Skip to contentSkip to conversation

NVMe: Failed controller initialization fixes

PATCH New Keith Busch 1 participant 2 messages Jun 18, 2015

Subsystems (2)
Block layer NVMe
lore
Share & tools
lore

Keith Busch

author
#0 lore raw

PATCH 1/2 NVMe: Failed controller initialization fixes

This fixes an infinite device reset loop that may occur on devices that fail initialization. If the drive fails to become ready for any reason that does not involve an admin command timeout, the probe task should assume the drive is unavailable and remove it from the topology. In the case an admin command times out during device probing, the driver's existing reset action will handle removing the drive.

Signed-off-by Keith Busch <keith.busch at intel.com>

---
Dependent on these other fixes:

http://lists.infradead.org/pipermail/linux-nvme/2015-June/001915.html

 drivers/block/nvme-core.c |   33 +++++++++++++++------------------
 1 file changed, 15 insertions(+), 18 deletions(-)
drivers/block/nvme-core.c +15 -18
index 318dcdb..9bc0b06 100644
@@ -2930,6 +2930,18 @@ static int nvme_dev_resume(struct nvme_dev *dev)
 	return 0;
 }
 
+static void nvme_dead_ctrl(struct nvme_dev *dev)
+{
+	dev_warn(dev->dev, "Device failed to resume\n");
+	kref_get(&dev->kref);
+	if (IS_ERR(kthread_run(nvme_remove_dead_ctrl, dev, "nvme%d",
+						dev->instance))) {
+		dev_err(dev->dev,
+			"Failed to start controller remove task\n");
+		kref_put(&dev->kref, nvme_free_dev);
+	}
+}
+
 static void nvme_dev_reset(struct nvme_dev *dev)
 {
 	bool in_probe = work_busy(&dev->probe_work);

@@ -2943,14 +2955,7 @@ static void nvme_dev_reset(struct nvme_dev *dev)
 	/* Fail this device if reset occured during probe to avoid
 	 * infinite initialization loops. */
 	if (in_probe) {
-		dev_warn(dev->dev, "Device failed to resume\n");
-		kref_get(&dev->kref);
-		if (IS_ERR(kthread_run(nvme_remove_dead_ctrl, dev, "nvme%d",
-							dev->instance))) {
-			dev_err(dev->dev,
-				"Failed to start controller remove task\n");
-			kref_put(&dev->kref, nvme_free_dev);
-		}
+		nvme_dead_ctrl(dev);
 		return;
 	}
 	/* Schedule device resume asynchronously so the reset work is available

@@ -3085,16 +3090,8 @@ static void nvme_async_probe(struct work_struct *work)
 {
 	struct nvme_dev *dev = container_of(work, struct nvme_dev, probe_work);
 
-	if (nvme_dev_resume(dev))
-		goto reset;
-	return;
- reset:
-	spin_lock(&dev_list_lock);
-	if (!work_busy(&dev->reset_work)) {
-		dev->reset_workfn = nvme_reset_failed_dev;
-		queue_work(nvme_workq, &dev->reset_work);
-	}
-	spin_unlock(&dev_list_lock);
+	if (nvme_dev_resume(dev) && !work_busy(&dev->reset_work))
+		nvme_dead_ctrl(dev);
 }
 
 static void nvme_reset_notify(struct pci_dev *pdev, bool prepare)
Signature · 1 line
1.7.10.4
author
#1 lore raw

PATCH 2/2 NVMe: Fix filesystem deadlock on removal

Move gendisk deletion before controller shutdown so filesystem may sync dirty pages. Before, this would deadlock trying to allocate requests on frozen queues that are about to be deleted.

Signed-off-by Keith Busch <keith.busch at intel.com>
---
 drivers/block/nvme-core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/block/nvme-core.c +1 -1
index 9bc0b06..a230b90 100644
@@ -3123,8 +3123,8 @@ static void nvme_remove(struct pci_dev *pdev)
 	flush_work(&dev->reset_work);
 	flush_work(&dev->scan_work);
 	device_remove_file(dev->device, &dev_attr_reset_controller);
-	nvme_dev_shutdown(dev);
 	nvme_dev_remove(dev);
+	nvme_dev_shutdown(dev);
 	nvme_dev_remove_admin(dev);
 	device_destroy(nvme_class, MKDEV(nvme_char_major, dev->instance));
 	nvme_free_queues(dev, 0);
Signature · 1 line
1.7.10.4