHREQ_ASCR	= (1U << 18),
+	NVME_CONNECT_AUTHREQ_ATR	= (1U << 17),
 };
 
 struct nvmf_connect_data {

include/linux/pagemap.h

···
 	pgoff_t _index;
 	unsigned int _nr_pages;
 	unsigned int _batch_count;
+	bool _workingset;
+	unsigned long _pflags;
 };
 
 #define DEFINE_READAHEAD(ractl, f, r, m, i)				\

include/linux/sbitmap.h

···
  * sbitmap_queue_wake_up() - Wake up some of waiters in one waitqueue
  * on a &struct sbitmap_queue.
  * @sbq: Bitmap queue to wake up.
+ * @nr: Number of bits cleared.
  */
-void sbitmap_queue_wake_up(struct sbitmap_queue *sbq);
+void sbitmap_queue_wake_up(struct sbitmap_queue *sbq, int nr);
 
 /**
  * sbitmap_queue_show() - Dump &struct sbitmap_queue information to a &struct

include/linux/sed-opal.h

···
 	case IOC_OPAL_MBR_DONE:
 	case IOC_OPAL_WRITE_SHADOW_MBR:
 	case IOC_OPAL_GENERIC_TABLE_RW:
+	case IOC_OPAL_GET_STATUS:
 		return true;
 	}
 	return false;

include/scsi/scsi_host.h

···
 	 *
 	 * Status: OPTIONAL
 	 */
-	int (* map_queues)(struct Scsi_Host *shost);
+	void (* map_queues)(struct Scsi_Host *shost);
 
 	/*
 	 * SCSI interface of blk_poll - poll for IO completions.

include/uapi/linux/sed-opal.h

···
 	__u64 priv;
 };
 
+#define OPAL_FL_SUPPORTED		0x00000001
+#define OPAL_FL_LOCKING_SUPPORTED	0x00000002
+#define OPAL_FL_LOCKING_ENABLED		0x00000004
+#define OPAL_FL_LOCKED			0x00000008
+#define OPAL_FL_MBR_ENABLED		0x00000010
+#define OPAL_FL_MBR_DONE		0x00000020
+
+struct opal_status {
+	__u32 flags;
+	__u32 reserved;
+};
+
 #define IOC_OPAL_SAVE		    _IOW('p', 220, struct opal_lock_unlock)
 #define IOC_OPAL_LOCK_UNLOCK	    _IOW('p', 221, struct opal_lock_unlock)
 #define IOC_OPAL_TAKE_OWNERSHIP	    _IOW('p', 222, struct opal_key)
···
 #define IOC_OPAL_MBR_DONE           _IOW('p', 233, struct opal_mbr_done)
 #define IOC_OPAL_WRITE_SHADOW_MBR   _IOW('p', 234, struct opal_shadow_mbr)
 #define IOC_OPAL_GENERIC_TABLE_RW   _IOW('p', 235, struct opal_read_write_table)
+#define IOC_OPAL_GET_STATUS         _IOR('p', 236, struct opal_status)
 
 #endif /* _UAPI_SED_OPAL_H */

include/uapi/linux/ublk_cmd.h

···
 #define	UBLK_CMD_STOP_DEV	0x07
 #define	UBLK_CMD_SET_PARAMS	0x08
 #define	UBLK_CMD_GET_PARAMS	0x09
-
+#define	UBLK_CMD_START_USER_RECOVERY	0x10
+#define	UBLK_CMD_END_USER_RECOVERY	0x11
 /*
  * IO commands, issued by ublk server, and handled by ublk driver.
  *
···
  */
 #define UBLK_F_NEED_GET_DATA (1UL << 2)
 
+#define UBLK_F_USER_RECOVERY	(1UL << 3)
+
+#define UBLK_F_USER_RECOVERY_REISSUE	(1UL << 4)
+
 /* device state */
 #define UBLK_S_DEV_DEAD	0
 #define UBLK_S_DEV_LIVE	1
+#define UBLK_S_DEV_QUIESCED	2
 
 /* shipped via sqe->cmd of io_uring command */
 struct ublksrv_ctrl_cmd {

io_uring/io_uring.c

···
 
 static bool io_bdev_nowait(struct block_device *bdev)
 {
-	return !bdev || blk_queue_nowait(bdev_get_queue(bdev));
+	return !bdev || bdev_nowait(bdev);
 }
 
 /*

kernel/sched/psi.c

···
 
 	rq_unlock_irq(rq, &rf);
 }
+EXPORT_SYMBOL_GPL(psi_memstall_enter);
 
 /**
  * psi_memstall_leave - mark the end of an memory stall section
···
 
 	rq_unlock_irq(rq, &rf);
 }
+EXPORT_SYMBOL_GPL(psi_memstall_leave);
 
 #ifdef CONFIG_CGROUPS
 int psi_cgroup_alloc(struct cgroup *cgroup)

lib/sbitmap.c

···
 		nr = find_first_zero_bit(&map->word, map_depth);
 		if (nr + nr_tags <= map_depth) {
 			atomic_long_t *ptr = (atomic_long_t *) &map->word;
-			int map_tags = min_t(int, nr_tags, map_depth);
-			unsigned long val, ret;
+			unsigned long val;
 
-			get_mask = ((1UL << map_tags) - 1) << nr;
+			get_mask = ((1UL << nr_tags) - 1) << nr;
+			val = READ_ONCE(map->word);
 			do {
-				val = READ_ONCE(map->word);
 				if ((val & ~get_mask) != val)
 					goto next;
-				ret = atomic_long_cmpxchg(ptr, val, get_mask | val);
-			} while (ret != val);
-			get_mask = (get_mask & ~ret) >> nr;
+			} while (!atomic_long_try_cmpxchg(ptr, &val,
+							  get_mask | val));
+			get_mask = (get_mask & ~val) >> nr;
 			if (get_mask) {
 				*offset = nr + (index << sb->shift);
 				update_alloc_hint_after_get(sb, depth, hint,
-							*offset + map_tags - 1);
+							*offset + nr_tags - 1);
 				return get_mask;
 			}
 		}
···
 	for (i = 0; i < SBQ_WAIT_QUEUES; i++) {
 		struct sbq_wait_state *ws = &sbq->ws[wake_index];
 
-		if (waitqueue_active(&ws->wait)) {
+		if (waitqueue_active(&ws->wait) && atomic_read(&ws->wait_cnt)) {
 			if (wake_index != atomic_read(&sbq->wake_index))
 				atomic_set(&sbq->wake_index, wake_index);
 			return ws;
···
 	return NULL;
 }
 
-static bool __sbq_wake_up(struct sbitmap_queue *sbq)
+static bool __sbq_wake_up(struct sbitmap_queue *sbq, int *nr)
 {
 	struct sbq_wait_state *ws;
 	unsigned int wake_batch;
-	int wait_cnt;
+	int wait_cnt, cur, sub;
+	bool ret;
+
+	if (*nr <= 0)
+		return false;
 
 	ws = sbq_wake_ptr(sbq);
 	if (!ws)
 		return false;
 
-	wait_cnt = atomic_dec_return(&ws->wait_cnt);
-	if (wait_cnt <= 0) {
-		int ret;
-
-		wake_batch = READ_ONCE(sbq->wake_batch);
-
+	cur = atomic_read(&ws->wait_cnt);
+	do {
 		/*
-		 * Pairs with the memory barrier in sbitmap_queue_resize() to
-		 * ensure that we see the batch size update before the wait
-		 * count is reset.
+		 * For concurrent callers of this, callers should call this
+		 * function again to wakeup a new batch on a different 'ws'.
 		 */
-		smp_mb__before_atomic();
+		if (cur == 0)
+			return true;
+		sub = min(*nr, cur);
+		wait_cnt = cur - sub;
+	} while (!atomic_try_cmpxchg(&ws->wait_cnt, &cur, wait_cnt));
 
-		/*
-		 * For concurrent callers of this, the one that failed the
-		 * atomic_cmpxhcg() race should call this function again
-		 * to wakeup a new batch on a different 'ws'.
-		 */
-		ret = atomic_cmpxchg(&ws->wait_cnt, wait_cnt, wake_batch);
-		if (ret == wait_cnt) {
-			sbq_index_atomic_inc(&sbq->wake_index);
-			wake_up_nr(&ws->wait, wake_batch);
-			return false;
-		}
+	/*
+	 * If we decremented queue without waiters, retry to avoid lost
+	 * wakeups.
+	 */
+	if (wait_cnt > 0)
+		return !waitqueue_active(&ws->wait);
 
-		return true;
-	}
+	*nr -= sub;
 
-	return false;
+	/*
+	 * When wait_cnt == 0, we have to be particularly careful as we are
+	 * responsible to reset wait_cnt regardless whether we've actually
+	 * woken up anybody. But in case we didn't wakeup anybody, we still
+	 * need to retry.
+	 */
+	ret = !waitqueue_active(&ws->wait);
+	wake_batch = READ_ONCE(sbq->wake_batch);
+
+	/*
+	 * Wake up first in case that concurrent callers decrease wait_cnt
+	 * while waitqueue is empty.
+	 */
+	wake_up_nr(&ws->wait, wake_batch);
+
+	/*
+	 * Pairs with the memory barrier in sbitmap_queue_resize() to
+	 * ensure that we see the batch size update before the wait
+	 * count is reset.
+	 *
+	 * Also pairs with the implicit barrier between decrementing wait_cnt
+	 * and checking for waitqueue_active() to make sure waitqueue_active()
+	 * sees result of the wakeup if atomic_dec_return() has seen the result
+	 * of atomic_set().
+	 */
+	smp_mb__before_atomic();
+
+	/*
+	 * Increase wake_index before updating wait_cnt, otherwise concurrent
+	 * callers can see valid wait_cnt in old waitqueue, which can cause
+	 * invalid wakeup on the old waitqueue.
+	 */
+	sbq_index_atomic_inc(&sbq->wake_index);
+	atomic_set(&ws->wait_cnt, wake_batch);
+
+	return ret || *nr;
 }
 
-void sbitmap_queue_wake_up(struct sbitmap_queue *sbq)
+void sbitmap_queue_wake_up(struct sbitmap_queue *sbq, int nr)
 {
-	while (__sbq_wake_up(sbq))
+	while (__sbq_wake_up(sbq, &nr))
 		;
 }
 EXPORT_SYMBOL_GPL(sbitmap_queue_wake_up);
···
 		atomic_long_andnot(mask, (atomic_long_t *) addr);
 
 	smp_mb__after_atomic();
-	sbitmap_queue_wake_up(sbq);
+	sbitmap_queue_wake_up(sbq, nr_tags);
 	sbitmap_update_cpu_hint(&sbq->sb, raw_smp_processor_id(),
 					tags[nr_tags - 1] - offset);
 }
···
 	 * waiter. See the comment on waitqueue_active().
 	 */
 	smp_mb__after_atomic();
-	sbitmap_queue_wake_up(sbq);
+	sbitmap_queue_wake_up(sbq, 1);
 	sbitmap_update_cpu_hint(&sbq->sb, cpu, nr);
 }
 EXPORT_SYMBOL_GPL(sbitmap_queue_clear);

mm/filemap.c

···
 static int filemap_read_folio(struct file *file, filler_t filler,
 		struct folio *folio)
 {
+	bool workingset = folio_test_workingset(folio);
+	unsigned long pflags;
 	int error;
 
 	/*
···
 	 * fails.
 	 */
 	folio_clear_error(folio);
+
 	/* Start the actual read. The read will unlock the page. */
+	if (unlikely(workingset))
+		psi_memstall_enter(&pflags);
 	error = filler(file, folio);
+	if (unlikely(workingset))
+		psi_memstall_leave(&pflags);
 	if (error)
 		return error;
 

mm/readahead.c

···
 #include <linux/task_io_accounting_ops.h>
 #include <linux/pagevec.h>
 #include <linux/pagemap.h>
+#include <linux/psi.h>
 #include <linux/syscalls.h>
 #include <linux/file.h>
 #include <linux/mm_inline.h>
···
 	if (!readahead_count(rac))
 		return;
 
+	if (unlikely(rac->_workingset))
+		psi_memstall_enter(&rac->_pflags);
 	blk_start_plug(&plug);
 
 	if (aops->readahead) {
···
 	}
 
 	blk_finish_plug(&plug);
+	if (unlikely(rac->_workingset))
+		psi_memstall_leave(&rac->_pflags);
+	rac->_workingset = false;
 
 	BUG_ON(readahead_count(rac));
 }
···
 		}
 		if (i == nr_to_read - lookahead_size)
 			folio_set_readahead(folio);
+		ractl->_workingset |= folio_test_workingset(folio);
 		ractl->_nr_pages++;
 	}
 
···
 	if (index == mark)
 		folio_set_readahead(folio);
 	err = filemap_add_folio(ractl->mapping, folio, index, gfp);
-	if (err)
+	if (err) {
 		folio_put(folio);
-	else
-		ractl->_nr_pages += 1UL << order;
-	return err;
+		return err;
+	}
+
+	ractl->_nr_pages += 1UL << order;
+	ractl->_workingset |= folio_test_workingset(folio);
+	return 0;
 }
 
 void page_cache_ra_order(struct readahead_control *ractl,
···
 		if (add_to_page_cache_lru(page, mapping, index, gfp_mask) < 0) {
 			put_page(page);
 			return;
+		}
+		if (unlikely(PageWorkingset(page)) && !ractl->_workingset) {
+			ractl->_workingset = true;
+			psi_memstall_enter(&ractl->_pflags);
 		}
 		ractl->_nr_pages++;
 		if (ra) {

mm/swapfile.c

···
 	plist_for_each_entry_safe(si, next, &swap_avail_heads[nid],
 				  avail_lists[nid]) {
 		if (si->bdev) {
-			blkcg_schedule_throttle(bdev_get_queue(si->bdev), true);
+			blkcg_schedule_throttle(si->bdev->bd_disk, true);
 			break;
 		}
 	}