Merge tag 'for-4.21/block-20181221' of git://git.kernel.dk/linux-block

Pull block updates from Jens Axboe:
 "This is the main pull request for block/storage for 4.21.

  Larger than usual, it was a busy round with lots of goodies queued up.
  Most notable is the removal of the old IO stack, which has been a long
  time coming. No new features for a while, everything coming in this
  week has all been fixes for things that were previously merged.

  This contains:

   - Use atomic counters instead of semaphores for mtip32xx (Arnd)

   - Cleanup of the mtip32xx request setup (Christoph)

   - Fix for circular locking dependency in loop (Jan, Tetsuo)

   - bcache (Coly, Guoju, Shenghui)
      * Optimizations for writeback caching
      * Various fixes and improvements

   - nvme (Chaitanya, Christoph, Sagi, Jay, me, Keith)
      * host and target support for NVMe over TCP
      * Error log page support
      * Support for separate read/write/poll queues
      * Much improved polling
      * discard OOM fallback
      * Tracepoint improvements

   - lightnvm (Hans, Hua, Igor, Matias, Javier)
      * Igor added packed metadata to pblk. Now drives without metadata
        per LBA can be used as well.
      * Fix from Geert on uninitialized value on chunk metadata reads.
      * Fixes from Hans and Javier to pblk recovery and write path.
      * Fix from Hua Su to fix a race condition in the pblk recovery
        code.
      * Scan optimization added to pblk recovery from Zhoujie.
      * Small geometry cleanup from me.

   - Conversion of the last few drivers that used the legacy path to
     blk-mq (me)

   - Removal of legacy IO path in SCSI (me, Christoph)

   - Removal of legacy IO stack and schedulers (me)

   - Support for much better polling, now without interrupts at all.
     blk-mq adds support for multiple queue maps, which enables us to
     have a map per type. This in turn enables nvme to have separate
     completion queues for polling, which can then be interrupt-less.
     Also means we're ready for async polled IO, which is hopefully
     coming in the next release.

   - Killing of (now) unused block exports (Christoph)

   - Unification of the blk-rq-qos and blk-wbt wait handling (Josef)

   - Support for zoned testing with null_blk (Masato)

   - sx8 conversion to per-host tag sets (Christoph)

   - IO priority improvements (Damien)

   - mq-deadline zoned fix (Damien)

   - Ref count blkcg series (Dennis)

   - Lots of blk-mq improvements and speedups (me)

   - sbitmap scalability improvements (me)

   - Make core inflight IO accounting per-cpu (Mikulas)

   - Export timeout setting in sysfs (Weiping)

   - Cleanup the direct issue path (Jianchao)

   - Export blk-wbt internals in block debugfs for easier debugging
     (Ming)

   - Lots of other fixes and improvements"

* tag 'for-4.21/block-20181221' of git://git.kernel.dk/linux-block: (364 commits)
  kyber: use sbitmap add_wait_queue/list_del wait helpers
  sbitmap: add helpers for add/del wait queue handling
  block: save irq state in blkg_lookup_create()
  dm: don't reuse bio for flushes
  nvme-pci: trace SQ status on completions
  nvme-rdma: implement polling queue map
  nvme-fabrics: allow user to pass in nr_poll_queues
  nvme-fabrics: allow nvmf_connect_io_queue to poll
  nvme-core: optionally poll sync commands
  block: make request_to_qc_t public
  nvme-tcp: fix spelling mistake "attepmpt" -> "attempt"
  nvme-tcp: fix endianess annotations
  nvmet-tcp: fix endianess annotations
  nvme-pci: refactor nvme_poll_irqdisable to make sparse happy
  nvme-pci: only set nr_maps to 2 if poll queues are supported
  nvmet: use a macro for default error location
  nvmet: fix comparison of a u16 with -1
  blk-mq: enable IO poll if .nr_queues of type poll > 0
  blk-mq: change blk_mq_queue_busy() to blk_mq_queue_inflight()
  blk-mq: skip zero-queue maps in blk_mq_map_swqueue
  ...

23 files changed, 709 insertions, 767 deletions

diff --git a/drivers/block/aoe/aoe.h b/drivers/block/aoe/aoe.h
index 7ca76ed2e71a..84d0fcebd6af 100644
--- a/drivers/block/aoe/aoe.h
+++ b/drivers/block/aoe/aoe.h
@@ -100,6 +100,10 @@ enum {
 	MAX_TAINT = 1000,	/* cap on aoetgt taint */
 };
 
+struct aoe_req {
+	unsigned long nr_bios;
+};
+
 struct buf {
 	ulong nframesout;
 	struct bio *bio;
diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c
index ed26b7287256..e2c6aae2d636 100644
--- a/drivers/block/aoe/aoeblk.c
+++ b/drivers/block/aoe/aoeblk.c
@@ -387,6 +387,7 @@ aoeblk_gdalloc(void *vp)
 
 	set = &d->tag_set;
 	set->ops = &aoeblk_mq_ops;
+	set->cmd_size = sizeof(struct aoe_req);
 	set->nr_hw_queues = 1;
 	set->queue_depth = 128;
 	set->numa_node = NUMA_NO_NODE;
diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c
index bb2fba651bd2..3cf9bc5d8d95 100644
--- a/drivers/block/aoe/aoecmd.c
+++ b/drivers/block/aoe/aoecmd.c
@@ -822,17 +822,6 @@ out:
 	spin_unlock_irqrestore(&d->lock, flags);
 }
 
-static unsigned long
-rqbiocnt(struct request *r)
-{
-	struct bio *bio;
-	unsigned long n = 0;
-
-	__rq_for_each_bio(bio, r)
-		n++;
-	return n;
-}
-
 static void
 bufinit(struct buf *buf, struct request *rq, struct bio *bio)
 {
@@ -847,6 +836,7 @@ nextbuf(struct aoedev *d)
 {
 	struct request *rq;
 	struct request_queue *q;
+	struct aoe_req *req;
 	struct buf *buf;
 	struct bio *bio;
 
@@ -865,7 +855,11 @@ nextbuf(struct aoedev *d)
 		blk_mq_start_request(rq);
 		d->ip.rq = rq;
 		d->ip.nxbio = rq->bio;
-		rq->special = (void *) rqbiocnt(rq);
+
+		req = blk_mq_rq_to_pdu(rq);
+		req->nr_bios = 0;
+		__rq_for_each_bio(bio, rq)
+			req->nr_bios++;
 	}
 	buf = mempool_alloc(d->bufpool, GFP_ATOMIC);
 	if (buf == NULL) {
@@ -1069,16 +1063,13 @@ aoe_end_request(struct aoedev *d, struct request *rq, int fastfail)
 static void
 aoe_end_buf(struct aoedev *d, struct buf *buf)
 {
-	struct request *rq;
-	unsigned long n;
+	struct request *rq = buf->rq;
+	struct aoe_req *req = blk_mq_rq_to_pdu(rq);
 
 	if (buf == d->ip.buf)
 		d->ip.buf = NULL;
-	rq = buf->rq;
 	mempool_free(buf, d->bufpool);
-	n = (unsigned long) rq->special;
-	rq->special = (void *) --n;
-	if (n == 0)
+	if (--req->nr_bios == 0)
 		aoe_end_request(d, rq, 0);
 }
 
diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c
index 9063f8efbd3b..5b49f1b33ebe 100644
--- a/drivers/block/aoe/aoedev.c
+++ b/drivers/block/aoe/aoedev.c
@@ -160,21 +160,22 @@ static void
 aoe_failip(struct aoedev *d)
 {
 	struct request *rq;
+	struct aoe_req *req;
 	struct bio *bio;
-	unsigned long n;
 
 	aoe_failbuf(d, d->ip.buf);
-
 	rq = d->ip.rq;
 	if (rq == NULL)
 		return;
+
+	req = blk_mq_rq_to_pdu(rq);
 	while ((bio = d->ip.nxbio)) {
 		bio->bi_status = BLK_STS_IOERR;
 		d->ip.nxbio = bio->bi_next;
-		n = (unsigned long) rq->special;
-		rq->special = (void *) --n;
+		req->nr_bios--;
 	}
-	if ((unsigned long) rq->special == 0)
+
+	if (!req->nr_bios)
 		aoe_end_request(d, rq, 0);
 }
 
diff --git a/drivers/block/aoe/aoemain.c b/drivers/block/aoe/aoemain.c
index 251482066977..1e4e2971171c 100644
--- a/drivers/block/aoe/aoemain.c
+++ b/drivers/block/aoe/aoemain.c
@@ -24,7 +24,7 @@ static void discover_timer(struct timer_list *t)
 	aoecmd_cfg(0xffff, 0xff);
 }
 
-static void
+static void __exit
 aoe_exit(void)
 {
 	del_timer_sync(&timer);
diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c
index f88b4c26d422..b0dbbdfeb33e 100644
--- a/drivers/block/ataflop.c
+++ b/drivers/block/ataflop.c
@@ -1471,6 +1471,15 @@ static void setup_req_params( int drive )
 			ReqTrack, ReqSector, (unsigned long)ReqData ));
 }
 
+static void ataflop_commit_rqs(struct blk_mq_hw_ctx *hctx)
+{
+	spin_lock_irq(&ataflop_lock);
+	atari_disable_irq(IRQ_MFP_FDC);
+	finish_fdc();
+	atari_enable_irq(IRQ_MFP_FDC);
+	spin_unlock_irq(&ataflop_lock);
+}
+
 static blk_status_t ataflop_queue_rq(struct blk_mq_hw_ctx *hctx,
 				     const struct blk_mq_queue_data *bd)
 {
@@ -1947,6 +1956,7 @@ static const struct block_device_operations floppy_fops = {
 
 static const struct blk_mq_ops ataflop_mq_ops = {
 	.queue_rq = ataflop_queue_rq,
+	.commit_rqs = ataflop_commit_rqs,
 };
 
 static struct kobject *floppy_find(dev_t dev, int *part, void *data)
@@ -1982,6 +1992,7 @@ static int __init atari_floppy_init (void)
 							   &ataflop_mq_ops, 2,
 							   BLK_MQ_F_SHOULD_MERGE);
 		if (IS_ERR(unit[i].disk->queue)) {
+			put_disk(unit[i].disk);
 			ret = PTR_ERR(unit[i].disk->queue);
 			unit[i].disk->queue = NULL;
 			goto err;
@@ -2033,18 +2044,13 @@ static int __init atari_floppy_init (void)
 	return 0;
 
 err:
-	do {
+	while (--i >= 0) {
 		struct gendisk *disk = unit[i].disk;
 
-		if (disk) {
-			if (disk->queue) {
-				blk_cleanup_queue(disk->queue);
-				disk->queue = NULL;
-			}
-			blk_mq_free_tag_set(&unit[i].tag_set);
-			put_disk(unit[i].disk);
-		}
-	} while (i--);
+		blk_cleanup_queue(disk->queue);
+		blk_mq_free_tag_set(&unit[i].tag_set);
+		put_disk(unit[i].disk);
+	}
 
 	unregister_blkdev(FLOPPY_MAJOR, "fd");
 	return ret;
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index fa8204214ac0..f973a2a845c8 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -2792,7 +2792,7 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
 
 	drbd_init_set_defaults(device);
 
-	q = blk_alloc_queue_node(GFP_KERNEL, NUMA_NO_NODE, &resource->req_lock);
+	q = blk_alloc_queue_node(GFP_KERNEL, NUMA_NO_NODE);
 	if (!q)
 		goto out_no_q;
 	device->rq_queue = q;
diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index fb23578e9a41..6f2856c6d0f2 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -2231,7 +2231,6 @@ static void request_done(int uptodate)
 {
 	struct request *req = current_req;
 	struct request_queue *q;
-	unsigned long flags;
 	int block;
 	char msg[sizeof("request done ") + sizeof(int) * 3];
 
@@ -2254,10 +2253,7 @@ static void request_done(int uptodate)
 		if (block > _floppy->sect)
 			DRS->maxtrack = 1;
 
-		/* unlock chained buffers */
-		spin_lock_irqsave(q->queue_lock, flags);
 		floppy_end_request(req, 0);
-		spin_unlock_irqrestore(q->queue_lock, flags);
 	} else {
 		if (rq_data_dir(req) == WRITE) {
 			/* record write error information */
@@ -2269,9 +2265,7 @@ static void request_done(int uptodate)
 			DRWE->last_error_sector = blk_rq_pos(req);
 			DRWE->last_error_generation = DRS->generation;
 		}
-		spin_lock_irqsave(q->queue_lock, flags);
 		floppy_end_request(req, BLK_STS_IOERR);
-		spin_unlock_irqrestore(q->queue_lock, flags);
 	}
 }
 
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index cb0cc8685076..0939f36548c9 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -77,13 +77,14 @@
 #include <linux/falloc.h>
 #include <linux/uio.h>
 #include <linux/ioprio.h>
+#include <linux/blk-cgroup.h>
 
 #include "loop.h"
 
 #include <linux/uaccess.h>
 
 static DEFINE_IDR(loop_index_idr);
-static DEFINE_MUTEX(loop_index_mutex);
+static DEFINE_MUTEX(loop_ctl_mutex);
 
 static int max_part;
 static int part_shift;
@@ -630,18 +631,7 @@ static void loop_reread_partitions(struct loop_device *lo,
 {
 	int rc;
 
-	/*
-	 * bd_mutex has been held already in release path, so don't
-	 * acquire it if this function is called in such case.
-	 *
-	 * If the reread partition isn't from release path, lo_refcnt
-	 * must be at least one and it can only become zero when the
-	 * current holder is released.
-	 */
-	if (!atomic_read(&lo->lo_refcnt))
-		rc = __blkdev_reread_part(bdev);
-	else
-		rc = blkdev_reread_part(bdev);
+	rc = blkdev_reread_part(bdev);
 	if (rc)
 		pr_warn("%s: partition scan of loop%d (%s) failed (rc=%d)\n",
 			__func__, lo->lo_number, lo->lo_file_name, rc);
@@ -688,26 +678,30 @@ static int loop_validate_file(struct file *file, struct block_device *bdev)
 static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
 			  unsigned int arg)
 {
-	struct file	*file, *old_file;
+	struct file	*file = NULL, *old_file;
 	int		error;
+	bool		partscan;
 
+	error = mutex_lock_killable(&loop_ctl_mutex);
+	if (error)
+		return error;
 	error = -ENXIO;
 	if (lo->lo_state != Lo_bound)
-		goto out;
+		goto out_err;
 
 	/* the loop device has to be read-only */
 	error = -EINVAL;
 	if (!(lo->lo_flags & LO_FLAGS_READ_ONLY))
-		goto out;
+		goto out_err;
 
 	error = -EBADF;
 	file = fget(arg);
 	if (!file)
-		goto out;
+		goto out_err;
 
 	error = loop_validate_file(file, bdev);
 	if (error)
-		goto out_putf;
+		goto out_err;
 
 	old_file = lo->lo_backing_file;
 
@@ -715,7 +709,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
 
 	/* size of the new backing store needs to be the same */
 	if (get_loop_size(lo, file) != get_loop_size(lo, old_file))
-		goto out_putf;
+		goto out_err;
 
 	/* and ... switch */
 	blk_mq_freeze_queue(lo->lo_queue);
@@ -726,15 +720,22 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev,
 			     lo->old_gfp_mask & ~(__GFP_IO|__GFP_FS));
 	loop_update_dio(lo);
 	blk_mq_unfreeze_queue(lo->lo_queue);
-
+	partscan = lo->lo_flags & LO_FLAGS_PARTSCAN;
+	mutex_unlock(&loop_ctl_mutex);
+	/*
+	 * We must drop file reference outside of loop_ctl_mutex as dropping
+	 * the file ref can take bd_mutex which creates circular locking
+	 * dependency.
+	 */
 	fput(old_file);
-	if (lo->lo_flags & LO_FLAGS_PARTSCAN)
+	if (partscan)
 		loop_reread_partitions(lo, bdev);
 	return 0;
 
- out_putf:
-	fput(file);
- out:
+out_err:
+	mutex_unlock(&loop_ctl_mutex);
+	if (file)
+		fput(file);
 	return error;
 }
 
@@ -909,6 +910,7 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode,
 	int		lo_flags = 0;
 	int		error;
 	loff_t		size;
+	bool		partscan;
 
 	/* This is safe, since we have a reference from open(). */
 	__module_get(THIS_MODULE);
@@ -918,13 +920,17 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode,
 	if (!file)
 		goto out;
 
+	error = mutex_lock_killable(&loop_ctl_mutex);
+	if (error)
+		goto out_putf;
+
 	error = -EBUSY;
 	if (lo->lo_state != Lo_unbound)
-		goto out_putf;
+		goto out_unlock;
 
 	error = loop_validate_file(file, bdev);
 	if (error)
-		goto out_putf;
+		goto out_unlock;
 
 	mapping = file->f_mapping;
 	inode = mapping->host;
@@ -936,10 +942,10 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode,
 	error = -EFBIG;
 	size = get_loop_size(lo, file);
 	if ((loff_t)(sector_t)size != size)
-		goto out_putf;
+		goto out_unlock;
 	error = loop_prepare_queue(lo);
 	if (error)
-		goto out_putf;
+		goto out_unlock;
 
 	error = 0;
 
@@ -971,18 +977,22 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode,
 	lo->lo_state = Lo_bound;
 	if (part_shift)
 		lo->lo_flags |= LO_FLAGS_PARTSCAN;
-	if (lo->lo_flags & LO_FLAGS_PARTSCAN)
-		loop_reread_partitions(lo, bdev);
+	partscan = lo->lo_flags & LO_FLAGS_PARTSCAN;
 
 	/* Grab the block_device to prevent its destruction after we
-	 * put /dev/loopXX inode. Later in loop_clr_fd() we bdput(bdev).
+	 * put /dev/loopXX inode. Later in __loop_clr_fd() we bdput(bdev).
 	 */
 	bdgrab(bdev);
+	mutex_unlock(&loop_ctl_mutex);
+	if (partscan)
+		loop_reread_partitions(lo, bdev);
 	return 0;
 
- out_putf:
+out_unlock:
+	mutex_unlock(&loop_ctl_mutex);
+out_putf:
 	fput(file);
- out:
+out:
 	/* This is safe: open() is still holding a reference. */
 	module_put(THIS_MODULE);
 	return error;
@@ -1025,39 +1035,31 @@ loop_init_xfer(struct loop_device *lo, struct loop_func_table *xfer,
 	return err;
 }
 
-static int loop_clr_fd(struct loop_device *lo)
+static int __loop_clr_fd(struct loop_device *lo, bool release)
 {
-	struct file *filp = lo->lo_backing_file;
+	struct file *filp = NULL;
 	gfp_t gfp = lo->old_gfp_mask;
 	struct block_device *bdev = lo->lo_device;
+	int err = 0;
+	bool partscan = false;
+	int lo_number;
 
-	if (lo->lo_state != Lo_bound)
-		return -ENXIO;
-
-	/*
-	 * If we've explicitly asked to tear down the loop device,
-	 * and it has an elevated reference count, set it for auto-teardown when
-	 * the last reference goes away. This stops $!~#$@ udev from
-	 * preventing teardown because it decided that it needs to run blkid on
-	 * the loopback device whenever they appear. xfstests is notorious for
-	 * failing tests because blkid via udev races with a losetup
-	 * <dev>/do something like mkfs/losetup -d <dev> causing the losetup -d
-	 * command to fail with EBUSY.
-	 */
-	if (atomic_read(&lo->lo_refcnt) > 1) {
-		lo->lo_flags |= LO_FLAGS_AUTOCLEAR;
-		mutex_unlock(&lo->lo_ctl_mutex);
-		return 0;
+	mutex_lock(&loop_ctl_mutex);
+	if (WARN_ON_ONCE(lo->lo_state != Lo_rundown)) {
+		err = -ENXIO;
+		goto out_unlock;
 	}
 
-	if (filp == NULL)
-		return -EINVAL;
+	filp = lo->lo_backing_file;
+	if (filp == NULL) {
+		err = -EINVAL;
+		goto out_unlock;
+	}
 
 	/* freeze request queue during the transition */
 	blk_mq_freeze_queue(lo->lo_queue);
 
 	spin_lock_irq(&lo->lo_lock);
-	lo->lo_state = Lo_rundown;
 	lo->lo_backing_file = NULL;
 	spin_unlock_irq(&lo->lo_lock);
 
@@ -1093,21 +1095,73 @@ static int loop_clr_fd(struct loop_device *lo)
 	module_put(THIS_MODULE);
 	blk_mq_unfreeze_queue(lo->lo_queue);
 
-	if (lo->lo_flags & LO_FLAGS_PARTSCAN && bdev)
-		loop_reread_partitions(lo, bdev);
+	partscan = lo->lo_flags & LO_FLAGS_PARTSCAN && bdev;
+	lo_number = lo->lo_number;
 	lo->lo_flags = 0;
 	if (!part_shift)
 		lo->lo_disk->flags |= GENHD_FL_NO_PART_SCAN;
 	loop_unprepare_queue(lo);
-	mutex_unlock(&lo->lo_ctl_mutex);
+out_unlock:
+	mutex_unlock(&loop_ctl_mutex);
+	if (partscan) {
+		/*
+		 * bd_mutex has been held already in release path, so don't
+		 * acquire it if this function is called in such case.
+		 *
+		 * If the reread partition isn't from release path, lo_refcnt
+		 * must be at least one and it can only become zero when the
+		 * current holder is released.
+		 */
+		if (release)
+			err = __blkdev_reread_part(bdev);
+		else
+			err = blkdev_reread_part(bdev);
+		pr_warn("%s: partition scan of loop%d failed (rc=%d)\n",
+			__func__, lo_number, err);
+		/* Device is gone, no point in returning error */
+		err = 0;
+	}
 	/*
-	 * Need not hold lo_ctl_mutex to fput backing file.
-	 * Calling fput holding lo_ctl_mutex triggers a circular
+	 * Need not hold loop_ctl_mutex to fput backing file.
+	 * Calling fput holding loop_ctl_mutex triggers a circular
 	 * lock dependency possibility warning as fput can take
-	 * bd_mutex which is usually taken before lo_ctl_mutex.
+	 * bd_mutex which is usually taken before loop_ctl_mutex.
 	 */
-	fput(filp);
-	return 0;
+	if (filp)
+		fput(filp);
+	return err;
+}
+
+static int loop_clr_fd(struct loop_device *lo)
+{
+	int err;
+
+	err = mutex_lock_killable(&loop_ctl_mutex);
+	if (err)
+		return err;
+	if (lo->lo_state != Lo_bound) {
+		mutex_unlock(&loop_ctl_mutex);
+		return -ENXIO;
+	}
+	/*
+	 * If we've explicitly asked to tear down the loop device,
+	 * and it has an elevated reference count, set it for auto-teardown when
+	 * the last reference goes away. This stops $!~#$@ udev from
+	 * preventing teardown because it decided that it needs to run blkid on
+	 * the loopback device whenever they appear. xfstests is notorious for
+	 * failing tests because blkid via udev races with a losetup
+	 * <dev>/do something like mkfs/losetup -d <dev> causing the losetup -d
+	 * command to fail with EBUSY.
+	 */
+	if (atomic_read(&lo->lo_refcnt) > 1) {
+		lo->lo_flags |= LO_FLAGS_AUTOCLEAR;
+		mutex_unlock(&loop_ctl_mutex);
+		return 0;
+	}
+	lo->lo_state = Lo_rundown;
+	mutex_unlock(&loop_ctl_mutex);
+
+	return __loop_clr_fd(lo, false);
 }
 
 static int
@@ -1116,47 +1170,58 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
 	int err;
 	struct loop_func_table *xfer;
 	kuid_t uid = current_uid();
+	struct block_device *bdev;
+	bool partscan = false;
 
+	err = mutex_lock_killable(&loop_ctl_mutex);
+	if (err)
+		return err;
 	if (lo->lo_encrypt_key_size &&
 	    !uid_eq(lo->lo_key_owner, uid) &&
-	    !capable(CAP_SYS_ADMIN))
-		return -EPERM;
-	if (lo->lo_state != Lo_bound)
-		return -ENXIO;
-	if ((unsigned int) info->lo_encrypt_key_size > LO_KEY_SIZE)
-		return -EINVAL;
+	    !capable(CAP_SYS_ADMIN)) {
+		err = -EPERM;
+		goto out_unlock;
+	}
+	if (lo->lo_state != Lo_bound) {
+		err = -ENXIO;
+		goto out_unlock;
+	}
+	if ((unsigned int) info->lo_encrypt_key_size > LO_KEY_SIZE) {
+		err = -EINVAL;
+		goto out_unlock;
+	}
 
 	/* I/O need to be drained during transfer transition */
 	blk_mq_freeze_queue(lo->lo_queue);
 
 	err = loop_release_xfer(lo);
 	if (err)
-		goto exit;
+		goto out_unfreeze;
 
 	if (info->lo_encrypt_type) {
 		unsigned int type = info->lo_encrypt_type;
 
 		if (type >= MAX_LO_CRYPT) {
 			err = -EINVAL;
-			goto exit;
+			goto out_unfreeze;
 		}
 		xfer = xfer_funcs[type];
 		if (xfer == NULL) {
 			err = -EINVAL;
-			goto exit;
+			goto out_unfreeze;
 		}
 	} else
 		xfer = NULL;
 
 	err = loop_init_xfer(lo, xfer, info);
 	if (err)
-		goto exit;
+		goto out_unfreeze;
 
 	if (lo->lo_offset != info->lo_offset ||
 	    lo->lo_sizelimit != info->lo_sizelimit) {
 		if (figure_loop_size(lo, info->lo_offset, info->lo_sizelimit)) {
 			err = -EFBIG;
-			goto exit;
+			goto out_unfreeze;
 		}
 	}
 
@@ -1188,15 +1253,20 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
 	/* update dio if lo_offset or transfer is changed */
 	__loop_update_dio(lo, lo->use_dio);
 
- exit:
+out_unfreeze:
 	blk_mq_unfreeze_queue(lo->lo_queue);
 
 	if (!err && (info->lo_flags & LO_FLAGS_PARTSCAN) &&
 	     !(lo->lo_flags & LO_FLAGS_PARTSCAN)) {
 		lo->lo_flags |= LO_FLAGS_PARTSCAN;
 		lo->lo_disk->flags &= ~GENHD_FL_NO_PART_SCAN;
-		loop_reread_partitions(lo, lo->lo_device);
+		bdev = lo->lo_device;
+		partscan = true;
 	}
+out_unlock:
+	mutex_unlock(&loop_ctl_mutex);
+	if (partscan)
+		loop_reread_partitions(lo, bdev);
 
 	return err;
 }
@@ -1204,12 +1274,15 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
 static int
 loop_get_status(struct loop_device *lo, struct loop_info64 *info)
 {
-	struct file *file;
+	struct path path;
 	struct kstat stat;
 	int ret;
 
+	ret = mutex_lock_killable(&loop_ctl_mutex);
+	if (ret)
+		return ret;
 	if (lo->lo_state != Lo_bound) {
-		mutex_unlock(&lo->lo_ctl_mutex);
+		mutex_unlock(&loop_ctl_mutex);
 		return -ENXIO;
 	}
 
@@ -1228,17 +1301,17 @@ loop_get_status(struct loop_device *lo, struct loop_info64 *info)
 		       lo->lo_encrypt_key_size);
 	}
 
-	/* Drop lo_ctl_mutex while we call into the filesystem. */
-	file = get_file(lo->lo_backing_file);
-	mutex_unlock(&lo->lo_ctl_mutex);
-	ret = vfs_getattr(&file->f_path, &stat, STATX_INO,
-			  AT_STATX_SYNC_AS_STAT);
+	/* Drop loop_ctl_mutex while we call into the filesystem. */
+	path = lo->lo_backing_file->f_path;
+	path_get(&path);
+	mutex_unlock(&loop_ctl_mutex);
+	ret = vfs_getattr(&path, &stat, STATX_INO, AT_STATX_SYNC_AS_STAT);
 	if (!ret) {
 		info->lo_device = huge_encode_dev(stat.dev);
 		info->lo_inode = stat.ino;
 		info->lo_rdevice = huge_encode_dev(stat.rdev);
 	}
-	fput(file);
+	path_put(&path);
 	return ret;
 }
 
@@ -1322,10 +1395,8 @@ loop_get_status_old(struct loop_device *lo, struct loop_info __user *arg) {
 	struct loop_info64 info64;
 	int err;
 
-	if (!arg) {
-		mutex_unlock(&lo->lo_ctl_mutex);
+	if (!arg)
 		return -EINVAL;
-	}
 	err = loop_get_status(lo, &info64);
 	if (!err)
 		err = loop_info64_to_old(&info64, &info);
@@ -1340,10 +1411,8 @@ loop_get_status64(struct loop_device *lo, struct loop_info64 __user *arg) {
 	struct loop_info64 info64;
 	int err;
 
-	if (!arg) {
-		mutex_unlock(&lo->lo_ctl_mutex);
+	if (!arg)
 		return -EINVAL;
-	}
 	err = loop_get_status(lo, &info64);
 	if (!err && copy_to_user(arg, &info64, sizeof(info64)))
 		err = -EFAULT;
@@ -1393,70 +1462,73 @@ static int loop_set_block_size(struct loop_device *lo, unsigned long arg)
 	return 0;
 }
 
+static int lo_simple_ioctl(struct loop_device *lo, unsigned int cmd,
+			   unsigned long arg)
+{
+	int err;
+
+	err = mutex_lock_killable(&loop_ctl_mutex);
+	if (err)
+		return err;
+	switch (cmd) {
+	case LOOP_SET_CAPACITY:
+		err = loop_set_capacity(lo);
+		break;
+	case LOOP_SET_DIRECT_IO:
+		err = loop_set_dio(lo, arg);
+		break;
+	case LOOP_SET_BLOCK_SIZE:
+		err = loop_set_block_size(lo, arg);
+		break;
+	default:
+		err = lo->ioctl ? lo->ioctl(lo, cmd, arg) : -EINVAL;
+	}
+	mutex_unlock(&loop_ctl_mutex);
+	return err;
+}
+
 static int lo_ioctl(struct block_device *bdev, fmode_t mode,
 	unsigned int cmd, unsigned long arg)
 {
 	struct loop_device *lo = bdev->bd_disk->private_data;
 	int err;
 
-	err = mutex_lock_killable_nested(&lo->lo_ctl_mutex, 1);
-	if (err)
-		goto out_unlocked;
-
 	switch (cmd) {
 	case LOOP_SET_FD:
-		err = loop_set_fd(lo, mode, bdev, arg);
-		break;
+		return loop_set_fd(lo, mode, bdev, arg);
 	case LOOP_CHANGE_FD:
-		err = loop_change_fd(lo, bdev, arg);
-		break;
+		return loop_change_fd(lo, bdev, arg);
 	case LOOP_CLR_FD:
-		/* loop_clr_fd would have unlocked lo_ctl_mutex on success */
-		err = loop_clr_fd(lo);
-		if (!err)
-			goto out_unlocked;
-		break;
+		return loop_clr_fd(lo);
 	case LOOP_SET_STATUS:
 		err = -EPERM;
-		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN))
+		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN)) {
 			err = loop_set_status_old(lo,
 					(struct loop_info __user *)arg);
+		}
 		break;
 	case LOOP_GET_STATUS:
-		err = loop_get_status_old(lo, (struct loop_info __user *) arg);
-		/* loop_get_status() unlocks lo_ctl_mutex */
-		goto out_unlocked;
+		return loop_get_status_old(lo, (struct loop_info __user *) arg);
 	case LOOP_SET_STATUS64:
 		err = -EPERM;
-		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN))
+		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN)) {
 			err = loop_set_status64(lo,
 					(struct loop_info64 __user *) arg);
+		}
 		break;
 	case LOOP_GET_STATUS64:
-		err = loop_get_status64(lo, (struct loop_info64 __user *) arg);
-		/* loop_get_status() unlocks lo_ctl_mutex */
-		goto out_unlocked;
+		return loop_get_status64(lo, (struct loop_info64 __user *) arg);
 	case LOOP_SET_CAPACITY:
-		err = -EPERM;
-		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN))
-			err = loop_set_capacity(lo);
-		break;
 	case LOOP_SET_DIRECT_IO:
-		err = -EPERM;
-		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN))
-			err = loop_set_dio(lo, arg);
-		break;
 	case LOOP_SET_BLOCK_SIZE:
-		err = -EPERM;
-		if ((mode & FMODE_WRITE) || capable(CAP_SYS_ADMIN))
-			err = loop_set_block_size(lo, arg);
-		break;
+		if (!(mode & FMODE_WRITE) && !capable(CAP_SYS_ADMIN))
+			return -EPERM;
+		/* Fall through */
 	default:
-		err = lo->ioctl ? lo->ioctl(lo, cmd, arg) : -EINVAL;
+		err = lo_simple_ioctl(lo, cmd, arg);
+		break;
 	}
-	mutex_unlock(&lo->lo_ctl_mutex);
 
-out_unlocked:
 	return err;
 }
 
@@ -1570,10 +1642,8 @@ loop_get_status_compat(struct loop_device *lo,
 	struct loop_info64 info64;
 	int err;
 
-	if (!arg) {
-		mutex_unlock(&lo->lo_ctl_mutex);
+	if (!arg)
 		return -EINVAL;
-	}
 	err = loop_get_status(lo, &info64);
 	if (!err)
 		err = loop_info64_to_compat(&info64, arg);
@@ -1588,20 +1658,12 @@ static int lo_compat_ioctl(struct block_device *bdev, fmode_t mode,
 
 	switch(cmd) {
 	case LOOP_SET_STATUS:
-		err = mutex_lock_killable(&lo->lo_ctl_mutex);
-		if (!err) {
-			err = loop_set_status_compat(lo,
-						     (const struct compat_loop_info __user *)arg);
-			mutex_unlock(&lo->lo_ctl_mutex);
-		}
+		err = loop_set_status_compat(lo,
+			     (const struct compat_loop_info __user *)arg);
 		break;
 	case LOOP_GET_STATUS:
-		err = mutex_lock_killable(&lo->lo_ctl_mutex);
-		if (!err) {
-			err = loop_get_status_compat(lo,
-						     (struct compat_loop_info __user *)arg);
-			/* loop_get_status() unlocks lo_ctl_mutex */
-		}
+		err = loop_get_status_compat(lo,
+				     (struct compat_loop_info __user *)arg);
 		break;
 	case LOOP_SET_CAPACITY:
 	case LOOP_CLR_FD:
@@ -1625,9 +1687,11 @@ static int lo_compat_ioctl(struct block_device *bdev, fmode_t mode,
 static int lo_open(struct block_device *bdev, fmode_t mode)
 {
 	struct loop_device *lo;
-	int err = 0;
+	int err;
 
-	mutex_lock(&loop_index_mutex);
+	err = mutex_lock_killable(&loop_ctl_mutex);
+	if (err)
+		return err;
 	lo = bdev->bd_disk->private_data;
 	if (!lo) {
 		err = -ENXIO;
@@ -1636,26 +1700,30 @@ static int lo_open(struct block_device *bdev, fmode_t mode)
 
 	atomic_inc(&lo->lo_refcnt);
 out:
-	mutex_unlock(&loop_index_mutex);
+	mutex_unlock(&loop_ctl_mutex);
 	return err;
 }
 
-static void __lo_release(struct loop_device *lo)
+static void lo_release(struct gendisk *disk, fmode_t mode)
 {
-	int err;
+	struct loop_device *lo;
 
+	mutex_lock(&loop_ctl_mutex);
+	lo = disk->private_data;
 	if (atomic_dec_return(&lo->lo_refcnt))
-		return;
+		goto out_unlock;
 
-	mutex_lock(&lo->lo_ctl_mutex);
 	if (lo->lo_flags & LO_FLAGS_AUTOCLEAR) {
+		if (lo->lo_state != Lo_bound)
+			goto out_unlock;
+		lo->lo_state = Lo_rundown;
+		mutex_unlock(&loop_ctl_mutex);
 		/*
 		 * In autoclear mode, stop the loop thread
 		 * and remove configuration after last close.
 		 */
-		err = loop_clr_fd(lo);
-		if (!err)
-			return;
+		__loop_clr_fd(lo, true);
+		return;
 	} else if (lo->lo_state == Lo_bound) {
 		/*
 		 * Otherwise keep thread (if running) and config,
@@ -1665,14 +1733,8 @@ static void __lo_release(struct loop_device *lo)
 		blk_mq_unfreeze_queue(lo->lo_queue);
 	}
 
-	mutex_unlock(&lo->lo_ctl_mutex);
-}
-
-static void lo_release(struct gendisk *disk, fmode_t mode)
-{
-	mutex_lock(&loop_index_mutex);
-	__lo_release(disk->private_data);
-	mutex_unlock(&loop_index_mutex);
+out_unlock:
+	mutex_unlock(&loop_ctl_mutex);
 }
 
 static const struct block_device_operations lo_fops = {
@@ -1711,10 +1773,10 @@ static int unregister_transfer_cb(int id, void *ptr, void *data)
 	struct loop_device *lo = ptr;
 	struct loop_func_table *xfer = data;
 
-	mutex_lock(&lo->lo_ctl_mutex);
+	mutex_lock(&loop_ctl_mutex);
 	if (lo->lo_encryption == xfer)
 		loop_release_xfer(lo);
-	mutex_unlock(&lo->lo_ctl_mutex);
+	mutex_unlock(&loop_ctl_mutex);
 	return 0;
 }
 
@@ -1759,8 +1821,8 @@ static blk_status_t loop_queue_rq(struct blk_mq_hw_ctx *hctx,
 
 	/* always use the first bio's css */
 #ifdef CONFIG_BLK_CGROUP
-	if (cmd->use_aio && rq->bio && rq->bio->bi_css) {
-		cmd->css = rq->bio->bi_css;
+	if (cmd->use_aio && rq->bio && rq->bio->bi_blkg) {
+		cmd->css = &bio_blkcg(rq->bio)->css;
 		css_get(cmd->css);
 	} else
 #endif
@@ -1853,7 +1915,7 @@ static int loop_add(struct loop_device **l, int i)
 		goto out_free_idr;
 
 	lo->lo_queue = blk_mq_init_queue(&lo->tag_set);
-	if (IS_ERR_OR_NULL(lo->lo_queue)) {
+	if (IS_ERR(lo->lo_queue)) {
 		err = PTR_ERR(lo->lo_queue);
 		goto out_cleanup_tags;
 	}
@@ -1895,7 +1957,6 @@ static int loop_add(struct loop_device **l, int i)
 	if (!part_shift)
 		disk->flags |= GENHD_FL_NO_PART_SCAN;
 	disk->flags |= GENHD_FL_EXT_DEVT;
-	mutex_init(&lo->lo_ctl_mutex);
 	atomic_set(&lo->lo_refcnt, 0);
 	lo->lo_number		= i;
 	spin_lock_init(&lo->lo_lock);
@@ -1974,7 +2035,7 @@ static struct kobject *loop_probe(dev_t dev, int *part, void *data)
 	struct kobject *kobj;
 	int err;
 
-	mutex_lock(&loop_index_mutex);
+	mutex_lock(&loop_ctl_mutex);
 	err = loop_lookup(&lo, MINOR(dev) >> part_shift);
 	if (err < 0)
 		err = loop_add(&lo, MINOR(dev) >> part_shift);
@@ -1982,7 +2043,7 @@ static struct kobject *loop_probe(dev_t dev, int *part, void *data)
 		kobj = NULL;
 	else
 		kobj = get_disk_and_module(lo->lo_disk);
-	mutex_unlock(&loop_index_mutex);
+	mutex_unlock(&loop_ctl_mutex);
 
 	*part = 0;
 	return kobj;
@@ -1992,9 +2053,13 @@ static long loop_control_ioctl(struct file *file, unsigned int cmd,
 			       unsigned long parm)
 {
 	struct loop_device *lo;
-	int ret = -ENOSYS;
+	int ret;
 
-	mutex_lock(&loop_index_mutex);
+	ret = mutex_lock_killable(&loop_ctl_mutex);
+	if (ret)
+		return ret;
+
+	ret = -ENOSYS;
 	switch (cmd) {
 	case LOOP_CTL_ADD:
 		ret = loop_lookup(&lo, parm);
@@ -2008,21 +2073,15 @@ static long loop_control_ioctl(struct file *file, unsigned int cmd,
 		ret = loop_lookup(&lo, parm);
 		if (ret < 0)
 			break;
-		ret = mutex_lock_killable(&lo->lo_ctl_mutex);
-		if (ret)
-			break;
 		if (lo->lo_state != Lo_unbound) {
 			ret = -EBUSY;
-			mutex_unlock(&lo->lo_ctl_mutex);
 			break;
 		}
 		if (atomic_read(&lo->lo_refcnt) > 0) {
 			ret = -EBUSY;
-			mutex_unlock(&lo->lo_ctl_mutex);
 			break;
 		}
 		lo->lo_disk->private_data = NULL;
-		mutex_unlock(&lo->lo_ctl_mutex);
 		idr_remove(&loop_index_idr, lo->lo_number);
 		loop_remove(lo);
 		break;
@@ -2032,7 +2091,7 @@ static long loop_control_ioctl(struct file *file, unsigned int cmd,
 			break;
 		ret = loop_add(&lo, -1);
 	}
-	mutex_unlock(&loop_index_mutex);
+	mutex_unlock(&loop_ctl_mutex);
 
 	return ret;
 }
@@ -2116,10 +2175,10 @@ static int __init loop_init(void)
 				  THIS_MODULE, loop_probe, NULL, NULL);
 
 	/* pre-create number of devices given by config or max_loop */
-	mutex_lock(&loop_index_mutex);
+	mutex_lock(&loop_ctl_mutex);
 	for (i = 0; i < nr; i++)
 		loop_add(&lo, i);
-	mutex_unlock(&loop_index_mutex);
+	mutex_unlock(&loop_ctl_mutex);
 
 	printk(KERN_INFO "loop: module loaded\n");
 	return 0;
diff --git a/drivers/block/loop.h b/drivers/block/loop.h
index 4d42c7af7de7..af75a5ee4094 100644
--- a/drivers/block/loop.h
+++ b/drivers/block/loop.h
@@ -54,7 +54,6 @@ struct loop_device {
 
 	spinlock_t		lo_lock;
 	int			lo_state;
-	struct mutex		lo_ctl_mutex;
 	struct kthread_worker	worker;
 	struct task_struct	*worker_task;
 	bool			use_dio;
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index a7daa8acbab3..88e8440e75c3 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -168,41 +168,6 @@ static bool mtip_check_surprise_removal(struct pci_dev *pdev)
 	return false; /* device present */
 }
 
-/* we have to use runtime tag to setup command header */
-static void mtip_init_cmd_header(struct request *rq)
-{
-	struct driver_data *dd = rq->q->queuedata;
-	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
-
-	/* Point the command headers at the command tables. */
-	cmd->command_header = dd->port->command_list +
-				(sizeof(struct mtip_cmd_hdr) * rq->tag);
-	cmd->command_header_dma = dd->port->command_list_dma +
-				(sizeof(struct mtip_cmd_hdr) * rq->tag);
-
-	if (test_bit(MTIP_PF_HOST_CAP_64, &dd->port->flags))
-		cmd->command_header->ctbau = __force_bit2int cpu_to_le32((cmd->command_dma >> 16) >> 16);
-
-	cmd->command_header->ctba = __force_bit2int cpu_to_le32(cmd->command_dma & 0xFFFFFFFF);
-}
-
-static struct mtip_cmd *mtip_get_int_command(struct driver_data *dd)
-{
-	struct request *rq;
-
-	if (mtip_check_surprise_removal(dd->pdev))
-		return NULL;
-
-	rq = blk_mq_alloc_request(dd->queue, REQ_OP_DRV_IN, BLK_MQ_REQ_RESERVED);
-	if (IS_ERR(rq))
-		return NULL;
-
-	/* Internal cmd isn't submitted via .queue_rq */
-	mtip_init_cmd_header(rq);
-
-	return blk_mq_rq_to_pdu(rq);
-}
-
 static struct mtip_cmd *mtip_cmd_from_tag(struct driver_data *dd,
 					  unsigned int tag)
 {
@@ -1023,13 +988,14 @@ static int mtip_exec_internal_command(struct mtip_port *port,
 		return -EFAULT;
 	}
 
-	int_cmd = mtip_get_int_command(dd);
-	if (!int_cmd) {
+	if (mtip_check_surprise_removal(dd->pdev))
+		return -EFAULT;
+
+	rq = blk_mq_alloc_request(dd->queue, REQ_OP_DRV_IN, BLK_MQ_REQ_RESERVED);
+	if (IS_ERR(rq)) {
 		dbg_printk(MTIP_DRV_NAME "Unable to allocate tag for PIO cmd\n");
 		return -EFAULT;
 	}
-	rq = blk_mq_rq_from_pdu(int_cmd);
-	rq->special = &icmd;
 
 	set_bit(MTIP_PF_IC_ACTIVE_BIT, &port->flags);
 
@@ -1050,6 +1016,8 @@ static int mtip_exec_internal_command(struct mtip_port *port,
 	}
 
 	/* Copy the command to the command table */
+	int_cmd = blk_mq_rq_to_pdu(rq);
+	int_cmd->icmd = &icmd;
 	memcpy(int_cmd->command, fis, fis_len*4);
 
 	rq->timeout = timeout;
@@ -1423,23 +1391,19 @@ static int mtip_get_smart_attr(struct mtip_port *port, unsigned int id,
  * @dd		pointer to driver_data structure
  * @lba		starting lba
  * @len		# of 512b sectors to trim
- *
- * return value
- *      -ENOMEM		Out of dma memory
- *      -EINVAL		Invalid parameters passed in, trim not supported
- *      -EIO		Error submitting trim request to hw
  */
-static int mtip_send_trim(struct driver_data *dd, unsigned int lba,
-				unsigned int len)
+static blk_status_t mtip_send_trim(struct driver_data *dd, unsigned int lba,
+		unsigned int len)
 {
-	int i, rv = 0;
 	u64 tlba, tlen, sect_left;
 	struct mtip_trim_entry *buf;
 	dma_addr_t dma_addr;
 	struct host_to_dev_fis fis;
+	blk_status_t ret = BLK_STS_OK;
+	int i;
 
 	if (!len || dd->trim_supp == false)
-		return -EINVAL;
+		return BLK_STS_IOERR;
 
 	/* Trim request too big */
 	WARN_ON(len > (MTIP_MAX_TRIM_ENTRY_LEN * MTIP_MAX_TRIM_ENTRIES));
@@ -1454,7 +1418,7 @@ static int mtip_send_trim(struct driver_data *dd, unsigned int lba,
 	buf = dmam_alloc_coherent(&dd->pdev->dev, ATA_SECT_SIZE, &dma_addr,
 								GFP_KERNEL);
 	if (!buf)
-		return -ENOMEM;
+		return BLK_STS_RESOURCE;
 	memset(buf, 0, ATA_SECT_SIZE);
 
 	for (i = 0, sect_left = len, tlba = lba;
@@ -1463,8 +1427,8 @@ static int mtip_send_trim(struct driver_data *dd, unsigned int lba,
 		tlen = (sect_left >= MTIP_MAX_TRIM_ENTRY_LEN ?
 					MTIP_MAX_TRIM_ENTRY_LEN :
 					sect_left);
-		buf[i].lba = __force_bit2int cpu_to_le32(tlba);
-		buf[i].range = __force_bit2int cpu_to_le16(tlen);
+		buf[i].lba = cpu_to_le32(tlba);
+		buf[i].range = cpu_to_le16(tlen);
 		tlba += tlen;
 		sect_left -= tlen;
 	}
@@ -1486,10 +1450,10 @@ static int mtip_send_trim(struct driver_data *dd, unsigned int lba,
 					ATA_SECT_SIZE,
 					0,
 					MTIP_TRIM_TIMEOUT_MS) < 0)
-		rv = -EIO;
+		ret = BLK_STS_IOERR;
 
 	dmam_free_coherent(&dd->pdev->dev, ATA_SECT_SIZE, buf, dma_addr);
-	return rv;
+	return ret;
 }
 
 /*
@@ -1585,23 +1549,20 @@ static inline void fill_command_sg(struct driver_data *dd,
 	int n;
 	unsigned int dma_len;
 	struct mtip_cmd_sg *command_sg;
-	struct scatterlist *sg = command->sg;
+	struct scatterlist *sg;
 
 	command_sg = command->command + AHCI_CMD_TBL_HDR_SZ;
 
-	for (n = 0; n < nents; n++) {
+	for_each_sg(command->sg, sg, nents, n) {
 		dma_len = sg_dma_len(sg);
 		if (dma_len > 0x400000)
 			dev_err(&dd->pdev->dev,
 				"DMA segment length truncated\n");
-		command_sg->info = __force_bit2int
-			cpu_to_le32((dma_len-1) & 0x3FFFFF);
-		command_sg->dba	= __force_bit2int
-			cpu_to_le32(sg_dma_address(sg));
-		command_sg->dba_upper = __force_bit2int
+		command_sg->info = cpu_to_le32((dma_len-1) & 0x3FFFFF);
+		command_sg->dba	=  cpu_to_le32(sg_dma_address(sg));
+		command_sg->dba_upper =
 			cpu_to_le32((sg_dma_address(sg) >> 16) >> 16);
 		command_sg++;
-		sg++;
 	}
 }
 
@@ -2171,7 +2132,6 @@ static int mtip_hw_ioctl(struct driver_data *dd, unsigned int cmd,
  * @dd       Pointer to the driver data structure.
  * @start    First sector to read.
  * @nsect    Number of sectors to read.
- * @nents    Number of entries in scatter list for the read command.
  * @tag      The tag of this read command.
  * @callback Pointer to the function that should be called
  *	     when the read completes.
@@ -2183,16 +2143,20 @@ static int mtip_hw_ioctl(struct driver_data *dd, unsigned int cmd,
  *	None
  */
 static void mtip_hw_submit_io(struct driver_data *dd, struct request *rq,
-			      struct mtip_cmd *command, int nents,
+			      struct mtip_cmd *command,
 			      struct blk_mq_hw_ctx *hctx)
 {
+	struct mtip_cmd_hdr *hdr =
+		dd->port->command_list + sizeof(struct mtip_cmd_hdr) * rq->tag;
 	struct host_to_dev_fis	*fis;
 	struct mtip_port *port = dd->port;
 	int dma_dir = rq_data_dir(rq) == READ ? DMA_FROM_DEVICE : DMA_TO_DEVICE;
 	u64 start = blk_rq_pos(rq);
 	unsigned int nsect = blk_rq_sectors(rq);
+	unsigned int nents;
 
 	/* Map the scatter list for DMA access */
+	nents = blk_rq_map_sg(hctx->queue, rq, command->sg);
 	nents = dma_map_sg(&dd->pdev->dev, command->sg, nents, dma_dir);
 
 	prefetch(&port->flags);
@@ -2233,10 +2197,11 @@ static void mtip_hw_submit_io(struct driver_data *dd, struct request *rq,
 		fis->device |= 1 << 7;
 
 	/* Populate the command header */
-	command->command_header->opts =
-			__force_bit2int cpu_to_le32(
-				(nents << 16) | 5 | AHCI_CMD_PREFETCH);
-	command->command_header->byte_count = 0;
+	hdr->ctba = cpu_to_le32(command->command_dma & 0xFFFFFFFF);
+	if (test_bit(MTIP_PF_HOST_CAP_64, &dd->port->flags))
+		hdr->ctbau = cpu_to_le32((command->command_dma >> 16) >> 16);
+	hdr->opts = cpu_to_le32((nents << 16) | 5 | AHCI_CMD_PREFETCH);
+	hdr->byte_count = 0;
 
 	command->direction = dma_dir;
 
@@ -2715,12 +2680,12 @@ static void mtip_softirq_done_fn(struct request *rq)
 							cmd->direction);
 
 	if (unlikely(cmd->unaligned))
-		up(&dd->port->cmd_slot_unal);
+		atomic_inc(&dd->port->cmd_slot_unal);
 
 	blk_mq_end_request(rq, cmd->status);
 }
 
-static void mtip_abort_cmd(struct request *req, void *data, bool reserved)
+static bool mtip_abort_cmd(struct request *req, void *data, bool reserved)
 {
 	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(req);
 	struct driver_data *dd = data;
@@ -2730,14 +2695,16 @@ static void mtip_abort_cmd(struct request *req, void *data, bool reserved)
 	clear_bit(req->tag, dd->port->cmds_to_issue);
 	cmd->status = BLK_STS_IOERR;
 	mtip_softirq_done_fn(req);
+	return true;
 }
 
-static void mtip_queue_cmd(struct request *req, void *data, bool reserved)
+static bool mtip_queue_cmd(struct request *req, void *data, bool reserved)
 {
 	struct driver_data *dd = data;
 
 	set_bit(req->tag, dd->port->cmds_to_issue);
 	blk_abort_request(req);
+	return true;
 }
 
 /*
@@ -2803,10 +2770,7 @@ restart_eh:
 
 			blk_mq_quiesce_queue(dd->queue);
 
-			spin_lock(dd->queue->queue_lock);
-			blk_mq_tagset_busy_iter(&dd->tags,
-							mtip_queue_cmd, dd);
-			spin_unlock(dd->queue->queue_lock);
+			blk_mq_tagset_busy_iter(&dd->tags, mtip_queue_cmd, dd);
 
 			set_bit(MTIP_PF_ISSUE_CMDS_BIT, &dd->port->flags);
 
@@ -3026,7 +2990,7 @@ static int mtip_hw_init(struct driver_data *dd)
 	else
 		dd->unal_qdepth = 0;
 
-	sema_init(&dd->port->cmd_slot_unal, dd->unal_qdepth);
+	atomic_set(&dd->port->cmd_slot_unal, dd->unal_qdepth);
 
 	/* Spinlock to prevent concurrent issue */
 	for (i = 0; i < MTIP_MAX_SLOT_GROUPS; i++)
@@ -3531,58 +3495,24 @@ static inline bool is_se_active(struct driver_data *dd)
 	return false;
 }
 
-/*
- * Block layer make request function.
- *
- * This function is called by the kernel to process a BIO for
- * the P320 device.
- *
- * @queue Pointer to the request queue. Unused other than to obtain
- *              the driver data structure.
- * @rq    Pointer to the request.
- *
- */
-static int mtip_submit_request(struct blk_mq_hw_ctx *hctx, struct request *rq)
+static inline bool is_stopped(struct driver_data *dd, struct request *rq)
 {
-	struct driver_data *dd = hctx->queue->queuedata;
-	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
-	unsigned int nents;
-
-	if (is_se_active(dd))
-		return -ENODATA;
-
-	if (unlikely(dd->dd_flag & MTIP_DDF_STOP_IO)) {
-		if (unlikely(test_bit(MTIP_DDF_REMOVE_PENDING_BIT,
-							&dd->dd_flag))) {
-			return -ENXIO;
-		}
-		if (unlikely(test_bit(MTIP_DDF_OVER_TEMP_BIT, &dd->dd_flag))) {
-			return -ENODATA;
-		}
-		if (unlikely(test_bit(MTIP_DDF_WRITE_PROTECT_BIT,
-							&dd->dd_flag) &&
-				rq_data_dir(rq))) {
-			return -ENODATA;
-		}
-		if (unlikely(test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag) ||
-			test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag)))
-			return -ENODATA;
-	}
-
-	if (req_op(rq) == REQ_OP_DISCARD) {
-		int err;
-
-		err = mtip_send_trim(dd, blk_rq_pos(rq), blk_rq_sectors(rq));
-		blk_mq_end_request(rq, err ? BLK_STS_IOERR : BLK_STS_OK);
-		return 0;
-	}
+	if (likely(!(dd->dd_flag & MTIP_DDF_STOP_IO)))
+		return false;
 
-	/* Create the scatter list for this request. */
-	nents = blk_rq_map_sg(hctx->queue, rq, cmd->sg);
+	if (test_bit(MTIP_DDF_REMOVE_PENDING_BIT, &dd->dd_flag))
+		return true;
+	if (test_bit(MTIP_DDF_OVER_TEMP_BIT, &dd->dd_flag))
+		return true;
+	if (test_bit(MTIP_DDF_WRITE_PROTECT_BIT, &dd->dd_flag) &&
+	    rq_data_dir(rq))
+		return true;
+	if (test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag))
+		return true;
+	if (test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag))
+		return true;
 
-	/* Issue the read/write. */
-	mtip_hw_submit_io(dd, rq, cmd, nents, hctx);
-	return 0;
+	return false;
 }
 
 static bool mtip_check_unal_depth(struct blk_mq_hw_ctx *hctx,
@@ -3603,7 +3533,7 @@ static bool mtip_check_unal_depth(struct blk_mq_hw_ctx *hctx,
 			cmd->unaligned = 1;
 	}
 
-	if (cmd->unaligned && down_trylock(&dd->port->cmd_slot_unal))
+	if (cmd->unaligned && atomic_dec_if_positive(&dd->port->cmd_slot_unal) >= 0)
 		return true;
 
 	return false;
@@ -3613,32 +3543,33 @@ static blk_status_t mtip_issue_reserved_cmd(struct blk_mq_hw_ctx *hctx,
 		struct request *rq)
 {
 	struct driver_data *dd = hctx->queue->queuedata;
-	struct mtip_int_cmd *icmd = rq->special;
 	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
+	struct mtip_int_cmd *icmd = cmd->icmd;
+	struct mtip_cmd_hdr *hdr =
+		dd->port->command_list + sizeof(struct mtip_cmd_hdr) * rq->tag;
 	struct mtip_cmd_sg *command_sg;
 
 	if (mtip_commands_active(dd->port))
-		return BLK_STS_RESOURCE;
+		return BLK_STS_DEV_RESOURCE;
 
+	hdr->ctba = cpu_to_le32(cmd->command_dma & 0xFFFFFFFF);
+	if (test_bit(MTIP_PF_HOST_CAP_64, &dd->port->flags))
+		hdr->ctbau = cpu_to_le32((cmd->command_dma >> 16) >> 16);
 	/* Populate the SG list */
-	cmd->command_header->opts =
-		 __force_bit2int cpu_to_le32(icmd->opts | icmd->fis_len);
+	hdr->opts = cpu_to_le32(icmd->opts | icmd->fis_len);
 	if (icmd->buf_len) {
 		command_sg = cmd->command + AHCI_CMD_TBL_HDR_SZ;
 
-		command_sg->info =
-			__force_bit2int cpu_to_le32((icmd->buf_len-1) & 0x3FFFFF);
-		command_sg->dba	=
-			__force_bit2int cpu_to_le32(icmd->buffer & 0xFFFFFFFF);
+		command_sg->info = cpu_to_le32((icmd->buf_len-1) & 0x3FFFFF);
+		command_sg->dba	= cpu_to_le32(icmd->buffer & 0xFFFFFFFF);
 		command_sg->dba_upper =
-			__force_bit2int cpu_to_le32((icmd->buffer >> 16) >> 16);
+			cpu_to_le32((icmd->buffer >> 16) >> 16);
 
-		cmd->command_header->opts |=
-			__force_bit2int cpu_to_le32((1 << 16));
+		hdr->opts |= cpu_to_le32((1 << 16));
 	}
 
 	/* Populate the command header */
-	cmd->command_header->byte_count = 0;
+	hdr->byte_count = 0;
 
 	blk_mq_start_request(rq);
 	mtip_issue_non_ncq_command(dd->port, rq->tag);
@@ -3648,23 +3579,25 @@ static blk_status_t mtip_issue_reserved_cmd(struct blk_mq_hw_ctx *hctx,
 static blk_status_t mtip_queue_rq(struct blk_mq_hw_ctx *hctx,
 			 const struct blk_mq_queue_data *bd)
 {
+	struct driver_data *dd = hctx->queue->queuedata;
 	struct request *rq = bd->rq;
-	int ret;
-
-	mtip_init_cmd_header(rq);
+	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
 
 	if (blk_rq_is_passthrough(rq))
 		return mtip_issue_reserved_cmd(hctx, rq);
 
 	if (unlikely(mtip_check_unal_depth(hctx, rq)))
-		return BLK_STS_RESOURCE;
+		return BLK_STS_DEV_RESOURCE;
+
+	if (is_se_active(dd) || is_stopped(dd, rq))
+		return BLK_STS_IOERR;
 
 	blk_mq_start_request(rq);
 
-	ret = mtip_submit_request(hctx, rq);
-	if (likely(!ret))
-		return BLK_STS_OK;
-	return BLK_STS_IOERR;
+	if (req_op(rq) == REQ_OP_DISCARD)
+		return mtip_send_trim(dd, blk_rq_pos(rq), blk_rq_sectors(rq));
+	mtip_hw_submit_io(dd, rq, cmd, hctx);
+	return BLK_STS_OK;
 }
 
 static void mtip_free_cmd(struct blk_mq_tag_set *set, struct request *rq,
@@ -3920,12 +3853,13 @@ protocol_init_error:
 	return rv;
 }
 
-static void mtip_no_dev_cleanup(struct request *rq, void *data, bool reserv)
+static bool mtip_no_dev_cleanup(struct request *rq, void *data, bool reserv)
 {
 	struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
 
 	cmd->status = BLK_STS_IOERR;
 	blk_mq_complete_request(rq);
+	return true;
 }
 
 /*
diff --git a/drivers/block/mtip32xx/mtip32xx.h b/drivers/block/mtip32xx/mtip32xx.h
index e20e55dab443..abce25f27f57 100644
--- a/drivers/block/mtip32xx/mtip32xx.h
+++ b/drivers/block/mtip32xx/mtip32xx.h
@@ -126,8 +126,6 @@
 
 #define MTIP_DFS_MAX_BUF_SIZE 1024
 
-#define __force_bit2int (unsigned int __force)
-
 enum {
 	/* below are bit numbers in 'flags' defined in mtip_port */
 	MTIP_PF_IC_ACTIVE_BIT       = 0, /* pio/ioctl */
@@ -174,10 +172,10 @@ enum {
 
 struct smart_attr {
 	u8 attr_id;
-	u16 flags;
+	__le16 flags;
 	u8 cur;
 	u8 worst;
-	u32 data;
+	__le32 data;
 	u8 res[3];
 } __packed;
 
@@ -200,9 +198,9 @@ struct mtip_work {
 #define MTIP_MAX_TRIM_ENTRY_LEN		0xfff8
 
 struct mtip_trim_entry {
-	u32 lba;   /* starting lba of region */
-	u16 rsvd;  /* unused */
-	u16 range; /* # of 512b blocks to trim */
+	__le32 lba;   /* starting lba of region */
+	__le16 rsvd;  /* unused */
+	__le16 range; /* # of 512b blocks to trim */
 } __packed;
 
 struct mtip_trim {
@@ -278,24 +276,24 @@ struct mtip_cmd_hdr {
 	 * - Bit 5 Unused in this implementation.
 	 * - Bits 4:0 Length of the command FIS in DWords (DWord = 4 bytes).
 	 */
-	unsigned int opts;
+	__le32 opts;
 	/* This field is unsed when using NCQ. */
 	union {
-		unsigned int byte_count;
-		unsigned int status;
+		__le32 byte_count;
+		__le32 status;
 	};
 	/*
 	 * Lower 32 bits of the command table address associated with this
 	 * header. The command table addresses must be 128 byte aligned.
 	 */
-	unsigned int ctba;
+	__le32 ctba;
 	/*
 	 * If 64 bit addressing is used this field is the upper 32 bits
 	 * of the command table address associated with this command.
 	 */
-	unsigned int ctbau;
+	__le32 ctbau;
 	/* Reserved and unused. */
-	unsigned int res[4];
+	u32 res[4];
 };
 
 /* Command scatter gather structure (PRD). */
@@ -305,31 +303,28 @@ struct mtip_cmd_sg {
 	 * address must be 8 byte aligned signified by bits 2:0 being
 	 * set to 0.
 	 */
-	unsigned int dba;
+	__le32 dba;
 	/*
 	 * When 64 bit addressing is used this field is the upper
 	 * 32 bits of the data buffer address.
 	 */
-	unsigned int dba_upper;
+	__le32 dba_upper;
 	/* Unused. */
-	unsigned int reserved;
+	__le32 reserved;
 	/*
 	 * Bit 31: interrupt when this data block has been transferred.
 	 * Bits 30..22: reserved
 	 * Bits 21..0: byte count (minus 1).  For P320 the byte count must be
 	 * 8 byte aligned signified by bits 2:0 being set to 1.
 	 */
-	unsigned int info;
+	__le32 info;
 };
 struct mtip_port;
 
+struct mtip_int_cmd;
+
 /* Structure used to describe a command. */
 struct mtip_cmd {
-
-	struct mtip_cmd_hdr *command_header; /* ptr to command header entry */
-
-	dma_addr_t command_header_dma; /* corresponding physical address */
-
 	void *command; /* ptr to command table entry */
 
 	dma_addr_t command_dma; /* corresponding physical address */
@@ -338,7 +333,10 @@ struct mtip_cmd {
 
 	int unaligned; /* command is unaligned on 4k boundary */
 
-	struct scatterlist sg[MTIP_MAX_SG]; /* Scatter list entries */
+	union {
+		struct scatterlist sg[MTIP_MAX_SG]; /* Scatter list entries */
+		struct mtip_int_cmd *icmd;
+	};
 
 	int retries; /* The number of retries left for this command. */
 
@@ -435,8 +433,8 @@ struct mtip_port {
 	 */
 	unsigned long ic_pause_timer;
 
-	/* Semaphore to control queue depth of unaligned IOs */
-	struct semaphore cmd_slot_unal;
+	/* Counter to control queue depth of unaligned IOs */
+	atomic_t cmd_slot_unal;
 
 	/* Spinlock for working around command-issue bug. */
 	spinlock_t cmd_issue_lock[MTIP_MAX_SLOT_GROUPS];
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 4d4d6129ff66..08696f5f00bb 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -734,12 +734,13 @@ static void recv_work(struct work_struct *work)
 	kfree(args);
 }
 
-static void nbd_clear_req(struct request *req, void *data, bool reserved)
+static bool nbd_clear_req(struct request *req, void *data, bool reserved)
 {
 	struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
 
 	cmd->status = BLK_STS_IOERR;
 	blk_mq_complete_request(req);
+	return true;
 }
 
 static void nbd_clear_que(struct nbd_device *nbd)
diff --git a/drivers/block/null_blk.h b/drivers/block/null_blk.h
index 7685df43f1ef..b3df2793e7cd 100644
--- a/drivers/block/null_blk.h
+++ b/drivers/block/null_blk.h
@@ -49,6 +49,7 @@ struct nullb_device {
 	unsigned long completion_nsec; /* time in ns to complete a request */
 	unsigned long cache_size; /* disk cache size in MB */
 	unsigned long zone_size; /* zone size in MB if device is zoned */
+	unsigned int zone_nr_conv; /* number of conventional zones */
 	unsigned int submit_queues; /* number of submission queues */
 	unsigned int home_node; /* home node for the device */
 	unsigned int queue_mode; /* block interface */
diff --git a/drivers/block/null_blk_main.c b/drivers/block/null_blk_main.c
index 09339203dfba..62c9654b9ce8 100644
--- a/drivers/block/null_blk_main.c
+++ b/drivers/block/null_blk_main.c
@@ -188,6 +188,10 @@ static unsigned long g_zone_size = 256;
 module_param_named(zone_size, g_zone_size, ulong, S_IRUGO);
 MODULE_PARM_DESC(zone_size, "Zone size in MB when block device is zoned. Must be power-of-two: Default: 256");
 
+static unsigned int g_zone_nr_conv;
+module_param_named(zone_nr_conv, g_zone_nr_conv, uint, 0444);
+MODULE_PARM_DESC(zone_nr_conv, "Number of conventional zones when block device is zoned. Default: 0");
+
 static struct nullb_device *null_alloc_dev(void);
 static void null_free_dev(struct nullb_device *dev);
 static void null_del_dev(struct nullb *nullb);
@@ -293,6 +297,7 @@ NULLB_DEVICE_ATTR(mbps, uint);
 NULLB_DEVICE_ATTR(cache_size, ulong);
 NULLB_DEVICE_ATTR(zoned, bool);
 NULLB_DEVICE_ATTR(zone_size, ulong);
+NULLB_DEVICE_ATTR(zone_nr_conv, uint);
 
 static ssize_t nullb_device_power_show(struct config_item *item, char *page)
 {
@@ -407,6 +412,7 @@ static struct configfs_attribute *nullb_device_attrs[] = {
 	&nullb_device_attr_badblocks,
 	&nullb_device_attr_zoned,
 	&nullb_device_attr_zone_size,
+	&nullb_device_attr_zone_nr_conv,
 	NULL,
 };
 
@@ -520,6 +526,7 @@ static struct nullb_device *null_alloc_dev(void)
 	dev->use_per_node_hctx = g_use_per_node_hctx;
 	dev->zoned = g_zoned;
 	dev->zone_size = g_zone_size;
+	dev->zone_nr_conv = g_zone_nr_conv;
 	return dev;
 }
 
@@ -635,14 +642,9 @@ static void null_cmd_end_timer(struct nullb_cmd *cmd)
 	hrtimer_start(&cmd->timer, kt, HRTIMER_MODE_REL);
 }
 
-static void null_softirq_done_fn(struct request *rq)
+static void null_complete_rq(struct request *rq)
 {
-	struct nullb *nullb = rq->q->queuedata;
-
-	if (nullb->dev->queue_mode == NULL_Q_MQ)
-		end_cmd(blk_mq_rq_to_pdu(rq));
-	else
-		end_cmd(rq->special);
+	end_cmd(blk_mq_rq_to_pdu(rq));
 }
 
 static struct nullb_page *null_alloc_page(gfp_t gfp_flags)
@@ -1350,7 +1352,7 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,
 
 static const struct blk_mq_ops null_mq_ops = {
 	.queue_rq       = null_queue_rq,
-	.complete	= null_softirq_done_fn,
+	.complete	= null_complete_rq,
 	.timeout	= null_timeout_rq,
 };
 
@@ -1657,8 +1659,7 @@ static int null_add_dev(struct nullb_device *dev)
 		}
 		null_init_queues(nullb);
 	} else if (dev->queue_mode == NULL_Q_BIO) {
-		nullb->q = blk_alloc_queue_node(GFP_KERNEL, dev->home_node,
-						NULL);
+		nullb->q = blk_alloc_queue_node(GFP_KERNEL, dev->home_node);
 		if (!nullb->q) {
 			rv = -ENOMEM;
 			goto out_cleanup_queues;
diff --git a/drivers/block/null_blk_zoned.c b/drivers/block/null_blk_zoned.c
index c0b0e4a3fa8f..5d1c261a2cfd 100644
--- a/drivers/block/null_blk_zoned.c
+++ b/drivers/block/null_blk_zoned.c
@@ -29,7 +29,25 @@ int null_zone_init(struct nullb_device *dev)
 	if (!dev->zones)
 		return -ENOMEM;
 
-	for (i = 0; i < dev->nr_zones; i++) {
+	if (dev->zone_nr_conv >= dev->nr_zones) {
+		dev->zone_nr_conv = dev->nr_zones - 1;
+		pr_info("null_blk: changed the number of conventional zones to %u",
+			dev->zone_nr_conv);
+	}
+
+	for (i = 0; i <  dev->zone_nr_conv; i++) {
+		struct blk_zone *zone = &dev->zones[i];
+
+		zone->start = sector;
+		zone->len = dev->zone_size_sects;
+		zone->wp = zone->start + zone->len;
+		zone->type = BLK_ZONE_TYPE_CONVENTIONAL;
+		zone->cond = BLK_ZONE_COND_NOT_WP;
+
+		sector += dev->zone_size_sects;
+	}
+
+	for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
 		struct blk_zone *zone = &dev->zones[i];
 
 		zone->start = zone->wp = sector;
@@ -98,6 +116,8 @@ void null_zone_write(struct nullb_cmd *cmd, sector_t sector,
 		if (zone->wp == zone->start + zone->len)
 			zone->cond = BLK_ZONE_COND_FULL;
 		break;
+	case BLK_ZONE_COND_NOT_WP:
+		break;
 	default:
 		/* Invalid zone condition */
 		cmd->error = BLK_STS_IOERR;
@@ -111,6 +131,11 @@ void null_zone_reset(struct nullb_cmd *cmd, sector_t sector)
 	unsigned int zno = null_zone_no(dev, sector);
 	struct blk_zone *zone = &dev->zones[zno];
 
+	if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL) {
+		cmd->error = BLK_STS_IOERR;
+		return;
+	}
+
 	zone->cond = BLK_ZONE_COND_EMPTY;
 	zone->wp = zone->start;
 }
diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c
index ae4971e5d9a8..0ff9b12d0e35 100644
--- a/drivers/block/paride/pd.c
+++ b/drivers/block/paride/pd.c
@@ -242,6 +242,11 @@ struct pd_unit {
 
 static struct pd_unit pd[PD_UNITS];
 
+struct pd_req {
+	/* for REQ_OP_DRV_IN: */
+	enum action (*func)(struct pd_unit *disk);
+};
+
 static char pd_scratch[512];	/* scratch block buffer */
 
 static char *pd_errs[17] = { "ERR", "INDEX", "ECC", "DRQ", "SEEK", "WRERR",
@@ -502,8 +507,9 @@ static enum action do_pd_io_start(void)
 
 static enum action pd_special(void)
 {
-	enum action (*func)(struct pd_unit *) = pd_req->special;
-	return func(pd_current);
+	struct pd_req *req = blk_mq_rq_to_pdu(pd_req);
+
+	return req->func(pd_current);
 }
 
 static int pd_next_buf(void)
@@ -767,12 +773,14 @@ static int pd_special_command(struct pd_unit *disk,
 		      enum action (*func)(struct pd_unit *disk))
 {
 	struct request *rq;
+	struct pd_req *req;
 
 	rq = blk_get_request(disk->gd->queue, REQ_OP_DRV_IN, 0);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
+	req = blk_mq_rq_to_pdu(rq);
 
-	rq->special = func;
+	req->func = func;
 	blk_execute_rq(disk->gd->queue, disk->gd, rq, 0);
 	blk_put_request(rq);
 	return 0;
@@ -892,9 +900,21 @@ static void pd_probe_drive(struct pd_unit *disk)
 	disk->gd = p;
 	p->private_data = disk;
 
-	p->queue = blk_mq_init_sq_queue(&disk->tag_set, &pd_mq_ops, 2,
-				BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING);
+	memset(&disk->tag_set, 0, sizeof(disk->tag_set));
+	disk->tag_set.ops = &pd_mq_ops;
+	disk->tag_set.cmd_size = sizeof(struct pd_req);
+	disk->tag_set.nr_hw_queues = 1;
+	disk->tag_set.nr_maps = 1;
+	disk->tag_set.queue_depth = 2;
+	disk->tag_set.numa_node = NUMA_NO_NODE;
+	disk->tag_set.flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING;
+
+	if (blk_mq_alloc_tag_set(&disk->tag_set))
+		return;
+
+	p->queue = blk_mq_init_queue(&disk->tag_set);
 	if (IS_ERR(p->queue)) {
+		blk_mq_free_tag_set(&disk->tag_set);
 		p->queue = NULL;
 		return;
 	}
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index 9381f4e3b221..f5a71023f76c 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -2203,9 +2203,7 @@ static int pkt_open_dev(struct pktcdvd_device *pd, fmode_t write)
 		 * Some CDRW drives can not handle writes larger than one packet,
 		 * even if the size is a multiple of the packet size.
 		 */
-		spin_lock_irq(q->queue_lock);
 		blk_queue_max_hw_sectors(q, pd->settings.size);
-		spin_unlock_irq(q->queue_lock);
 		set_bit(PACKET_WRITABLE, &pd->flags);
 	} else {
 		pkt_set_speed(pd, MAX_SPEED, MAX_SPEED);
diff --git a/drivers/block/skd_main.c b/drivers/block/skd_main.c
index 2459dcc04b1c..a10d5736d8f7 100644
--- a/drivers/block/skd_main.c
+++ b/drivers/block/skd_main.c
@@ -181,6 +181,7 @@ struct skd_request_context {
 	struct fit_completion_entry_v1 completion;
 
 	struct fit_comp_error_info err_info;
+	int retries;
 
 	blk_status_t status;
 };
@@ -382,11 +383,12 @@ static void skd_log_skreq(struct skd_device *skdev,
  * READ/WRITE REQUESTS
  *****************************************************************************
  */
-static void skd_inc_in_flight(struct request *rq, void *data, bool reserved)
+static bool skd_inc_in_flight(struct request *rq, void *data, bool reserved)
 {
 	int *count = data;
 
 	count++;
+	return true;
 }
 
 static int skd_in_flight(struct skd_device *skdev)
@@ -494,6 +496,11 @@ static blk_status_t skd_mq_queue_rq(struct blk_mq_hw_ctx *hctx,
 	if (unlikely(skdev->state != SKD_DRVR_STATE_ONLINE))
 		return skd_fail_all(q) ? BLK_STS_IOERR : BLK_STS_RESOURCE;
 
+	if (!(req->rq_flags & RQF_DONTPREP)) {
+		skreq->retries = 0;
+		req->rq_flags |= RQF_DONTPREP;
+	}
+
 	blk_mq_start_request(req);
 
 	WARN_ONCE(tag >= skd_max_queue_depth, "%#x > %#x (nr_requests = %lu)\n",
@@ -1425,7 +1432,7 @@ static void skd_resolve_req_exception(struct skd_device *skdev,
 		break;
 
 	case SKD_CHECK_STATUS_REQUEUE_REQUEST:
-		if ((unsigned long) ++req->special < SKD_MAX_RETRIES) {
+		if (++skreq->retries < SKD_MAX_RETRIES) {
 			skd_log_skreq(skdev, skreq, "retry");
 			blk_mq_requeue_request(req, true);
 			break;
@@ -1887,13 +1894,13 @@ static void skd_isr_fwstate(struct skd_device *skdev)
 		skd_skdev_state_to_str(skdev->state), skdev->state);
 }
 
-static void skd_recover_request(struct request *req, void *data, bool reserved)
+static bool skd_recover_request(struct request *req, void *data, bool reserved)
 {
 	struct skd_device *const skdev = data;
 	struct skd_request_context *skreq = blk_mq_rq_to_pdu(req);
 
 	if (skreq->state != SKD_REQ_STATE_BUSY)
-		return;
+		return true;
 
 	skd_log_skreq(skdev, skreq, "recover");
 
@@ -1904,6 +1911,7 @@ static void skd_recover_request(struct request *req, void *data, bool reserved)
 	skreq->state = SKD_REQ_STATE_IDLE;
 	skreq->status = BLK_STS_IOERR;
 	blk_mq_complete_request(req);
+	return true;
 }
 
 static void skd_recover_requests(struct skd_device *skdev)
diff --git a/drivers/block/sunvdc.c b/drivers/block/sunvdc.c
index b54fa6726303..9c0553dd13e7 100644
--- a/drivers/block/sunvdc.c
+++ b/drivers/block/sunvdc.c
@@ -6,7 +6,7 @@
 #include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/types.h>
-#include <linux/blkdev.h>
+#include <linux/blk-mq.h>
 #include <linux/hdreg.h>
 #include <linux/genhd.h>
 #include <linux/cdrom.h>
@@ -45,6 +45,8 @@ MODULE_VERSION(DRV_MODULE_VERSION);
 #define WAITING_FOR_GEN_CMD	0x04
 #define WAITING_FOR_ANY		-1
 
+#define	VDC_MAX_RETRIES	10
+
 static struct workqueue_struct *sunvdc_wq;
 
 struct vdc_req_entry {
@@ -66,9 +68,10 @@ struct vdc_port {
 
 	u64			max_xfer_size;
 	u32			vdisk_block_size;
+	u32			drain;
 
 	u64			ldc_timeout;
-	struct timer_list	ldc_reset_timer;
+	struct delayed_work	ldc_reset_timer_work;
 	struct work_struct	ldc_reset_work;
 
 	/* The server fills these in for us in the disk attribute
@@ -80,12 +83,14 @@ struct vdc_port {
 	u8			vdisk_mtype;
 	u32			vdisk_phys_blksz;
 
+	struct blk_mq_tag_set	tag_set;
+
 	char			disk_name[32];
 };
 
 static void vdc_ldc_reset(struct vdc_port *port);
 static void vdc_ldc_reset_work(struct work_struct *work);
-static void vdc_ldc_reset_timer(struct timer_list *t);
+static void vdc_ldc_reset_timer_work(struct work_struct *work);
 
 static inline struct vdc_port *to_vdc_port(struct vio_driver_state *vio)
 {
@@ -175,11 +180,8 @@ static void vdc_blk_queue_start(struct vdc_port *port)
 	 * handshake completes, so check for initial handshake before we've
 	 * allocated a disk.
 	 */
-	if (port->disk && blk_queue_stopped(port->disk->queue) &&
-	    vdc_tx_dring_avail(dr) * 100 / VDC_TX_RING_SIZE >= 50) {
-		blk_start_queue(port->disk->queue);
-	}
-
+	if (port->disk && vdc_tx_dring_avail(dr) * 100 / VDC_TX_RING_SIZE >= 50)
+		blk_mq_start_hw_queues(port->disk->queue);
 }
 
 static void vdc_finish(struct vio_driver_state *vio, int err, int waiting_for)
@@ -197,7 +199,7 @@ static void vdc_handshake_complete(struct vio_driver_state *vio)
 {
 	struct vdc_port *port = to_vdc_port(vio);
 
-	del_timer(&port->ldc_reset_timer);
+	cancel_delayed_work(&port->ldc_reset_timer_work);
 	vdc_finish(vio, 0, WAITING_FOR_LINK_UP);
 	vdc_blk_queue_start(port);
 }
@@ -320,7 +322,7 @@ static void vdc_end_one(struct vdc_port *port, struct vio_dring_state *dr,
 
 	rqe->req = NULL;
 
-	__blk_end_request(req, (desc->status ? BLK_STS_IOERR : 0), desc->size);
+	blk_mq_end_request(req, desc->status ? BLK_STS_IOERR : 0);
 
 	vdc_blk_queue_start(port);
 }
@@ -431,6 +433,7 @@ static int __vdc_tx_trigger(struct vdc_port *port)
 		.end_idx		= dr->prod,
 	};
 	int err, delay;
+	int retries = 0;
 
 	hdr.seq = dr->snd_nxt;
 	delay = 1;
@@ -443,6 +446,8 @@ static int __vdc_tx_trigger(struct vdc_port *port)
 		udelay(delay);
 		if ((delay <<= 1) > 128)
 			delay = 128;
+		if (retries++ > VDC_MAX_RETRIES)
+			break;
 	} while (err == -EAGAIN);
 
 	if (err == -ENOTCONN)
@@ -525,29 +530,40 @@ static int __send_request(struct request *req)
 	return err;
 }
 
-static void do_vdc_request(struct request_queue *rq)
+static blk_status_t vdc_queue_rq(struct blk_mq_hw_ctx *hctx,
+				 const struct blk_mq_queue_data *bd)
 {
-	struct request *req;
+	struct vdc_port *port = hctx->queue->queuedata;
+	struct vio_dring_state *dr;
+	unsigned long flags;
 
-	while ((req = blk_peek_request(rq)) != NULL) {
-		struct vdc_port *port;
-		struct vio_dring_state *dr;
+	dr = &port->vio.drings[VIO_DRIVER_TX_RING];
 
-		port = req->rq_disk->private_data;
-		dr = &port->vio.drings[VIO_DRIVER_TX_RING];
-		if (unlikely(vdc_tx_dring_avail(dr) < 1))
-			goto wait;
+	blk_mq_start_request(bd->rq);
 
-		blk_start_request(req);
+	spin_lock_irqsave(&port->vio.lock, flags);
 
-		if (__send_request(req) < 0) {
-			blk_requeue_request(rq, req);
-wait:
-			/* Avoid pointless unplugs. */
-			blk_stop_queue(rq);
-			break;
-		}
+	/*
+	 * Doing drain, just end the request in error
+	 */
+	if (unlikely(port->drain)) {
+		spin_unlock_irqrestore(&port->vio.lock, flags);
+		return BLK_STS_IOERR;
+	}
+
+	if (unlikely(vdc_tx_dring_avail(dr) < 1)) {
+		spin_unlock_irqrestore(&port->vio.lock, flags);
+		blk_mq_stop_hw_queue(hctx);
+		return BLK_STS_DEV_RESOURCE;
+	}
+
+	if (__send_request(bd->rq) < 0) {
+		spin_unlock_irqrestore(&port->vio.lock, flags);
+		return BLK_STS_IOERR;
 	}
+
+	spin_unlock_irqrestore(&port->vio.lock, flags);
+	return BLK_STS_OK;
 }
 
 static int generic_request(struct vdc_port *port, u8 op, void *buf, int len)
@@ -759,6 +775,31 @@ static void vdc_port_down(struct vdc_port *port)
 	vio_ldc_free(&port->vio);
 }
 
+static const struct blk_mq_ops vdc_mq_ops = {
+	.queue_rq	= vdc_queue_rq,
+};
+
+static void cleanup_queue(struct request_queue *q)
+{
+	struct vdc_port *port = q->queuedata;
+
+	blk_cleanup_queue(q);
+	blk_mq_free_tag_set(&port->tag_set);
+}
+
+static struct request_queue *init_queue(struct vdc_port *port)
+{
+	struct request_queue *q;
+
+	q = blk_mq_init_sq_queue(&port->tag_set, &vdc_mq_ops, VDC_TX_RING_SIZE,
+					BLK_MQ_F_SHOULD_MERGE);
+	if (IS_ERR(q))
+		return q;
+
+	q->queuedata = port;
+	return q;
+}
+
 static int probe_disk(struct vdc_port *port)
 {
 	struct request_queue *q;
@@ -796,17 +837,17 @@ static int probe_disk(struct vdc_port *port)
 				    (u64)geom.num_sec);
 	}
 
-	q = blk_init_queue(do_vdc_request, &port->vio.lock);
-	if (!q) {
+	q = init_queue(port);
+	if (IS_ERR(q)) {
 		printk(KERN_ERR PFX "%s: Could not allocate queue.\n",
 		       port->vio.name);
-		return -ENOMEM;
+		return PTR_ERR(q);
 	}
 	g = alloc_disk(1 << PARTITION_SHIFT);
 	if (!g) {
 		printk(KERN_ERR PFX "%s: Could not allocate gendisk.\n",
 		       port->vio.name);
-		blk_cleanup_queue(q);
+		cleanup_queue(q);
 		return -ENOMEM;
 	}
 
@@ -981,7 +1022,7 @@ static int vdc_port_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 	 */
 	ldc_timeout = mdesc_get_property(hp, vdev->mp, "vdc-timeout", NULL);
 	port->ldc_timeout = ldc_timeout ? *ldc_timeout : 0;
-	timer_setup(&port->ldc_reset_timer, vdc_ldc_reset_timer, 0);
+	INIT_DELAYED_WORK(&port->ldc_reset_timer_work, vdc_ldc_reset_timer_work);
 	INIT_WORK(&port->ldc_reset_work, vdc_ldc_reset_work);
 
 	err = vio_driver_init(&port->vio, vdev, VDEV_DISK,
@@ -1034,18 +1075,14 @@ static int vdc_port_remove(struct vio_dev *vdev)
 	struct vdc_port *port = dev_get_drvdata(&vdev->dev);
 
 	if (port) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&port->vio.lock, flags);
-		blk_stop_queue(port->disk->queue);
-		spin_unlock_irqrestore(&port->vio.lock, flags);
+		blk_mq_stop_hw_queues(port->disk->queue);
 
 		flush_work(&port->ldc_reset_work);
-		del_timer_sync(&port->ldc_reset_timer);
+		cancel_delayed_work_sync(&port->ldc_reset_timer_work);
 		del_timer_sync(&port->vio.timer);
 
 		del_gendisk(port->disk);
-		blk_cleanup_queue(port->disk->queue);
+		cleanup_queue(port->disk->queue);
 		put_disk(port->disk);
 		port->disk = NULL;
 
@@ -1080,32 +1117,46 @@ static void vdc_requeue_inflight(struct vdc_port *port)
 		}
 
 		rqe->req = NULL;
-		blk_requeue_request(port->disk->queue, req);
+		blk_mq_requeue_request(req, false);
 	}
 }
 
 static void vdc_queue_drain(struct vdc_port *port)
 {
-	struct request *req;
+	struct request_queue *q = port->disk->queue;
+
+	/*
+	 * Mark the queue as draining, then freeze/quiesce to ensure
+	 * that all existing requests are seen in ->queue_rq() and killed
+	 */
+	port->drain = 1;
+	spin_unlock_irq(&port->vio.lock);
 
-	while ((req = blk_fetch_request(port->disk->queue)) != NULL)
-		__blk_end_request_all(req, BLK_STS_IOERR);
+	blk_mq_freeze_queue(q);
+	blk_mq_quiesce_queue(q);
+
+	spin_lock_irq(&port->vio.lock);
+	port->drain = 0;
+	blk_mq_unquiesce_queue(q);
+	blk_mq_unfreeze_queue(q);
 }
 
-static void vdc_ldc_reset_timer(struct timer_list *t)
+static void vdc_ldc_reset_timer_work(struct work_struct *work)
 {
-	struct vdc_port *port = from_timer(port, t, ldc_reset_timer);
-	struct vio_driver_state *vio = &port->vio;
-	unsigned long flags;
+	struct vdc_port *port;
+	struct vio_driver_state *vio;
 
-	spin_lock_irqsave(&vio->lock, flags);
+	port = container_of(work, struct vdc_port, ldc_reset_timer_work.work);
+	vio = &port->vio;
+
+	spin_lock_irq(&vio->lock);
 	if (!(port->vio.hs_state & VIO_HS_COMPLETE)) {
 		pr_warn(PFX "%s ldc down %llu seconds, draining queue\n",
 			port->disk_name, port->ldc_timeout);
 		vdc_queue_drain(port);
 		vdc_blk_queue_start(port);
 	}
-	spin_unlock_irqrestore(&vio->lock, flags);
+	spin_unlock_irq(&vio->lock);
 }
 
 static void vdc_ldc_reset_work(struct work_struct *work)
@@ -1129,7 +1180,7 @@ static void vdc_ldc_reset(struct vdc_port *port)
 	assert_spin_locked(&port->vio.lock);
 
 	pr_warn(PFX "%s ldc link reset\n", port->disk_name);
-	blk_stop_queue(port->disk->queue);
+	blk_mq_stop_hw_queues(port->disk->queue);
 	vdc_requeue_inflight(port);
 	vdc_port_down(port);
 
@@ -1146,7 +1197,7 @@ static void vdc_ldc_reset(struct vdc_port *port)
 	}
 
 	if (port->ldc_timeout)
-		mod_timer(&port->ldc_reset_timer,
+		mod_delayed_work(system_wq, &port->ldc_reset_timer_work,
 			  round_jiffies(jiffies + HZ * port->ldc_timeout));
 	mod_timer(&port->vio.timer, round_jiffies(jiffies + HZ));
 	return;
diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c
index 064b8c5c7a32..4478eb7efee0 100644
--- a/drivers/block/sx8.c
+++ b/drivers/block/sx8.c
@@ -243,7 +243,6 @@ struct carm_port {
 	unsigned int			port_no;
 	struct gendisk			*disk;
 	struct carm_host		*host;
-	struct blk_mq_tag_set		tag_set;
 
 	/* attached device characteristics */
 	u64				capacity;
@@ -254,13 +253,10 @@ struct carm_port {
 };
 
 struct carm_request {
-	unsigned int			tag;
 	int				n_elem;
 	unsigned int			msg_type;
 	unsigned int			msg_subtype;
 	unsigned int			msg_bucket;
-	struct request			*rq;
-	struct carm_port		*port;
 	struct scatterlist		sg[CARM_MAX_REQ_SG];
 };
 
@@ -291,9 +287,6 @@ struct carm_host {
 	unsigned int			wait_q_cons;
 	struct request_queue		*wait_q[CARM_MAX_WAIT_Q];
 
-	unsigned int			n_msgs;
-	u64				msg_alloc;
-	struct carm_request		req[CARM_MAX_REQ];
 	void				*msg_base;
 	dma_addr_t			msg_dma;
 
@@ -478,10 +471,10 @@ static inline dma_addr_t carm_ref_msg_dma(struct carm_host *host,
 }
 
 static int carm_send_msg(struct carm_host *host,
-			 struct carm_request *crq)
+			 struct carm_request *crq, unsigned tag)
 {
 	void __iomem *mmio = host->mmio;
-	u32 msg = (u32) carm_ref_msg_dma(host, crq->tag);
+	u32 msg = (u32) carm_ref_msg_dma(host, tag);
 	u32 cm_bucket = crq->msg_bucket;
 	u32 tmp;
 	int rc = 0;
@@ -506,99 +499,24 @@ static int carm_send_msg(struct carm_host *host,
 	return rc;
 }
 
-static struct carm_request *carm_get_request(struct carm_host *host)
-{
-	unsigned int i;
-
-	/* obey global hardware limit on S/G entries */
-	if (host->hw_sg_used >= (CARM_MAX_HOST_SG - CARM_MAX_REQ_SG))
-		return NULL;
-
-	for (i = 0; i < max_queue; i++)
-		if ((host->msg_alloc & (1ULL << i)) == 0) {
-			struct carm_request *crq = &host->req[i];
-			crq->port = NULL;
-			crq->n_elem = 0;
-
-			host->msg_alloc |= (1ULL << i);
-			host->n_msgs++;
-
-			assert(host->n_msgs <= CARM_MAX_REQ);
-			sg_init_table(crq->sg, CARM_MAX_REQ_SG);
-			return crq;
-		}
-
-	DPRINTK("no request available, returning NULL\n");
-	return NULL;
-}
-
-static int carm_put_request(struct carm_host *host, struct carm_request *crq)
-{
-	assert(crq->tag < max_queue);
-
-	if (unlikely((host->msg_alloc & (1ULL << crq->tag)) == 0))
-		return -EINVAL; /* tried to clear a tag that was not active */
-
-	assert(host->hw_sg_used >= crq->n_elem);
-
-	host->msg_alloc &= ~(1ULL << crq->tag);
-	host->hw_sg_used -= crq->n_elem;
-	host->n_msgs--;
-
-	return 0;
-}
-
-static struct carm_request *carm_get_special(struct carm_host *host)
-{
-	unsigned long flags;
-	struct carm_request *crq = NULL;
-	struct request *rq;
-	int tries = 5000;
-
-	while (tries-- > 0) {
-		spin_lock_irqsave(&host->lock, flags);
-		crq = carm_get_request(host);
-		spin_unlock_irqrestore(&host->lock, flags);
-
-		if (crq)
-			break;
-		msleep(10);
-	}
-
-	if (!crq)
-		return NULL;
-
-	rq = blk_get_request(host->oob_q, REQ_OP_DRV_OUT, 0);
-	if (IS_ERR(rq)) {
-		spin_lock_irqsave(&host->lock, flags);
-		carm_put_request(host, crq);
-		spin_unlock_irqrestore(&host->lock, flags);
-		return NULL;
-	}
-
-	crq->rq = rq;
-	return crq;
-}
-
 static int carm_array_info (struct carm_host *host, unsigned int array_idx)
 {
 	struct carm_msg_ioctl *ioc;
-	unsigned int idx;
 	u32 msg_data;
 	dma_addr_t msg_dma;
 	struct carm_request *crq;
+	struct request *rq;
 	int rc;
 
-	crq = carm_get_special(host);
-	if (!crq) {
+	rq = blk_mq_alloc_request(host->oob_q, REQ_OP_DRV_OUT, 0);
+	if (IS_ERR(rq)) {
 		rc = -ENOMEM;
 		goto err_out;
 	}
+	crq = blk_mq_rq_to_pdu(rq);
 
-	idx = crq->tag;
-
-	ioc = carm_ref_msg(host, idx);
-	msg_dma = carm_ref_msg_dma(host, idx);
+	ioc = carm_ref_msg(host, rq->tag);
+	msg_dma = carm_ref_msg_dma(host, rq->tag);
 	msg_data = (u32) (msg_dma + sizeof(struct carm_array_info));
 
 	crq->msg_type = CARM_MSG_ARRAY;
@@ -612,7 +530,7 @@ static int carm_array_info (struct carm_host *host, unsigned int array_idx)
 	ioc->type	= CARM_MSG_ARRAY;
 	ioc->subtype	= CARM_ARRAY_INFO;
 	ioc->array_id	= (u8) array_idx;
-	ioc->handle	= cpu_to_le32(TAG_ENCODE(idx));
+	ioc->handle	= cpu_to_le32(TAG_ENCODE(rq->tag));
 	ioc->data_addr	= cpu_to_le32(msg_data);
 
 	spin_lock_irq(&host->lock);
@@ -620,9 +538,8 @@ static int carm_array_info (struct carm_host *host, unsigned int array_idx)
 	       host->state == HST_DEV_SCAN);
 	spin_unlock_irq(&host->lock);
 
-	DPRINTK("blk_execute_rq_nowait, tag == %u\n", idx);
-	crq->rq->special = crq;
-	blk_execute_rq_nowait(host->oob_q, NULL, crq->rq, true, NULL);
+	DPRINTK("blk_execute_rq_nowait, tag == %u\n", rq->tag);
+	blk_execute_rq_nowait(host->oob_q, NULL, rq, true, NULL);
 
 	return 0;
 
@@ -637,21 +554,21 @@ typedef unsigned int (*carm_sspc_t)(struct carm_host *, unsigned int, void *);
 
 static int carm_send_special (struct carm_host *host, carm_sspc_t func)
 {
+	struct request *rq;
 	struct carm_request *crq;
 	struct carm_msg_ioctl *ioc;
 	void *mem;
-	unsigned int idx, msg_size;
+	unsigned int msg_size;
 	int rc;
 
-	crq = carm_get_special(host);
-	if (!crq)
+	rq = blk_mq_alloc_request(host->oob_q, REQ_OP_DRV_OUT, 0);
+	if (IS_ERR(rq))
 		return -ENOMEM;
+	crq = blk_mq_rq_to_pdu(rq);
 
-	idx = crq->tag;
+	mem = carm_ref_msg(host, rq->tag);
 
-	mem = carm_ref_msg(host, idx);
-
-	msg_size = func(host, idx, mem);
+	msg_size = func(host, rq->tag, mem);
 
 	ioc = mem;
 	crq->msg_type = ioc->type;
@@ -660,9 +577,8 @@ static int carm_send_special (struct carm_host *host, carm_sspc_t func)
 	BUG_ON(rc < 0);
 	crq->msg_bucket = (u32) rc;
 
-	DPRINTK("blk_execute_rq_nowait, tag == %u\n", idx);
-	crq->rq->special = crq;
-	blk_execute_rq_nowait(host->oob_q, NULL, crq->rq, true, NULL);
+	DPRINTK("blk_execute_rq_nowait, tag == %u\n", rq->tag);
+	blk_execute_rq_nowait(host->oob_q, NULL, rq, true, NULL);
 
 	return 0;
 }
@@ -744,19 +660,6 @@ static unsigned int carm_fill_get_fw_ver(struct carm_host *host,
 	       sizeof(struct carm_fw_ver);
 }
 
-static inline void carm_end_request_queued(struct carm_host *host,
-					   struct carm_request *crq,
-					   blk_status_t error)
-{
-	struct request *req = crq->rq;
-	int rc;
-
-	blk_mq_end_request(req, error);
-
-	rc = carm_put_request(host, crq);
-	assert(rc == 0);
-}
-
 static inline void carm_push_q (struct carm_host *host, struct request_queue *q)
 {
 	unsigned int idx = host->wait_q_prod % CARM_MAX_WAIT_Q;
@@ -791,101 +694,50 @@ static inline void carm_round_robin(struct carm_host *host)
 	}
 }
 
-static inline void carm_end_rq(struct carm_host *host, struct carm_request *crq,
-			       blk_status_t error)
-{
-	carm_end_request_queued(host, crq, error);
-	if (max_queue == 1)
-		carm_round_robin(host);
-	else if ((host->n_msgs <= CARM_MSG_LOW_WATER) &&
-		 (host->hw_sg_used <= CARM_SG_LOW_WATER)) {
-		carm_round_robin(host);
-	}
-}
-
-static blk_status_t carm_oob_queue_rq(struct blk_mq_hw_ctx *hctx,
-				      const struct blk_mq_queue_data *bd)
+static inline enum dma_data_direction carm_rq_dir(struct request *rq)
 {
-	struct request_queue *q = hctx->queue;
-	struct carm_host *host = q->queuedata;
-	struct carm_request *crq;
-	int rc;
-
-	blk_mq_start_request(bd->rq);
-
-	spin_lock_irq(&host->lock);
-
-	crq = bd->rq->special;
-	assert(crq != NULL);
-	assert(crq->rq == bd->rq);
-
-	crq->n_elem = 0;
-
-	DPRINTK("send req\n");
-	rc = carm_send_msg(host, crq);
-	if (rc) {
-		carm_push_q(host, q);
-		spin_unlock_irq(&host->lock);
-		return BLK_STS_DEV_RESOURCE;
-	}
-
-	spin_unlock_irq(&host->lock);
-	return BLK_STS_OK;
+	return op_is_write(req_op(rq)) ? DMA_TO_DEVICE : DMA_FROM_DEVICE;
 }
 
 static blk_status_t carm_queue_rq(struct blk_mq_hw_ctx *hctx,
 				  const struct blk_mq_queue_data *bd)
 {
 	struct request_queue *q = hctx->queue;
+	struct request *rq = bd->rq;
 	struct carm_port *port = q->queuedata;
 	struct carm_host *host = port->host;
+	struct carm_request *crq = blk_mq_rq_to_pdu(rq);
 	struct carm_msg_rw *msg;
-	struct carm_request *crq;
-	struct request *rq = bd->rq;
 	struct scatterlist *sg;
-	int writing = 0, pci_dir, i, n_elem, rc;
-	u32 tmp;
+	int i, n_elem = 0, rc;
 	unsigned int msg_size;
+	u32 tmp;
+
+	crq->n_elem = 0;
+	sg_init_table(crq->sg, CARM_MAX_REQ_SG);
 
 	blk_mq_start_request(rq);
 
 	spin_lock_irq(&host->lock);
-
-	crq = carm_get_request(host);
-	if (!crq) {
-		carm_push_q(host, q);
-		spin_unlock_irq(&host->lock);
-		return BLK_STS_DEV_RESOURCE;
-	}
-	crq->rq = rq;
-
-	if (rq_data_dir(rq) == WRITE) {
-		writing = 1;
-		pci_dir = DMA_TO_DEVICE;
-	} else {
-		pci_dir = DMA_FROM_DEVICE;
-	}
+	if (req_op(rq) == REQ_OP_DRV_OUT)
+		goto send_msg;
 
 	/* get scatterlist from block layer */
 	sg = &crq->sg[0];
 	n_elem = blk_rq_map_sg(q, rq, sg);
-	if (n_elem <= 0) {
-		/* request with no s/g entries? */
-		carm_end_rq(host, crq, BLK_STS_IOERR);
-		spin_unlock_irq(&host->lock);
-		return BLK_STS_IOERR;
-	}
+	if (n_elem <= 0)
+		goto out_ioerr;
 
 	/* map scatterlist to PCI bus addresses */
-	n_elem = dma_map_sg(&host->pdev->dev, sg, n_elem, pci_dir);
-	if (n_elem <= 0) {
-		/* request with no s/g entries? */
-		carm_end_rq(host, crq, BLK_STS_IOERR);
-		spin_unlock_irq(&host->lock);
-		return BLK_STS_IOERR;
-	}
+	n_elem = dma_map_sg(&host->pdev->dev, sg, n_elem, carm_rq_dir(rq));
+	if (n_elem <= 0)
+		goto out_ioerr;
+
+	/* obey global hardware limit on S/G entries */
+	if (host->hw_sg_used >= CARM_MAX_HOST_SG - n_elem)
+		goto out_resource;
+
 	crq->n_elem = n_elem;
-	crq->port = port;
 	host->hw_sg_used += n_elem;
 
 	/*
@@ -893,9 +745,9 @@ static blk_status_t carm_queue_rq(struct blk_mq_hw_ctx *hctx,
 	 */
 
 	VPRINTK("build msg\n");
-	msg = (struct carm_msg_rw *) carm_ref_msg(host, crq->tag);
+	msg = (struct carm_msg_rw *) carm_ref_msg(host, rq->tag);
 
-	if (writing) {
+	if (rq_data_dir(rq) == WRITE) {
 		msg->type = CARM_MSG_WRITE;
 		crq->msg_type = CARM_MSG_WRITE;
 	} else {
@@ -906,7 +758,7 @@ static blk_status_t carm_queue_rq(struct blk_mq_hw_ctx *hctx,
 	msg->id		= port->port_no;
 	msg->sg_count	= n_elem;
 	msg->sg_type	= SGT_32BIT;
-	msg->handle	= cpu_to_le32(TAG_ENCODE(crq->tag));
+	msg->handle	= cpu_to_le32(TAG_ENCODE(rq->tag));
 	msg->lba	= cpu_to_le32(blk_rq_pos(rq) & 0xffffffff);
 	tmp		= (blk_rq_pos(rq) >> 16) >> 16;
 	msg->lba_high	= cpu_to_le16( (u16) tmp );
@@ -923,22 +775,28 @@ static blk_status_t carm_queue_rq(struct blk_mq_hw_ctx *hctx,
 	rc = carm_lookup_bucket(msg_size);
 	BUG_ON(rc < 0);
 	crq->msg_bucket = (u32) rc;
-
+send_msg:
 	/*
 	 * queue read/write message to hardware
 	 */
-
-	VPRINTK("send msg, tag == %u\n", crq->tag);
-	rc = carm_send_msg(host, crq);
+	VPRINTK("send msg, tag == %u\n", rq->tag);
+	rc = carm_send_msg(host, crq, rq->tag);
 	if (rc) {
-		carm_put_request(host, crq);
-		carm_push_q(host, q);
-		spin_unlock_irq(&host->lock);
-		return BLK_STS_DEV_RESOURCE;
+		host->hw_sg_used -= n_elem;
+		goto out_resource;
 	}
 
 	spin_unlock_irq(&host->lock);
 	return BLK_STS_OK;
+out_resource:
+	dma_unmap_sg(&host->pdev->dev, &crq->sg[0], n_elem, carm_rq_dir(rq));
+	carm_push_q(host, q);
+	spin_unlock_irq(&host->lock);
+	return BLK_STS_DEV_RESOURCE;
+out_ioerr:
+	carm_round_robin(host);
+	spin_unlock_irq(&host->lock);
+	return BLK_STS_IOERR;
 }
 
 static void carm_handle_array_info(struct carm_host *host,
@@ -954,8 +812,6 @@ static void carm_handle_array_info(struct carm_host *host,
 
 	DPRINTK("ENTER\n");
 
-	carm_end_rq(host, crq, error);
-
 	if (error)
 		goto out;
 	if (le32_to_cpu(desc->array_status) & ARRAY_NO_EXIST)
@@ -1011,8 +867,6 @@ static void carm_handle_scan_chan(struct carm_host *host,
 
 	DPRINTK("ENTER\n");
 
-	carm_end_rq(host, crq, error);
-
 	if (error) {
 		new_state = HST_ERROR;
 		goto out;
@@ -1040,8 +894,6 @@ static void carm_handle_generic(struct carm_host *host,
 {
 	DPRINTK("ENTER\n");
 
-	carm_end_rq(host, crq, error);
-
 	assert(host->state == cur_state);
 	if (error)
 		host->state = HST_ERROR;
@@ -1050,28 +902,12 @@ static void carm_handle_generic(struct carm_host *host,
 	schedule_work(&host->fsm_task);
 }
 
-static inline void carm_handle_rw(struct carm_host *host,
-				  struct carm_request *crq, blk_status_t error)
-{
-	int pci_dir;
-
-	VPRINTK("ENTER\n");
-
-	if (rq_data_dir(crq->rq) == WRITE)
-		pci_dir = DMA_TO_DEVICE;
-	else
-		pci_dir = DMA_FROM_DEVICE;
-
-	dma_unmap_sg(&host->pdev->dev, &crq->sg[0], crq->n_elem, pci_dir);
-
-	carm_end_rq(host, crq, error);
-}
-
 static inline void carm_handle_resp(struct carm_host *host,
 				    __le32 ret_handle_le, u32 status)
 {
 	u32 handle = le32_to_cpu(ret_handle_le);
 	unsigned int msg_idx;
+	struct request *rq;
 	struct carm_request *crq;
 	blk_status_t error = (status == RMSG_OK) ? 0 : BLK_STS_IOERR;
 	u8 *mem;
@@ -1087,13 +923,15 @@ static inline void carm_handle_resp(struct carm_host *host,
 	msg_idx = TAG_DECODE(handle);
 	VPRINTK("tag == %u\n", msg_idx);
 
-	crq = &host->req[msg_idx];
+	rq = blk_mq_tag_to_rq(host->tag_set.tags[0], msg_idx);
+	crq = blk_mq_rq_to_pdu(rq);
 
 	/* fast path */
 	if (likely(crq->msg_type == CARM_MSG_READ ||
 		   crq->msg_type == CARM_MSG_WRITE)) {
-		carm_handle_rw(host, crq, error);
-		return;
+		dma_unmap_sg(&host->pdev->dev, &crq->sg[0], crq->n_elem,
+			     carm_rq_dir(rq));
+		goto done;
 	}
 
 	mem = carm_ref_msg(host, msg_idx);
@@ -1103,7 +941,7 @@ static inline void carm_handle_resp(struct carm_host *host,
 		switch (crq->msg_subtype) {
 		case CARM_IOC_SCAN_CHAN:
 			carm_handle_scan_chan(host, crq, mem, error);
-			break;
+			goto done;
 		default:
 			/* unknown / invalid response */
 			goto err_out;
@@ -1116,11 +954,11 @@ static inline void carm_handle_resp(struct carm_host *host,
 		case MISC_ALLOC_MEM:
 			carm_handle_generic(host, crq, error,
 					    HST_ALLOC_BUF, HST_SYNC_TIME);
-			break;
+			goto done;
 		case MISC_SET_TIME:
 			carm_handle_generic(host, crq, error,
 					    HST_SYNC_TIME, HST_GET_FW_VER);
-			break;
+			goto done;
 		case MISC_GET_FW_VER: {
 			struct carm_fw_ver *ver = (struct carm_fw_ver *)
 				(mem + sizeof(struct carm_msg_get_fw_ver));
@@ -1130,7 +968,7 @@ static inline void carm_handle_resp(struct carm_host *host,
 			}
 			carm_handle_generic(host, crq, error,
 					    HST_GET_FW_VER, HST_PORT_SCAN);
-			break;
+			goto done;
 		}
 		default:
 			/* unknown / invalid response */
@@ -1161,7 +999,13 @@ static inline void carm_handle_resp(struct carm_host *host,
 err_out:
 	printk(KERN_WARNING DRV_NAME "(%s): BUG: unhandled message type %d/%d\n",
 	       pci_name(host->pdev), crq->msg_type, crq->msg_subtype);
-	carm_end_rq(host, crq, BLK_STS_IOERR);
+	error = BLK_STS_IOERR;
+done:
+	host->hw_sg_used -= crq->n_elem;
+	blk_mq_end_request(blk_mq_rq_from_pdu(crq), error);
+
+	if (host->hw_sg_used <= CARM_SG_LOW_WATER)
+		carm_round_robin(host);
 }
 
 static inline void carm_handle_responses(struct carm_host *host)
@@ -1491,78 +1335,56 @@ static int carm_init_host(struct carm_host *host)
 	return 0;
 }
 
-static const struct blk_mq_ops carm_oob_mq_ops = {
-	.queue_rq	= carm_oob_queue_rq,
-};
-
 static const struct blk_mq_ops carm_mq_ops = {
 	.queue_rq	= carm_queue_rq,
 };
 
-static int carm_init_disks(struct carm_host *host)
+static int carm_init_disk(struct carm_host *host, unsigned int port_no)
 {
-	unsigned int i;
-	int rc = 0;
+	struct carm_port *port = &host->port[port_no];
+	struct gendisk *disk;
+	struct request_queue *q;
 
-	for (i = 0; i < CARM_MAX_PORTS; i++) {
-		struct gendisk *disk;
-		struct request_queue *q;
-		struct carm_port *port;
+	port->host = host;
+	port->port_no = port_no;
 
-		port = &host->port[i];
-		port->host = host;
-		port->port_no = i;
+	disk = alloc_disk(CARM_MINORS_PER_MAJOR);
+	if (!disk)
+		return -ENOMEM;
 
-		disk = alloc_disk(CARM_MINORS_PER_MAJOR);
-		if (!disk) {
-			rc = -ENOMEM;
-			break;
-		}
+	port->disk = disk;
+	sprintf(disk->disk_name, DRV_NAME "/%u",
+		(unsigned int)host->id * CARM_MAX_PORTS + port_no);
+	disk->major = host->major;
+	disk->first_minor = port_no * CARM_MINORS_PER_MAJOR;
+	disk->fops = &carm_bd_ops;
+	disk->private_data = port;
 
-		port->disk = disk;
-		sprintf(disk->disk_name, DRV_NAME "/%u",
-			(unsigned int) (host->id * CARM_MAX_PORTS) + i);
-		disk->major = host->major;
-		disk->first_minor = i * CARM_MINORS_PER_MAJOR;
-		disk->fops = &carm_bd_ops;
-		disk->private_data = port;
-
-		q = blk_mq_init_sq_queue(&port->tag_set, &carm_mq_ops,
-					 max_queue, BLK_MQ_F_SHOULD_MERGE);
-		if (IS_ERR(q)) {
-			rc = PTR_ERR(q);
-			break;
-		}
-		disk->queue = q;
-		blk_queue_max_segments(q, CARM_MAX_REQ_SG);
-		blk_queue_segment_boundary(q, CARM_SG_BOUNDARY);
+	q = blk_mq_init_queue(&host->tag_set);
+	if (IS_ERR(q))
+		return PTR_ERR(q);
 
-		q->queuedata = port;
-	}
+	blk_queue_max_segments(q, CARM_MAX_REQ_SG);
+	blk_queue_segment_boundary(q, CARM_SG_BOUNDARY);
 
-	return rc;
+	q->queuedata = port;
+	disk->queue = q;
+	return 0;
 }
 
-static void carm_free_disks(struct carm_host *host)
+static void carm_free_disk(struct carm_host *host, unsigned int port_no)
 {
-	unsigned int i;
-
-	for (i = 0; i < CARM_MAX_PORTS; i++) {
-		struct carm_port *port = &host->port[i];
-		struct gendisk *disk = port->disk;
+	struct carm_port *port = &host->port[port_no];
+	struct gendisk *disk = port->disk;
 
-		if (disk) {
-			struct request_queue *q = disk->queue;
+	if (!disk)
+		return;
 
-			if (disk->flags & GENHD_FL_UP)
-				del_gendisk(disk);
-			if (q) {
-				blk_mq_free_tag_set(&port->tag_set);
-				blk_cleanup_queue(q);
-			}
-			put_disk(disk);
-		}
-	}
+	if (disk->flags & GENHD_FL_UP)
+		del_gendisk(disk);
+	if (disk->queue)
+		blk_cleanup_queue(disk->queue);
+	put_disk(disk);
 }
 
 static int carm_init_shm(struct carm_host *host)
@@ -1618,9 +1440,6 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	INIT_WORK(&host->fsm_task, carm_fsm_task);
 	init_completion(&host->probe_comp);
 
-	for (i = 0; i < ARRAY_SIZE(host->req); i++)
-		host->req[i].tag = i;
-
 	host->mmio = ioremap(pci_resource_start(pdev, 0),
 			     pci_resource_len(pdev, 0));
 	if (!host->mmio) {
@@ -1637,14 +1456,26 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto err_out_iounmap;
 	}
 
-	q = blk_mq_init_sq_queue(&host->tag_set, &carm_oob_mq_ops, 1,
-					BLK_MQ_F_NO_SCHED);
+	memset(&host->tag_set, 0, sizeof(host->tag_set));
+	host->tag_set.ops = &carm_mq_ops;
+	host->tag_set.cmd_size = sizeof(struct carm_request);
+	host->tag_set.nr_hw_queues = 1;
+	host->tag_set.nr_maps = 1;
+	host->tag_set.queue_depth = max_queue;
+	host->tag_set.numa_node = NUMA_NO_NODE;
+	host->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
+
+	rc = blk_mq_alloc_tag_set(&host->tag_set);
+	if (rc)
+		goto err_out_dma_free;
+
+	q = blk_mq_init_queue(&host->tag_set);
 	if (IS_ERR(q)) {
-		printk(KERN_ERR DRV_NAME "(%s): OOB queue alloc failure\n",
-		       pci_name(pdev));
 		rc = PTR_ERR(q);
+		blk_mq_free_tag_set(&host->tag_set);
 		goto err_out_dma_free;
 	}
+
 	host->oob_q = q;
 	q->queuedata = host;
 
@@ -1667,9 +1498,11 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (host->flags & FL_DYN_MAJOR)
 		host->major = rc;
 
-	rc = carm_init_disks(host);
-	if (rc)
-		goto err_out_blkdev_disks;
+	for (i = 0; i < CARM_MAX_PORTS; i++) {
+		rc = carm_init_disk(host, i);
+		if (rc)
+			goto err_out_blkdev_disks;
+	}
 
 	pci_set_master(pdev);
 
@@ -1699,7 +1532,8 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 err_out_free_irq:
 	free_irq(pdev->irq, host);
 err_out_blkdev_disks:
-	carm_free_disks(host);
+	for (i = 0; i < CARM_MAX_PORTS; i++)
+		carm_free_disk(host, i);
 	unregister_blkdev(host->major, host->name);
 err_out_free_majors:
 	if (host->major == 160)
@@ -1724,6 +1558,7 @@ err_out:
 static void carm_remove_one (struct pci_dev *pdev)
 {
 	struct carm_host *host = pci_get_drvdata(pdev);
+	unsigned int i;
 
 	if (!host) {
 		printk(KERN_ERR PFX "BUG: no host data for PCI(%s)\n",
@@ -1732,7 +1567,8 @@ static void carm_remove_one (struct pci_dev *pdev)
 	}
 
 	free_irq(pdev->irq, host);
-	carm_free_disks(host);
+	for (i = 0; i < CARM_MAX_PORTS; i++)
+		carm_free_disk(host, i);
 	unregister_blkdev(host->major, host->name);
 	if (host->major == 160)
 		clear_bit(0, &carm_major_alloc);
diff --git a/drivers/block/umem.c b/drivers/block/umem.c
index be3e3ab79950..aa035cf8a51d 100644
--- a/drivers/block/umem.c
+++ b/drivers/block/umem.c
@@ -888,8 +888,7 @@ static int mm_pci_probe(struct pci_dev *dev, const struct pci_device_id *id)
 	card->biotail = &card->bio;
 	spin_lock_init(&card->lock);
 
-	card->queue = blk_alloc_queue_node(GFP_KERNEL, NUMA_NO_NODE,
-					   &card->lock);
+	card->queue = blk_alloc_queue_node(GFP_KERNEL, NUMA_NO_NODE);
 	if (!card->queue)
 		goto failed_alloc;
 
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 086c6bb12baa..912c4265e592 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -214,6 +214,20 @@ static void virtblk_done(struct virtqueue *vq)
 	spin_unlock_irqrestore(&vblk->vqs[qid].lock, flags);
 }
 
+static void virtio_commit_rqs(struct blk_mq_hw_ctx *hctx)
+{
+	struct virtio_blk *vblk = hctx->queue->queuedata;
+	struct virtio_blk_vq *vq = &vblk->vqs[hctx->queue_num];
+	bool kick;
+
+	spin_lock_irq(&vq->lock);
+	kick = virtqueue_kick_prepare(vq->vq);
+	spin_unlock_irq(&vq->lock);
+
+	if (kick)
+		virtqueue_notify(vq->vq);
+}
+
 static blk_status_t virtio_queue_rq(struct blk_mq_hw_ctx *hctx,
 			   const struct blk_mq_queue_data *bd)
 {
@@ -624,7 +638,7 @@ static int virtblk_map_queues(struct blk_mq_tag_set *set)
 {
 	struct virtio_blk *vblk = set->driver_data;
 
-	return blk_mq_virtio_map_queues(set, vblk->vdev, 0);
+	return blk_mq_virtio_map_queues(&set->map[0], vblk->vdev, 0);
 }
 
 #ifdef CONFIG_VIRTIO_BLK_SCSI
@@ -638,6 +652,7 @@ static void virtblk_initialize_rq(struct request *req)
 
 static const struct blk_mq_ops virtio_mq_ops = {
 	.queue_rq	= virtio_queue_rq,
+	.commit_rqs	= virtio_commit_rqs,
 	.complete	= virtblk_request_done,
 	.init_request	= virtblk_init_request,
 #ifdef CONFIG_VIRTIO_BLK_SCSI