NVMe: Skip async events for degraded controllers

If the controller is degraded, the driver should stay out of the way so
the user can recover the drive. This patch skips driver initiated async
event requests when the drive is in this state.

Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 008c9eec437a38c325503d50e8150ad636b3461b..7508a0ae57ff696ce3fcb8c27005186442acd158 100644 (file)
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1855,8 +1855,16 @@ static void nvme_reset_work(struct work_struct *work)
        if (result)
                goto out;
 
-       dev->ctrl.event_limit = NVME_NR_AEN_COMMANDS;
-       queue_work(nvme_workq, &dev->async_work);
+       /*
+        * A controller that can not execute IO typically requires user
+        * intervention to correct. For such degraded controllers, the driver
+        * should not submit commands the user did not request, so skip
+        * registering for asynchronous event notification on this condition.
+        */
+       if (dev->online_queues > 1) {
+               dev->ctrl.event_limit = NVME_NR_AEN_COMMANDS;
+               queue_work(nvme_workq, &dev->async_work);
+       }
 
        mod_timer(&dev->watchdog_timer, round_jiffies(jiffies + HZ));