1 files changed, 869 insertions, 341 deletions

diff --git a/drivers/nvme/target/core.c b/drivers/nvme/target/core.c
index 88d260f31835..cc88e5a28c8a 100644
--- a/drivers/nvme/target/core.c
+++ b/drivers/nvme/target/core.c
@@ -1,28 +1,32 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Common code for the NVMe target.
  * Copyright (c) 2015-2016 HGST, a Western Digital Company.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
  */
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 #include <linux/module.h>
 #include <linux/random.h>
 #include <linux/rculist.h>
 #include <linux/pci-p2pdma.h>
+#include <linux/scatterlist.h>
+
+#include <generated/utsrelease.h>
+
+#define CREATE_TRACE_POINTS
+#include "trace.h"
 
 #include "nvmet.h"
+#include "debugfs.h"
 
+struct kmem_cache *nvmet_bvec_cache;
 struct workqueue_struct *buffered_io_wq;
+struct workqueue_struct *zbd_wq;
 static const struct nvmet_fabrics_ops *nvmet_transports[NVMF_TRTYPE_MAX];
 static DEFINE_IDA(cntlid_ida);
 
+struct workqueue_struct *nvmet_wq;
+EXPORT_SYMBOL_GPL(nvmet_wq);
+
 /*
  * This read/write semaphore is used to synchronize access to configuration
  * information on a target system that will result in discovery log page
@@ -36,7 +40,7 @@ static DEFINE_IDA(cntlid_ida);
  *  - the nvmet_transports array
  *
  * When updating any of those lists/structures write lock should be obtained,
- * while when reading (popolating discovery log page or checking host-subsystem
+ * while when reading (populating discovery log page or checking host-subsystem
  * link) read lock is obtained to allow concurrent reads.
  */
 DECLARE_RWSEM(nvmet_config_sem);
@@ -47,40 +51,36 @@ DECLARE_RWSEM(nvmet_ana_sem);
 
 inline u16 errno_to_nvme_status(struct nvmet_req *req, int errno)
 {
-	u16 status;
-
 	switch (errno) {
+	case 0:
+		return NVME_SC_SUCCESS;
 	case -ENOSPC:
 		req->error_loc = offsetof(struct nvme_rw_command, length);
-		status = NVME_SC_CAP_EXCEEDED | NVME_SC_DNR;
-		break;
+		return NVME_SC_CAP_EXCEEDED | NVME_STATUS_DNR;
 	case -EREMOTEIO:
 		req->error_loc = offsetof(struct nvme_rw_command, slba);
-		status = NVME_SC_LBA_RANGE | NVME_SC_DNR;
-		break;
+		return  NVME_SC_LBA_RANGE | NVME_STATUS_DNR;
 	case -EOPNOTSUPP:
 		req->error_loc = offsetof(struct nvme_common_command, opcode);
-		switch (req->cmd->common.opcode) {
-		case nvme_cmd_dsm:
-		case nvme_cmd_write_zeroes:
-			status = NVME_SC_ONCS_NOT_SUPPORTED | NVME_SC_DNR;
-			break;
-		default:
-			status = NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
-		}
-		break;
+		return NVME_SC_INVALID_OPCODE | NVME_STATUS_DNR;
 	case -ENODATA:
 		req->error_loc = offsetof(struct nvme_rw_command, nsid);
-		status = NVME_SC_ACCESS_DENIED;
-		break;
+		return NVME_SC_ACCESS_DENIED;
 	case -EIO:
-		/* FALLTHRU */
+		fallthrough;
 	default:
 		req->error_loc = offsetof(struct nvme_common_command, opcode);
-		status = NVME_SC_INTERNAL | NVME_SC_DNR;
+		return NVME_SC_INTERNAL | NVME_STATUS_DNR;
 	}
+}
 
-	return status;
+u16 nvmet_report_invalid_opcode(struct nvmet_req *req)
+{
+	pr_debug("unhandled cmd %d on qid %d\n", req->cmd->common.opcode,
+		 req->sq->qid);
+
+	req->error_loc = offsetof(struct nvme_common_command, opcode);
+	return NVME_SC_INVALID_OPCODE | NVME_STATUS_DNR;
 }
 
 static struct nvmet_subsys *nvmet_find_get_subsys(struct nvmet_port *port,
@@ -91,7 +91,7 @@ u16 nvmet_copy_to_sgl(struct nvmet_req *req, off_t off, const void *buf,
 {
 	if (sg_pcopy_from_buffer(req->sg, req->sg_cnt, buf, len, off) != len) {
 		req->error_loc = offsetof(struct nvme_common_command, dptr);
-		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+		return NVME_SC_SGL_INVALID_DATA | NVME_STATUS_DNR;
 	}
 	return 0;
 }
@@ -100,7 +100,7 @@ u16 nvmet_copy_from_sgl(struct nvmet_req *req, off_t off, void *buf, size_t len)
 {
 	if (sg_pcopy_to_buffer(req->sg, req->sg_cnt, buf, len, off) != len) {
 		req->error_loc = offsetof(struct nvme_common_command, dptr);
-		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+		return NVME_SC_SGL_INVALID_DATA | NVME_STATUS_DNR;
 	}
 	return 0;
 }
@@ -109,20 +109,21 @@ u16 nvmet_zero_sgl(struct nvmet_req *req, off_t off, size_t len)
 {
 	if (sg_zero_buffer(req->sg, req->sg_cnt, len, off) != len) {
 		req->error_loc = offsetof(struct nvme_common_command, dptr);
-		return NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR;
+		return NVME_SC_SGL_INVALID_DATA | NVME_STATUS_DNR;
 	}
 	return 0;
 }
 
-static unsigned int nvmet_max_nsid(struct nvmet_subsys *subsys)
+static u32 nvmet_max_nsid(struct nvmet_subsys *subsys)
 {
-	struct nvmet_ns *ns;
+	struct nvmet_ns *cur;
+	unsigned long idx;
+	u32 nsid = 0;
 
-	if (list_empty(&subsys->namespaces))
-		return 0;
+	nvmet_for_each_enabled_ns(&subsys->namespaces, idx, cur)
+		nsid = cur->nsid;
 
-	ns = list_last_entry(&subsys->namespaces, struct nvmet_ns, dev_link);
-	return ns->nsid;
+	return nsid;
 }
 
 static u32 nvmet_async_event_result(struct nvmet_async_event *aen)
@@ -130,39 +131,29 @@ static u32 nvmet_async_event_result(struct nvmet_async_event *aen)
 	return aen->event_type | (aen->event_info << 8) | (aen->log_page << 16);
 }
 
-static void nvmet_async_events_free(struct nvmet_ctrl *ctrl)
+static void nvmet_async_events_failall(struct nvmet_ctrl *ctrl)
 {
 	struct nvmet_req *req;
 
-	while (1) {
-		mutex_lock(&ctrl->lock);
-		if (!ctrl->nr_async_event_cmds) {
-			mutex_unlock(&ctrl->lock);
-			return;
-		}
-
+	mutex_lock(&ctrl->lock);
+	while (ctrl->nr_async_event_cmds) {
 		req = ctrl->async_event_cmds[--ctrl->nr_async_event_cmds];
 		mutex_unlock(&ctrl->lock);
-		nvmet_req_complete(req, NVME_SC_INTERNAL | NVME_SC_DNR);
+		nvmet_req_complete(req, NVME_SC_INTERNAL | NVME_STATUS_DNR);
+		mutex_lock(&ctrl->lock);
 	}
+	mutex_unlock(&ctrl->lock);
 }
 
-static void nvmet_async_event_work(struct work_struct *work)
+static void nvmet_async_events_process(struct nvmet_ctrl *ctrl)
 {
-	struct nvmet_ctrl *ctrl =
-		container_of(work, struct nvmet_ctrl, async_event_work);
 	struct nvmet_async_event *aen;
 	struct nvmet_req *req;
 
-	while (1) {
-		mutex_lock(&ctrl->lock);
-		aen = list_first_entry_or_null(&ctrl->async_events,
-				struct nvmet_async_event, entry);
-		if (!aen || !ctrl->nr_async_event_cmds) {
-			mutex_unlock(&ctrl->lock);
-			return;
-		}
-
+	mutex_lock(&ctrl->lock);
+	while (ctrl->nr_async_event_cmds && !list_empty(&ctrl->async_events)) {
+		aen = list_first_entry(&ctrl->async_events,
+				       struct nvmet_async_event, entry);
 		req = ctrl->async_event_cmds[--ctrl->nr_async_event_cmds];
 		nvmet_set_result(req, nvmet_async_event_result(aen));
 
@@ -170,8 +161,31 @@ static void nvmet_async_event_work(struct work_struct *work)
 		kfree(aen);
 
 		mutex_unlock(&ctrl->lock);
+		trace_nvmet_async_event(ctrl, req->cqe->result.u32);
 		nvmet_req_complete(req, 0);
+		mutex_lock(&ctrl->lock);
 	}
+	mutex_unlock(&ctrl->lock);
+}
+
+static void nvmet_async_events_free(struct nvmet_ctrl *ctrl)
+{
+	struct nvmet_async_event *aen, *tmp;
+
+	mutex_lock(&ctrl->lock);
+	list_for_each_entry_safe(aen, tmp, &ctrl->async_events, entry) {
+		list_del(&aen->entry);
+		kfree(aen);
+	}
+	mutex_unlock(&ctrl->lock);
+}
+
+static void nvmet_async_event_work(struct work_struct *work)
+{
+	struct nvmet_ctrl *ctrl =
+		container_of(work, struct nvmet_ctrl, async_event_work);
+
+	nvmet_async_events_process(ctrl);
 }
 
 void nvmet_add_async_event(struct nvmet_ctrl *ctrl, u8 event_type,
@@ -191,7 +205,7 @@ void nvmet_add_async_event(struct nvmet_ctrl *ctrl, u8 event_type,
 	list_add_tail(&aen->entry, &ctrl->async_events);
 	mutex_unlock(&ctrl->lock);
 
-	schedule_work(&ctrl->async_event_work);
+	queue_work(nvmet_wq, &ctrl->async_event_work);
 }
 
 static void nvmet_add_to_changed_ns_log(struct nvmet_ctrl *ctrl, __le32 nsid)
@@ -222,11 +236,13 @@ void nvmet_ns_changed(struct nvmet_subsys *subsys, u32 nsid)
 {
 	struct nvmet_ctrl *ctrl;
 
+	lockdep_assert_held(&subsys->lock);
+
 	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry) {
 		nvmet_add_to_changed_ns_log(ctrl, cpu_to_le32(nsid));
 		if (nvmet_aen_bit_disabled(ctrl, NVME_AEN_BIT_NS_ATTR))
 			continue;
-		nvmet_add_async_event(ctrl, NVME_AER_TYPE_NOTICE,
+		nvmet_add_async_event(ctrl, NVME_AER_NOTICE,
 				NVME_AER_NOTICE_NS_CHANGED,
 				NVME_LOG_CHANGED_NS);
 	}
@@ -243,7 +259,7 @@ void nvmet_send_ana_event(struct nvmet_subsys *subsys,
 			continue;
 		if (nvmet_aen_bit_disabled(ctrl, NVME_AEN_BIT_ANA_CHANGE))
 			continue;
-		nvmet_add_async_event(ctrl, NVME_AER_TYPE_NOTICE,
+		nvmet_add_async_event(ctrl, NVME_AER_NOTICE,
 				NVME_AER_NOTICE_ANA, NVME_LOG_ANA);
 	}
 	mutex_unlock(&subsys->lock);
@@ -282,6 +298,18 @@ void nvmet_unregister_transport(const struct nvmet_fabrics_ops *ops)
 }
 EXPORT_SYMBOL_GPL(nvmet_unregister_transport);
 
+void nvmet_port_del_ctrls(struct nvmet_port *port, struct nvmet_subsys *subsys)
+{
+	struct nvmet_ctrl *ctrl;
+
+	mutex_lock(&subsys->lock);
+	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry) {
+		if (ctrl->port == port)
+			ctrl->ops->delete_ctrl(ctrl);
+	}
+	mutex_unlock(&subsys->lock);
+}
+
 int nvmet_enable_port(struct nvmet_port *port)
 {
 	const struct nvmet_fabrics_ops *ops;
@@ -289,6 +317,9 @@ int nvmet_enable_port(struct nvmet_port *port)
 
 	lockdep_assert_held(&nvmet_config_sem);
 
+	if (port->disc_addr.trtype == NVMF_TRTYPE_MAX)
+		return -EINVAL;
+
 	ops = nvmet_transports[port->disc_addr.trtype];
 	if (!ops) {
 		up_write(&nvmet_config_sem);
@@ -305,18 +336,44 @@ int nvmet_enable_port(struct nvmet_port *port)
 	if (!try_module_get(ops->owner))
 		return -EINVAL;
 
-	ret = ops->add_port(port);
-	if (ret) {
-		module_put(ops->owner);
-		return ret;
+	/*
+	 * If the user requested PI support and the transport isn't pi capable,
+	 * don't enable the port.
+	 */
+	if (port->pi_enable && !(ops->flags & NVMF_METADATA_SUPPORTED)) {
+		pr_err("T10-PI is not supported by transport type %d\n",
+		       port->disc_addr.trtype);
+		ret = -EINVAL;
+		goto out_put;
 	}
 
+	ret = ops->add_port(port);
+	if (ret)
+		goto out_put;
+
 	/* If the transport didn't set inline_data_size, then disable it. */
 	if (port->inline_data_size < 0)
 		port->inline_data_size = 0;
 
+	/*
+	 * If the transport didn't set the max_queue_size properly, then clamp
+	 * it to the target limits. Also set default values in case the
+	 * transport didn't set it at all.
+	 */
+	if (port->max_queue_size < 0)
+		port->max_queue_size = NVMET_MAX_QUEUE_SIZE;
+	else
+		port->max_queue_size = clamp_t(int, port->max_queue_size,
+					       NVMET_MIN_QUEUE_SIZE,
+					       NVMET_MAX_QUEUE_SIZE);
+
 	port->enabled = true;
+	port->tr_ops = ops;
 	return 0;
+
+out_put:
+	module_put(ops->owner);
+	return ret;
 }
 
 void nvmet_disable_port(struct nvmet_port *port)
@@ -326,6 +383,7 @@ void nvmet_disable_port(struct nvmet_port *port)
 	lockdep_assert_held(&nvmet_config_sem);
 
 	port->enabled = false;
+	port->tr_ops = NULL;
 
 	ops = nvmet_transports[port->disc_addr.trtype];
 	ops->remove_port(port);
@@ -336,13 +394,13 @@ static void nvmet_keep_alive_timer(struct work_struct *work)
 {
 	struct nvmet_ctrl *ctrl = container_of(to_delayed_work(work),
 			struct nvmet_ctrl, ka_work);
-	bool cmd_seen = ctrl->cmd_seen;
+	bool reset_tbkas = ctrl->reset_tbkas;
 
-	ctrl->cmd_seen = false;
-	if (cmd_seen) {
+	ctrl->reset_tbkas = false;
+	if (reset_tbkas) {
 		pr_debug("ctrl %d reschedule traffic based keep-alive timer\n",
 			ctrl->cntlid);
-		schedule_delayed_work(&ctrl->ka_work, ctrl->kato * HZ);
+		queue_delayed_work(nvmet_wq, &ctrl->ka_work, ctrl->kato * HZ);
 		return;
 	}
 
@@ -352,46 +410,45 @@ static void nvmet_keep_alive_timer(struct work_struct *work)
 	nvmet_ctrl_fatal_error(ctrl);
 }
 
-static void nvmet_start_keep_alive_timer(struct nvmet_ctrl *ctrl)
+void nvmet_start_keep_alive_timer(struct nvmet_ctrl *ctrl)
 {
+	if (unlikely(ctrl->kato == 0))
+		return;
+
 	pr_debug("ctrl %d start keep-alive timer for %d secs\n",
 		ctrl->cntlid, ctrl->kato);
 
-	INIT_DELAYED_WORK(&ctrl->ka_work, nvmet_keep_alive_timer);
-	schedule_delayed_work(&ctrl->ka_work, ctrl->kato * HZ);
+	queue_delayed_work(nvmet_wq, &ctrl->ka_work, ctrl->kato * HZ);
 }
 
-static void nvmet_stop_keep_alive_timer(struct nvmet_ctrl *ctrl)
+void nvmet_stop_keep_alive_timer(struct nvmet_ctrl *ctrl)
 {
+	if (unlikely(ctrl->kato == 0))
+		return;
+
 	pr_debug("ctrl %d stop keep-alive\n", ctrl->cntlid);
 
 	cancel_delayed_work_sync(&ctrl->ka_work);
 }
 
-static struct nvmet_ns *__nvmet_find_namespace(struct nvmet_ctrl *ctrl,
-		__le32 nsid)
+u16 nvmet_req_find_ns(struct nvmet_req *req)
 {
-	struct nvmet_ns *ns;
+	u32 nsid = le32_to_cpu(req->cmd->common.nsid);
+	struct nvmet_subsys *subsys = nvmet_req_subsys(req);
 
-	list_for_each_entry_rcu(ns, &ctrl->subsys->namespaces, dev_link) {
-		if (ns->nsid == le32_to_cpu(nsid))
-			return ns;
-	}
-
-	return NULL;
-}
-
-struct nvmet_ns *nvmet_find_namespace(struct nvmet_ctrl *ctrl, __le32 nsid)
-{
-	struct nvmet_ns *ns;
+	req->ns = xa_load(&subsys->namespaces, nsid);
+	if (unlikely(!req->ns || !req->ns->enabled)) {
+		req->error_loc = offsetof(struct nvme_common_command, nsid);
+		if (!req->ns) /* ns doesn't exist! */
+			return NVME_SC_INVALID_NS | NVME_STATUS_DNR;
 
-	rcu_read_lock();
-	ns = __nvmet_find_namespace(ctrl, nsid);
-	if (ns)
-		percpu_ref_get(&ns->ref);
-	rcu_read_unlock();
+		/* ns exists but it's disabled */
+		req->ns = NULL;
+		return NVME_SC_INTERNAL_PATH_ERROR;
+	}
 
-	return ns;
+	percpu_ref_get(&req->ns->ref);
+	return NVME_SC_SUCCESS;
 }
 
 static void nvmet_destroy_namespace(struct percpu_ref *ref)
@@ -425,7 +482,7 @@ static int nvmet_p2pmem_ns_enable(struct nvmet_ns *ns)
 		return -EINVAL;
 	}
 
-	if (!blk_queue_pci_p2pdma(ns->bdev->bd_queue)) {
+	if (!blk_queue_pci_p2pdma(ns->bdev->bd_disk->queue)) {
 		pr_err("peer-to-peer DMA is not supported by the driver of %s\n",
 		       ns->device_path);
 		return -EINVAL;
@@ -456,9 +513,6 @@ static int nvmet_p2pmem_ns_enable(struct nvmet_ns *ns)
 	return 0;
 }
 
-/*
- * Note: ctrl->subsys->lock should be held when calling this function
- */
 static void nvmet_p2pmem_ns_add_p2p(struct nvmet_ctrl *ctrl,
 				    struct nvmet_ns *ns)
 {
@@ -466,6 +520,8 @@ static void nvmet_p2pmem_ns_add_p2p(struct nvmet_ctrl *ctrl,
 	struct pci_dev *p2p_dev;
 	int ret;
 
+	lockdep_assert_held(&ctrl->subsys->lock);
+
 	if (!ctrl->p2p_client || !ns->use_p2pmem)
 		return;
 
@@ -495,6 +551,18 @@ static void nvmet_p2pmem_ns_add_p2p(struct nvmet_ctrl *ctrl,
 		ns->nsid);
 }
 
+bool nvmet_ns_revalidate(struct nvmet_ns *ns)
+{
+	loff_t oldsize = ns->size;
+
+	if (ns->bdev)
+		nvmet_bdev_ns_revalidate(ns);
+	else
+		nvmet_file_ns_revalidate(ns);
+
+	return oldsize != ns->size;
+}
+
 int nvmet_ns_enable(struct nvmet_ns *ns)
 {
 	struct nvmet_subsys *subsys = ns->subsys;
@@ -502,10 +570,13 @@ int nvmet_ns_enable(struct nvmet_ns *ns)
 	int ret;
 
 	mutex_lock(&subsys->lock);
-	ret = -EMFILE;
-	if (subsys->nr_namespaces == NVMET_MAX_NAMESPACES)
-		goto out_unlock;
 	ret = 0;
+
+	if (nvmet_is_passthru_subsys(subsys)) {
+		pr_info("cannot enable both passthru and regular namespaces for a single subsystem");
+		goto out_unlock;
+	}
+
 	if (ns->enabled)
 		goto out_unlock;
 
@@ -517,48 +588,34 @@ int nvmet_ns_enable(struct nvmet_ns *ns)
 
 	ret = nvmet_p2pmem_ns_enable(ns);
 	if (ret)
-		goto out_unlock;
+		goto out_dev_disable;
 
 	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry)
 		nvmet_p2pmem_ns_add_p2p(ctrl, ns);
 
-	ret = percpu_ref_init(&ns->ref, nvmet_destroy_namespace,
-				0, GFP_KERNEL);
-	if (ret)
-		goto out_dev_put;
-
-	if (ns->nsid > subsys->max_nsid)
-		subsys->max_nsid = ns->nsid;
-
-	/*
-	 * The namespaces list needs to be sorted to simplify the implementation
-	 * of the Identify Namepace List subcommand.
-	 */
-	if (list_empty(&subsys->namespaces)) {
-		list_add_tail_rcu(&ns->dev_link, &subsys->namespaces);
-	} else {
-		struct nvmet_ns *old;
-
-		list_for_each_entry_rcu(old, &subsys->namespaces, dev_link) {
-			BUG_ON(ns->nsid == old->nsid);
-			if (ns->nsid < old->nsid)
-				break;
-		}
-
-		list_add_tail_rcu(&ns->dev_link, &old->dev_link);
+	if (ns->pr.enable) {
+		ret = nvmet_pr_init_ns(ns);
+		if (ret)
+			goto out_dev_put;
 	}
-	subsys->nr_namespaces++;
+
+	if (percpu_ref_init(&ns->ref, nvmet_destroy_namespace, 0, GFP_KERNEL))
+		goto out_pr_exit;
 
 	nvmet_ns_changed(subsys, ns->nsid);
 	ns->enabled = true;
+	xa_set_mark(&subsys->namespaces, ns->nsid, NVMET_NS_ENABLED);
 	ret = 0;
 out_unlock:
 	mutex_unlock(&subsys->lock);
 	return ret;
+out_pr_exit:
+	if (ns->pr.enable)
+		nvmet_pr_exit_ns(ns);
 out_dev_put:
 	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry)
 		pci_dev_put(radix_tree_delete(&ctrl->p2p_ns_map, ns->nsid));
-
+out_dev_disable:
 	nvmet_ns_dev_disable(ns);
 	goto out_unlock;
 }
@@ -573,9 +630,7 @@ void nvmet_ns_disable(struct nvmet_ns *ns)
 		goto out_unlock;
 
 	ns->enabled = false;
-	list_del_rcu(&ns->dev_link);
-	if (ns->nsid == subsys->max_nsid)
-		subsys->max_nsid = nvmet_max_nsid(subsys);
+	xa_clear_mark(&subsys->namespaces, ns->nsid, NVMET_NS_ENABLED);
 
 	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry)
 		pci_dev_put(radix_tree_delete(&ctrl->p2p_ns_map, ns->nsid));
@@ -586,7 +641,7 @@ void nvmet_ns_disable(struct nvmet_ns *ns)
 	 * Now that we removed the namespaces from the lookup list, we
 	 * can kill the per_cpu ref and wait for any remaining references
 	 * to be dropped, as well as a RCU grace period for anyone only
-	 * using the namepace under rcu_read_lock().  Note that we can't
+	 * using the namespace under rcu_read_lock().  Note that we can't
 	 * use call_rcu here as we need to ensure the namespaces have
 	 * been fully destroyed before unloading the module.
 	 */
@@ -595,9 +650,10 @@ void nvmet_ns_disable(struct nvmet_ns *ns)
 	wait_for_completion(&ns->disable_done);
 	percpu_ref_exit(&ns->ref);
 
-	mutex_lock(&subsys->lock);
+	if (ns->pr.enable)
+		nvmet_pr_exit_ns(ns);
 
-	subsys->nr_namespaces--;
+	mutex_lock(&subsys->lock);
 	nvmet_ns_changed(subsys, ns->nsid);
 	nvmet_ns_dev_disable(ns);
 out_unlock:
@@ -606,8 +662,19 @@ out_unlock:
 
 void nvmet_ns_free(struct nvmet_ns *ns)
 {
+	struct nvmet_subsys *subsys = ns->subsys;
+
 	nvmet_ns_disable(ns);
 
+	mutex_lock(&subsys->lock);
+
+	xa_erase(&subsys->namespaces, ns->nsid);
+	if (ns->nsid == subsys->max_nsid)
+		subsys->max_nsid = nvmet_max_nsid(subsys);
+
+	subsys->nr_namespaces--;
+	mutex_unlock(&subsys->lock);
+
 	down_write(&nvmet_ana_sem);
 	nvmet_ana_group_enabled[ns->anagrpid]--;
 	up_write(&nvmet_ana_sem);
@@ -620,16 +687,30 @@ struct nvmet_ns *nvmet_ns_alloc(struct nvmet_subsys *subsys, u32 nsid)
 {
 	struct nvmet_ns *ns;
 
+	mutex_lock(&subsys->lock);
+
+	if (subsys->nr_namespaces == NVMET_MAX_NAMESPACES)
+		goto out_unlock;
+
 	ns = kzalloc(sizeof(*ns), GFP_KERNEL);
 	if (!ns)
-		return NULL;
+		goto out_unlock;
 
-	INIT_LIST_HEAD(&ns->dev_link);
 	init_completion(&ns->disable_done);
 
 	ns->nsid = nsid;
 	ns->subsys = subsys;
 
+	if (ns->nsid > subsys->max_nsid)
+		subsys->max_nsid = nsid;
+
+	if (xa_insert(&subsys->namespaces, ns->nsid, ns, GFP_KERNEL))
+		goto out_exit;
+
+	subsys->nr_namespaces++;
+
+	mutex_unlock(&subsys->lock);
+
 	down_write(&nvmet_ana_sem);
 	ns->anagrpid = NVMET_DEFAULT_ANA_GRPID;
 	nvmet_ana_group_enabled[ns->anagrpid]++;
@@ -637,8 +718,15 @@ struct nvmet_ns *nvmet_ns_alloc(struct nvmet_subsys *subsys, u32 nsid)
 
 	uuid_gen(&ns->uuid);
 	ns->buffered_io = false;
+	ns->csi = NVME_CSI_NVM;
 
 	return ns;
+out_exit:
+	subsys->max_nsid = nvmet_max_nsid(subsys);
+	kfree(ns);
+out_unlock:
+	mutex_unlock(&subsys->lock);
+	return NULL;
 }
 
 static void nvmet_update_sq_head(struct nvmet_req *req)
@@ -646,13 +734,12 @@ static void nvmet_update_sq_head(struct nvmet_req *req)
 	if (req->sq->size) {
 		u32 old_sqhd, new_sqhd;
 
+		old_sqhd = READ_ONCE(req->sq->sqhd);
 		do {
-			old_sqhd = req->sq->sqhd;
 			new_sqhd = (old_sqhd + 1) % req->sq->size;
-		} while (cmpxchg(&req->sq->sqhd, old_sqhd, new_sqhd) !=
-					old_sqhd);
+		} while (!try_cmpxchg(&req->sq->sqhd, &old_sqhd, new_sqhd));
 	}
-	req->rsp->sq_head = cpu_to_le16(req->sq->sqhd & 0x0000FFFF);
+	req->cqe->sq_head = cpu_to_le16(req->sq->sqhd & 0x0000FFFF);
 }
 
 static void nvmet_set_error(struct nvmet_req *req, u16 status)
@@ -661,7 +748,7 @@ static void nvmet_set_error(struct nvmet_req *req, u16 status)
 	struct nvme_error_slot *new_error_slot;
 	unsigned long flags;
 
-	req->rsp->status = cpu_to_le16(status << 1);
+	req->cqe->status = cpu_to_le16(status << 1);
 
 	if (!ctrl || req->error_loc == NVMET_NO_ERROR_LOC)
 		return;
@@ -681,30 +768,60 @@ static void nvmet_set_error(struct nvmet_req *req, u16 status)
 	spin_unlock_irqrestore(&ctrl->error_lock, flags);
 
 	/* set the more bit for this request */
-	req->rsp->status |= cpu_to_le16(1 << 14);
+	req->cqe->status |= cpu_to_le16(1 << 14);
 }
 
 static void __nvmet_req_complete(struct nvmet_req *req, u16 status)
 {
+	struct nvmet_ns *ns = req->ns;
+	struct nvmet_pr_per_ctrl_ref *pc_ref = req->pc_ref;
+
 	if (!req->sq->sqhd_disabled)
 		nvmet_update_sq_head(req);
-	req->rsp->sq_id = cpu_to_le16(req->sq->qid);
-	req->rsp->command_id = req->cmd->common.command_id;
+	req->cqe->sq_id = cpu_to_le16(req->sq->qid);
+	req->cqe->command_id = req->cmd->common.command_id;
 
 	if (unlikely(status))
 		nvmet_set_error(req, status);
-	if (req->ns)
-		nvmet_put_namespace(req->ns);
+
+	trace_nvmet_req_complete(req);
+
 	req->ops->queue_response(req);
+
+	if (pc_ref)
+		nvmet_pr_put_ns_pc_ref(pc_ref);
+	if (ns)
+		nvmet_put_namespace(ns);
 }
 
 void nvmet_req_complete(struct nvmet_req *req, u16 status)
 {
+	struct nvmet_sq *sq = req->sq;
+
 	__nvmet_req_complete(req, status);
-	percpu_ref_put(&req->sq->ref);
+	percpu_ref_put(&sq->ref);
 }
 EXPORT_SYMBOL_GPL(nvmet_req_complete);
 
+void nvmet_cq_init(struct nvmet_cq *cq)
+{
+	refcount_set(&cq->ref, 1);
+}
+EXPORT_SYMBOL_GPL(nvmet_cq_init);
+
+bool nvmet_cq_get(struct nvmet_cq *cq)
+{
+	return refcount_inc_not_zero(&cq->ref);
+}
+EXPORT_SYMBOL_GPL(nvmet_cq_get);
+
+void nvmet_cq_put(struct nvmet_cq *cq)
+{
+	if (refcount_dec_and_test(&cq->ref))
+		nvmet_cq_destroy(cq);
+}
+EXPORT_SYMBOL_GPL(nvmet_cq_put);
+
 void nvmet_cq_setup(struct nvmet_ctrl *ctrl, struct nvmet_cq *cq,
 		u16 qid, u16 size)
 {
@@ -714,6 +831,17 @@ void nvmet_cq_setup(struct nvmet_ctrl *ctrl, struct nvmet_cq *cq,
 	ctrl->cqs[qid] = cq;
 }
 
+void nvmet_cq_destroy(struct nvmet_cq *cq)
+{
+	struct nvmet_ctrl *ctrl = cq->ctrl;
+
+	if (ctrl) {
+		ctrl->cqs[cq->qid] = NULL;
+		nvmet_ctrl_put(cq->ctrl);
+		cq->ctrl = NULL;
+	}
+}
+
 void nvmet_sq_setup(struct nvmet_ctrl *ctrl, struct nvmet_sq *sq,
 		u16 qid, u16 size)
 {
@@ -731,21 +859,135 @@ static void nvmet_confirm_sq(struct percpu_ref *ref)
 	complete(&sq->confirm_done);
 }
 
+u16 nvmet_check_cqid(struct nvmet_ctrl *ctrl, u16 cqid, bool create)
+{
+	if (!ctrl->cqs)
+		return NVME_SC_INTERNAL | NVME_STATUS_DNR;
+
+	if (cqid > ctrl->subsys->max_qid)
+		return NVME_SC_QID_INVALID | NVME_STATUS_DNR;
+
+	if ((create && ctrl->cqs[cqid]) || (!create && !ctrl->cqs[cqid]))
+		return NVME_SC_QID_INVALID | NVME_STATUS_DNR;
+
+	return NVME_SC_SUCCESS;
+}
+
+u16 nvmet_check_io_cqid(struct nvmet_ctrl *ctrl, u16 cqid, bool create)
+{
+	if (!cqid)
+		return NVME_SC_QID_INVALID | NVME_STATUS_DNR;
+	return nvmet_check_cqid(ctrl, cqid, create);
+}
+
+bool nvmet_cq_in_use(struct nvmet_cq *cq)
+{
+	return refcount_read(&cq->ref) > 1;
+}
+EXPORT_SYMBOL_GPL(nvmet_cq_in_use);
+
+u16 nvmet_cq_create(struct nvmet_ctrl *ctrl, struct nvmet_cq *cq,
+		    u16 qid, u16 size)
+{
+	u16 status;
+
+	status = nvmet_check_cqid(ctrl, qid, true);
+	if (status != NVME_SC_SUCCESS)
+		return status;
+
+	if (!kref_get_unless_zero(&ctrl->ref))
+		return NVME_SC_INTERNAL | NVME_STATUS_DNR;
+	cq->ctrl = ctrl;
+
+	nvmet_cq_init(cq);
+	nvmet_cq_setup(ctrl, cq, qid, size);
+
+	return NVME_SC_SUCCESS;
+}
+EXPORT_SYMBOL_GPL(nvmet_cq_create);
+
+u16 nvmet_check_sqid(struct nvmet_ctrl *ctrl, u16 sqid,
+		     bool create)
+{
+	if (!ctrl->sqs)
+		return NVME_SC_INTERNAL | NVME_STATUS_DNR;
+
+	if (sqid > ctrl->subsys->max_qid)
+		return NVME_SC_QID_INVALID | NVME_STATUS_DNR;
+
+	if ((create && ctrl->sqs[sqid]) ||
+	    (!create && !ctrl->sqs[sqid]))
+		return NVME_SC_QID_INVALID | NVME_STATUS_DNR;
+
+	return NVME_SC_SUCCESS;
+}
+
+u16 nvmet_sq_create(struct nvmet_ctrl *ctrl, struct nvmet_sq *sq,
+		    struct nvmet_cq *cq, u16 sqid, u16 size)
+{
+	u16 status;
+	int ret;
+
+	if (!kref_get_unless_zero(&ctrl->ref))
+		return NVME_SC_INTERNAL | NVME_STATUS_DNR;
+
+	status = nvmet_check_sqid(ctrl, sqid, true);
+	if (status != NVME_SC_SUCCESS)
+		return status;
+
+	ret = nvmet_sq_init(sq, cq);
+	if (ret) {
+		status = NVME_SC_INTERNAL | NVME_STATUS_DNR;
+		goto ctrl_put;
+	}
+
+	nvmet_sq_setup(ctrl, sq, sqid, size);
+	sq->ctrl = ctrl;
+
+	return NVME_SC_SUCCESS;
+
+ctrl_put:
+	nvmet_ctrl_put(ctrl);
+	return status;
+}
+EXPORT_SYMBOL_GPL(nvmet_sq_create);
+
 void nvmet_sq_destroy(struct nvmet_sq *sq)
 {
+	struct nvmet_ctrl *ctrl = sq->ctrl;
+
 	/*
 	 * If this is the admin queue, complete all AERs so that our
 	 * queue doesn't have outstanding requests on it.
 	 */
-	if (sq->ctrl && sq->ctrl->sqs && sq->ctrl->sqs[0] == sq)
-		nvmet_async_events_free(sq->ctrl);
+	if (ctrl && ctrl->sqs && ctrl->sqs[0] == sq)
+		nvmet_async_events_failall(ctrl);
 	percpu_ref_kill_and_confirm(&sq->ref, nvmet_confirm_sq);
 	wait_for_completion(&sq->confirm_done);
 	wait_for_completion(&sq->free_done);
 	percpu_ref_exit(&sq->ref);
+	nvmet_auth_sq_free(sq);
+	nvmet_cq_put(sq->cq);
 
-	if (sq->ctrl) {
-		nvmet_ctrl_put(sq->ctrl);
+	/*
+	 * we must reference the ctrl again after waiting for inflight IO
+	 * to complete. Because admin connect may have sneaked in after we
+	 * store sq->ctrl locally, but before we killed the percpu_ref. the
+	 * admin connect allocates and assigns sq->ctrl, which now needs a
+	 * final ref put, as this ctrl is going away.
+	 */
+	ctrl = sq->ctrl;
+
+	if (ctrl) {
+		/*
+		 * The teardown flow may take some time, and the host may not
+		 * send us keep-alive during this period, hence reset the
+		 * traffic based keep-alive timer so we don't trigger a
+		 * controller teardown as a result of a keep-alive expiration.
+		 */
+		ctrl->reset_tbkas = true;
+		sq->ctrl->sqs[sq->qid] = NULL;
+		nvmet_ctrl_put(ctrl);
 		sq->ctrl = NULL; /* allows reusing the queue later */
 	}
 }
@@ -758,17 +1000,23 @@ static void nvmet_sq_free(struct percpu_ref *ref)
 	complete(&sq->free_done);
 }
 
-int nvmet_sq_init(struct nvmet_sq *sq)
+int nvmet_sq_init(struct nvmet_sq *sq, struct nvmet_cq *cq)
 {
 	int ret;
 
+	if (!nvmet_cq_get(cq))
+		return -EINVAL;
+
 	ret = percpu_ref_init(&sq->ref, nvmet_sq_free, 0, GFP_KERNEL);
 	if (ret) {
 		pr_err("percpu_ref init failed!\n");
+		nvmet_cq_put(cq);
 		return ret;
 	}
 	init_completion(&sq->free_done);
 	init_completion(&sq->confirm_done);
+	nvmet_auth_sq_init(sq);
+	sq->cq = cq;
 
 	return 0;
 }
@@ -803,20 +1051,55 @@ static inline u16 nvmet_io_cmd_check_access(struct nvmet_req *req)
 	return 0;
 }
 
+static u32 nvmet_io_cmd_transfer_len(struct nvmet_req *req)
+{
+	struct nvme_command *cmd = req->cmd;
+	u32 metadata_len = 0;
+
+	if (nvme_is_fabrics(cmd))
+		return nvmet_fabrics_io_cmd_data_len(req);
+
+	if (!req->ns)
+		return 0;
+
+	switch (req->cmd->common.opcode) {
+	case nvme_cmd_read:
+	case nvme_cmd_write:
+	case nvme_cmd_zone_append:
+		if (req->sq->ctrl->pi_support && nvmet_ns_has_pi(req->ns))
+			metadata_len = nvmet_rw_metadata_len(req);
+		return nvmet_rw_data_len(req) + metadata_len;
+	case nvme_cmd_dsm:
+		return nvmet_dsm_len(req);
+	case nvme_cmd_zone_mgmt_recv:
+		return (le32_to_cpu(req->cmd->zmr.numd) + 1) << 2;
+	default:
+		return 0;
+	}
+}
+
 static u16 nvmet_parse_io_cmd(struct nvmet_req *req)
 {
 	struct nvme_command *cmd = req->cmd;
 	u16 ret;
 
-	ret = nvmet_check_ctrl_status(req, cmd);
+	if (nvme_is_fabrics(cmd))
+		return nvmet_parse_fabrics_io_cmd(req);
+
+	if (unlikely(!nvmet_check_auth_status(req)))
+		return NVME_SC_AUTH_REQUIRED | NVME_STATUS_DNR;
+
+	ret = nvmet_check_ctrl_status(req);
+	if (unlikely(ret))
+		return ret;
+
+	if (nvmet_is_passthru_req(req))
+		return nvmet_parse_passthru_io_cmd(req);
+
+	ret = nvmet_req_find_ns(req);
 	if (unlikely(ret))
 		return ret;
 
-	req->ns = nvmet_find_namespace(req->sq->ctrl, cmd->rw.nsid);
-	if (unlikely(!req->ns)) {
-		req->error_loc = offsetof(struct nvme_common_command, nsid);
-		return NVME_SC_INVALID_NS | NVME_SC_DNR;
-	}
 	ret = nvmet_check_ana_state(req->port, req->ns);
 	if (unlikely(ret)) {
 		req->error_loc = offsetof(struct nvme_common_command, nsid);
@@ -828,70 +1111,105 @@ static u16 nvmet_parse_io_cmd(struct nvmet_req *req)
 		return ret;
 	}
 
-	if (req->ns->file)
-		return nvmet_file_parse_io_cmd(req);
-	else
-		return nvmet_bdev_parse_io_cmd(req);
+	if (req->ns->pr.enable) {
+		ret = nvmet_parse_pr_cmd(req);
+		if (!ret)
+			return ret;
+	}
+
+	switch (req->ns->csi) {
+	case NVME_CSI_NVM:
+		if (req->ns->file)
+			ret = nvmet_file_parse_io_cmd(req);
+		else
+			ret = nvmet_bdev_parse_io_cmd(req);
+		break;
+	case NVME_CSI_ZNS:
+		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED))
+			ret = nvmet_bdev_zns_parse_io_cmd(req);
+		else
+			ret = NVME_SC_INVALID_IO_CMD_SET;
+		break;
+	default:
+		ret = NVME_SC_INVALID_IO_CMD_SET;
+	}
+	if (ret)
+		return ret;
+
+	if (req->ns->pr.enable) {
+		ret = nvmet_pr_check_cmd_access(req);
+		if (ret)
+			return ret;
+
+		ret = nvmet_pr_get_ns_pc_ref(req);
+	}
+	return ret;
 }
 
-bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq,
-		struct nvmet_sq *sq, const struct nvmet_fabrics_ops *ops)
+bool nvmet_req_init(struct nvmet_req *req, struct nvmet_sq *sq,
+		const struct nvmet_fabrics_ops *ops)
 {
 	u8 flags = req->cmd->common.flags;
 	u16 status;
 
-	req->cq = cq;
+	req->cq = sq->cq;
 	req->sq = sq;
 	req->ops = ops;
 	req->sg = NULL;
+	req->metadata_sg = NULL;
 	req->sg_cnt = 0;
+	req->metadata_sg_cnt = 0;
 	req->transfer_len = 0;
-	req->rsp->status = 0;
-	req->rsp->sq_head = 0;
+	req->metadata_len = 0;
+	req->cqe->result.u64 = 0;
+	req->cqe->status = 0;
+	req->cqe->sq_head = 0;
 	req->ns = NULL;
 	req->error_loc = NVMET_NO_ERROR_LOC;
 	req->error_slba = 0;
+	req->pc_ref = NULL;
 
 	/* 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);
-		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
+		status = NVME_SC_INVALID_FIELD | NVME_STATUS_DNR;
 		goto fail;
 	}
 
 	/*
 	 * For fabrics, PSDT field shall describe metadata pointer (MPTR) that
 	 * contains an address of a single contiguous physical buffer that is
-	 * byte aligned.
+	 * byte aligned. For PCI controllers, this is optional so not enforced.
 	 */
 	if (unlikely((flags & NVME_CMD_SGL_ALL) != NVME_CMD_SGL_METABUF)) {
-		req->error_loc = offsetof(struct nvme_common_command, flags);
-		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
-		goto fail;
+		if (!req->sq->ctrl || !nvmet_is_pci_ctrl(req->sq->ctrl)) {
+			req->error_loc =
+				offsetof(struct nvme_common_command, flags);
+			status = NVME_SC_INVALID_FIELD | NVME_STATUS_DNR;
+			goto fail;
+		}
 	}
 
 	if (unlikely(!req->sq->ctrl))
-		/* will return an error for any Non-connect command: */
+		/* will return an error for any non-connect command: */
 		status = nvmet_parse_connect_cmd(req);
 	else if (likely(req->sq->qid != 0))
 		status = nvmet_parse_io_cmd(req);
-	else if (req->cmd->common.opcode == nvme_fabrics_command)
-		status = nvmet_parse_fabrics_cmd(req);
-	else if (req->sq->ctrl->subsys->type == NVME_NQN_DISC)
-		status = nvmet_parse_discovery_cmd(req);
 	else
 		status = nvmet_parse_admin_cmd(req);
 
 	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;
+		status = NVME_SC_INVALID_FIELD | NVME_STATUS_DNR;
 		goto fail;
 	}
 
 	if (sq->ctrl)
-		sq->ctrl->cmd_seen = true;
+		sq->ctrl->reset_tbkas = true;
 
 	return true;
 
@@ -904,110 +1222,173 @@ EXPORT_SYMBOL_GPL(nvmet_req_init);
 void nvmet_req_uninit(struct nvmet_req *req)
 {
 	percpu_ref_put(&req->sq->ref);
+	if (req->pc_ref)
+		nvmet_pr_put_ns_pc_ref(req->pc_ref);
 	if (req->ns)
 		nvmet_put_namespace(req->ns);
 }
 EXPORT_SYMBOL_GPL(nvmet_req_uninit);
 
-void nvmet_req_execute(struct nvmet_req *req)
+size_t nvmet_req_transfer_len(struct nvmet_req *req)
 {
-	if (unlikely(req->data_len != req->transfer_len)) {
-		req->error_loc = offsetof(struct nvme_common_command, dptr);
-		nvmet_req_complete(req, NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR);
-	} else
-		req->execute(req);
+	if (likely(req->sq->qid != 0))
+		return nvmet_io_cmd_transfer_len(req);
+	if (unlikely(!req->sq->ctrl))
+		return nvmet_connect_cmd_data_len(req);
+	return nvmet_admin_cmd_data_len(req);
 }
-EXPORT_SYMBOL_GPL(nvmet_req_execute);
+EXPORT_SYMBOL_GPL(nvmet_req_transfer_len);
 
-int nvmet_req_alloc_sgl(struct nvmet_req *req)
+bool nvmet_check_transfer_len(struct nvmet_req *req, size_t len)
 {
-	struct pci_dev *p2p_dev = NULL;
+	if (unlikely(len != req->transfer_len)) {
+		u16 status;
 
-	if (IS_ENABLED(CONFIG_PCI_P2PDMA)) {
-		if (req->sq->ctrl && req->ns)
-			p2p_dev = radix_tree_lookup(&req->sq->ctrl->p2p_ns_map,
-						    req->ns->nsid);
-
-		req->p2p_dev = NULL;
-		if (req->sq->qid && p2p_dev) {
-			req->sg = pci_p2pmem_alloc_sgl(p2p_dev, &req->sg_cnt,
-						       req->transfer_len);
-			if (req->sg) {
-				req->p2p_dev = p2p_dev;
-				return 0;
-			}
-		}
-
-		/*
-		 * If no P2P memory was available we fallback to using
-		 * regular memory
-		 */
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
+		if (req->cmd->common.flags & NVME_CMD_SGL_ALL)
+			status = NVME_SC_SGL_INVALID_DATA;
+		else
+			status = NVME_SC_INVALID_FIELD;
+		nvmet_req_complete(req, status | NVME_STATUS_DNR);
+		return false;
 	}
 
-	req->sg = sgl_alloc(req->transfer_len, GFP_KERNEL, &req->sg_cnt);
-	if (!req->sg)
-		return -ENOMEM;
-
-	return 0;
+	return true;
 }
-EXPORT_SYMBOL_GPL(nvmet_req_alloc_sgl);
+EXPORT_SYMBOL_GPL(nvmet_check_transfer_len);
 
-void nvmet_req_free_sgl(struct nvmet_req *req)
+bool nvmet_check_data_len_lte(struct nvmet_req *req, size_t data_len)
 {
-	if (req->p2p_dev)
-		pci_p2pmem_free_sgl(req->p2p_dev, req->sg);
-	else
-		sgl_free(req->sg);
+	if (unlikely(data_len > req->transfer_len)) {
+		u16 status;
 
-	req->sg = NULL;
-	req->sg_cnt = 0;
-}
-EXPORT_SYMBOL_GPL(nvmet_req_free_sgl);
+		req->error_loc = offsetof(struct nvme_common_command, dptr);
+		if (req->cmd->common.flags & NVME_CMD_SGL_ALL)
+			status = NVME_SC_SGL_INVALID_DATA;
+		else
+			status = NVME_SC_INVALID_FIELD;
+		nvmet_req_complete(req, status | NVME_STATUS_DNR);
+		return false;
+	}
 
-static inline bool nvmet_cc_en(u32 cc)
-{
-	return (cc >> NVME_CC_EN_SHIFT) & 0x1;
+	return true;
 }
 
-static inline u8 nvmet_cc_css(u32 cc)
+static unsigned int nvmet_data_transfer_len(struct nvmet_req *req)
 {
-	return (cc >> NVME_CC_CSS_SHIFT) & 0x7;
+	return req->transfer_len - req->metadata_len;
 }
 
-static inline u8 nvmet_cc_mps(u32 cc)
+static int nvmet_req_alloc_p2pmem_sgls(struct pci_dev *p2p_dev,
+		struct nvmet_req *req)
 {
-	return (cc >> NVME_CC_MPS_SHIFT) & 0xf;
+	req->sg = pci_p2pmem_alloc_sgl(p2p_dev, &req->sg_cnt,
+			nvmet_data_transfer_len(req));
+	if (!req->sg)
+		goto out_err;
+
+	if (req->metadata_len) {
+		req->metadata_sg = pci_p2pmem_alloc_sgl(p2p_dev,
+				&req->metadata_sg_cnt, req->metadata_len);
+		if (!req->metadata_sg)
+			goto out_free_sg;
+	}
+
+	req->p2p_dev = p2p_dev;
+
+	return 0;
+out_free_sg:
+	pci_p2pmem_free_sgl(req->p2p_dev, req->sg);
+out_err:
+	return -ENOMEM;
 }
 
-static inline u8 nvmet_cc_ams(u32 cc)
+static struct pci_dev *nvmet_req_find_p2p_dev(struct nvmet_req *req)
 {
-	return (cc >> NVME_CC_AMS_SHIFT) & 0x7;
+	if (!IS_ENABLED(CONFIG_PCI_P2PDMA) ||
+	    !req->sq->ctrl || !req->sq->qid || !req->ns)
+		return NULL;
+	return radix_tree_lookup(&req->sq->ctrl->p2p_ns_map, req->ns->nsid);
 }
 
-static inline u8 nvmet_cc_shn(u32 cc)
+int nvmet_req_alloc_sgls(struct nvmet_req *req)
 {
-	return (cc >> NVME_CC_SHN_SHIFT) & 0x3;
+	struct pci_dev *p2p_dev = nvmet_req_find_p2p_dev(req);
+
+	if (p2p_dev && !nvmet_req_alloc_p2pmem_sgls(p2p_dev, req))
+		return 0;
+
+	req->sg = sgl_alloc(nvmet_data_transfer_len(req), GFP_KERNEL,
+			    &req->sg_cnt);
+	if (unlikely(!req->sg))
+		goto out;
+
+	if (req->metadata_len) {
+		req->metadata_sg = sgl_alloc(req->metadata_len, GFP_KERNEL,
+					     &req->metadata_sg_cnt);
+		if (unlikely(!req->metadata_sg))
+			goto out_free;
+	}
+
+	return 0;
+out_free:
+	sgl_free(req->sg);
+out:
+	return -ENOMEM;
 }
+EXPORT_SYMBOL_GPL(nvmet_req_alloc_sgls);
 
-static inline u8 nvmet_cc_iosqes(u32 cc)
+void nvmet_req_free_sgls(struct nvmet_req *req)
 {
-	return (cc >> NVME_CC_IOSQES_SHIFT) & 0xf;
+	if (req->p2p_dev) {
+		pci_p2pmem_free_sgl(req->p2p_dev, req->sg);
+		if (req->metadata_sg)
+			pci_p2pmem_free_sgl(req->p2p_dev, req->metadata_sg);
+		req->p2p_dev = NULL;
+	} else {
+		sgl_free(req->sg);
+		if (req->metadata_sg)
+			sgl_free(req->metadata_sg);
+	}
+
+	req->sg = NULL;
+	req->metadata_sg = NULL;
+	req->sg_cnt = 0;
+	req->metadata_sg_cnt = 0;
 }
+EXPORT_SYMBOL_GPL(nvmet_req_free_sgls);
 
-static inline u8 nvmet_cc_iocqes(u32 cc)
+static inline bool nvmet_css_supported(u8 cc_css)
 {
-	return (cc >> NVME_CC_IOCQES_SHIFT) & 0xf;
+	switch (cc_css << NVME_CC_CSS_SHIFT) {
+	case NVME_CC_CSS_NVM:
+	case NVME_CC_CSS_CSI:
+		return true;
+	default:
+		return false;
+	}
 }
 
 static void nvmet_start_ctrl(struct nvmet_ctrl *ctrl)
 {
 	lockdep_assert_held(&ctrl->lock);
 
-	if (nvmet_cc_iosqes(ctrl->cc) != NVME_NVM_IOSQES ||
-	    nvmet_cc_iocqes(ctrl->cc) != NVME_NVM_IOCQES ||
-	    nvmet_cc_mps(ctrl->cc) != 0 ||
+	/*
+	 * Only I/O controllers should verify iosqes,iocqes.
+	 * Strictly speaking, the spec says a discovery controller
+	 * should verify iosqes,iocqes are zeroed, however that
+	 * would break backwards compatibility, so don't enforce it.
+	 */
+	if (!nvmet_is_disc_subsys(ctrl->subsys) &&
+	    (nvmet_cc_iosqes(ctrl->cc) != NVME_NVM_IOSQES ||
+	     nvmet_cc_iocqes(ctrl->cc) != NVME_NVM_IOCQES)) {
+		ctrl->csts = NVME_CSTS_CFS;
+		return;
+	}
+
+	if (nvmet_cc_mps(ctrl->cc) != 0 ||
 	    nvmet_cc_ams(ctrl->cc) != 0 ||
-	    nvmet_cc_css(ctrl->cc) != 0) {
+	    !nvmet_css_supported(nvmet_cc_css(ctrl->cc))) {
 		ctrl->csts = NVME_CSTS_CFS;
 		return;
 	}
@@ -1020,7 +1401,8 @@ static void nvmet_start_ctrl(struct nvmet_ctrl *ctrl)
 	 * in case a host died before it enabled the controller.  Hence, simply
 	 * reset the keep alive timer when the controller is enabled.
 	 */
-	mod_delayed_work(system_wq, &ctrl->ka_work, ctrl->kato * HZ);
+	if (ctrl->kato)
+		mod_delayed_work(nvmet_wq, &ctrl->ka_work, ctrl->kato * HZ);
 }
 
 static void nvmet_clear_ctrl(struct nvmet_ctrl *ctrl)
@@ -1052,30 +1434,40 @@ void nvmet_update_cc(struct nvmet_ctrl *ctrl, u32 new)
 		ctrl->csts &= ~NVME_CSTS_SHST_CMPLT;
 	mutex_unlock(&ctrl->lock);
 }
+EXPORT_SYMBOL_GPL(nvmet_update_cc);
 
 static void nvmet_init_cap(struct nvmet_ctrl *ctrl)
 {
 	/* command sets supported: NVMe command set: */
 	ctrl->cap = (1ULL << 37);
+	/* Controller supports one or more I/O Command Sets */
+	ctrl->cap |= (1ULL << 43);
 	/* CC.EN timeout in 500msec units: */
 	ctrl->cap |= (15ULL << 24);
 	/* maximum queue entries supported: */
-	ctrl->cap |= NVMET_QUEUE_SIZE - 1;
+	if (ctrl->ops->get_max_queue_size)
+		ctrl->cap |= min_t(u16, ctrl->ops->get_max_queue_size(ctrl),
+				   ctrl->port->max_queue_size) - 1;
+	else
+		ctrl->cap |= ctrl->port->max_queue_size - 1;
+
+	if (nvmet_is_passthru_subsys(ctrl->subsys))
+		nvmet_passthrough_override_cap(ctrl);
 }
 
-u16 nvmet_ctrl_find_get(const char *subsysnqn, const char *hostnqn, u16 cntlid,
-		struct nvmet_req *req, struct nvmet_ctrl **ret)
+struct nvmet_ctrl *nvmet_ctrl_find_get(const char *subsysnqn,
+				       const char *hostnqn, u16 cntlid,
+				       struct nvmet_req *req)
 {
+	struct nvmet_ctrl *ctrl = NULL;
 	struct nvmet_subsys *subsys;
-	struct nvmet_ctrl *ctrl;
-	u16 status = 0;
 
 	subsys = nvmet_find_get_subsys(req->port, subsysnqn);
 	if (!subsys) {
 		pr_warn("connect request for invalid subsystem %s!\n",
 			subsysnqn);
-		req->rsp->result.u32 = IPO_IATTR_CONNECT_DATA(subsysnqn);
-		return NVME_SC_CONNECT_INVALID_PARAM | NVME_SC_DNR;
+		req->cqe->result.u32 = IPO_IATTR_CONNECT_DATA(subsysnqn);
+		goto out;
 	}
 
 	mutex_lock(&subsys->lock);
@@ -1088,34 +1480,40 @@ u16 nvmet_ctrl_find_get(const char *subsysnqn, const char *hostnqn, u16 cntlid,
 			if (!kref_get_unless_zero(&ctrl->ref))
 				continue;
 
-			*ret = ctrl;
-			goto out;
+			/* ctrl found */
+			goto found;
 		}
 	}
 
+	ctrl = NULL; /* ctrl not found */
 	pr_warn("could not find controller %d for subsys %s / host %s\n",
 		cntlid, subsysnqn, hostnqn);
-	req->rsp->result.u32 = IPO_IATTR_CONNECT_DATA(cntlid);
-	status = NVME_SC_CONNECT_INVALID_PARAM | NVME_SC_DNR;
+	req->cqe->result.u32 = IPO_IATTR_CONNECT_DATA(cntlid);
 
-out:
+found:
 	mutex_unlock(&subsys->lock);
 	nvmet_subsys_put(subsys);
-	return status;
+out:
+	return ctrl;
 }
 
-u16 nvmet_check_ctrl_status(struct nvmet_req *req, struct nvme_command *cmd)
+u16 nvmet_check_ctrl_status(struct nvmet_req *req)
 {
 	if (unlikely(!(req->sq->ctrl->cc & NVME_CC_ENABLE))) {
 		pr_err("got cmd %d while CC.EN == 0 on qid = %d\n",
-		       cmd->common.opcode, req->sq->qid);
-		return NVME_SC_CMD_SEQ_ERROR | NVME_SC_DNR;
+		       req->cmd->common.opcode, req->sq->qid);
+		return NVME_SC_CMD_SEQ_ERROR | NVME_STATUS_DNR;
 	}
 
 	if (unlikely(!(req->sq->ctrl->csts & NVME_CSTS_RDY))) {
 		pr_err("got cmd %d while CSTS.RDY == 0 on qid = %d\n",
-		       cmd->common.opcode, req->sq->qid);
-		return NVME_SC_CMD_SEQ_ERROR | NVME_SC_DNR;
+		       req->cmd->common.opcode, req->sq->qid);
+		return NVME_SC_CMD_SEQ_ERROR | NVME_STATUS_DNR;
+	}
+
+	if (unlikely(!nvmet_check_auth_status(req))) {
+		pr_warn("qid %d not authenticated\n", req->sq->qid);
+		return NVME_SC_AUTH_REQUIRED | NVME_STATUS_DNR;
 	}
 	return 0;
 }
@@ -1129,7 +1527,7 @@ bool nvmet_host_allowed(struct nvmet_subsys *subsys, const char *hostnqn)
 	if (subsys->allow_any_host)
 		return true;
 
-	if (subsys->type == NVME_NQN_DISC) /* allow all access to disc subsys */
+	if (nvmet_is_disc_subsys(subsys)) /* allow all access to disc subsys */
 		return true;
 
 	list_for_each_entry(p, &subsys->hosts, entry) {
@@ -1140,85 +1538,102 @@ bool nvmet_host_allowed(struct nvmet_subsys *subsys, const char *hostnqn)
 	return false;
 }
 
-/*
- * Note: ctrl->subsys->lock should be held when calling this function
- */
 static void nvmet_setup_p2p_ns_map(struct nvmet_ctrl *ctrl,
-		struct nvmet_req *req)
+		struct device *p2p_client)
 {
 	struct nvmet_ns *ns;
+	unsigned long idx;
 
-	if (!req->p2p_client)
+	lockdep_assert_held(&ctrl->subsys->lock);
+
+	if (!p2p_client)
 		return;
 
-	ctrl->p2p_client = get_device(req->p2p_client);
+	ctrl->p2p_client = get_device(p2p_client);
 
-	list_for_each_entry_rcu(ns, &ctrl->subsys->namespaces, dev_link)
+	nvmet_for_each_enabled_ns(&ctrl->subsys->namespaces, idx, ns)
 		nvmet_p2pmem_ns_add_p2p(ctrl, ns);
 }
 
-/*
- * Note: ctrl->subsys->lock should be held when calling this function
- */
 static void nvmet_release_p2p_ns_map(struct nvmet_ctrl *ctrl)
 {
 	struct radix_tree_iter iter;
 	void __rcu **slot;
 
+	lockdep_assert_held(&ctrl->subsys->lock);
+
 	radix_tree_for_each_slot(slot, &ctrl->p2p_ns_map, &iter, 0)
 		pci_dev_put(radix_tree_deref_slot(slot));
 
 	put_device(ctrl->p2p_client);
 }
 
-u16 nvmet_alloc_ctrl(const char *subsysnqn, const char *hostnqn,
-		struct nvmet_req *req, u32 kato, struct nvmet_ctrl **ctrlp)
+static void nvmet_fatal_error_handler(struct work_struct *work)
+{
+	struct nvmet_ctrl *ctrl =
+			container_of(work, struct nvmet_ctrl, fatal_err_work);
+
+	pr_err("ctrl %d fatal error occurred!\n", ctrl->cntlid);
+	ctrl->ops->delete_ctrl(ctrl);
+}
+
+struct nvmet_ctrl *nvmet_alloc_ctrl(struct nvmet_alloc_ctrl_args *args)
 {
 	struct nvmet_subsys *subsys;
 	struct nvmet_ctrl *ctrl;
+	u32 kato = args->kato;
+	u8 dhchap_status;
 	int ret;
-	u16 status;
 
-	status = NVME_SC_CONNECT_INVALID_PARAM | NVME_SC_DNR;
-	subsys = nvmet_find_get_subsys(req->port, subsysnqn);
+	args->status = NVME_SC_CONNECT_INVALID_PARAM | NVME_STATUS_DNR;
+	subsys = nvmet_find_get_subsys(args->port, args->subsysnqn);
 	if (!subsys) {
 		pr_warn("connect request for invalid subsystem %s!\n",
-			subsysnqn);
-		req->rsp->result.u32 = IPO_IATTR_CONNECT_DATA(subsysnqn);
-		goto out;
+			args->subsysnqn);
+		args->result = IPO_IATTR_CONNECT_DATA(subsysnqn);
+		args->error_loc = offsetof(struct nvme_common_command, dptr);
+		return NULL;
 	}
 
-	status = NVME_SC_CONNECT_INVALID_PARAM | NVME_SC_DNR;
 	down_read(&nvmet_config_sem);
-	if (!nvmet_host_allowed(subsys, hostnqn)) {
+	if (!nvmet_host_allowed(subsys, args->hostnqn)) {
 		pr_info("connect by host %s for subsystem %s not allowed\n",
-			hostnqn, subsysnqn);
-		req->rsp->result.u32 = IPO_IATTR_CONNECT_DATA(hostnqn);
+			args->hostnqn, args->subsysnqn);
+		args->result = IPO_IATTR_CONNECT_DATA(hostnqn);
 		up_read(&nvmet_config_sem);
-		status = NVME_SC_CONNECT_INVALID_HOST | NVME_SC_DNR;
+		args->status = NVME_SC_CONNECT_INVALID_HOST | NVME_STATUS_DNR;
+		args->error_loc = offsetof(struct nvme_common_command, dptr);
 		goto out_put_subsystem;
 	}
 	up_read(&nvmet_config_sem);
 
-	status = NVME_SC_INTERNAL;
+	args->status = NVME_SC_INTERNAL;
 	ctrl = kzalloc(sizeof(*ctrl), GFP_KERNEL);
 	if (!ctrl)
 		goto out_put_subsystem;
 	mutex_init(&ctrl->lock);
 
-	nvmet_init_cap(ctrl);
+	ctrl->port = args->port;
+	ctrl->ops = args->ops;
 
-	ctrl->port = req->port;
+#ifdef CONFIG_NVME_TARGET_PASSTHRU
+	/* By default, set loop targets to clear IDS by default */
+	if (ctrl->port->disc_addr.trtype == NVMF_TRTYPE_LOOP)
+		subsys->clear_ids = 1;
+#endif
 
 	INIT_WORK(&ctrl->async_event_work, nvmet_async_event_work);
 	INIT_LIST_HEAD(&ctrl->async_events);
 	INIT_RADIX_TREE(&ctrl->p2p_ns_map, GFP_KERNEL);
+	INIT_WORK(&ctrl->fatal_err_work, nvmet_fatal_error_handler);
+	INIT_DELAYED_WORK(&ctrl->ka_work, nvmet_keep_alive_timer);
 
-	memcpy(ctrl->subsysnqn, subsysnqn, NVMF_NQN_SIZE);
-	memcpy(ctrl->hostnqn, hostnqn, NVMF_NQN_SIZE);
+	memcpy(ctrl->hostnqn, args->hostnqn, NVMF_NQN_SIZE);
 
 	kref_init(&ctrl->ref);
 	ctrl->subsys = subsys;
+	ctrl->pi_support = ctrl->port->pi_enable && ctrl->subsys->pi_support;
+	nvmet_init_cap(ctrl);
 	WRITE_ONCE(ctrl->aen_enabled, NVMET_AEN_CFG_OPTIONAL);
 
 	ctrl->changed_ns_list = kmalloc_array(NVME_MAX_CHANGED_NAMESPACES,
@@ -1226,34 +1641,31 @@ u16 nvmet_alloc_ctrl(const char *subsysnqn, const char *hostnqn,
 	if (!ctrl->changed_ns_list)
 		goto out_free_ctrl;
 
-	ctrl->cqs = kcalloc(subsys->max_qid + 1,
-			sizeof(struct nvmet_cq *),
-			GFP_KERNEL);
-	if (!ctrl->cqs)
-		goto out_free_changed_ns_list;
-
 	ctrl->sqs = kcalloc(subsys->max_qid + 1,
 			sizeof(struct nvmet_sq *),
 			GFP_KERNEL);
 	if (!ctrl->sqs)
-		goto out_free_cqs;
+		goto out_free_changed_ns_list;
 
-	ret = ida_simple_get(&cntlid_ida,
-			     NVME_CNTLID_MIN, NVME_CNTLID_MAX,
+	ctrl->cqs = kcalloc(subsys->max_qid + 1, sizeof(struct nvmet_cq *),
+			   GFP_KERNEL);
+	if (!ctrl->cqs)
+		goto out_free_sqs;
+
+	ret = ida_alloc_range(&cntlid_ida,
+			     subsys->cntlid_min, subsys->cntlid_max,
 			     GFP_KERNEL);
 	if (ret < 0) {
-		status = NVME_SC_CONNECT_CTRL_BUSY | NVME_SC_DNR;
-		goto out_free_sqs;
+		args->status = NVME_SC_CONNECT_CTRL_BUSY | NVME_STATUS_DNR;
+		goto out_free_cqs;
 	}
 	ctrl->cntlid = ret;
 
-	ctrl->ops = req->ops;
-
 	/*
 	 * Discovery controllers may use some arbitrary high value
 	 * in order to cleanup stale discovery sessions
 	 */
-	if ((ctrl->subsys->type == NVME_NQN_DISC) && !kato)
+	if (nvmet_is_disc_subsys(ctrl->subsys) && !kato)
 		kato = NVMET_DISC_KATO_MS;
 
 	/* keep-alive timeout in seconds */
@@ -1265,26 +1677,58 @@ u16 nvmet_alloc_ctrl(const char *subsysnqn, const char *hostnqn,
 	nvmet_start_keep_alive_timer(ctrl);
 
 	mutex_lock(&subsys->lock);
+	ret = nvmet_ctrl_init_pr(ctrl);
+	if (ret)
+		goto init_pr_fail;
 	list_add_tail(&ctrl->subsys_entry, &subsys->ctrls);
-	nvmet_setup_p2p_ns_map(ctrl, req);
+	nvmet_setup_p2p_ns_map(ctrl, args->p2p_client);
+	nvmet_debugfs_ctrl_setup(ctrl);
 	mutex_unlock(&subsys->lock);
 
-	*ctrlp = ctrl;
-	return 0;
+	if (args->hostid)
+		uuid_copy(&ctrl->hostid, args->hostid);
+
+	dhchap_status = nvmet_setup_auth(ctrl, args->sq);
+	if (dhchap_status) {
+		pr_err("Failed to setup authentication, dhchap status %u\n",
+		       dhchap_status);
+		nvmet_ctrl_put(ctrl);
+		if (dhchap_status == NVME_AUTH_DHCHAP_FAILURE_FAILED)
+			args->status =
+				NVME_SC_CONNECT_INVALID_HOST | NVME_STATUS_DNR;
+		else
+			args->status = NVME_SC_INTERNAL;
+		return NULL;
+	}
 
-out_free_sqs:
-	kfree(ctrl->sqs);
+	args->status = NVME_SC_SUCCESS;
+
+	pr_info("Created %s controller %d for subsystem %s for NQN %s%s%s%s.\n",
+		nvmet_is_disc_subsys(ctrl->subsys) ? "discovery" : "nvm",
+		ctrl->cntlid, ctrl->subsys->subsysnqn, ctrl->hostnqn,
+		ctrl->pi_support ? " T10-PI is enabled" : "",
+		nvmet_has_auth(ctrl, args->sq) ? " with DH-HMAC-CHAP" : "",
+		nvmet_queue_tls_keyid(args->sq) ? ", TLS" : "");
+
+	return ctrl;
+
+init_pr_fail:
+	mutex_unlock(&subsys->lock);
+	nvmet_stop_keep_alive_timer(ctrl);
+	ida_free(&cntlid_ida, ctrl->cntlid);
 out_free_cqs:
 	kfree(ctrl->cqs);
+out_free_sqs:
+	kfree(ctrl->sqs);
 out_free_changed_ns_list:
 	kfree(ctrl->changed_ns_list);
 out_free_ctrl:
 	kfree(ctrl);
 out_put_subsystem:
 	nvmet_subsys_put(subsys);
-out:
-	return status;
+	return NULL;
 }
+EXPORT_SYMBOL_GPL(nvmet_alloc_ctrl);
 
 static void nvmet_ctrl_free(struct kref *ref)
 {
@@ -1292,6 +1736,7 @@ static void nvmet_ctrl_free(struct kref *ref)
 	struct nvmet_subsys *subsys = ctrl->subsys;
 
 	mutex_lock(&subsys->lock);
+	nvmet_ctrl_destroy_pr(ctrl);
 	nvmet_release_p2p_ns_map(ctrl);
 	list_del(&ctrl->subsys_entry);
 	mutex_unlock(&subsys->lock);
@@ -1301,8 +1746,13 @@ static void nvmet_ctrl_free(struct kref *ref)
 	flush_work(&ctrl->async_event_work);
 	cancel_work_sync(&ctrl->fatal_err_work);
 
-	ida_simple_remove(&cntlid_ida, ctrl->cntlid);
+	nvmet_destroy_auth(ctrl);
+
+	nvmet_debugfs_ctrl_free(ctrl);
 
+	ida_free(&cntlid_ida, ctrl->cntlid);
+
+	nvmet_async_events_free(ctrl);
 	kfree(ctrl->sqs);
 	kfree(ctrl->cqs);
 	kfree(ctrl->changed_ns_list);
@@ -1315,28 +1765,27 @@ void nvmet_ctrl_put(struct nvmet_ctrl *ctrl)
 {
 	kref_put(&ctrl->ref, nvmet_ctrl_free);
 }
-
-static void nvmet_fatal_error_handler(struct work_struct *work)
-{
-	struct nvmet_ctrl *ctrl =
-			container_of(work, struct nvmet_ctrl, fatal_err_work);
-
-	pr_err("ctrl %d fatal error occurred!\n", ctrl->cntlid);
-	ctrl->ops->delete_ctrl(ctrl);
-}
+EXPORT_SYMBOL_GPL(nvmet_ctrl_put);
 
 void nvmet_ctrl_fatal_error(struct nvmet_ctrl *ctrl)
 {
 	mutex_lock(&ctrl->lock);
 	if (!(ctrl->csts & NVME_CSTS_CFS)) {
 		ctrl->csts |= NVME_CSTS_CFS;
-		INIT_WORK(&ctrl->fatal_err_work, nvmet_fatal_error_handler);
-		schedule_work(&ctrl->fatal_err_work);
+		queue_work(nvmet_wq, &ctrl->fatal_err_work);
 	}
 	mutex_unlock(&ctrl->lock);
 }
 EXPORT_SYMBOL_GPL(nvmet_ctrl_fatal_error);
 
+ssize_t nvmet_ctrl_host_traddr(struct nvmet_ctrl *ctrl,
+		char *traddr, size_t traddr_len)
+{
+	if (!ctrl->ops->host_traddr)
+		return -EOPNOTSUPP;
+	return ctrl->ops->host_traddr(ctrl, traddr, traddr_len);
+}
+
 static struct nvmet_subsys *nvmet_find_get_subsys(struct nvmet_port *port,
 		const char *subsysnqn)
 {
@@ -1352,6 +1801,13 @@ static struct nvmet_subsys *nvmet_find_get_subsys(struct nvmet_port *port,
 	}
 
 	down_read(&nvmet_config_sem);
+	if (!strncmp(nvmet_disc_subsys->subsysnqn, subsysnqn,
+				NVMF_NQN_SIZE)) {
+		if (kref_get_unless_zero(&nvmet_disc_subsys->ref)) {
+			up_read(&nvmet_config_sem);
+			return nvmet_disc_subsys;
+		}
+	}
 	list_for_each_entry(p, &port->subsystems, entry) {
 		if (!strncmp(p->subsys->subsysnqn, subsysnqn,
 				NVMF_NQN_SIZE)) {
@@ -1369,43 +1825,76 @@ struct nvmet_subsys *nvmet_subsys_alloc(const char *subsysnqn,
 		enum nvme_subsys_type type)
 {
 	struct nvmet_subsys *subsys;
+	char serial[NVMET_SN_MAX_SIZE / 2];
+	int ret;
 
 	subsys = kzalloc(sizeof(*subsys), GFP_KERNEL);
 	if (!subsys)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
-	subsys->ver = NVME_VS(1, 3, 0); /* NVMe 1.3.0 */
+	subsys->ver = NVMET_DEFAULT_VS;
 	/* generate a random serial number as our controllers are ephemeral: */
-	get_random_bytes(&subsys->serial, sizeof(subsys->serial));
+	get_random_bytes(&serial, sizeof(serial));
+	bin2hex(subsys->serial, &serial, sizeof(serial));
+
+	subsys->model_number = kstrdup(NVMET_DEFAULT_CTRL_MODEL, GFP_KERNEL);
+	if (!subsys->model_number) {
+		ret = -ENOMEM;
+		goto free_subsys;
+	}
+
+	subsys->ieee_oui = 0;
+
+	subsys->firmware_rev = kstrndup(UTS_RELEASE, NVMET_FR_MAX_SIZE, GFP_KERNEL);
+	if (!subsys->firmware_rev) {
+		ret = -ENOMEM;
+		goto free_mn;
+	}
 
 	switch (type) {
 	case NVME_NQN_NVME:
 		subsys->max_qid = NVMET_NR_QUEUES;
 		break;
 	case NVME_NQN_DISC:
+	case NVME_NQN_CURR:
 		subsys->max_qid = 0;
 		break;
 	default:
 		pr_err("%s: Unknown Subsystem type - %d\n", __func__, type);
-		kfree(subsys);
-		return NULL;
+		ret = -EINVAL;
+		goto free_fr;
 	}
 	subsys->type = type;
 	subsys->subsysnqn = kstrndup(subsysnqn, NVMF_NQN_SIZE,
 			GFP_KERNEL);
 	if (!subsys->subsysnqn) {
-		kfree(subsys);
-		return NULL;
+		ret = -ENOMEM;
+		goto free_fr;
 	}
-
+	subsys->cntlid_min = NVME_CNTLID_MIN;
+	subsys->cntlid_max = NVME_CNTLID_MAX;
 	kref_init(&subsys->ref);
 
 	mutex_init(&subsys->lock);
-	INIT_LIST_HEAD(&subsys->namespaces);
+	xa_init(&subsys->namespaces);
 	INIT_LIST_HEAD(&subsys->ctrls);
 	INIT_LIST_HEAD(&subsys->hosts);
 
+	ret = nvmet_debugfs_subsys_setup(subsys);
+	if (ret)
+		goto free_subsysnqn;
+
 	return subsys;
+
+free_subsysnqn:
+	kfree(subsys->subsysnqn);
+free_fr:
+	kfree(subsys->firmware_rev);
+free_mn:
+	kfree(subsys->model_number);
+free_subsys:
+	kfree(subsys);
+	return ERR_PTR(ret);
 }
 
 static void nvmet_subsys_free(struct kref *ref)
@@ -1413,9 +1902,18 @@ static void nvmet_subsys_free(struct kref *ref)
 	struct nvmet_subsys *subsys =
 		container_of(ref, struct nvmet_subsys, ref);
 
-	WARN_ON_ONCE(!list_empty(&subsys->namespaces));
+	WARN_ON_ONCE(!list_empty(&subsys->ctrls));
+	WARN_ON_ONCE(!list_empty(&subsys->hosts));
+	WARN_ON_ONCE(!xa_empty(&subsys->namespaces));
+
+	nvmet_debugfs_subsys_free(subsys);
+
+	xa_destroy(&subsys->namespaces);
+	nvmet_passthru_subsys_free(subsys);
 
 	kfree(subsys->subsysnqn);
+	kfree(subsys->model_number);
+	kfree(subsys->firmware_rev);
 	kfree(subsys);
 }
 
@@ -1436,31 +1934,56 @@ void nvmet_subsys_put(struct nvmet_subsys *subsys)
 
 static int __init nvmet_init(void)
 {
-	int error;
+	int error = -ENOMEM;
 
 	nvmet_ana_group_enabled[NVMET_DEFAULT_ANA_GRPID] = 1;
 
+	nvmet_bvec_cache = kmem_cache_create("nvmet-bvec",
+			NVMET_MAX_MPOOL_BVEC * sizeof(struct bio_vec), 0,
+			SLAB_HWCACHE_ALIGN, NULL);
+	if (!nvmet_bvec_cache)
+		return -ENOMEM;
+
+	zbd_wq = alloc_workqueue("nvmet-zbd-wq", WQ_MEM_RECLAIM, 0);
+	if (!zbd_wq)
+		goto out_destroy_bvec_cache;
+
 	buffered_io_wq = alloc_workqueue("nvmet-buffered-io-wq",
 			WQ_MEM_RECLAIM, 0);
-	if (!buffered_io_wq) {
-		error = -ENOMEM;
-		goto out;
-	}
+	if (!buffered_io_wq)
+		goto out_free_zbd_work_queue;
+
+	nvmet_wq = alloc_workqueue("nvmet-wq",
+			WQ_MEM_RECLAIM | WQ_UNBOUND | WQ_SYSFS, 0);
+	if (!nvmet_wq)
+		goto out_free_buffered_work_queue;
+
+	error = nvmet_init_debugfs();
+	if (error)
+		goto out_free_nvmet_work_queue;
 
 	error = nvmet_init_discovery();
 	if (error)
-		goto out_free_work_queue;
+		goto out_exit_debugfs;
 
 	error = nvmet_init_configfs();
 	if (error)
 		goto out_exit_discovery;
+
 	return 0;
 
 out_exit_discovery:
 	nvmet_exit_discovery();
-out_free_work_queue:
+out_exit_debugfs:
+	nvmet_exit_debugfs();
+out_free_nvmet_work_queue:
+	destroy_workqueue(nvmet_wq);
+out_free_buffered_work_queue:
 	destroy_workqueue(buffered_io_wq);
-out:
+out_free_zbd_work_queue:
+	destroy_workqueue(zbd_wq);
+out_destroy_bvec_cache:
+	kmem_cache_destroy(nvmet_bvec_cache);
 	return error;
 }
 
@@ -1468,8 +1991,12 @@ static void __exit nvmet_exit(void)
 {
 	nvmet_exit_configfs();
 	nvmet_exit_discovery();
+	nvmet_exit_debugfs();
 	ida_destroy(&cntlid_ida);
+	destroy_workqueue(nvmet_wq);
 	destroy_workqueue(buffered_io_wq);
+	destroy_workqueue(zbd_wq);
+	kmem_cache_destroy(nvmet_bvec_cache);
 
 	BUILD_BUG_ON(sizeof(struct nvmf_disc_rsp_page_entry) != 1024);
 	BUILD_BUG_ON(sizeof(struct nvmf_disc_rsp_page_hdr) != 1024);
@@ -1478,4 +2005,5 @@ static void __exit nvmet_exit(void)
 module_init(nvmet_init);
 module_exit(nvmet_exit);
 
+MODULE_DESCRIPTION("NVMe target core framework");
 MODULE_LICENSE("GPL v2");