Merge tag 'for-4.19/block-20180812' of git://git.kernel.dk/linux-block

Pull block updates from Jens Axboe:
 "First pull request for this merge window, there will also be a
  followup request with some stragglers.

  This pull request contains:

   - Fix for a thundering heard issue in the wbt block code (Anchal
     Agarwal)

   - A few NVMe pull requests:
      * Improved tracepoints (Keith)
      * Larger inline data support for RDMA (Steve Wise)
      * RDMA setup/teardown fixes (Sagi)
      * Effects log suppor for NVMe target (Chaitanya Kulkarni)
      * Buffered IO suppor for NVMe target (Chaitanya Kulkarni)
      * TP4004 (ANA) support (Christoph)
      * Various NVMe fixes

   - Block io-latency controller support. Much needed support for
     properly containing block devices. (Josef)

   - Series improving how we handle sense information on the stack
     (Kees)

   - Lightnvm fixes and updates/improvements (Mathias/Javier et al)

   - Zoned device support for null_blk (Matias)

   - AIX partition fixes (Mauricio Faria de Oliveira)

   - DIF checksum code made generic (Max Gurtovoy)

   - Add support for discard in iostats (Michael Callahan / Tejun)

   - Set of updates for BFQ (Paolo)

   - Removal of async write support for bsg (Christoph)

   - Bio page dirtying and clone fixups (Christoph)

   - Set of bcache fix/changes (via Coly)

   - Series improving blk-mq queue setup/teardown speed (Ming)

   - Series improving merging performance on blk-mq (Ming)

   - Lots of other fixes and cleanups from a slew of folks"

* tag 'for-4.19/block-20180812' of git://git.kernel.dk/linux-block: (190 commits)
  blkcg: Make blkg_root_lookup() work for queues in bypass mode
  bcache: fix error setting writeback_rate through sysfs interface
  null_blk: add lock drop/acquire annotation
  Blk-throttle: reduce tail io latency when iops limit is enforced
  block: paride: pd: mark expected switch fall-throughs
  block: Ensure that a request queue is dissociated from the cgroup controller
  block: Introduce blk_exit_queue()
  blkcg: Introduce blkg_root_lookup()
  block: Remove two superfluous #include directives
  blk-mq: count the hctx as active before allocating tag
  block: bvec_nr_vecs() returns value for wrong slab
  bcache: trivial - remove tailing backslash in macro BTREE_FLAG
  bcache: make the pr_err statement used for ENOENT only in sysfs_attatch section
  bcache: set max writeback rate when I/O request is idle
  bcache: add code comments for bset.c
  bcache: fix mistaken comments in request.c
  bcache: fix mistaken code comments in bcache.h
  bcache: add a comment in super.c
  bcache: avoid unncessary cache prefetch bch_btree_node_get()
  bcache: display rate debug parameters to 0 when writeback is not running
  ...

drivers/nvme/host/core.c

@@ -252,7 +252,8 @@ void nvme_complete_rq(struct request *req)
 	trace_nvme_complete_rq(req);
 
 	if (unlikely(status != BLK_STS_OK && nvme_req_needs_retry(req))) {
-		if (nvme_req_needs_failover(req, status)) {
+		if ((req->cmd_flags & REQ_NVME_MPATH) &&
+		    blk_path_error(status)) {
 			nvme_failover_req(req);
 			return;
 		}
@@ -617,6 +618,8 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
 			if (WARN_ON_ONCE(!nvme_ns_has_pi(ns)))
 				return BLK_STS_NOTSUPP;
 			control |= NVME_RW_PRINFO_PRACT;
+		} else if (req_op(req) == REQ_OP_WRITE) {
+			t10_pi_prepare(req, ns->pi_type);
 		}
 
 		switch (ns->pi_type) {
@@ -627,8 +630,7 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
 		case NVME_NS_DPS_PI_TYPE2:
 			control |= NVME_RW_PRINFO_PRCHK_GUARD |
 					NVME_RW_PRINFO_PRCHK_REF;
-			cmnd->rw.reftag = cpu_to_le32(
-					nvme_block_nr(ns, blk_rq_pos(req)));
+			cmnd->rw.reftag = cpu_to_le32(t10_pi_ref_tag(req));
 			break;
 		}
 	}
@@ -638,6 +640,22 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
 	return 0;
 }
 
+void nvme_cleanup_cmd(struct request *req)
+{
+	if (blk_integrity_rq(req) && req_op(req) == REQ_OP_READ &&
+	    nvme_req(req)->status == 0) {
+		struct nvme_ns *ns = req->rq_disk->private_data;
+
+		t10_pi_complete(req, ns->pi_type,
+				blk_rq_bytes(req) >> ns->lba_shift);
+	}
+	if (req->rq_flags & RQF_SPECIAL_PAYLOAD) {
+		kfree(page_address(req->special_vec.bv_page) +
+		      req->special_vec.bv_offset);
+	}
+}
+EXPORT_SYMBOL_GPL(nvme_cleanup_cmd);
+
 blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
 		struct nvme_command *cmd)
 {
@@ -668,10 +686,7 @@ blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
 	}
 
 	cmd->common.command_id = req->tag;
-	if (ns)
-		trace_nvme_setup_nvm_cmd(req->q->id, cmd);
-	else
-		trace_nvme_setup_admin_cmd(cmd);
+	trace_nvme_setup_cmd(req, cmd);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(nvme_setup_cmd);
@@ -864,9 +879,6 @@ static void nvme_start_keep_alive(struct nvme_ctrl *ctrl)
 	if (unlikely(ctrl->kato == 0))
 		return;
 
-	INIT_DELAYED_WORK(&ctrl->ka_work, nvme_keep_alive_work);
-	memset(&ctrl->ka_cmd, 0, sizeof(ctrl->ka_cmd));
-	ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive;
 	schedule_delayed_work(&ctrl->ka_work, ctrl->kato * HZ);
 }
 
@@ -1056,7 +1068,7 @@ int nvme_set_queue_count(struct nvme_ctrl *ctrl, int *count)
 EXPORT_SYMBOL_GPL(nvme_set_queue_count);
 
 #define NVME_AEN_SUPPORTED \
-	(NVME_AEN_CFG_NS_ATTR | NVME_AEN_CFG_FW_ACT)
+	(NVME_AEN_CFG_NS_ATTR | NVME_AEN_CFG_FW_ACT | NVME_AEN_CFG_ANA_CHANGE)
 
 static void nvme_enable_aen(struct nvme_ctrl *ctrl)
 {
@@ -1472,6 +1484,12 @@ static void nvme_update_disk_info(struct gendisk *disk,
 
 	set_capacity(disk, capacity);
 	nvme_config_discard(ns);
+
+	if (id->nsattr & (1 << 0))
+		set_disk_ro(disk, true);
+	else
+		set_disk_ro(disk, false);
+
 	blk_mq_unfreeze_queue(disk->queue);
 }
 
@@ -2270,21 +2288,16 @@ out_unlock:
 	return ret;
 }
 
-int nvme_get_log_ext(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
-		     u8 log_page, void *log,
-		     size_t size, u64 offset)
+int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp,
+		void *log, size_t size, u64 offset)
 {
 	struct nvme_command c = { };
 	unsigned long dwlen = size / 4 - 1;
 
 	c.get_log_page.opcode = nvme_admin_get_log_page;
-
-	if (ns)
-		c.get_log_page.nsid = cpu_to_le32(ns->head->ns_id);
-	else
-		c.get_log_page.nsid = cpu_to_le32(NVME_NSID_ALL);
-
+	c.get_log_page.nsid = cpu_to_le32(nsid);
 	c.get_log_page.lid = log_page;
+	c.get_log_page.lsp = lsp;
 	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(lower_32_bits(offset));
@@ -2293,12 +2306,6 @@ int nvme_get_log_ext(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
 	return nvme_submit_sync_cmd(ctrl->admin_q, &c, log, size);
 }
 
-static int nvme_get_log(struct nvme_ctrl *ctrl, u8 log_page, void *log,
-			size_t size)
-{
-	return nvme_get_log_ext(ctrl, NULL, log_page, log, size, 0);
-}
-
 static int nvme_get_effects_log(struct nvme_ctrl *ctrl)
 {
 	int ret;
@@ -2309,8 +2316,8 @@ static int nvme_get_effects_log(struct nvme_ctrl *ctrl)
 	if (!ctrl->effects)
 		return 0;
 
-	ret = nvme_get_log(ctrl, NVME_LOG_CMD_EFFECTS, ctrl->effects,
-					sizeof(*ctrl->effects));
+	ret = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_CMD_EFFECTS, 0,
+			ctrl->effects, sizeof(*ctrl->effects), 0);
 	if (ret) {
 		kfree(ctrl->effects);
 		ctrl->effects = NULL;
@@ -2401,6 +2408,7 @@ int nvme_init_identify(struct nvme_ctrl *ctrl)
 	nvme_set_queue_limits(ctrl, ctrl->admin_q);
 	ctrl->sgls = le32_to_cpu(id->sgls);
 	ctrl->kas = le16_to_cpu(id->kas);
+	ctrl->max_namespaces = le32_to_cpu(id->mnan);
 
 	if (id->rtd3e) {
 		/* us -> s */
@@ -2460,8 +2468,12 @@ int nvme_init_identify(struct nvme_ctrl *ctrl)
 		ctrl->hmmaxd = le16_to_cpu(id->hmmaxd);
 	}
 
+	ret = nvme_mpath_init(ctrl, id);
 	kfree(id);
 
+	if (ret < 0)
+		return ret;
+
 	if (ctrl->apst_enabled && !prev_apst_enabled)
 		dev_pm_qos_expose_latency_tolerance(ctrl->device);
 	else if (!ctrl->apst_enabled && prev_apst_enabled)
@@ -2680,6 +2692,10 @@ static struct attribute *nvme_ns_id_attrs[] = {
 	&dev_attr_nguid.attr,
 	&dev_attr_eui.attr,
 	&dev_attr_nsid.attr,
+#ifdef CONFIG_NVME_MULTIPATH
+	&dev_attr_ana_grpid.attr,
+	&dev_attr_ana_state.attr,
+#endif
 	NULL,
 };
 
@@ -2702,6 +2718,14 @@ static umode_t nvme_ns_id_attrs_are_visible(struct kobject *kobj,
 		if (!memchr_inv(ids->eui64, 0, sizeof(ids->eui64)))
 			return 0;
 	}
+#ifdef CONFIG_NVME_MULTIPATH
+	if (a == &dev_attr_ana_grpid.attr || a == &dev_attr_ana_state.attr) {
+		if (dev_to_disk(dev)->fops != &nvme_fops) /* per-path attr */
+			return 0;
+		if (!nvme_ctrl_use_ana(nvme_get_ns_from_dev(dev)->ctrl))
+			return 0;
+	}
+#endif
 	return a->mode;
 }
 
@@ -3075,8 +3099,6 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
 
 	nvme_get_ctrl(ctrl);
 
-	kfree(id);
-
 	device_add_disk(ctrl->device, ns->disk);
 	if (sysfs_create_group(&disk_to_dev(ns->disk)->kobj,
 					&nvme_ns_id_attr_group))
@@ -3086,8 +3108,10 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
 		pr_warn("%s: failed to register lightnvm sysfs group for identification\n",
 			ns->disk->disk_name);
 
-	nvme_mpath_add_disk(ns->head);
+	nvme_mpath_add_disk(ns, id);
 	nvme_fault_inject_init(ns);
+	kfree(id);
+
 	return;
  out_unlink_ns:
 	mutex_lock(&ctrl->subsys->lock);
@@ -3229,7 +3253,8 @@ static void nvme_clear_changed_ns_log(struct nvme_ctrl *ctrl)
 	 * raced with us in reading the log page, which could cause us to miss
 	 * updates.
 	 */
-	error = nvme_get_log(ctrl, NVME_LOG_CHANGED_NS, log, log_size);
+	error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_CHANGED_NS, 0, log,
+			log_size, 0);
 	if (error)
 		dev_warn(ctrl->device,
 			"reading changed ns log failed: %d\n", error);
@@ -3346,9 +3371,9 @@ static void nvme_get_fw_slot_info(struct nvme_ctrl *ctrl)
 	if (!log)
 		return;
 
-	if (nvme_get_log(ctrl, NVME_LOG_FW_SLOT, log, sizeof(*log)))
-		dev_warn(ctrl->device,
-				"Get FW SLOT INFO log error\n");
+	if (nvme_get_log(ctrl, NVME_NSID_ALL, 0, NVME_LOG_FW_SLOT, log,
+			sizeof(*log), 0))
+		dev_warn(ctrl->device, "Get FW SLOT INFO log error\n");
 	kfree(log);
 }
 
@@ -3394,6 +3419,13 @@ static void nvme_handle_aen_notice(struct nvme_ctrl *ctrl, u32 result)
 	case NVME_AER_NOTICE_FW_ACT_STARTING:
 		queue_work(nvme_wq, &ctrl->fw_act_work);
 		break;
+#ifdef CONFIG_NVME_MULTIPATH
+	case NVME_AER_NOTICE_ANA:
+		if (!ctrl->ana_log_buf)
+			break;
+		queue_work(nvme_wq, &ctrl->ana_work);
+		break;
+#endif
 	default:
 		dev_warn(ctrl->device, "async event result %08x\n", result);
 	}
@@ -3426,6 +3458,7 @@ EXPORT_SYMBOL_GPL(nvme_complete_async_event);
 
 void nvme_stop_ctrl(struct nvme_ctrl *ctrl)
 {
+	nvme_mpath_stop(ctrl);
 	nvme_stop_keep_alive(ctrl);
 	flush_work(&ctrl->async_event_work);
 	flush_work(&ctrl->scan_work);
@@ -3463,6 +3496,7 @@ static void nvme_free_ctrl(struct device *dev)
 
 	ida_simple_remove(&nvme_instance_ida, ctrl->instance);
 	kfree(ctrl->effects);
+	nvme_mpath_uninit(ctrl);
 
 	if (subsys) {
 		mutex_lock(&subsys->lock);
@@ -3499,6 +3533,10 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev,
 	INIT_WORK(&ctrl->fw_act_work, nvme_fw_act_work);
 	INIT_WORK(&ctrl->delete_work, nvme_delete_ctrl_work);
 
+	INIT_DELAYED_WORK(&ctrl->ka_work, nvme_keep_alive_work);
+	memset(&ctrl->ka_cmd, 0, sizeof(ctrl->ka_cmd));
+	ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive;
+
 	ret = ida_simple_get(&nvme_instance_ida, 0, 0, GFP_KERNEL);
 	if (ret < 0)
 		goto out;

drivers/nvme/host/fabrics.c

@@ -474,7 +474,7 @@ EXPORT_SYMBOL_GPL(nvmf_connect_io_queue);
 
 bool nvmf_should_reconnect(struct nvme_ctrl *ctrl)
 {
-	if (ctrl->opts->max_reconnects != -1 &&
+	if (ctrl->opts->max_reconnects == -1 ||
 	    ctrl->nr_reconnects < ctrl->opts->max_reconnects)
 		return true;
 

drivers/nvme/host/fc.c

@@ -1737,6 +1737,7 @@ nvme_fc_init_request(struct blk_mq_tag_set *set, struct request *rq,
 	int queue_idx = (set == &ctrl->tag_set) ? hctx_idx + 1 : 0;
 	struct nvme_fc_queue *queue = &ctrl->queues[queue_idx];
 
+	nvme_req(rq)->ctrl = &ctrl->ctrl;
 	return __nvme_fc_init_request(ctrl, queue, op, rq, queue->rqcnt++);
 }
 

drivers/nvme/host/lightnvm.c

@@ -414,12 +414,6 @@ static int nvme_nvm_setup_20(struct nvme_nvm_id20 *id,
 	/* Set compacted version for upper layers */
 	geo->version = NVM_OCSSD_SPEC_20;
 
-	if (!(geo->major_ver_id == 2 && geo->minor_ver_id == 0)) {
-		pr_err("nvm: OCSSD version not supported (v%d.%d)\n",
-				geo->major_ver_id, geo->minor_ver_id);
-		return -EINVAL;
-	}
-
 	geo->num_ch = le16_to_cpu(id->num_grp);
 	geo->num_lun = le16_to_cpu(id->num_pu);
 	geo->all_luns = geo->num_ch * geo->num_lun;
@@ -583,7 +577,13 @@ static int nvme_nvm_get_chk_meta(struct nvm_dev *ndev,
 	struct ppa_addr ppa;
 	size_t left = nchks * sizeof(struct nvme_nvm_chk_meta);
 	size_t log_pos, offset, len;
-	int ret, i;
+	int ret, i, max_len;
+
+	/*
+	 * limit requests to maximum 256K to avoid issuing arbitrary large
+	 * requests when the device does not specific a maximum transfer size.
+	 */
+	max_len = min_t(unsigned int, ctrl->max_hw_sectors << 9, 256 * 1024);
 
 	/* Normalize lba address space to obtain log offset */
 	ppa.ppa = slba;
@@ -596,10 +596,11 @@ static int nvme_nvm_get_chk_meta(struct nvm_dev *ndev,
 	offset = log_pos * sizeof(struct nvme_nvm_chk_meta);
 
 	while (left) {
-		len = min_t(unsigned int, left, ctrl->max_hw_sectors << 9);
+		len = min_t(unsigned int, left, max_len);
 
-		ret = nvme_get_log_ext(ctrl, ns, NVME_NVM_LOG_REPORT_CHUNK,
-				dev_meta, len, offset);
+		ret = nvme_get_log(ctrl, ns->head->ns_id,
+				NVME_NVM_LOG_REPORT_CHUNK, 0, dev_meta, len,
+				offset);
 		if (ret) {
 			dev_err(ctrl->device, "Get REPORT CHUNK log error\n");
 			break;
@@ -662,12 +663,10 @@ static struct request *nvme_nvm_alloc_request(struct request_queue *q,
 
 	rq->cmd_flags &= ~REQ_FAILFAST_DRIVER;
 
-	if (rqd->bio) {
+	if (rqd->bio)
 		blk_init_request_from_bio(rq, rqd->bio);
-	} else {
+	else
 		rq->ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, IOPRIO_NORM);
-		rq->__data_len = 0;
-	}
 
 	return rq;
 }

drivers/nvme/host/multipath.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017 Christoph Hellwig.
+ * Copyright (c) 2017-2018 Christoph Hellwig.
  *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms and conditions of the GNU General Public License,
@@ -20,6 +20,11 @@ module_param(multipath, bool, 0444);
 MODULE_PARM_DESC(multipath,
 	"turn on native support for multiple controllers per subsystem");
 
+inline bool nvme_ctrl_use_ana(struct nvme_ctrl *ctrl)
+{
+	return multipath && ctrl->subsys && (ctrl->subsys->cmic & (1 << 3));
+}
+
 /*
  * If multipathing is enabled we need to always use the subsystem instance
  * number for numbering our devices to avoid conflicts between subsystems that
@@ -45,6 +50,7 @@ void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
 void nvme_failover_req(struct request *req)
 {
 	struct nvme_ns *ns = req->q->queuedata;
+	u16 status = nvme_req(req)->status;
 	unsigned long flags;
 
 	spin_lock_irqsave(&ns->head->requeue_lock, flags);
@@ -52,15 +58,35 @@ void nvme_failover_req(struct request *req)
 	spin_unlock_irqrestore(&ns->head->requeue_lock, flags);
 	blk_mq_end_request(req, 0);
 
-	nvme_reset_ctrl(ns->ctrl);
-	kblockd_schedule_work(&ns->head->requeue_work);
-}
+	switch (status & 0x7ff) {
+	case NVME_SC_ANA_TRANSITION:
+	case NVME_SC_ANA_INACCESSIBLE:
+	case NVME_SC_ANA_PERSISTENT_LOSS:
+		/*
+		 * If we got back an ANA error we know the controller is alive,
+		 * but not ready to serve this namespaces.  The spec suggests
+		 * we should update our general state here, but due to the fact
+		 * that the admin and I/O queues are not serialized that is
+		 * fundamentally racy.  So instead just clear the current path,
+		 * mark the the path as pending and kick of a re-read of the ANA
+		 * log page ASAP.
+		 */
+		nvme_mpath_clear_current_path(ns);
+		if (ns->ctrl->ana_log_buf) {
+			set_bit(NVME_NS_ANA_PENDING, &ns->flags);
+			queue_work(nvme_wq, &ns->ctrl->ana_work);
+		}
+		break;
+	default:
+		/*
+		 * Reset the controller for any non-ANA error as we don't know
+		 * what caused the error.
+		 */
+		nvme_reset_ctrl(ns->ctrl);
+		break;
+	}
 
-bool nvme_req_needs_failover(struct request *req, blk_status_t error)
-{
-	if (!(req->cmd_flags & REQ_NVME_MPATH))
-		return false;
-	return blk_path_error(error);
+	kblockd_schedule_work(&ns->head->requeue_work);
 }
 
 void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
@@ -75,25 +101,51 @@ void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
 	up_read(&ctrl->namespaces_rwsem);
 }
 
+static const char *nvme_ana_state_names[] = {
+	[0]				= "invalid state",
+	[NVME_ANA_OPTIMIZED]		= "optimized",
+	[NVME_ANA_NONOPTIMIZED]		= "non-optimized",
+	[NVME_ANA_INACCESSIBLE]		= "inaccessible",
+	[NVME_ANA_PERSISTENT_LOSS]	= "persistent-loss",
+	[NVME_ANA_CHANGE]		= "change",
+};
+
 static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head)
 {
-	struct nvme_ns *ns;
+	struct nvme_ns *ns, *fallback = NULL;
 
 	list_for_each_entry_rcu(ns, &head->list, siblings) {
-		if (ns->ctrl->state == NVME_CTRL_LIVE) {
+		if (ns->ctrl->state != NVME_CTRL_LIVE ||
+		    test_bit(NVME_NS_ANA_PENDING, &ns->flags))
+			continue;
+		switch (ns->ana_state) {
+		case NVME_ANA_OPTIMIZED:
 			rcu_assign_pointer(head->current_path, ns);
 			return ns;
+		case NVME_ANA_NONOPTIMIZED:
+			fallback = ns;
+			break;
+		default:
+			break;
 		}
 	}
 
-	return NULL;
+	if (fallback)
+		rcu_assign_pointer(head->current_path, fallback);
+	return fallback;
+}
+
+static inline bool nvme_path_is_optimized(struct nvme_ns *ns)
+{
+	return ns->ctrl->state == NVME_CTRL_LIVE &&
+		ns->ana_state == NVME_ANA_OPTIMIZED;
 }
 
 inline struct nvme_ns *nvme_find_path(struct nvme_ns_head *head)
 {
 	struct nvme_ns *ns = srcu_dereference(head->current_path, &head->srcu);
 
-	if (unlikely(!ns || ns->ctrl->state != NVME_CTRL_LIVE))
+	if (unlikely(!ns || !nvme_path_is_optimized(ns)))
 		ns = __nvme_find_path(head);
 	return ns;
 }
@@ -142,7 +194,7 @@ static bool nvme_ns_head_poll(struct request_queue *q, blk_qc_t qc)
 
 	srcu_idx = srcu_read_lock(&head->srcu);
 	ns = srcu_dereference(head->current_path, &head->srcu);
-	if (likely(ns && ns->ctrl->state == NVME_CTRL_LIVE))
+	if (likely(ns && nvme_path_is_optimized(ns)))
 		found = ns->queue->poll_fn(q, qc);
 	srcu_read_unlock(&head->srcu, srcu_idx);
 	return found;
@@ -176,6 +228,7 @@ int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl, struct nvme_ns_head *head)
 	struct request_queue *q;
 	bool vwc = false;
 
+	mutex_init(&head->lock);
 	bio_list_init(&head->requeue_list);
 	spin_lock_init(&head->requeue_lock);
 	INIT_WORK(&head->requeue_work, nvme_requeue_work);
@@ -220,29 +273,232 @@ out:
 	return -ENOMEM;
 }
 
-void nvme_mpath_add_disk(struct nvme_ns_head *head)
+static void nvme_mpath_set_live(struct nvme_ns *ns)
 {
+	struct nvme_ns_head *head = ns->head;
+
+	lockdep_assert_held(&ns->head->lock);
+
 	if (!head->disk)
 		return;
 
-	mutex_lock(&head->subsys->lock);
 	if (!(head->disk->flags & GENHD_FL_UP)) {
 		device_add_disk(&head->subsys->dev, head->disk);
 		if (sysfs_create_group(&disk_to_dev(head->disk)->kobj,
 				&nvme_ns_id_attr_group))
-			pr_warn("%s: failed to create sysfs group for identification\n",
-				head->disk->disk_name);
+			dev_warn(&head->subsys->dev,
+				 "failed to create id group.\n");
+	}
+
+	kblockd_schedule_work(&ns->head->requeue_work);
+}
+
+static int nvme_parse_ana_log(struct nvme_ctrl *ctrl, void *data,
+		int (*cb)(struct nvme_ctrl *ctrl, struct nvme_ana_group_desc *,
+			void *))
+{
+	void *base = ctrl->ana_log_buf;
+	size_t offset = sizeof(struct nvme_ana_rsp_hdr);
+	int error, i;
+
+	lockdep_assert_held(&ctrl->ana_lock);
+
+	for (i = 0; i < le16_to_cpu(ctrl->ana_log_buf->ngrps); i++) {
+		struct nvme_ana_group_desc *desc = base + offset;
+		u32 nr_nsids = le32_to_cpu(desc->nnsids);
+		size_t nsid_buf_size = nr_nsids * sizeof(__le32);
+
+		if (WARN_ON_ONCE(desc->grpid == 0))
+			return -EINVAL;
+		if (WARN_ON_ONCE(le32_to_cpu(desc->grpid) > ctrl->anagrpmax))
+			return -EINVAL;
+		if (WARN_ON_ONCE(desc->state == 0))
+			return -EINVAL;
+		if (WARN_ON_ONCE(desc->state > NVME_ANA_CHANGE))
+			return -EINVAL;
+
+		offset += sizeof(*desc);
+		if (WARN_ON_ONCE(offset > ctrl->ana_log_size - nsid_buf_size))
+			return -EINVAL;
+
+		error = cb(ctrl, desc, data);
+		if (error)
+			return error;
+
+		offset += nsid_buf_size;
+		if (WARN_ON_ONCE(offset > ctrl->ana_log_size - sizeof(*desc)))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+static inline bool nvme_state_is_live(enum nvme_ana_state state)
+{
+	return state == NVME_ANA_OPTIMIZED || state == NVME_ANA_NONOPTIMIZED;
+}
+
+static void nvme_update_ns_ana_state(struct nvme_ana_group_desc *desc,
+		struct nvme_ns *ns)
+{
+	enum nvme_ana_state old;
+
+	mutex_lock(&ns->head->lock);
+	old = ns->ana_state;
+	ns->ana_grpid = le32_to_cpu(desc->grpid);
+	ns->ana_state = desc->state;
+	clear_bit(NVME_NS_ANA_PENDING, &ns->flags);
+
+	if (nvme_state_is_live(ns->ana_state) && !nvme_state_is_live(old))
+		nvme_mpath_set_live(ns);
+	mutex_unlock(&ns->head->lock);
+}
+
+static int nvme_update_ana_state(struct nvme_ctrl *ctrl,
+		struct nvme_ana_group_desc *desc, void *data)
+{
+	u32 nr_nsids = le32_to_cpu(desc->nnsids), n = 0;
+	unsigned *nr_change_groups = data;
+	struct nvme_ns *ns;
+
+	dev_info(ctrl->device, "ANA group %d: %s.\n",
+			le32_to_cpu(desc->grpid),
+			nvme_ana_state_names[desc->state]);
+
+	if (desc->state == NVME_ANA_CHANGE)
+		(*nr_change_groups)++;
+
+	if (!nr_nsids)
+		return 0;
+
+	down_write(&ctrl->namespaces_rwsem);
+	list_for_each_entry(ns, &ctrl->namespaces, list) {
+		if (ns->head->ns_id != le32_to_cpu(desc->nsids[n]))
+			continue;
+		nvme_update_ns_ana_state(desc, ns);
+		if (++n == nr_nsids)
+			break;
+	}
+	up_write(&ctrl->namespaces_rwsem);
+	WARN_ON_ONCE(n < nr_nsids);
+	return 0;
+}
+
+static int nvme_read_ana_log(struct nvme_ctrl *ctrl, bool groups_only)
+{
+	u32 nr_change_groups = 0;
+	int error;
+
+	mutex_lock(&ctrl->ana_lock);
+	error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_ANA,
+			groups_only ? NVME_ANA_LOG_RGO : 0,
+			ctrl->ana_log_buf, ctrl->ana_log_size, 0);
+	if (error) {
+		dev_warn(ctrl->device, "Failed to get ANA log: %d\n", error);
+		goto out_unlock;
+	}
+
+	error = nvme_parse_ana_log(ctrl, &nr_change_groups,
+			nvme_update_ana_state);
+	if (error)
+		goto out_unlock;
+
+	/*
+	 * In theory we should have an ANATT timer per group as they might enter
+	 * the change state at different times.  But that is a lot of overhead
+	 * just to protect against a target that keeps entering new changes
+	 * states while never finishing previous ones.  But we'll still
+	 * eventually time out once all groups are in change state, so this
+	 * isn't a big deal.
+	 *
+	 * We also double the ANATT value to provide some slack for transports
+	 * or AEN processing overhead.
+	 */
+	if (nr_change_groups)
+		mod_timer(&ctrl->anatt_timer, ctrl->anatt * HZ * 2 + jiffies);
+	else
+		del_timer_sync(&ctrl->anatt_timer);
+out_unlock:
+	mutex_unlock(&ctrl->ana_lock);
+	return error;
+}
+
+static void nvme_ana_work(struct work_struct *work)
+{
+	struct nvme_ctrl *ctrl = container_of(work, struct nvme_ctrl, ana_work);
+
+	nvme_read_ana_log(ctrl, false);
+}
+
+static void nvme_anatt_timeout(struct timer_list *t)
+{
+	struct nvme_ctrl *ctrl = from_timer(ctrl, t, anatt_timer);
+
+	dev_info(ctrl->device, "ANATT timeout, resetting controller.\n");
+	nvme_reset_ctrl(ctrl);
+}
+
+void nvme_mpath_stop(struct nvme_ctrl *ctrl)
+{
+	if (!nvme_ctrl_use_ana(ctrl))
+		return;
+	del_timer_sync(&ctrl->anatt_timer);
+	cancel_work_sync(&ctrl->ana_work);
+}
+
+static ssize_t ana_grpid_show(struct device *dev, struct device_attribute *attr,
+		char *buf)
+{
+	return sprintf(buf, "%d\n", nvme_get_ns_from_dev(dev)->ana_grpid);
+}
+DEVICE_ATTR_RO(ana_grpid);
+
+static ssize_t ana_state_show(struct device *dev, struct device_attribute *attr,
+		char *buf)
+{
+	struct nvme_ns *ns = nvme_get_ns_from_dev(dev);
+
+	return sprintf(buf, "%s\n", nvme_ana_state_names[ns->ana_state]);
+}
+DEVICE_ATTR_RO(ana_state);
+
+static int nvme_set_ns_ana_state(struct nvme_ctrl *ctrl,
+		struct nvme_ana_group_desc *desc, void *data)
+{
+	struct nvme_ns *ns = data;
+
+	if (ns->ana_grpid == le32_to_cpu(desc->grpid)) {
+		nvme_update_ns_ana_state(desc, ns);
+		return -ENXIO; /* just break out of the loop */
+	}
+
+	return 0;
+}
+
+void nvme_mpath_add_disk(struct nvme_ns *ns, struct nvme_id_ns *id)
+{
+	if (nvme_ctrl_use_ana(ns->ctrl)) {
+		mutex_lock(&ns->ctrl->ana_lock);
+		ns->ana_grpid = le32_to_cpu(id->anagrpid);
+		nvme_parse_ana_log(ns->ctrl, ns, nvme_set_ns_ana_state);
+		mutex_unlock(&ns->ctrl->ana_lock);
+	} else {
+		mutex_lock(&ns->head->lock);
+		ns->ana_state = NVME_ANA_OPTIMIZED; 
+		nvme_mpath_set_live(ns);
+		mutex_unlock(&ns->head->lock);
 	}
-	mutex_unlock(&head->subsys->lock);
 }
 
 void nvme_mpath_remove_disk(struct nvme_ns_head *head)
 {
 	if (!head->disk)
 		return;
-	sysfs_remove_group(&disk_to_dev(head->disk)->kobj,
-			   &nvme_ns_id_attr_group);
-	del_gendisk(head->disk);
+	if (head->disk->flags & GENHD_FL_UP) {
+		sysfs_remove_group(&disk_to_dev(head->disk)->kobj,
+				   &nvme_ns_id_attr_group);
+		del_gendisk(head->disk);
+	}
 	blk_set_queue_dying(head->disk->queue);
 	/* make sure all pending bios are cleaned up */
 	kblockd_schedule_work(&head->requeue_work);
@@ -250,3 +506,52 @@ void nvme_mpath_remove_disk(struct nvme_ns_head *head)
 	blk_cleanup_queue(head->disk->queue);
 	put_disk(head->disk);
 }
+
+int nvme_mpath_init(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id)
+{
+	int error;
+
+	if (!nvme_ctrl_use_ana(ctrl))
+		return 0;
+
+	ctrl->anacap = id->anacap;
+	ctrl->anatt = id->anatt;
+	ctrl->nanagrpid = le32_to_cpu(id->nanagrpid);
+	ctrl->anagrpmax = le32_to_cpu(id->anagrpmax);
+
+	mutex_init(&ctrl->ana_lock);
+	timer_setup(&ctrl->anatt_timer, nvme_anatt_timeout, 0);
+	ctrl->ana_log_size = sizeof(struct nvme_ana_rsp_hdr) +
+		ctrl->nanagrpid * sizeof(struct nvme_ana_group_desc);
+	if (!(ctrl->anacap & (1 << 6)))
+		ctrl->ana_log_size += ctrl->max_namespaces * sizeof(__le32);
+
+	if (ctrl->ana_log_size > ctrl->max_hw_sectors << SECTOR_SHIFT) {
+		dev_err(ctrl->device,
+			"ANA log page size (%zd) larger than MDTS (%d).\n",
+			ctrl->ana_log_size,
+			ctrl->max_hw_sectors << SECTOR_SHIFT);
+		dev_err(ctrl->device, "disabling ANA support.\n");
+		return 0;
+	}
+
+	INIT_WORK(&ctrl->ana_work, nvme_ana_work);
+	ctrl->ana_log_buf = kmalloc(ctrl->ana_log_size, GFP_KERNEL);
+	if (!ctrl->ana_log_buf)
+		goto out;
+
+	error = nvme_read_ana_log(ctrl, true);
+	if (error)
+		goto out_free_ana_log_buf;
+	return 0;
+out_free_ana_log_buf:
+	kfree(ctrl->ana_log_buf);
+out:
+	return -ENOMEM;
+}
+
+void nvme_mpath_uninit(struct nvme_ctrl *ctrl)
+{
+	kfree(ctrl->ana_log_buf);
+}
+

drivers/nvme/host/nvme.h

@@ -102,6 +102,7 @@ struct nvme_request {
 	u8			retries;
 	u8			flags;
 	u16			status;
+	struct nvme_ctrl	*ctrl;
 };
 
 /*
@@ -119,6 +120,13 @@ static inline struct nvme_request *nvme_req(struct request *req)
 	return blk_mq_rq_to_pdu(req);
 }
 
+static inline u16 nvme_req_qid(struct request *req)
+{
+	if (!req->rq_disk)
+		return 0;
+	return blk_mq_unique_tag_to_hwq(blk_mq_unique_tag(req)) + 1;
+}
+
 /* The below value is the specific amount of delay needed before checking
  * readiness in case of the PCI_DEVICE(0x1c58, 0x0003), which needs the
  * NVME_QUIRK_DELAY_BEFORE_CHK_RDY quirk enabled. The value (in ms) was
@@ -175,6 +183,7 @@ struct nvme_ctrl {
 	u16 oacs;
 	u16 nssa;
 	u16 nr_streams;
+	u32 max_namespaces;
 	atomic_t abort_limit;
 	u8 vwc;
 	u32 vs;
@@ -197,6 +206,19 @@ struct nvme_ctrl {
 	struct work_struct fw_act_work;
 	unsigned long events;
 
+#ifdef CONFIG_NVME_MULTIPATH
+	/* asymmetric namespace access: */
+	u8 anacap;
+	u8 anatt;
+	u32 anagrpmax;
+	u32 nanagrpid;
+	struct mutex ana_lock;
+	struct nvme_ana_rsp_hdr *ana_log_buf;
+	size_t ana_log_size;
+	struct timer_list anatt_timer;
+	struct work_struct ana_work;
+#endif
+
 	/* Power saving configuration */
 	u64 ps_max_latency_us;
 	bool apst_enabled;
@@ -261,6 +283,7 @@ struct nvme_ns_head {
 	struct bio_list		requeue_list;
 	spinlock_t		requeue_lock;
 	struct work_struct	requeue_work;
+	struct mutex		lock;
 #endif
 	struct list_head	list;
 	struct srcu_struct      srcu;
@@ -287,6 +310,10 @@ struct nvme_ns {
 	struct nvme_ctrl *ctrl;
 	struct request_queue *queue;
 	struct gendisk *disk;
+#ifdef CONFIG_NVME_MULTIPATH
+	enum nvme_ana_state ana_state;
+	u32 ana_grpid;
+#endif
 	struct list_head siblings;
 	struct nvm_dev *ndev;
 	struct kref kref;
@@ -299,8 +326,9 @@ struct nvme_ns {
 	bool ext;
 	u8 pi_type;
 	unsigned long flags;
-#define NVME_NS_REMOVING 0
-#define NVME_NS_DEAD     1
+#define NVME_NS_REMOVING	0
+#define NVME_NS_DEAD     	1
+#define NVME_NS_ANA_PENDING	2
 	u16 noiob;
 
 #ifdef CONFIG_FAULT_INJECTION_DEBUG_FS
@@ -356,14 +384,6 @@ static inline u64 nvme_block_nr(struct nvme_ns *ns, sector_t sector)
 	return (sector >> (ns->lba_shift - 9));
 }
 
-static inline void nvme_cleanup_cmd(struct request *req)
-{
-	if (req->rq_flags & RQF_SPECIAL_PAYLOAD) {
-		kfree(page_address(req->special_vec.bv_page) +
-		      req->special_vec.bv_offset);
-	}
-}
-
 static inline void nvme_end_request(struct request *req, __le16 status,
 		union nvme_result result)
 {
@@ -420,6 +440,7 @@ void nvme_start_freeze(struct nvme_ctrl *ctrl);
 #define NVME_QID_ANY -1
 struct request *nvme_alloc_request(struct request_queue *q,
 		struct nvme_command *cmd, blk_mq_req_flags_t flags, int qid);
+void nvme_cleanup_cmd(struct request *req);
 blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
 		struct nvme_command *cmd);
 int nvme_submit_sync_cmd(struct request_queue *q, struct nvme_command *cmd,
@@ -435,21 +456,24 @@ int nvme_reset_ctrl_sync(struct nvme_ctrl *ctrl);
 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, u64 offset);
+int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp,
+		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;
 
 #ifdef CONFIG_NVME_MULTIPATH
+bool nvme_ctrl_use_ana(struct nvme_ctrl *ctrl);
 void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
 			struct nvme_ctrl *ctrl, int *flags);
 void nvme_failover_req(struct request *req);
-bool nvme_req_needs_failover(struct request *req, blk_status_t error);
 void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl);
 int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,struct nvme_ns_head *head);
-void nvme_mpath_add_disk(struct nvme_ns_head *head);
+void nvme_mpath_add_disk(struct nvme_ns *ns, struct nvme_id_ns *id);
 void nvme_mpath_remove_disk(struct nvme_ns_head *head);
+int nvme_mpath_init(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id);
+void nvme_mpath_uninit(struct nvme_ctrl *ctrl);
+void nvme_mpath_stop(struct nvme_ctrl *ctrl);
 
 static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
 {
@@ -468,7 +492,14 @@ static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
 		kblockd_schedule_work(&head->requeue_work);
 }
 
+extern struct device_attribute dev_attr_ana_grpid;
+extern struct device_attribute dev_attr_ana_state;
+
 #else
+static inline bool nvme_ctrl_use_ana(struct nvme_ctrl *ctrl)
+{
+	return false;
+}
 /*
  * Without the multipath code enabled, multiple controller per subsystems are
  * visible as devices and thus we cannot use the subsystem instance.
@@ -482,11 +513,6 @@ static inline void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
 static inline void nvme_failover_req(struct request *req)
 {
 }
-static inline bool nvme_req_needs_failover(struct request *req,
-					   blk_status_t error)
-{
-	return false;
-}
 static inline void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
 {
 }
@@ -495,7 +521,8 @@ static inline int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,
 {
 	return 0;
 }
-static inline void nvme_mpath_add_disk(struct nvme_ns_head *head)
+static inline void nvme_mpath_add_disk(struct nvme_ns *ns,
+		struct nvme_id_ns *id)
 {
 }
 static inline void nvme_mpath_remove_disk(struct nvme_ns_head *head)
@@ -507,6 +534,17 @@ static inline void nvme_mpath_clear_current_path(struct nvme_ns *ns)
 static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
 {
 }
+static inline int nvme_mpath_init(struct nvme_ctrl *ctrl,
+		struct nvme_id_ctrl *id)
+{
+	return 0;
+}
+static inline void nvme_mpath_uninit(struct nvme_ctrl *ctrl)
+{
+}
+static inline void nvme_mpath_stop(struct nvme_ctrl *ctrl)
+{
+}
 #endif /* CONFIG_NVME_MULTIPATH */
 
 #ifdef CONFIG_NVM

drivers/nvme/host/pci.c

@@ -418,6 +418,8 @@ static int nvme_init_request(struct blk_mq_tag_set *set, struct request *req,
 
 	BUG_ON(!nvmeq);
 	iod->nvmeq = nvmeq;
+
+	nvme_req(req)->ctrl = &dev->ctrl;
 	return 0;
 }
 
@@ -535,73 +537,6 @@ static void nvme_free_iod(struct nvme_dev *dev, struct request *req)
 		mempool_free(iod->sg, dev->iod_mempool);
 }
 
-#ifdef CONFIG_BLK_DEV_INTEGRITY
-static void nvme_dif_prep(u32 p, u32 v, struct t10_pi_tuple *pi)
-{
-	if (be32_to_cpu(pi->ref_tag) == v)
-		pi->ref_tag = cpu_to_be32(p);
-}
-
-static void nvme_dif_complete(u32 p, u32 v, struct t10_pi_tuple *pi)
-{
-	if (be32_to_cpu(pi->ref_tag) == p)
-		pi->ref_tag = cpu_to_be32(v);
-}
-
-/**
- * nvme_dif_remap - remaps ref tags to bip seed and physical lba
- *
- * The virtual start sector is the one that was originally submitted by the
- * block layer.	Due to partitioning, MD/DM cloning, etc. the actual physical
- * start sector may be different. Remap protection information to match the
- * physical LBA on writes, and back to the original seed on reads.
- *
- * Type 0 and 3 do not have a ref tag, so no remapping required.
- */
-static void nvme_dif_remap(struct request *req,
-			void (*dif_swap)(u32 p, u32 v, struct t10_pi_tuple *pi))
-{
-	struct nvme_ns *ns = req->rq_disk->private_data;
-	struct bio_integrity_payload *bip;
-	struct t10_pi_tuple *pi;
-	void *p, *pmap;
-	u32 i, nlb, ts, phys, virt;
-
-	if (!ns->pi_type || ns->pi_type == NVME_NS_DPS_PI_TYPE3)
-		return;
-
-	bip = bio_integrity(req->bio);
-	if (!bip)
-		return;
-
-	pmap = kmap_atomic(bip->bip_vec->bv_page) + bip->bip_vec->bv_offset;
-
-	p = pmap;
-	virt = bip_get_seed(bip);
-	phys = nvme_block_nr(ns, blk_rq_pos(req));
-	nlb = (blk_rq_bytes(req) >> ns->lba_shift);
-	ts = ns->disk->queue->integrity.tuple_size;
-
-	for (i = 0; i < nlb; i++, virt++, phys++) {
-		pi = (struct t10_pi_tuple *)p;
-		dif_swap(phys, virt, pi);
-		p += ts;
-	}
-	kunmap_atomic(pmap);
-}
-#else /* CONFIG_BLK_DEV_INTEGRITY */
-static void nvme_dif_remap(struct request *req,
-			void (*dif_swap)(u32 p, u32 v, struct t10_pi_tuple *pi))
-{
-}
-static void nvme_dif_prep(u32 p, u32 v, struct t10_pi_tuple *pi)
-{
-}
-static void nvme_dif_complete(u32 p, u32 v, struct t10_pi_tuple *pi)
-{
-}
-#endif
-
 static void nvme_print_sgl(struct scatterlist *sgl, int nents)
 {
 	int i;
@@ -827,9 +762,6 @@ static blk_status_t nvme_map_data(struct nvme_dev *dev, struct request *req,
 		if (blk_rq_map_integrity_sg(q, req->bio, &iod->meta_sg) != 1)
 			goto out_unmap;
 
-		if (req_op(req) == REQ_OP_WRITE)
-			nvme_dif_remap(req, nvme_dif_prep);
-
 		if (!dma_map_sg(dev->dev, &iod->meta_sg, 1, dma_dir))
 			goto out_unmap;
 	}
@@ -852,11 +784,8 @@ static void nvme_unmap_data(struct nvme_dev *dev, struct request *req)
 
 	if (iod->nents) {
 		dma_unmap_sg(dev->dev, iod->sg, iod->nents, dma_dir);
-		if (blk_integrity_rq(req)) {
-			if (req_op(req) == REQ_OP_READ)
-				nvme_dif_remap(req, nvme_dif_complete);
+		if (blk_integrity_rq(req))
 			dma_unmap_sg(dev->dev, &iod->meta_sg, 1, dma_dir);
-		}
 	}
 
 	nvme_cleanup_cmd(req);

drivers/nvme/host/rdma.c

@@ -40,13 +40,14 @@
 
 #define NVME_RDMA_MAX_SEGMENTS		256
 
-#define NVME_RDMA_MAX_INLINE_SEGMENTS	1
+#define NVME_RDMA_MAX_INLINE_SEGMENTS	4
 
 struct nvme_rdma_device {
 	struct ib_device	*dev;
 	struct ib_pd		*pd;
 	struct kref		ref;
 	struct list_head	entry;
+	unsigned int		num_inline_segments;
 };
 
 struct nvme_rdma_qe {
@@ -117,6 +118,7 @@ struct nvme_rdma_ctrl {
 	struct sockaddr_storage src_addr;
 
 	struct nvme_ctrl	ctrl;
+	bool			use_inline_data;
 };
 
 static inline struct nvme_rdma_ctrl *to_rdma_ctrl(struct nvme_ctrl *ctrl)
@@ -249,7 +251,7 @@ static int nvme_rdma_create_qp(struct nvme_rdma_queue *queue, const int factor)
 	/* +1 for drain */
 	init_attr.cap.max_recv_wr = queue->queue_size + 1;
 	init_attr.cap.max_recv_sge = 1;
-	init_attr.cap.max_send_sge = 1 + NVME_RDMA_MAX_INLINE_SEGMENTS;
+	init_attr.cap.max_send_sge = 1 + dev->num_inline_segments;
 	init_attr.sq_sig_type = IB_SIGNAL_REQ_WR;
 	init_attr.qp_type = IB_QPT_RC;
 	init_attr.send_cq = queue->ib_cq;
@@ -286,6 +288,7 @@ static int nvme_rdma_init_request(struct blk_mq_tag_set *set,
 	struct ib_device *ibdev = dev->dev;
 	int ret;
 
+	nvme_req(rq)->ctrl = &ctrl->ctrl;
 	ret = nvme_rdma_alloc_qe(ibdev, &req->sqe, sizeof(struct nvme_command),
 			DMA_TO_DEVICE);
 	if (ret)
@@ -374,6 +377,8 @@ nvme_rdma_find_get_device(struct rdma_cm_id *cm_id)
 		goto out_free_pd;
 	}
 
+	ndev->num_inline_segments = min(NVME_RDMA_MAX_INLINE_SEGMENTS,
+					ndev->dev->attrs.max_sge - 1);
 	list_add(&ndev->entry, &device_list);
 out_unlock:
 	mutex_unlock(&device_list_mutex);
@@ -868,6 +873,31 @@ out_free_io_queues:
 	return ret;
 }
 
+static void nvme_rdma_teardown_admin_queue(struct nvme_rdma_ctrl *ctrl,
+		bool remove)
+{
+	blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
+	nvme_rdma_stop_queue(&ctrl->queues[0]);
+	blk_mq_tagset_busy_iter(&ctrl->admin_tag_set, nvme_cancel_request,
+			&ctrl->ctrl);
+	blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
+	nvme_rdma_destroy_admin_queue(ctrl, remove);
+}
+
+static void nvme_rdma_teardown_io_queues(struct nvme_rdma_ctrl *ctrl,
+		bool remove)
+{
+	if (ctrl->ctrl.queue_count > 1) {
+		nvme_stop_queues(&ctrl->ctrl);
+		nvme_rdma_stop_io_queues(ctrl);
+		blk_mq_tagset_busy_iter(&ctrl->tag_set, nvme_cancel_request,
+				&ctrl->ctrl);
+		if (remove)
+			nvme_start_queues(&ctrl->ctrl);
+		nvme_rdma_destroy_io_queues(ctrl, remove);
+	}
+}
+
 static void nvme_rdma_stop_ctrl(struct nvme_ctrl *nctrl)
 {
 	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);
@@ -912,21 +942,44 @@ static void nvme_rdma_reconnect_or_remove(struct nvme_rdma_ctrl *ctrl)
 	}
 }
 
-static void nvme_rdma_reconnect_ctrl_work(struct work_struct *work)
+static int nvme_rdma_setup_ctrl(struct nvme_rdma_ctrl *ctrl, bool new)
 {
-	struct nvme_rdma_ctrl *ctrl = container_of(to_delayed_work(work),
-			struct nvme_rdma_ctrl, reconnect_work);
+	int ret = -EINVAL;
 	bool changed;
-	int ret;
 
-	++ctrl->ctrl.nr_reconnects;
-
-	ret = nvme_rdma_configure_admin_queue(ctrl, false);
+	ret = nvme_rdma_configure_admin_queue(ctrl, new);
 	if (ret)
-		goto requeue;
+		return ret;
+
+	if (ctrl->ctrl.icdoff) {
+		dev_err(ctrl->ctrl.device, "icdoff is not supported!\n");
+		goto destroy_admin;
+	}
+
+	if (!(ctrl->ctrl.sgls & (1 << 2))) {
+		dev_err(ctrl->ctrl.device,
+			"Mandatory keyed sgls are not supported!\n");
+		goto destroy_admin;
+	}
+
+	if (ctrl->ctrl.opts->queue_size > ctrl->ctrl.sqsize + 1) {
+		dev_warn(ctrl->ctrl.device,
+			"queue_size %zu > ctrl sqsize %u, clamping down\n",
+			ctrl->ctrl.opts->queue_size, ctrl->ctrl.sqsize + 1);
+	}
+
+	if (ctrl->ctrl.sqsize + 1 > ctrl->ctrl.maxcmd) {
+		dev_warn(ctrl->ctrl.device,
+			"sqsize %u > ctrl maxcmd %u, clamping down\n",
+			ctrl->ctrl.sqsize + 1, ctrl->ctrl.maxcmd);
+		ctrl->ctrl.sqsize = ctrl->ctrl.maxcmd - 1;
+	}
+
+	if (ctrl->ctrl.sgls & (1 << 20))
+		ctrl->use_inline_data = true;
 
 	if (ctrl->ctrl.queue_count > 1) {
-		ret = nvme_rdma_configure_io_queues(ctrl, false);
+		ret = nvme_rdma_configure_io_queues(ctrl, new);
 		if (ret)
 			goto destroy_admin;
 	}
@@ -935,10 +988,31 @@ static void nvme_rdma_reconnect_ctrl_work(struct work_struct *work)
 	if (!changed) {
 		/* state change failure is ok if we're in DELETING state */
 		WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING);
-		return;
+		ret = -EINVAL;
+		goto destroy_io;
 	}
 
 	nvme_start_ctrl(&ctrl->ctrl);
+	return 0;
+
+destroy_io:
+	if (ctrl->ctrl.queue_count > 1)
+		nvme_rdma_destroy_io_queues(ctrl, new);
+destroy_admin:
+	nvme_rdma_stop_queue(&ctrl->queues[0]);
+	nvme_rdma_destroy_admin_queue(ctrl, new);
+	return ret;
+}
+
+static void nvme_rdma_reconnect_ctrl_work(struct work_struct *work)
+{
+	struct nvme_rdma_ctrl *ctrl = container_of(to_delayed_work(work),
+			struct nvme_rdma_ctrl, reconnect_work);
+
+	++ctrl->ctrl.nr_reconnects;
+
+	if (nvme_rdma_setup_ctrl(ctrl, false))
+		goto requeue;
 
 	dev_info(ctrl->ctrl.device, "Successfully reconnected (%d attempts)\n",
 			ctrl->ctrl.nr_reconnects);
@@ -947,9 +1021,6 @@ static void nvme_rdma_reconnect_ctrl_work(struct work_struct *work)
 
 	return;
 
-destroy_admin:
-	nvme_rdma_stop_queue(&ctrl->queues[0]);
-	nvme_rdma_destroy_admin_queue(ctrl, false);
 requeue:
 	dev_info(ctrl->ctrl.device, "Failed reconnect attempt %d\n",
 			ctrl->ctrl.nr_reconnects);
@@ -962,27 +1033,9 @@ static void nvme_rdma_error_recovery_work(struct work_struct *work)
 			struct nvme_rdma_ctrl, err_work);
 
 	nvme_stop_keep_alive(&ctrl->ctrl);
-
-	if (ctrl->ctrl.queue_count > 1) {
-		nvme_stop_queues(&ctrl->ctrl);
-		nvme_rdma_stop_io_queues(ctrl);
-		blk_mq_tagset_busy_iter(&ctrl->tag_set,
-					nvme_cancel_request, &ctrl->ctrl);
-		nvme_rdma_destroy_io_queues(ctrl, false);
-	}
-
-	blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
-	nvme_rdma_stop_queue(&ctrl->queues[0]);
-	blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
-				nvme_cancel_request, &ctrl->ctrl);
-	nvme_rdma_destroy_admin_queue(ctrl, false);
-
-	/*
-	 * queues are not a live anymore, so restart the queues to fail fast
-	 * new IO
-	 */
-	blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
+	nvme_rdma_teardown_io_queues(ctrl, false);
 	nvme_start_queues(&ctrl->ctrl);
+	nvme_rdma_teardown_admin_queue(ctrl, false);
 
 	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING)) {
 		/* state change failure is ok if we're in DELETING state */
@@ -1090,19 +1143,27 @@ static int nvme_rdma_set_sg_null(struct nvme_command *c)
 }
 
 static int nvme_rdma_map_sg_inline(struct nvme_rdma_queue *queue,
-		struct nvme_rdma_request *req, struct nvme_command *c)
+		struct nvme_rdma_request *req, struct nvme_command *c,
+		int count)
 {
 	struct nvme_sgl_desc *sg = &c->common.dptr.sgl;
+	struct scatterlist *sgl = req->sg_table.sgl;
+	struct ib_sge *sge = &req->sge[1];
+	u32 len = 0;
+	int i;
 
-	req->sge[1].addr = sg_dma_address(req->sg_table.sgl);
-	req->sge[1].length = sg_dma_len(req->sg_table.sgl);
-	req->sge[1].lkey = queue->device->pd->local_dma_lkey;
+	for (i = 0; i < count; i++, sgl++, sge++) {
+		sge->addr = sg_dma_address(sgl);
+		sge->length = sg_dma_len(sgl);
+		sge->lkey = queue->device->pd->local_dma_lkey;
+		len += sge->length;
+	}
 
 	sg->addr = cpu_to_le64(queue->ctrl->ctrl.icdoff);
-	sg->length = cpu_to_le32(sg_dma_len(req->sg_table.sgl));
+	sg->length = cpu_to_le32(len);
 	sg->type = (NVME_SGL_FMT_DATA_DESC << 4) | NVME_SGL_FMT_OFFSET;
 
-	req->num_sge++;
+	req->num_sge += count;
 	return 0;
 }
 
@@ -1195,15 +1256,16 @@ static int nvme_rdma_map_data(struct nvme_rdma_queue *queue,
 		goto out_free_table;
 	}
 
-	if (count == 1) {
+	if (count <= dev->num_inline_segments) {
 		if (rq_data_dir(rq) == WRITE && nvme_rdma_queue_idx(queue) &&
+		    queue->ctrl->use_inline_data &&
 		    blk_rq_payload_bytes(rq) <=
 				nvme_rdma_inline_data_size(queue)) {
-			ret = nvme_rdma_map_sg_inline(queue, req, c);
+			ret = nvme_rdma_map_sg_inline(queue, req, c, count);
 			goto out;
 		}
 
-		if (dev->pd->flags & IB_PD_UNSAFE_GLOBAL_RKEY) {
+		if (count == 1 && dev->pd->flags & IB_PD_UNSAFE_GLOBAL_RKEY) {
 			ret = nvme_rdma_map_sg_single(queue, req, c);
 			goto out;
 		}
@@ -1574,6 +1636,7 @@ static int nvme_rdma_cm_handler(struct rdma_cm_id *cm_id,
 	case RDMA_CM_EVENT_CONNECT_ERROR:
 	case RDMA_CM_EVENT_UNREACHABLE:
 		nvme_rdma_destroy_queue_ib(queue);
+		/* fall through */
 	case RDMA_CM_EVENT_ADDR_ERROR:
 		dev_dbg(queue->ctrl->ctrl.device,
 			"CM error event %d\n", ev->event);
@@ -1736,25 +1799,12 @@ static const struct blk_mq_ops nvme_rdma_admin_mq_ops = {
 
 static void nvme_rdma_shutdown_ctrl(struct nvme_rdma_ctrl *ctrl, bool shutdown)
 {
-	if (ctrl->ctrl.queue_count > 1) {
-		nvme_stop_queues(&ctrl->ctrl);
-		nvme_rdma_stop_io_queues(ctrl);
-		blk_mq_tagset_busy_iter(&ctrl->tag_set,
-					nvme_cancel_request, &ctrl->ctrl);
-		nvme_rdma_destroy_io_queues(ctrl, shutdown);
-	}
-
+	nvme_rdma_teardown_io_queues(ctrl, shutdown);
 	if (shutdown)
 		nvme_shutdown_ctrl(&ctrl->ctrl);
 	else
 		nvme_disable_ctrl(&ctrl->ctrl, ctrl->ctrl.cap);
-
-	blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
-	nvme_rdma_stop_queue(&ctrl->queues[0]);
-	blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
-				nvme_cancel_request, &ctrl->ctrl);
-	blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
-	nvme_rdma_destroy_admin_queue(ctrl, shutdown);
+	nvme_rdma_teardown_admin_queue(ctrl, shutdown);
 }
 
 static void nvme_rdma_delete_ctrl(struct nvme_ctrl *ctrl)
@@ -1766,8 +1816,6 @@ static void nvme_rdma_reset_ctrl_work(struct work_struct *work)
 {
 	struct nvme_rdma_ctrl *ctrl =
 		container_of(work, struct nvme_rdma_ctrl, ctrl.reset_work);
-	int ret;
-	bool changed;
 
 	nvme_stop_ctrl(&ctrl->ctrl);
 	nvme_rdma_shutdown_ctrl(ctrl, false);
@@ -1778,25 +1826,9 @@ static void nvme_rdma_reset_ctrl_work(struct work_struct *work)
 		return;
 	}
 
-	ret = nvme_rdma_configure_admin_queue(ctrl, false);
-	if (ret)
+	if (nvme_rdma_setup_ctrl(ctrl, false))
 		goto out_fail;
 
-	if (ctrl->ctrl.queue_count > 1) {
-		ret = nvme_rdma_configure_io_queues(ctrl, false);
-		if (ret)
-			goto out_fail;
-	}
-
-	changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
-	if (!changed) {
-		/* state change failure is ok if we're in DELETING state */
-		WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING);
-		return;
-	}
-
-	nvme_start_ctrl(&ctrl->ctrl);
-
 	return;
 
 out_fail:
@@ -1959,49 +1991,10 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev,
 	changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING);
 	WARN_ON_ONCE(!changed);
 
-	ret = nvme_rdma_configure_admin_queue(ctrl, true);
+	ret = nvme_rdma_setup_ctrl(ctrl, true);
 	if (ret)
 		goto out_uninit_ctrl;
 
-	/* sanity check icdoff */
-	if (ctrl->ctrl.icdoff) {
-		dev_err(ctrl->ctrl.device, "icdoff is not supported!\n");
-		ret = -EINVAL;
-		goto out_remove_admin_queue;
-	}
-
-	/* sanity check keyed sgls */
-	if (!(ctrl->ctrl.sgls & (1 << 2))) {
-		dev_err(ctrl->ctrl.device,
-			"Mandatory keyed sgls are not supported!\n");
-		ret = -EINVAL;
-		goto out_remove_admin_queue;
-	}
-
-	/* only warn if argument is too large here, will clamp later */
-	if (opts->queue_size > ctrl->ctrl.sqsize + 1) {
-		dev_warn(ctrl->ctrl.device,
-			"queue_size %zu > ctrl sqsize %u, clamping down\n",
-			opts->queue_size, ctrl->ctrl.sqsize + 1);
-	}
-
-	/* warn if maxcmd is lower than sqsize+1 */
-	if (ctrl->ctrl.sqsize + 1 > ctrl->ctrl.maxcmd) {
-		dev_warn(ctrl->ctrl.device,
-			"sqsize %u > ctrl maxcmd %u, clamping down\n",
-			ctrl->ctrl.sqsize + 1, ctrl->ctrl.maxcmd);
-		ctrl->ctrl.sqsize = ctrl->ctrl.maxcmd - 1;
-	}
-
-	if (opts->nr_io_queues) {
-		ret = nvme_rdma_configure_io_queues(ctrl, true);
-		if (ret)
-			goto out_remove_admin_queue;
-	}
-
-	changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
-	WARN_ON_ONCE(!changed);
-
 	dev_info(ctrl->ctrl.device, "new ctrl: NQN \"%s\", addr %pISpcs\n",
 		ctrl->ctrl.opts->subsysnqn, &ctrl->addr);
 
@@ -2011,13 +2004,8 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev,
 	list_add_tail(&ctrl->list, &nvme_rdma_ctrl_list);
 	mutex_unlock(&nvme_rdma_ctrl_mutex);
 
-	nvme_start_ctrl(&ctrl->ctrl);
-
 	return &ctrl->ctrl;
 
-out_remove_admin_queue:
-	nvme_rdma_stop_queue(&ctrl->queues[0]);
-	nvme_rdma_destroy_admin_queue(ctrl, true);
 out_uninit_ctrl:
 	nvme_uninit_ctrl(&ctrl->ctrl);
 	nvme_put_ctrl(&ctrl->ctrl);

drivers/nvme/host/trace.c

@@ -128,3 +128,14 @@ const char *nvme_trace_parse_nvm_cmd(struct trace_seq *p,
 		return nvme_trace_common(p, cdw10);
 	}
 }
+
+const char *nvme_trace_disk_name(struct trace_seq *p, char *name)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+
+	if (*name)
+		trace_seq_printf(p, "disk=%s, ", name);
+	trace_seq_putc(p, 0);
+
+	return ret;
+}

drivers/nvme/host/trace.h

@@ -50,13 +50,8 @@
 		nvme_admin_opcode_name(nvme_admin_security_recv),	\
 		nvme_admin_opcode_name(nvme_admin_sanitize_nvm))
 
-const char *nvme_trace_parse_admin_cmd(struct trace_seq *p, u8 opcode,
-				       u8 *cdw10);
-#define __parse_nvme_admin_cmd(opcode, cdw10) \
-	nvme_trace_parse_admin_cmd(p, opcode, cdw10)
-
 #define nvme_opcode_name(opcode)	{ opcode, #opcode }
-#define show_opcode_name(val)					\
+#define show_nvm_opcode_name(val)				\
 	__print_symbolic(val,					\
 		nvme_opcode_name(nvme_cmd_flush),		\
 		nvme_opcode_name(nvme_cmd_write),		\
@@ -70,85 +65,92 @@ const char *nvme_trace_parse_admin_cmd(struct trace_seq *p, u8 opcode,
 		nvme_opcode_name(nvme_cmd_resv_acquire),	\
 		nvme_opcode_name(nvme_cmd_resv_release))
 
+#define show_opcode_name(qid, opcode)					\
+	(qid ? show_nvm_opcode_name(opcode) : show_admin_opcode_name(opcode))
+
+const char *nvme_trace_parse_admin_cmd(struct trace_seq *p, u8 opcode,
+		u8 *cdw10);
 const char *nvme_trace_parse_nvm_cmd(struct trace_seq *p, u8 opcode,
-				     u8 *cdw10);
-#define __parse_nvme_cmd(opcode, cdw10) \
-	nvme_trace_parse_nvm_cmd(p, opcode, cdw10)
+		u8 *cdw10);
 
-TRACE_EVENT(nvme_setup_admin_cmd,
-	    TP_PROTO(struct nvme_command *cmd),
-	    TP_ARGS(cmd),
+#define parse_nvme_cmd(qid, opcode, cdw10) 			\
+	(qid ?							\
+	 nvme_trace_parse_nvm_cmd(p, opcode, cdw10) : 		\
+	 nvme_trace_parse_admin_cmd(p, opcode, cdw10))
+
+const char *nvme_trace_disk_name(struct trace_seq *p, char *name);
+#define __print_disk_name(name)				\
+	nvme_trace_disk_name(p, name)
+
+#ifndef TRACE_HEADER_MULTI_READ
+static inline void __assign_disk_name(char *name, struct gendisk *disk)
+{
+	if (disk)
+		memcpy(name, disk->disk_name, DISK_NAME_LEN);
+	else
+		memset(name, 0, DISK_NAME_LEN);
+}
+#endif
+
+TRACE_EVENT(nvme_setup_cmd,
+	    TP_PROTO(struct request *req, struct nvme_command *cmd),
+	    TP_ARGS(req, cmd),
 	    TP_STRUCT__entry(
-		    __field(u8, opcode)
-		    __field(u8, flags)
-		    __field(u16, cid)
-		    __field(u64, metadata)
-		    __array(u8, cdw10, 24)
+		__array(char, disk, DISK_NAME_LEN)
+		__field(int, ctrl_id)
+		__field(int, qid)
+		__field(u8, opcode)
+		__field(u8, flags)
+		__field(u16, cid)
+		__field(u32, nsid)
+		__field(u64, metadata)
+		__array(u8, cdw10, 24)
 	    ),
 	    TP_fast_assign(
-		    __entry->opcode = cmd->common.opcode;
-		    __entry->flags = cmd->common.flags;
-		    __entry->cid = cmd->common.command_id;
-		    __entry->metadata = le64_to_cpu(cmd->common.metadata);
-		    memcpy(__entry->cdw10, cmd->common.cdw10,
-			   sizeof(__entry->cdw10));
+		__entry->ctrl_id = nvme_req(req)->ctrl->instance;
+		__entry->qid = nvme_req_qid(req);
+		__entry->opcode = cmd->common.opcode;
+		__entry->flags = cmd->common.flags;
+		__entry->cid = cmd->common.command_id;
+		__entry->nsid = le32_to_cpu(cmd->common.nsid);
+		__entry->metadata = le64_to_cpu(cmd->common.metadata);
+		__assign_disk_name(__entry->disk, req->rq_disk);
+		memcpy(__entry->cdw10, cmd->common.cdw10,
+		       sizeof(__entry->cdw10));
 	    ),
-	    TP_printk(" cmdid=%u, flags=0x%x, meta=0x%llx, cmd=(%s %s)",
-		      __entry->cid, __entry->flags, __entry->metadata,
-		      show_admin_opcode_name(__entry->opcode),
-		      __parse_nvme_admin_cmd(__entry->opcode, __entry->cdw10))
-);
-
-
-TRACE_EVENT(nvme_setup_nvm_cmd,
-	    TP_PROTO(int qid, struct nvme_command *cmd),
-	    TP_ARGS(qid, cmd),
-	    TP_STRUCT__entry(
-		    __field(int, qid)
-		    __field(u8, opcode)
-		    __field(u8, flags)
-		    __field(u16, cid)
-		    __field(u32, nsid)
-		    __field(u64, metadata)
-		    __array(u8, cdw10, 24)
-	    ),
-	    TP_fast_assign(
-		    __entry->qid = qid;
-		    __entry->opcode = cmd->common.opcode;
-		    __entry->flags = cmd->common.flags;
-		    __entry->cid = cmd->common.command_id;
-		    __entry->nsid = le32_to_cpu(cmd->common.nsid);
-		    __entry->metadata = le64_to_cpu(cmd->common.metadata);
-		    memcpy(__entry->cdw10, cmd->common.cdw10,
-			   sizeof(__entry->cdw10));
-	    ),
-	    TP_printk("qid=%d, nsid=%u, cmdid=%u, flags=0x%x, meta=0x%llx, cmd=(%s %s)",
-		      __entry->qid, __entry->nsid, __entry->cid,
+	    TP_printk("nvme%d: %sqid=%d, cmdid=%u, nsid=%u, flags=0x%x, meta=0x%llx, cmd=(%s %s)",
+		      __entry->ctrl_id, __print_disk_name(__entry->disk),
+		      __entry->qid, __entry->cid, __entry->nsid,
 		      __entry->flags, __entry->metadata,
-		      show_opcode_name(__entry->opcode),
-		      __parse_nvme_cmd(__entry->opcode, __entry->cdw10))
+		      show_opcode_name(__entry->qid, __entry->opcode),
+		      parse_nvme_cmd(__entry->qid, __entry->opcode, __entry->cdw10))
 );
 
 TRACE_EVENT(nvme_complete_rq,
 	    TP_PROTO(struct request *req),
 	    TP_ARGS(req),
 	    TP_STRUCT__entry(
-		    __field(int, qid)
-		    __field(int, cid)
-		    __field(u64, result)
-		    __field(u8, retries)
-		    __field(u8, flags)
-		    __field(u16, status)
+		__array(char, disk, DISK_NAME_LEN)
+		__field(int, ctrl_id)
+		__field(int, qid)
+		__field(int, cid)
+		__field(u64, result)
+		__field(u8, retries)
+		__field(u8, flags)
+		__field(u16, status)
 	    ),
 	    TP_fast_assign(
-		    __entry->qid = req->q->id;
-		    __entry->cid = req->tag;
-		    __entry->result = le64_to_cpu(nvme_req(req)->result.u64);
-		    __entry->retries = nvme_req(req)->retries;
-		    __entry->flags = nvme_req(req)->flags;
-		    __entry->status = nvme_req(req)->status;
+		__entry->ctrl_id = nvme_req(req)->ctrl->instance;
+		__entry->qid = nvme_req_qid(req);
+		__entry->cid = req->tag;
+		__entry->result = le64_to_cpu(nvme_req(req)->result.u64);
+		__entry->retries = nvme_req(req)->retries;
+		__entry->flags = nvme_req(req)->flags;
+		__entry->status = nvme_req(req)->status;
+		__assign_disk_name(__entry->disk, req->rq_disk);
 	    ),
-	    TP_printk("qid=%d, cmdid=%u, res=%llu, retries=%u, flags=0x%x, status=%u",
+	    TP_printk("nvme%d: %sqid=%d, cmdid=%u, res=%llu, retries=%u, flags=0x%x, status=%u",
+		      __entry->ctrl_id, __print_disk_name(__entry->disk),
 		      __entry->qid, __entry->cid, __entry->result,
 		      __entry->retries, __entry->flags, __entry->status)