Merge tag 'block-5.10-2020-10-30' of git://git.kernel.dk/linux-block

Pull block fixes from Jens Axboe:

 - null_blk zone fixes (Damien, Kanchan)

 - NVMe pull request from Christoph:
       - improve zone revalidation (Keith Busch)
       - gracefully handle zero length messages in nvme-rdma (zhenwei pi)
       - nvme-fc error handling fixes (James Smart)
       - nvmet tracing NULL pointer dereference fix (Chaitanya Kulkarni)"

 - xsysace platform fixes (Andy)

 - scatterlist type cleanup (David)

 - blk-cgroup memory fixes (Gabriel)

 - nbd block size update fix (Ming)

 - Flush completion state fix (Ming)

 - bio_add_hw_page() iteration fix (Naohiro)

* tag 'block-5.10-2020-10-30' of git://git.kernel.dk/linux-block:
  blk-mq: mark flush request as IDLE in flush_end_io()
  lib/scatterlist: use consistent sg_copy_buffer() return type
  xsysace: use platform_get_resource() and platform_get_irq_optional()
  null_blk: Fix locking in zoned mode
  null_blk: Fix zone reset all tracing
  nbd: don't update block size after device is started
  block: advance iov_iter on bio_add_hw_page failure
  null_blk: synchronization fix for zoned device
  nvmet: fix a NULL pointer dereference when tracing the flush command
  nvme-fc: remove nvme_fc_terminate_io()
  nvme-fc: eliminate terminate_io use by nvme_fc_error_recovery
  nvme-fc: remove err_work work item
  nvme-fc: track error_recovery while connecting
  nvme-rdma: handle unexpected nvme completion data length
  nvme: ignore zone validate errors on subsequent scans
  blk-cgroup: Pre-allocate tree node on blkg_conf_prep
  blk-cgroup: Fix memleak on error path

block/bio.c

@@ -1044,6 +1044,7 @@ static int __bio_iov_append_get_pages(struct bio *bio, struct iov_iter *iter)
 	ssize_t size, left;
 	unsigned len, i;
 	size_t offset;
+	int ret = 0;
 
 	if (WARN_ON_ONCE(!max_append_sectors))
 		return 0;
@@ -1066,15 +1067,17 @@ static int __bio_iov_append_get_pages(struct bio *bio, struct iov_iter *iter)
 
 		len = min_t(size_t, PAGE_SIZE - offset, left);
 		if (bio_add_hw_page(q, bio, page, len, offset,
-				max_append_sectors, &same_page) != len)
-			return -EINVAL;
+				max_append_sectors, &same_page) != len) {
+			ret = -EINVAL;
+			break;
+		}
 		if (same_page)
 			put_page(page);
 		offset = 0;
 	}
 
-	iov_iter_advance(iter, size);
-	return 0;
+	iov_iter_advance(iter, size - left);
+	return ret;
 }
 
 /**

block/blk-cgroup.c

@@ -657,13 +657,20 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
 			goto fail;
 		}
 
+		if (radix_tree_preload(GFP_KERNEL)) {
+			blkg_free(new_blkg);
+			ret = -ENOMEM;
+			goto fail;
+		}
+
 		rcu_read_lock();
 		spin_lock_irq(&q->queue_lock);
 
 		blkg = blkg_lookup_check(pos, pol, q);
 		if (IS_ERR(blkg)) {
 			ret = PTR_ERR(blkg);
-			goto fail_unlock;
+			blkg_free(new_blkg);
+			goto fail_preloaded;
 		}
 
 		if (blkg) {
@@ -672,10 +679,12 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
 			blkg = blkg_create(pos, q, new_blkg);
 			if (IS_ERR(blkg)) {
 				ret = PTR_ERR(blkg);
-				goto fail_unlock;
+				goto fail_preloaded;
 			}
 		}
 
+		radix_tree_preload_end();
+
 		if (pos == blkcg)
 			goto success;
 	}
@@ -685,6 +694,8 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
 	ctx->body = input;
 	return 0;
 
+fail_preloaded:
+	radix_tree_preload_end();
 fail_unlock:
 	spin_unlock_irq(&q->queue_lock);
 	rcu_read_unlock();

block/blk-flush.c

@@ -225,6 +225,7 @@ static void flush_end_io(struct request *flush_rq, blk_status_t error)
 	/* release the tag's ownership to the req cloned from */
 	spin_lock_irqsave(&fq->mq_flush_lock, flags);
 
+	WRITE_ONCE(flush_rq->state, MQ_RQ_IDLE);
 	if (!refcount_dec_and_test(&flush_rq->ref)) {
 		fq->rq_status = error;
 		spin_unlock_irqrestore(&fq->mq_flush_lock, flags);

drivers/block/nbd.c

@@ -296,7 +296,7 @@ static void nbd_size_clear(struct nbd_device *nbd)
 	}
 }
 
-static void nbd_size_update(struct nbd_device *nbd)
+static void nbd_size_update(struct nbd_device *nbd, bool start)
 {
 	struct nbd_config *config = nbd->config;
 	struct block_device *bdev = bdget_disk(nbd->disk, 0);
@@ -313,7 +313,8 @@ static void nbd_size_update(struct nbd_device *nbd)
 	if (bdev) {
 		if (bdev->bd_disk) {
 			bd_set_nr_sectors(bdev, nr_sectors);
-			set_blocksize(bdev, config->blksize);
+			if (start)
+				set_blocksize(bdev, config->blksize);
 		} else
 			set_bit(GD_NEED_PART_SCAN, &nbd->disk->state);
 		bdput(bdev);
@@ -328,7 +329,7 @@ static void nbd_size_set(struct nbd_device *nbd, loff_t blocksize,
 	config->blksize = blocksize;
 	config->bytesize = blocksize * nr_blocks;
 	if (nbd->task_recv != NULL)
-		nbd_size_update(nbd);
+		nbd_size_update(nbd, false);
 }
 
 static void nbd_complete_rq(struct request *req)
@@ -1308,7 +1309,7 @@ static int nbd_start_device(struct nbd_device *nbd)
 		args->index = i;
 		queue_work(nbd->recv_workq, &args->work);
 	}
-	nbd_size_update(nbd);
+	nbd_size_update(nbd, true);
 	return error;
 }
 

drivers/block/null_blk.h

@@ -47,6 +47,8 @@ struct nullb_device {
 	unsigned int nr_zones_closed;
 	struct blk_zone *zones;
 	sector_t zone_size_sects;
+	spinlock_t zone_dev_lock;
+	unsigned long *zone_locks;
 
 	unsigned long size; /* device size in MB */
 	unsigned long completion_nsec; /* time in ns to complete a request */

drivers/block/null_blk_zoned.c

 // SPDX-License-Identifier: GPL-2.0
 #include <linux/vmalloc.h>
+#include <linux/bitmap.h>
 #include "null_blk.h"
 
 #define CREATE_TRACE_POINTS
@@ -45,6 +46,13 @@ int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q)
 	if (!dev->zones)
 		return -ENOMEM;
 
+	spin_lock_init(&dev->zone_dev_lock);
+	dev->zone_locks = bitmap_zalloc(dev->nr_zones, GFP_KERNEL);
+	if (!dev->zone_locks) {
+		kvfree(dev->zones);
+		return -ENOMEM;
+	}
+
 	if (dev->zone_nr_conv >= dev->nr_zones) {
 		dev->zone_nr_conv = dev->nr_zones - 1;
 		pr_info("changed the number of conventional zones to %u",
@@ -123,15 +131,26 @@ int null_register_zoned_dev(struct nullb *nullb)
 
 void null_free_zoned_dev(struct nullb_device *dev)
 {
+	bitmap_free(dev->zone_locks);
 	kvfree(dev->zones);
 }
 
+static inline void null_lock_zone(struct nullb_device *dev, unsigned int zno)
+{
+	wait_on_bit_lock_io(dev->zone_locks, zno, TASK_UNINTERRUPTIBLE);
+}
+
+static inline void null_unlock_zone(struct nullb_device *dev, unsigned int zno)
+{
+	clear_and_wake_up_bit(zno, dev->zone_locks);
+}
+
 int null_report_zones(struct gendisk *disk, sector_t sector,
 		unsigned int nr_zones, report_zones_cb cb, void *data)
 {
 	struct nullb *nullb = disk->private_data;
 	struct nullb_device *dev = nullb->dev;
-	unsigned int first_zone, i;
+	unsigned int first_zone, i, zno;
 	struct blk_zone zone;
 	int error;
 
@@ -142,15 +161,18 @@ int null_report_zones(struct gendisk *disk, sector_t sector,
 	nr_zones = min(nr_zones, dev->nr_zones - first_zone);
 	trace_nullb_report_zones(nullb, nr_zones);
 
-	for (i = 0; i < nr_zones; i++) {
+	zno = first_zone;
+	for (i = 0; i < nr_zones; i++, zno++) {
 		/*
 		 * Stacked DM target drivers will remap the zone information by
 		 * modifying the zone information passed to the report callback.
 		 * So use a local copy to avoid corruption of the device zone
 		 * array.
 		 */
-		memcpy(&zone, &dev->zones[first_zone + i],
-		       sizeof(struct blk_zone));
+		null_lock_zone(dev, zno);
+		memcpy(&zone, &dev->zones[zno], sizeof(struct blk_zone));
+		null_unlock_zone(dev, zno);
+
 		error = cb(&zone, i, data);
 		if (error)
 			return error;
@@ -159,6 +181,10 @@ int null_report_zones(struct gendisk *disk, sector_t sector,
 	return nr_zones;
 }
 
+/*
+ * This is called in the case of memory backing from null_process_cmd()
+ * with the target zone already locked.
+ */
 size_t null_zone_valid_read_len(struct nullb *nullb,
 				sector_t sector, unsigned int len)
 {
@@ -295,22 +321,27 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,
 	if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL)
 		return null_process_cmd(cmd, REQ_OP_WRITE, sector, nr_sectors);
 
+	null_lock_zone(dev, zno);
+	spin_lock(&dev->zone_dev_lock);
+
 	switch (zone->cond) {
 	case BLK_ZONE_COND_FULL:
 		/* Cannot write to a full zone */
-		return BLK_STS_IOERR;
+		ret = BLK_STS_IOERR;
+		goto unlock;
 	case BLK_ZONE_COND_EMPTY:
 	case BLK_ZONE_COND_CLOSED:
 		ret = null_check_zone_resources(dev, zone);
 		if (ret != BLK_STS_OK)
-			return ret;
+			goto unlock;
 		break;
 	case BLK_ZONE_COND_IMP_OPEN:
 	case BLK_ZONE_COND_EXP_OPEN:
 		break;
 	default:
 		/* Invalid zone condition */
-		return BLK_STS_IOERR;
+		ret = BLK_STS_IOERR;
+		goto unlock;
 	}
 
 	/*
@@ -326,11 +357,14 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,
 		else
 			cmd->rq->__sector = sector;
 	} else if (sector != zone->wp) {
-		return BLK_STS_IOERR;
+		ret = BLK_STS_IOERR;
+		goto unlock;
 	}
 
-	if (zone->wp + nr_sectors > zone->start + zone->capacity)
-		return BLK_STS_IOERR;
+	if (zone->wp + nr_sectors > zone->start + zone->capacity) {
+		ret = BLK_STS_IOERR;
+		goto unlock;
+	}
 
 	if (zone->cond == BLK_ZONE_COND_CLOSED) {
 		dev->nr_zones_closed--;
@@ -341,9 +375,11 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,
 	if (zone->cond != BLK_ZONE_COND_EXP_OPEN)
 		zone->cond = BLK_ZONE_COND_IMP_OPEN;
 
+	spin_unlock(&dev->zone_dev_lock);
 	ret = null_process_cmd(cmd, REQ_OP_WRITE, sector, nr_sectors);
+	spin_lock(&dev->zone_dev_lock);
 	if (ret != BLK_STS_OK)
-		return ret;
+		goto unlock;
 
 	zone->wp += nr_sectors;
 	if (zone->wp == zone->start + zone->capacity) {
@@ -353,7 +389,13 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,
 			dev->nr_zones_imp_open--;
 		zone->cond = BLK_ZONE_COND_FULL;
 	}
-	return BLK_STS_OK;
+	ret = BLK_STS_OK;
+
+unlock:
+	spin_unlock(&dev->zone_dev_lock);
+	null_unlock_zone(dev, zno);
+
+	return ret;
 }
 
 static blk_status_t null_open_zone(struct nullb_device *dev, struct blk_zone *zone)
@@ -464,16 +506,33 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 				   sector_t sector)
 {
 	struct nullb_device *dev = cmd->nq->dev;
-	unsigned int zone_no = null_zone_no(dev, sector);
-	struct blk_zone *zone = &dev->zones[zone_no];
-	blk_status_t ret = BLK_STS_OK;
+	unsigned int zone_no;
+	struct blk_zone *zone;
+	blk_status_t ret;
 	size_t i;
 
+	if (op == REQ_OP_ZONE_RESET_ALL) {
+		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
+			null_lock_zone(dev, i);
+			zone = &dev->zones[i];
+			if (zone->cond != BLK_ZONE_COND_EMPTY) {
+				spin_lock(&dev->zone_dev_lock);
+				null_reset_zone(dev, zone);
+				spin_unlock(&dev->zone_dev_lock);
+				trace_nullb_zone_op(cmd, i, zone->cond);
+			}
+			null_unlock_zone(dev, i);
+		}
+		return BLK_STS_OK;
+	}
+
+	zone_no = null_zone_no(dev, sector);
+	zone = &dev->zones[zone_no];
+
+	null_lock_zone(dev, zone_no);
+	spin_lock(&dev->zone_dev_lock);
+
 	switch (op) {
-	case REQ_OP_ZONE_RESET_ALL:
-		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++)
-			null_reset_zone(dev, &dev->zones[i]);
-		break;
 	case REQ_OP_ZONE_RESET:
 		ret = null_reset_zone(dev, zone);
 		break;
@@ -487,30 +546,46 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 		ret = null_finish_zone(dev, zone);
 		break;
 	default:
-		return BLK_STS_NOTSUPP;
+		ret = BLK_STS_NOTSUPP;
+		break;
 	}
 
+	spin_unlock(&dev->zone_dev_lock);
+
 	if (ret == BLK_STS_OK)
 		trace_nullb_zone_op(cmd, zone_no, zone->cond);
 
+	null_unlock_zone(dev, zone_no);
+
 	return ret;
 }
 
 blk_status_t null_process_zoned_cmd(struct nullb_cmd *cmd, enum req_opf op,
 				    sector_t sector, sector_t nr_sectors)
 {
+	struct nullb_device *dev = cmd->nq->dev;
+	unsigned int zno = null_zone_no(dev, sector);
+	blk_status_t sts;
+
 	switch (op) {
 	case REQ_OP_WRITE:
-		return null_zone_write(cmd, sector, nr_sectors, false);
+		sts = null_zone_write(cmd, sector, nr_sectors, false);
+		break;
 	case REQ_OP_ZONE_APPEND:
-		return null_zone_write(cmd, sector, nr_sectors, true);
+		sts = null_zone_write(cmd, sector, nr_sectors, true);
+		break;
 	case REQ_OP_ZONE_RESET:
 	case REQ_OP_ZONE_RESET_ALL:
 	case REQ_OP_ZONE_OPEN:
 	case REQ_OP_ZONE_CLOSE:
 	case REQ_OP_ZONE_FINISH:
-		return null_zone_mgmt(cmd, op, sector);
+		sts = null_zone_mgmt(cmd, op, sector);
+		break;
 	default:
-		return null_process_cmd(cmd, op, sector, nr_sectors);
+		null_lock_zone(dev, zno);
+		sts = null_process_cmd(cmd, op, sector, nr_sectors);
+		null_unlock_zone(dev, zno);
 	}
+
+	return sts;
 }

drivers/block/xsysace.c

@@ -443,22 +443,27 @@ static void ace_fix_driveid(u16 *id)
 #define ACE_FSM_NUM_STATES              11
 
 /* Set flag to exit FSM loop and reschedule tasklet */
-static inline void ace_fsm_yield(struct ace_device *ace)
+static inline void ace_fsm_yieldpoll(struct ace_device *ace)
 {
-	dev_dbg(ace->dev, "ace_fsm_yield()\n");
 	tasklet_schedule(&ace->fsm_tasklet);
 	ace->fsm_continue_flag = 0;
 }
 
+static inline void ace_fsm_yield(struct ace_device *ace)
+{
+	dev_dbg(ace->dev, "%s()\n", __func__);
+	ace_fsm_yieldpoll(ace);
+}
+
 /* Set flag to exit FSM loop and wait for IRQ to reschedule tasklet */
 static inline void ace_fsm_yieldirq(struct ace_device *ace)
 {
 	dev_dbg(ace->dev, "ace_fsm_yieldirq()\n");
 
-	if (!ace->irq)
-		/* No IRQ assigned, so need to poll */
-		tasklet_schedule(&ace->fsm_tasklet);
-	ace->fsm_continue_flag = 0;
+	if (ace->irq > 0)
+		ace->fsm_continue_flag = 0;
+	else
+		ace_fsm_yieldpoll(ace);
 }
 
 static bool ace_has_next_request(struct request_queue *q)
@@ -1053,12 +1058,12 @@ static int ace_setup(struct ace_device *ace)
 		ACE_CTRL_DATABUFRDYIRQ | ACE_CTRL_ERRORIRQ);
 
 	/* Now we can hook up the irq handler */
-	if (ace->irq) {
+	if (ace->irq > 0) {
 		rc = request_irq(ace->irq, ace_interrupt, 0, "systemace", ace);
 		if (rc) {
 			/* Failure - fall back to polled mode */
 			dev_err(ace->dev, "request_irq failed\n");
-			ace->irq = 0;
+			ace->irq = rc;
 		}
 	}
 
@@ -1110,7 +1115,7 @@ static void ace_teardown(struct ace_device *ace)
 
 	tasklet_kill(&ace->fsm_tasklet);
 
-	if (ace->irq)
+	if (ace->irq > 0)
 		free_irq(ace->irq, ace);
 
 	iounmap(ace->baseaddr);
@@ -1123,11 +1128,6 @@ static int ace_alloc(struct device *dev, int id, resource_size_t physaddr,
 	int rc;
 	dev_dbg(dev, "ace_alloc(%p)\n", dev);
 
-	if (!physaddr) {
-		rc = -ENODEV;
-		goto err_noreg;
-	}
-
 	/* Allocate and initialize the ace device structure */
 	ace = kzalloc(sizeof(struct ace_device), GFP_KERNEL);
 	if (!ace) {
@@ -1153,7 +1153,6 @@ static int ace_alloc(struct device *dev, int id, resource_size_t physaddr,
 	dev_set_drvdata(dev, NULL);
 	kfree(ace);
 err_alloc:
-err_noreg:
 	dev_err(dev, "could not initialize device, err=%i\n", rc);
 	return rc;
 }
@@ -1176,10 +1175,11 @@ static void ace_free(struct device *dev)
 
 static int ace_probe(struct platform_device *dev)
 {
-	resource_size_t physaddr = 0;
 	int bus_width = ACE_BUS_WIDTH_16; /* FIXME: should not be hard coded */
+	resource_size_t physaddr;
+	struct resource *res;
 	u32 id = dev->id;
-	int irq = 0;
+	int irq;
 	int i;
 
 	dev_dbg(&dev->dev, "ace_probe(%p)\n", dev);
@@ -1190,12 +1190,15 @@ static int ace_probe(struct platform_device *dev)
 	if (of_find_property(dev->dev.of_node, "8-bit", NULL))
 		bus_width = ACE_BUS_WIDTH_8;
 
-	for (i = 0; i < dev->num_resources; i++) {
-		if (dev->resource[i].flags & IORESOURCE_MEM)
-			physaddr = dev->resource[i].start;
-		if (dev->resource[i].flags & IORESOURCE_IRQ)
-			irq = dev->resource[i].start;
-	}
+	res = platform_get_resource(dev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
+
+	physaddr = res->start;
+	if (!physaddr)
+		return -ENODEV;
+
+	irq = platform_get_irq_optional(dev, 0);
 
 	/* Call the bus-independent setup code */
 	return ace_alloc(&dev->dev, id, physaddr, irq, bus_width);

drivers/nvme/host/core.c

@@ -2125,7 +2125,7 @@ static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
 
 	if (blk_queue_is_zoned(ns->queue)) {
 		ret = nvme_revalidate_zones(ns);
-		if (ret)
+		if (ret && !nvme_first_scan(ns->disk))
 			return ret;
 	}
 

drivers/nvme/host/fc.c

@@ -146,7 +146,8 @@ struct nvme_fc_rport {
 
 /* fc_ctrl flags values - specified as bit positions */
 #define ASSOC_ACTIVE		0
-#define FCCTRL_TERMIO		1
+#define ASSOC_FAILED		1
+#define FCCTRL_TERMIO		2
 
 struct nvme_fc_ctrl {
 	spinlock_t		lock;
@@ -157,7 +158,6 @@ struct nvme_fc_ctrl {
 	u32			cnum;
 
 	bool			ioq_live;
-	atomic_t		err_work_active;
 	u64			association_id;
 	struct nvmefc_ls_rcv_op	*rcv_disconn;
 
@@ -167,7 +167,6 @@ struct nvme_fc_ctrl {
 	struct blk_mq_tag_set	tag_set;
 
 	struct delayed_work	connect_work;
-	struct work_struct	err_work;
 
 	struct kref		ref;
 	unsigned long		flags;
@@ -2414,24 +2413,97 @@ nvme_fc_nvme_ctrl_freed(struct nvme_ctrl *nctrl)
 	nvme_fc_ctrl_put(ctrl);
 }
 
+/*
+ * This routine is used by the transport when it needs to find active
+ * io on a queue that is to be terminated. The transport uses
+ * blk_mq_tagset_busy_itr() to find the busy requests, which then invoke
+ * this routine to kill them on a 1 by 1 basis.
+ *
+ * As FC allocates FC exchange for each io, the transport must contact
+ * the LLDD to terminate the exchange, thus releasing the FC exchange.
+ * After terminating the exchange the LLDD will call the transport's
+ * normal io done path for the request, but it will have an aborted
+ * status. The done path will return the io request back to the block
+ * layer with an error status.
+ */
+static bool
+nvme_fc_terminate_exchange(struct request *req, void *data, bool reserved)
+{
+	struct nvme_ctrl *nctrl = data;
+	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
+	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(req);
+
+	__nvme_fc_abort_op(ctrl, op);
+	return true;
+}
+
+/*
+ * This routine runs through all outstanding commands on the association
+ * and aborts them.  This routine is typically be called by the
+ * delete_association routine. It is also called due to an error during
+ * reconnect. In that scenario, it is most likely a command that initializes
+ * the controller, including fabric Connect commands on io queues, that
+ * may have timed out or failed thus the io must be killed for the connect
+ * thread to see the error.
+ */
 static void
-nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg)
+__nvme_fc_abort_outstanding_ios(struct nvme_fc_ctrl *ctrl, bool start_queues)
 {
-	int active;
+	/*
+	 * If io queues are present, stop them and terminate all outstanding
+	 * ios on them. As FC allocates FC exchange for each io, the
+	 * transport must contact the LLDD to terminate the exchange,
+	 * thus releasing the FC exchange. We use blk_mq_tagset_busy_itr()
+	 * to tell us what io's are busy and invoke a transport routine
+	 * to kill them with the LLDD.  After terminating the exchange
+	 * the LLDD will call the transport's normal io done path, but it
+	 * will have an aborted status. The done path will return the
+	 * io requests back to the block layer as part of normal completions
+	 * (but with error status).
+	 */
+	if (ctrl->ctrl.queue_count > 1) {
+		nvme_stop_queues(&ctrl->ctrl);
+		blk_mq_tagset_busy_iter(&ctrl->tag_set,
+				nvme_fc_terminate_exchange, &ctrl->ctrl);
+		blk_mq_tagset_wait_completed_request(&ctrl->tag_set);
+		if (start_queues)
+			nvme_start_queues(&ctrl->ctrl);
+	}
+
+	/*
+	 * Other transports, which don't have link-level contexts bound
+	 * to sqe's, would try to gracefully shutdown the controller by
+	 * writing the registers for shutdown and polling (call
+	 * nvme_shutdown_ctrl()). Given a bunch of i/o was potentially
+	 * just aborted and we will wait on those contexts, and given
+	 * there was no indication of how live the controlelr is on the
+	 * link, don't send more io to create more contexts for the
+	 * shutdown. Let the controller fail via keepalive failure if
+	 * its still present.
+	 */
 
 	/*
-	 * if an error (io timeout, etc) while (re)connecting,
-	 * it's an error on creating the new association.
-	 * Start the error recovery thread if it hasn't already
-	 * been started. It is expected there could be multiple
-	 * ios hitting this path before things are cleaned up.
+	 * clean up the admin queue. Same thing as above.
+	 */
+	blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
+	blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
+				nvme_fc_terminate_exchange, &ctrl->ctrl);
+	blk_mq_tagset_wait_completed_request(&ctrl->admin_tag_set);
+}
+
+static void
+nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg)
+{
+	/*
+	 * if an error (io timeout, etc) while (re)connecting, the remote
+	 * port requested terminating of the association (disconnect_ls)
+	 * or an error (timeout or abort) occurred on an io while creating
+	 * the controller.  Abort any ios on the association and let the
+	 * create_association error path resolve things.
 	 */
 	if (ctrl->ctrl.state == NVME_CTRL_CONNECTING) {
-		active = atomic_xchg(&ctrl->err_work_active, 1);
-		if (!active && !queue_work(nvme_fc_wq, &ctrl->err_work)) {
-			atomic_set(&ctrl->err_work_active, 0);
-			WARN_ON(1);
-		}
+		__nvme_fc_abort_outstanding_ios(ctrl, true);
+		set_bit(ASSOC_FAILED, &ctrl->flags);
 		return;
 	}
 
@@ -2745,30 +2817,6 @@ nvme_fc_complete_rq(struct request *rq)
 	nvme_fc_ctrl_put(ctrl);
 }
 
-/*
- * This routine is used by the transport when it needs to find active
- * io on a queue that is to be terminated. The transport uses
- * blk_mq_tagset_busy_itr() to find the busy requests, which then invoke
- * this routine to kill them on a 1 by 1 basis.
- *
- * As FC allocates FC exchange for each io, the transport must contact
- * the LLDD to terminate the exchange, thus releasing the FC exchange.
- * After terminating the exchange the LLDD will call the transport's
- * normal io done path for the request, but it will have an aborted
- * status. The done path will return the io request back to the block
- * layer with an error status.
- */
-static bool
-nvme_fc_terminate_exchange(struct request *req, void *data, bool reserved)
-{
-	struct nvme_ctrl *nctrl = data;
-	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
-	struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(req);
-
-	__nvme_fc_abort_op(ctrl, op);
-	return true;
-}
-
 
 static const struct blk_mq_ops nvme_fc_mq_ops = {
 	.queue_rq	= nvme_fc_queue_rq,
@@ -2988,6 +3036,8 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 		ctrl->cnum, ctrl->lport->localport.port_name,
 		ctrl->rport->remoteport.port_name, ctrl->ctrl.opts->subsysnqn);
 
+	clear_bit(ASSOC_FAILED, &ctrl->flags);
+
 	/*
 	 * Create the admin queue
 	 */
@@ -3016,7 +3066,7 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 	 */
 
 	ret = nvme_enable_ctrl(&ctrl->ctrl);
-	if (ret)
+	if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
 		goto out_disconnect_admin_queue;
 
 	ctrl->ctrl.max_segments = ctrl->lport->ops->max_sgl_segments;
@@ -3026,7 +3076,7 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 	blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
 
 	ret = nvme_init_identify(&ctrl->ctrl);
-	if (ret)
+	if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
 		goto out_disconnect_admin_queue;
 
 	/* sanity checks */
@@ -3071,9 +3121,9 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 			ret = nvme_fc_create_io_queues(ctrl);
 		else
 			ret = nvme_fc_recreate_io_queues(ctrl);
-		if (ret)
-			goto out_term_aen_ops;
 	}
+	if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
+		goto out_term_aen_ops;
 
 	changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
 
@@ -3107,60 +3157,6 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 }
 
 
-/*
- * This routine runs through all outstanding commands on the association
- * and aborts them.  This routine is typically be called by the
- * delete_association routine. It is also called due to an error during
- * reconnect. In that scenario, it is most likely a command that initializes
- * the controller, including fabric Connect commands on io queues, that
- * may have timed out or failed thus the io must be killed for the connect
- * thread to see the error.
- */
-static void
-__nvme_fc_abort_outstanding_ios(struct nvme_fc_ctrl *ctrl, bool start_queues)
-{
-	/*
-	 * If io queues are present, stop them and terminate all outstanding
-	 * ios on them. As FC allocates FC exchange for each io, the
-	 * transport must contact the LLDD to terminate the exchange,
-	 * thus releasing the FC exchange. We use blk_mq_tagset_busy_itr()
-	 * to tell us what io's are busy and invoke a transport routine
-	 * to kill them with the LLDD.  After terminating the exchange
-	 * the LLDD will call the transport's normal io done path, but it
-	 * will have an aborted status. The done path will return the
-	 * io requests back to the block layer as part of normal completions
-	 * (but with error status).
-	 */
-	if (ctrl->ctrl.queue_count > 1) {
-		nvme_stop_queues(&ctrl->ctrl);
-		blk_mq_tagset_busy_iter(&ctrl->tag_set,
-				nvme_fc_terminate_exchange, &ctrl->ctrl);
-		blk_mq_tagset_wait_completed_request(&ctrl->tag_set);
-		if (start_queues)
-			nvme_start_queues(&ctrl->ctrl);
-	}
-
-	/*
-	 * Other transports, which don't have link-level contexts bound
-	 * to sqe's, would try to gracefully shutdown the controller by
-	 * writing the registers for shutdown and polling (call
-	 * nvme_shutdown_ctrl()). Given a bunch of i/o was potentially
-	 * just aborted and we will wait on those contexts, and given
-	 * there was no indication of how live the controlelr is on the
-	 * link, don't send more io to create more contexts for the
-	 * shutdown. Let the controller fail via keepalive failure if
-	 * its still present.
-	 */
-
-	/*
-	 * clean up the admin queue. Same thing as above.
-	 */
-	blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
-	blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
-				nvme_fc_terminate_exchange, &ctrl->ctrl);
-	blk_mq_tagset_wait_completed_request(&ctrl->admin_tag_set);
-}
-
 /*
  * This routine stops operation of the controller on the host side.
  * On the host os stack side: Admin and IO queues are stopped,
@@ -3237,7 +3233,6 @@ nvme_fc_delete_ctrl(struct nvme_ctrl *nctrl)
 {
 	struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
 
-	cancel_work_sync(&ctrl->err_work);
 	cancel_delayed_work_sync(&ctrl->connect_work);
 	/*
 	 * kill the association on the link side.  this will block
@@ -3292,78 +3287,34 @@ nvme_fc_reconnect_or_delete(struct nvme_fc_ctrl *ctrl, int status)
 }
 
 static void
-__nvme_fc_terminate_io(struct nvme_fc_ctrl *ctrl)
+nvme_fc_reset_ctrl_work(struct work_struct *work)
 {
-	/*
-	 * if state is CONNECTING - the error occurred as part of a
-	 * reconnect attempt. Abort any ios on the association and
-	 * let the create_association error paths resolve things.
-	 */
-	if (ctrl->ctrl.state == NVME_CTRL_CONNECTING) {
-		__nvme_fc_abort_outstanding_ios(ctrl, true);
-		return;
-	}
-
-	/*
-	 * For any other state, kill the association. As this routine
-	 * is a common io abort routine for resetting and such, after
-	 * the association is terminated, ensure that the state is set
-	 * to CONNECTING.
-	 */
+	struct nvme_fc_ctrl *ctrl =
+		container_of(work, struct nvme_fc_ctrl, ctrl.reset_work);
 
-	nvme_stop_keep_alive(&ctrl->ctrl);
+	nvme_stop_ctrl(&ctrl->ctrl);
 
 	/* will block will waiting for io to terminate */
 	nvme_fc_delete_association(ctrl);
 
-	if (ctrl->ctrl.state != NVME_CTRL_CONNECTING &&
-	    !nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING))
+	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING))
 		dev_err(ctrl->ctrl.device,
 			"NVME-FC{%d}: error_recovery: Couldn't change state "
 			"to CONNECTING\n", ctrl->cnum);
-}
-
-static void
-nvme_fc_reset_ctrl_work(struct work_struct *work)
-{
-	struct nvme_fc_ctrl *ctrl =
-		container_of(work, struct nvme_fc_ctrl, ctrl.reset_work);
-	int ret;
 
-	__nvme_fc_terminate_io(ctrl);
-
-	nvme_stop_ctrl(&ctrl->ctrl);
-
-	if (ctrl->rport->remoteport.port_state == FC_OBJSTATE_ONLINE)
-		ret = nvme_fc_create_association(ctrl);
-	else
-		ret = -ENOTCONN;
-
-	if (ret)
-		nvme_fc_reconnect_or_delete(ctrl, ret);
-	else
-		dev_info(ctrl->ctrl.device,
-			"NVME-FC{%d}: controller reset complete\n",
-			ctrl->cnum);
+	if (ctrl->rport->remoteport.port_state == FC_OBJSTATE_ONLINE) {
+		if (!queue_delayed_work(nvme_wq, &ctrl->connect_work, 0)) {
+			dev_err(ctrl->ctrl.device,
+				"NVME-FC{%d}: failed to schedule connect "
+				"after reset\n", ctrl->cnum);
+		} else {
+			flush_delayed_work(&ctrl->connect_work);
+		}
+	} else {
+		nvme_fc_reconnect_or_delete(ctrl, -ENOTCONN);
+	}
 }
 
-static void
-nvme_fc_connect_err_work(struct work_struct *work)
-{
-	struct nvme_fc_ctrl *ctrl =
-			container_of(work, struct nvme_fc_ctrl, err_work);
-
-	__nvme_fc_terminate_io(ctrl);
-
-	atomic_set(&ctrl->err_work_active, 0);
-
-	/*
-	 * Rescheduling the connection after recovering
-	 * from the io error is left to the reconnect work
-	 * item, which is what should have stalled waiting on
-	 * the io that had the error that scheduled this work.
-	 */
-}
 
 static const struct nvme_ctrl_ops nvme_fc_ctrl_ops = {
 	.name			= "fc",
@@ -3491,7 +3442,6 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts,
 	ctrl->dev = lport->dev;
 	ctrl->cnum = idx;
 	ctrl->ioq_live = false;
-	atomic_set(&ctrl->err_work_active, 0);
 	init_waitqueue_head(&ctrl->ioabort_wait);
 
 	get_device(ctrl->dev);
@@ -3499,7 +3449,6 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts,
 
 	INIT_WORK(&ctrl->ctrl.reset_work, nvme_fc_reset_ctrl_work);
 	INIT_DELAYED_WORK(&ctrl->connect_work, nvme_fc_connect_ctrl_work);
-	INIT_WORK(&ctrl->err_work, nvme_fc_connect_err_work);
 	spin_lock_init(&ctrl->lock);
 
 	/* io queue count */
@@ -3592,7 +3541,6 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts,
 fail_ctrl:
 	nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_DELETING);
 	cancel_work_sync(&ctrl->ctrl.reset_work);
-	cancel_work_sync(&ctrl->err_work);
 	cancel_delayed_work_sync(&ctrl->connect_work);
 
 	ctrl->ctrl.opts = NULL;

drivers/nvme/host/rdma.c

@@ -1768,6 +1768,14 @@ static void nvme_rdma_recv_done(struct ib_cq *cq, struct ib_wc *wc)
 		return;
 	}
 
+	/* sanity checking for received data length */
+	if (unlikely(wc->byte_len < len)) {
+		dev_err(queue->ctrl->ctrl.device,
+			"Unexpected nvme completion length(%d)\n", wc->byte_len);
+		nvme_rdma_error_recovery(queue->ctrl);
+		return;
+	}
+
 	ib_dma_sync_single_for_cpu(ibdev, qe->dma, len, DMA_FROM_DEVICE);
 	/*
 	 * AEN requests are special as they don't time out and can

drivers/nvme/target/core.c

@@ -907,8 +907,6 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
 	req->error_loc = NVMET_NO_ERROR_LOC;
 	req->error_slba = 0;
 
-	trace_nvmet_req_init(req, req->cmd);
-
 	/* no support for fused commands yet */
 	if (unlikely(flags & (NVME_CMD_FUSE_FIRST | NVME_CMD_FUSE_SECOND))) {
 		req->error_loc = offsetof(struct nvme_common_command, flags);
@@ -938,6 +936,8 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
 	if (status)
 		goto fail;
 
+	trace_nvmet_req_init(req, req->cmd);
+
 	if (unlikely(!percpu_ref_tryget_live(&sq->ref))) {
 		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
 		goto fail;

drivers/nvme/target/trace.h

@@ -46,19 +46,12 @@ static inline struct nvmet_ctrl *nvmet_req_to_ctrl(struct nvmet_req *req)
 	return req->sq->ctrl;
 }
 
-static inline void __assign_disk_name(char *name, struct nvmet_req *req,
-		bool init)
+static inline void __assign_req_name(char *name, struct nvmet_req *req)
 {
-	struct nvmet_ctrl *ctrl = nvmet_req_to_ctrl(req);
-	struct nvmet_ns *ns;
-
-	if ((init && req->sq->qid) || (!init && req->cq->qid)) {
-		ns = nvmet_find_namespace(ctrl, req->cmd->rw.nsid);
-		strncpy(name, ns->device_path, DISK_NAME_LEN);
-		return;
-	}
-
-	memset(name, 0, DISK_NAME_LEN);
+	if (req->ns)
+		strncpy(name, req->ns->device_path, DISK_NAME_LEN);
+	else
+		memset(name, 0, DISK_NAME_LEN);
 }
 #endif
 
@@ -81,7 +74,7 @@ TRACE_EVENT(nvmet_req_init,
 	TP_fast_assign(
 		__entry->cmd = cmd;
 		__entry->ctrl = nvmet_req_to_ctrl(req);
-		__assign_disk_name(__entry->disk, req, true);
+		__assign_req_name(__entry->disk, req);
 		__entry->qid = req->sq->qid;
 		__entry->cid = cmd->common.command_id;
 		__entry->opcode = cmd->common.opcode;
@@ -121,7 +114,7 @@ TRACE_EVENT(nvmet_req_complete,
 		__entry->cid = req->cqe->command_id;
 		__entry->result = le64_to_cpu(req->cqe->result.u64);
 		__entry->status = le16_to_cpu(req->cqe->status) >> 1;
-		__assign_disk_name(__entry->disk, req, false);
+		__assign_req_name(__entry->disk, req);
 	),
 	TP_printk("nvmet%s: %sqid=%d, cmdid=%u, res=%#llx, status=%#x",
 		__print_ctrl_name(__entry->ctrl),

lib/scatterlist.c

@@ -933,7 +933,7 @@ size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents, void *buf,
 	sg_miter_start(&miter, sgl, nents, sg_flags);
 
 	if (!sg_miter_skip(&miter, skip))
-		return false;
+		return 0;
 
 	while ((offset < buflen) && sg_miter_next(&miter)) {
 		unsigned int len;