[PATCH v2 7/7] nvme_fc: add 'nvme_discovery' sysfs attribute to fc transport device

[jsmart2021@gmail.com (James Smart)]

The fc transport device should allow for a rediscovery, as userspace
might have lost the events. Example is udev events not handled during
system startup.

This patch add a sysfs entry 'nvme_discovery' on the fc class to
have it replay all udev discovery events for all local port/remote
port address pairs.

Signed-off-by: James Smart <james.smart at broadcom.com>
Reviewed-by: Hannes Reinecke <hare at suse.com>
Cc: Ewan Milne <emilne at redhat.com>
Cc: Johannes Thumshirn <jthumshirn at suse.com>

---
This is actually v4 - based on Hannes' v3 of the same patch:
[PATCHv3] nvme/fc: add 'discovery' sysfs attribute to fc transport device
http://lists.infradead.org/pipermail/linux-nvme/2017-December/014334.html

However, refcounting in v3 was still messed up when an rport_get() fails.
Modified the patch to add the unwind routine to correct it.

Also renamed the "discovery" attribute to "nvme_discovery" so that there
will not be collisions if/when we move to a shared fc transport and have
scsi there as well.

Note: this patch alone doesn't allow the simple udev script.  The patches
for the fc transport to move initial connect to use the asychronous
reconnect path is what allows the simple udev script. This patch simply
allows to catch up with connectivity if the udev events weren't handled.
---
 drivers/nvme/host/fc.c | 114 +++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 105 insertions(+), 9 deletions(-)

diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index 72bf01d5435a..ec68dd8821bf 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -122,6 +122,7 @@ struct nvme_fc_rport {
 	struct list_head		endp_list; /* for lport->endp_list */
 	struct list_head		ctrl_list;
 	struct list_head		ls_req_list;
+	struct list_head		disc_list;
 	struct device			*dev;	/* physical device for dma */
 	struct nvme_fc_lport		*lport;
 	spinlock_t			lock;
@@ -210,7 +211,6 @@ static DEFINE_IDA(nvme_fc_ctrl_cnt);
  * These items are short-term. They will eventually be moved into
  * a generic FC class. See comments in module init.
  */
-static struct class *fc_class;
 static struct device *fc_udev_device;
 
 
@@ -507,6 +507,7 @@ nvme_fc_free_rport(struct kref *ref)
 	list_del(&rport->endp_list);
 	spin_unlock_irqrestore(&nvme_fc_lock, flags);
 
+	WARN_ON(!list_empty(&rport->disc_list));
 	ida_simple_remove(&lport->endp_cnt, rport->remoteport.port_num);
 
 	kfree(rport);
@@ -694,6 +695,7 @@ nvme_fc_register_remoteport(struct nvme_fc_local_port *localport,
 	INIT_LIST_HEAD(&newrec->endp_list);
 	INIT_LIST_HEAD(&newrec->ctrl_list);
 	INIT_LIST_HEAD(&newrec->ls_req_list);
+	INIT_LIST_HEAD(&newrec->disc_list);
 	kref_init(&newrec->ref);
 	atomic_set(&newrec->act_ctrl_cnt, 0);
 	spin_lock_init(&newrec->lock);
@@ -3253,6 +3255,100 @@ static struct nvmf_transport_ops nvme_fc_transport = {
 	.create_ctrl	= nvme_fc_create_ctrl,
 };
 
+static void nvme_fc_discovery_unwind(struct list_head *lheadp)
+{
+	unsigned long flags;
+	struct nvme_fc_lport *lport;
+	struct nvme_fc_rport *rport, *tmp_rport;
+
+	list_for_each_entry_safe(rport, tmp_rport,
+				 lheadp, disc_list) {
+		spin_lock_irqsave(&nvme_fc_lock, flags);
+		list_del_init(&rport->disc_list);
+		spin_unlock_irqrestore(&nvme_fc_lock, flags);
+		lport = rport->lport;
+		/* signal discovery. Won't hurt if it repeats */
+		nvme_fc_signal_discovery_scan(lport, rport);
+		nvme_fc_rport_put(rport);
+		nvme_fc_lport_put(lport);
+	}
+}
+
+/* Arbitrary successive failures max. With lots of subsystems could be high */
+#define DISCOVERY_MAX_FAIL	20
+
+static ssize_t nvme_fc_nvme_discovery_store(struct device *dev,
+				       struct device_attribute *attr,
+				       const char *buf, size_t count)
+{
+	unsigned long flags;
+	LIST_HEAD(nvme_fc_disc_list);
+	struct nvme_fc_lport *lport;
+	struct nvme_fc_rport *rport, *tmp_rport;
+	int failcnt = 0;
+
+restart:
+	spin_lock_irqsave(&nvme_fc_lock, flags);
+	list_for_each_entry(lport, &nvme_fc_lport_list, port_list) {
+		list_for_each_entry(rport, &lport->endp_list, endp_list) {
+			if (!nvme_fc_lport_get(lport))
+				continue;
+			if (!nvme_fc_rport_get(rport)) {
+				/*
+				 * This is a temporary condition, so upon
+				 * restart this node will be gone from the
+				 * list.
+				 */
+				spin_unlock_irqrestore(&nvme_fc_lock, flags);
+				nvme_fc_lport_put(lport);
+				nvme_fc_discovery_unwind(&nvme_fc_disc_list);
+				if (failcnt++ < DISCOVERY_MAX_FAIL)
+					goto restart;
+				pr_err("nvme_discovery: too many reference "
+				       "failures\n");
+				return 0;
+			}
+			if (list_empty(&rport->disc_list))
+				list_add_tail(&rport->disc_list,
+					      &nvme_fc_disc_list);
+		}
+	}
+	spin_unlock_irqrestore(&nvme_fc_lock, flags);
+
+	list_for_each_entry_safe(rport, tmp_rport,
+				 &nvme_fc_disc_list, disc_list) {
+		spin_lock_irqsave(&nvme_fc_lock, flags);
+		list_del_init(&rport->disc_list);
+		spin_unlock_irqrestore(&nvme_fc_lock, flags);
+		lport = rport->lport;
+		nvme_fc_signal_discovery_scan(lport, rport);
+		nvme_fc_rport_put(rport);
+		nvme_fc_lport_put(lport);
+	}
+	return count;
+}
+static DEVICE_ATTR(nvme_discovery, 0200, NULL, nvme_fc_nvme_discovery_store);
+
+static struct attribute *nvme_fc_attrs[] = {
+	&dev_attr_nvme_discovery.attr,
+	NULL
+};
+
+static struct attribute_group nvme_fc_attr_group = {
+	.attrs = nvme_fc_attrs,
+};
+
+static const struct attribute_group *nvme_fc_attr_groups[] = {
+	&nvme_fc_attr_group,
+	NULL
+};
+
+static struct class fc_class = {
+	.name = "fc",
+	.dev_groups = nvme_fc_attr_groups,
+	.owner = THIS_MODULE,
+};
+
 static int __init nvme_fc_init_module(void)
 {
 	int ret;
@@ -3271,16 +3367,16 @@ static int __init nvme_fc_init_module(void)
 	 * put in place, this code will move to a more generic
 	 * location for the class.
 	 */
-	fc_class = class_create(THIS_MODULE, "fc");
-	if (IS_ERR(fc_class)) {
+	ret = class_register(&fc_class);
+	if (ret) {
 		pr_err("couldn't register class fc\n");
-		return PTR_ERR(fc_class);
+		return ret;
 	}
 
 	/*
 	 * Create a device for the FC-centric udev events
 	 */
-	fc_udev_device = device_create(fc_class, NULL, MKDEV(0, 0), NULL,
+	fc_udev_device = device_create(&fc_class, NULL, MKDEV(0, 0), NULL,
 				"fc_udev_device");
 	if (IS_ERR(fc_udev_device)) {
 		pr_err("couldn't create fc_udev device!\n");
@@ -3295,9 +3391,9 @@ static int __init nvme_fc_init_module(void)
 	return 0;
 
 out_destroy_device:
-	device_destroy(fc_class, MKDEV(0, 0));
+	device_destroy(&fc_class, MKDEV(0, 0));
 out_destroy_class:
-	class_destroy(fc_class);
+	class_unregister(&fc_class);
 	return ret;
 }
 
@@ -3312,8 +3408,8 @@ static void __exit nvme_fc_exit_module(void)
 	ida_destroy(&nvme_fc_local_port_cnt);
 	ida_destroy(&nvme_fc_ctrl_cnt);
 
-	device_destroy(fc_class, MKDEV(0, 0));
-	class_destroy(fc_class);
+	device_destroy(&fc_class, MKDEV(0, 0));
+	class_unregister(&fc_class);
 }
 
 module_init(nvme_fc_init_module);
-- 
2.13.1