Merge tag 'for-linus-20180413' of git://git.kernel.dk/linux-block

Pull block fixes from Jens Axboe:
 "Followup fixes for this merge window. This contains:

   - Series from Ming, fixing corner cases in our CPU <-> queue mapping.

     This triggered repeated warnings on especially s390, but I also hit
     it in cpu hot plug/unplug testing while doing IO on NVMe on x86-64.

   - Another fix from Ming, ensuring that we always order budget and
     driver tag identically, avoiding a deadlock on QD=1 devices.

   - Loop locking regression fix from this merge window, from Omar.

   - Another loop locking fix, this time missing an unlock, from Tetsuo
     Handa.

   - Fix for racing IO submission with device removal from Bart.

   - sr reference fix from me, fixing a case where disk change or
     getevents can race with device removal.

   - Set of nvme fixes by way of Keith, from various contributors"

* tag 'for-linus-20180413' of git://git.kernel.dk/linux-block: (28 commits)
  nvme: expand nvmf_check_if_ready checks
  nvme: Use admin command effects for admin commands
  nvmet: fix space padding in serial number
  nvme: check return value of init_srcu_struct function
  nvmet: Fix nvmet_execute_write_zeroes sector count
  nvme-pci: Separate IO and admin queue IRQ vectors
  nvme-pci: Remove unused queue parameter
  nvme-pci: Skip queue deletion if there are no queues
  nvme: target: fix buffer overflow
  nvme: don't send keep-alives to the discovery controller
  nvme: unexport nvme_start_keep_alive
  nvme-loop: fix kernel oops in case of unhandled command
  nvme: enforce 64bit offset for nvme_get_log_ext fn
  sr: get/drop reference to device in revalidate and check_events
  blk-mq: Revert "blk-mq: reimplement blk_mq_hw_queue_mapped"
  blk-mq: Avoid that submitting a bio concurrently with device removal triggers a crash
  backing: silence compiler warning using __printf
  blk-mq: remove code for dealing with remapping queue
  blk-mq: reimplement blk_mq_hw_queue_mapped
  blk-mq: don't check queue mapped in __blk_mq_delay_run_hw_queue()
  ...

7 files changed, 133 insertions, 81 deletions

diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 197a6ba9700f..9df4f71e58ca 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -376,6 +376,15 @@ static void nvme_put_ns(struct nvme_ns *ns)
 	kref_put(&ns->kref, nvme_free_ns);
 }
 
+static inline void nvme_clear_nvme_request(struct request *req)
+{
+	if (!(req->rq_flags & RQF_DONTPREP)) {
+		nvme_req(req)->retries = 0;
+		nvme_req(req)->flags = 0;
+		req->rq_flags |= RQF_DONTPREP;
+	}
+}
+
 struct request *nvme_alloc_request(struct request_queue *q,
 		struct nvme_command *cmd, blk_mq_req_flags_t flags, int qid)
 {
@@ -392,6 +401,7 @@ struct request *nvme_alloc_request(struct request_queue *q,
 		return req;
 
 	req->cmd_flags |= REQ_FAILFAST_DRIVER;
+	nvme_clear_nvme_request(req);
 	nvme_req(req)->cmd = cmd;
 
 	return req;
@@ -608,11 +618,7 @@ blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
 {
 	blk_status_t ret = BLK_STS_OK;
 
-	if (!(req->rq_flags & RQF_DONTPREP)) {
-		nvme_req(req)->retries = 0;
-		nvme_req(req)->flags = 0;
-		req->rq_flags |= RQF_DONTPREP;
-	}
+	nvme_clear_nvme_request(req);
 
 	switch (req_op(req)) {
 	case REQ_OP_DRV_IN:
@@ -742,6 +748,7 @@ static int nvme_submit_user_cmd(struct request_queue *q,
 		return PTR_ERR(req);
 
 	req->timeout = timeout ? timeout : ADMIN_TIMEOUT;
+	nvme_req(req)->flags |= NVME_REQ_USERCMD;
 
 	if (ubuffer && bufflen) {
 		ret = blk_rq_map_user(q, req, NULL, ubuffer, bufflen,
@@ -826,7 +833,7 @@ static void nvme_keep_alive_work(struct work_struct *work)
 	}
 }
 
-void nvme_start_keep_alive(struct nvme_ctrl *ctrl)
+static void nvme_start_keep_alive(struct nvme_ctrl *ctrl)
 {
 	if (unlikely(ctrl->kato == 0))
 		return;
@@ -836,7 +843,6 @@ void nvme_start_keep_alive(struct nvme_ctrl *ctrl)
 	ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive;
 	schedule_delayed_work(&ctrl->ka_work, ctrl->kato * HZ);
 }
-EXPORT_SYMBOL_GPL(nvme_start_keep_alive);
 
 void nvme_stop_keep_alive(struct nvme_ctrl *ctrl)
 {
@@ -1103,7 +1109,7 @@ static u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 	}
 
 	if (ctrl->effects)
-		effects = le32_to_cpu(ctrl->effects->iocs[opcode]);
+		effects = le32_to_cpu(ctrl->effects->acs[opcode]);
 	else
 		effects = nvme_known_admin_effects(opcode);
 
@@ -2220,7 +2226,7 @@ out_unlock:
 
 int nvme_get_log_ext(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 		     u8 log_page, void *log,
-		     size_t size, size_t offset)
+		     size_t size, u64 offset)
 {
 	struct nvme_command c = { };
 	unsigned long dwlen = size / 4 - 1;
@@ -2235,8 +2241,8 @@ int nvme_get_log_ext(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 	c.get_log_page.lid = log_page;
 	c.get_log_page.numdl = cpu_to_le16(dwlen & ((1 << 16) - 1));
 	c.get_log_page.numdu = cpu_to_le16(dwlen >> 16);
-	c.get_log_page.lpol = cpu_to_le32(offset & ((1ULL << 32) - 1));
-	c.get_log_page.lpou = cpu_to_le32(offset >> 32ULL);
+	c.get_log_page.lpol = cpu_to_le32(lower_32_bits(offset));
+	c.get_log_page.lpou = cpu_to_le32(upper_32_bits(offset));
 
 	return nvme_submit_sync_cmd(ctrl->admin_q, &c, log, size);
 }
@@ -2833,7 +2839,9 @@ static struct nvme_ns_head *nvme_alloc_ns_head(struct nvme_ctrl *ctrl,
 		goto out_free_head;
 	head->instance = ret;
 	INIT_LIST_HEAD(&head->list);
-	init_srcu_struct(&head->srcu);
+	ret = init_srcu_struct(&head->srcu);
+	if (ret)
+		goto out_ida_remove;
 	head->subsys = ctrl->subsys;
 	head->ns_id = nsid;
 	kref_init(&head->ref);
@@ -2855,6 +2863,7 @@ static struct nvme_ns_head *nvme_alloc_ns_head(struct nvme_ctrl *ctrl,
 	return head;
 out_cleanup_srcu:
 	cleanup_srcu_struct(&head->srcu);
+out_ida_remove:
 	ida_simple_remove(&ctrl->subsys->ns_ida, head->instance);
 out_free_head:
 	kfree(head);
diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index 8f0f34d06d46..124c458806df 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -536,6 +536,85 @@ static struct nvmf_transport_ops *nvmf_lookup_transport(
 	return NULL;
 }
 
+blk_status_t nvmf_check_if_ready(struct nvme_ctrl *ctrl, struct request *rq,
+		bool queue_live, bool is_connected)
+{
+	struct nvme_command *cmd = nvme_req(rq)->cmd;
+
+	if (likely(ctrl->state == NVME_CTRL_LIVE && is_connected))
+		return BLK_STS_OK;
+
+	switch (ctrl->state) {
+	case NVME_CTRL_DELETING:
+		goto reject_io;
+
+	case NVME_CTRL_NEW:
+	case NVME_CTRL_CONNECTING:
+		if (!is_connected)
+			/*
+			 * This is the case of starting a new
+			 * association but connectivity was lost
+			 * before it was fully created. We need to
+			 * error the commands used to initialize the
+			 * controller so the reconnect can go into a
+			 * retry attempt. The commands should all be
+			 * marked REQ_FAILFAST_DRIVER, which will hit
+			 * the reject path below. Anything else will
+			 * be queued while the state settles.
+			 */
+			goto reject_or_queue_io;
+
+		if ((queue_live &&
+		     !(nvme_req(rq)->flags & NVME_REQ_USERCMD)) ||
+		    (!queue_live && blk_rq_is_passthrough(rq) &&
+		     cmd->common.opcode == nvme_fabrics_command &&
+		     cmd->fabrics.fctype == nvme_fabrics_type_connect))
+			/*
+			 * If queue is live, allow only commands that
+			 * are internally generated pass through. These
+			 * are commands on the admin queue to initialize
+			 * the controller. This will reject any ioctl
+			 * admin cmds received while initializing.
+			 *
+			 * If the queue is not live, allow only a
+			 * connect command. This will reject any ioctl
+			 * admin cmd as well as initialization commands
+			 * if the controller reverted the queue to non-live.
+			 */
+			return BLK_STS_OK;
+
+		/*
+		 * fall-thru to the reject_or_queue_io clause
+		 */
+		break;
+
+	/* these cases fall-thru
+	 * case NVME_CTRL_LIVE:
+	 * case NVME_CTRL_RESETTING:
+	 */
+	default:
+		break;
+	}
+
+reject_or_queue_io:
+	/*
+	 * Any other new io is something we're not in a state to send
+	 * to the device. Default action is to busy it and retry it
+	 * after the controller state is recovered. However, anything
+	 * marked for failfast or nvme multipath is immediately failed.
+	 * Note: commands used to initialize the controller will be
+	 *  marked for failfast.
+	 * Note: nvme cli/ioctl commands are marked for failfast.
+	 */
+	if (!blk_noretry_request(rq) && !(rq->cmd_flags & REQ_NVME_MPATH))
+		return BLK_STS_RESOURCE;
+
+reject_io:
+	nvme_req(rq)->status = NVME_SC_ABORT_REQ;
+	return BLK_STS_IOERR;
+}
+EXPORT_SYMBOL_GPL(nvmf_check_if_ready);
+
 static const match_table_t opt_tokens = {
 	{ NVMF_OPT_TRANSPORT,		"transport=%s"		},
 	{ NVMF_OPT_TRADDR,		"traddr=%s"		},
@@ -608,8 +687,10 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 			opts->discovery_nqn =
 				!(strcmp(opts->subsysnqn,
 					 NVME_DISC_SUBSYS_NAME));
-			if (opts->discovery_nqn)
+			if (opts->discovery_nqn) {
+				opts->kato = 0;
 				opts->nr_io_queues = 0;
+			}
 			break;
 		case NVMF_OPT_TRADDR:
 			p = match_strdup(args);
diff --git a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h
index a3145d90c1d2..ef46c915b7b5 100644
--- a/drivers/nvme/host/fabrics.h
+++ b/drivers/nvme/host/fabrics.h
@@ -157,36 +157,7 @@ void nvmf_unregister_transport(struct nvmf_transport_ops *ops);
 void nvmf_free_options(struct nvmf_ctrl_options *opts);
 int nvmf_get_address(struct nvme_ctrl *ctrl, char *buf, int size);
 bool nvmf_should_reconnect(struct nvme_ctrl *ctrl);
-
-static inline blk_status_t nvmf_check_init_req(struct nvme_ctrl *ctrl,
-		struct request *rq)
-{
-	struct nvme_command *cmd = nvme_req(rq)->cmd;
-
-	/*
-	 * We cannot accept any other command until the connect command has
-	 * completed, so only allow connect to pass.
-	 */
-	if (!blk_rq_is_passthrough(rq) ||
-	    cmd->common.opcode != nvme_fabrics_command ||
-	    cmd->fabrics.fctype != nvme_fabrics_type_connect) {
-		/*
-		 * Connecting state means transport disruption or initial
-		 * establishment, which can take a long time and even might
-		 * fail permanently, fail fast to give upper layers a chance
-		 * to failover.
-		 * Deleting state means that the ctrl will never accept commands
-		 * again, fail it permanently.
-		 */
-		if (ctrl->state == NVME_CTRL_CONNECTING ||
-		    ctrl->state == NVME_CTRL_DELETING) {
-			nvme_req(rq)->status = NVME_SC_ABORT_REQ;
-			return BLK_STS_IOERR;
-		}
-		return BLK_STS_RESOURCE; /* try again later */
-	}
-
-	return BLK_STS_OK;
-}
+blk_status_t nvmf_check_if_ready(struct nvme_ctrl *ctrl,
+	struct request *rq, bool queue_live, bool is_connected);
 
 #endif /* _NVME_FABRICS_H */
diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index c6e719b2f3ca..6cb26bcf6ec0 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -2277,14 +2277,6 @@ nvme_fc_start_fcp_op(struct nvme_fc_ctrl *ctrl, struct nvme_fc_queue *queue,
 	return BLK_STS_OK;
 }
 
-static inline blk_status_t nvme_fc_is_ready(struct nvme_fc_queue *queue,
-		struct request *rq)
-{
-	if (unlikely(!test_bit(NVME_FC_Q_LIVE, &queue->flags)))
-		return nvmf_check_init_req(&queue->ctrl->ctrl, rq);
-	return BLK_STS_OK;
-}
-
 static blk_status_t
 nvme_fc_queue_rq(struct blk_mq_hw_ctx *hctx,
 			const struct blk_mq_queue_data *bd)
@@ -2300,7 +2292,9 @@ nvme_fc_queue_rq(struct blk_mq_hw_ctx *hctx,
 	u32 data_len;
 	blk_status_t ret;
 
-	ret = nvme_fc_is_ready(queue, rq);
+	ret = nvmf_check_if_ready(&queue->ctrl->ctrl, rq,
+		test_bit(NVME_FC_Q_LIVE, &queue->flags),
+		ctrl->rport->remoteport.port_state == FC_OBJSTATE_ONLINE);
 	if (unlikely(ret))
 		return ret;
 
diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index cf93690b3ffc..061fecfd44f5 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -105,6 +105,7 @@ struct nvme_request {
 
 enum {
 	NVME_REQ_CANCELLED		= (1 << 0),
+	NVME_REQ_USERCMD		= (1 << 1),
 };
 
 static inline struct nvme_request *nvme_req(struct request *req)
@@ -422,7 +423,6 @@ int __nvme_submit_sync_cmd(struct request_queue *q, struct nvme_command *cmd,
 		unsigned timeout, int qid, int at_head,
 		blk_mq_req_flags_t flags);
 int nvme_set_queue_count(struct nvme_ctrl *ctrl, int *count);
-void nvme_start_keep_alive(struct nvme_ctrl *ctrl);
 void nvme_stop_keep_alive(struct nvme_ctrl *ctrl);
 int nvme_reset_ctrl(struct nvme_ctrl *ctrl);
 int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl);
@@ -430,7 +430,7 @@ int nvme_delete_ctrl(struct nvme_ctrl *ctrl);
 int nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl);
 
 int nvme_get_log_ext(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
-		u8 log_page, void *log, size_t size, size_t offset);
+		u8 log_page, void *log, size_t size, u64 offset);
 
 extern const struct attribute_group nvme_ns_id_attr_group;
 extern const struct block_device_operations nvme_ns_head_ops;
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 295fbec1e5f2..fbc71fac6f1e 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -84,6 +84,7 @@ struct nvme_dev {
 	struct dma_pool *prp_small_pool;
 	unsigned online_queues;
 	unsigned max_qid;
+	unsigned int num_vecs;
 	int q_depth;
 	u32 db_stride;
 	void __iomem *bar;
@@ -414,7 +415,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
+	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
+			dev->num_vecs > 1 ? 1 /* admin queue */ : 0);
 }
 
 /**
@@ -1380,8 +1382,7 @@ static int nvme_alloc_sq_cmds(struct nvme_dev *dev, struct nvme_queue *nvmeq,
 	return 0;
 }
 
-static int nvme_alloc_queue(struct nvme_dev *dev, int qid,
-		int depth, int node)
+static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 {
 	struct nvme_queue *nvmeq = &dev->queues[qid];
 
@@ -1457,7 +1458,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 		nvmeq->sq_cmds_io = dev->cmb + offset;
 	}
 
-	nvmeq->cq_vector = qid - 1;
+	/*
+	 * A queue's vector matches the queue identifier unless the controller
+	 * has only one vector available.
+	 */
+	nvmeq->cq_vector = dev->num_vecs == 1 ? 0 : qid;
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		goto release_vector;
@@ -1596,8 +1601,7 @@ static int nvme_pci_configure_admin_queue(struct nvme_dev *dev)
 	if (result < 0)
 		return result;
 
-	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH,
-			dev_to_node(dev->dev));
+	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
 	if (result)
 		return result;
 
@@ -1630,9 +1634,7 @@ static int nvme_create_io_queues(struct nvme_dev *dev)
 	int ret = 0;
 
 	for (i = dev->ctrl.queue_count; i <= dev->max_qid; i++) {
-		/* vector == qid - 1, match nvme_create_queue */
-		if (nvme_alloc_queue(dev, i, dev->q_depth,
-		     pci_irq_get_node(to_pci_dev(dev->dev), i - 1))) {
+		if (nvme_alloc_queue(dev, i, dev->q_depth)) {
 			ret = -ENOMEM;
 			break;
 		}
@@ -1914,6 +1916,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	int result, nr_io_queues;
 	unsigned long size;
 
+	struct irq_affinity affd = {
+		.pre_vectors = 1
+	};
+
 	nr_io_queues = num_possible_cpus();
 	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
 	if (result < 0)
@@ -1949,11 +1955,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 * setting up the full range we need.
 	 */
 	pci_free_irq_vectors(pdev);
-	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
-			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
-	if (nr_io_queues <= 0)
+	result = pci_alloc_irq_vectors_affinity(pdev, 1, nr_io_queues + 1,
+			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
+	if (result <= 0)
 		return -EIO;
-	dev->max_qid = nr_io_queues;
+	dev->num_vecs = result;
+	dev->max_qid = max(result - 1, 1);
 
 	/*
 	 * Should investigate if there's a performance win from allocating
@@ -2201,7 +2208,7 @@ static void nvme_dev_disable(struct nvme_dev *dev, bool shutdown)
 
 	nvme_stop_queues(&dev->ctrl);
 
-	if (!dead) {
+	if (!dead && dev->ctrl.queue_count > 0) {
 		/*
 		 * If the controller is still alive tell it to stop using the
 		 * host memory buffer.  In theory the shutdown / reset should
diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index 758537e9ba07..1eb4438a8763 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -1601,17 +1601,6 @@ nvme_rdma_timeout(struct request *rq, bool reserved)
 	return BLK_EH_HANDLED;
 }
 
-/*
- * We cannot accept any other command until the Connect command has completed.
- */
-static inline blk_status_t
-nvme_rdma_is_ready(struct nvme_rdma_queue *queue, struct request *rq)
-{
-	if (unlikely(!test_bit(NVME_RDMA_Q_LIVE, &queue->flags)))
-		return nvmf_check_init_req(&queue->ctrl->ctrl, rq);
-	return BLK_STS_OK;
-}
-
 static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx,
 		const struct blk_mq_queue_data *bd)
 {
@@ -1627,7 +1616,8 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx,
 
 	WARN_ON_ONCE(rq->tag < 0);
 
-	ret = nvme_rdma_is_ready(queue, rq);
+	ret = nvmf_check_if_ready(&queue->ctrl->ctrl, rq,
+		test_bit(NVME_RDMA_Q_LIVE, &queue->flags), true);
 	if (unlikely(ret))
 		return ret;