ublk: quiesce request queue when aborting queue

So far aborting queue ends request when the ubq daemon is exiting, and
it can be run concurrently with ublk_queue_rq(), this way is fragile and
we depend on the tricky usage of UBLK_IO_FLAG_ABORTED for avoiding such
race.

Quiesce queue when aborting queue, and the two code paths can be run
completely exclusively, then it becomes easier to add new ublk feature,
such as relaxing single same task limit for each queue.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20231009093324.957829-6-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c
index 01cd2be1df0e5341ba11550a2e145cffa08b68b0..f80e48ab7ba0cc4ecf7e38e256c8c0d665590eae 100644 (file)
--- a/drivers/block/ublk_drv.c
+++ b/drivers/block/ublk_drv.c
@@ -1441,25 +1441,59 @@ static void ublk_abort_queue(struct ublk_device *ub, struct ublk_queue *ubq)
        }
 }
 
-static void ublk_daemon_monitor_work(struct work_struct *work)
+static bool ublk_abort_requests(struct ublk_device *ub)
 {
-       struct ublk_device *ub =
-               container_of(work, struct ublk_device, monitor_work.work);
+       struct gendisk *disk;
        int i;
 
+       spin_lock(&ub->lock);
+       disk = ub->ub_disk;
+       if (disk)
+               get_device(disk_to_dev(disk));
+       spin_unlock(&ub->lock);
+
+       /* Our disk has been dead */
+       if (!disk)
+               return false;
+
+       /* Now we are serialized with ublk_queue_rq() */
+       blk_mq_quiesce_queue(disk->queue);
        for (i = 0; i < ub->dev_info.nr_hw_queues; i++) {
                struct ublk_queue *ubq = ublk_get_queue(ub, i);
 
                if (ubq_daemon_is_dying(ubq)) {
-                       if (ublk_queue_can_use_recovery(ubq))
-                               schedule_work(&ub->quiesce_work);
-                       else
-                               schedule_work(&ub->stop_work);
-
                        /* abort queue is for making forward progress */
                        ublk_abort_queue(ub, ubq);
                }
        }
+       blk_mq_unquiesce_queue(disk->queue);
+       put_device(disk_to_dev(disk));
+
+       return true;
+}
+
+static void ublk_daemon_monitor_work(struct work_struct *work)
+{
+       struct ublk_device *ub =
+               container_of(work, struct ublk_device, monitor_work.work);
+       int i;
+
+       for (i = 0; i < ub->dev_info.nr_hw_queues; i++) {
+               struct ublk_queue *ubq = ublk_get_queue(ub, i);
+
+               if (ubq_daemon_is_dying(ubq))
+                       goto found;
+       }
+       return;
+
+found:
+       if (!ublk_abort_requests(ub))
+               return;
+
+       if (ublk_can_use_recovery(ub))
+               schedule_work(&ub->quiesce_work);
+       else
+               schedule_work(&ub->stop_work);
 
        /*
         * We can't schedule monitor work after ub's state is not UBLK_S_DEV_LIVE.
@@ -1594,6 +1628,8 @@ static void ublk_unquiesce_dev(struct ublk_device *ub)
 
 static void ublk_stop_dev(struct ublk_device *ub)
 {
+       struct gendisk *disk;
+
        mutex_lock(&ub->mutex);
        if (ub->dev_info.state == UBLK_S_DEV_DEAD)
                goto unlock;
@@ -1603,10 +1639,15 @@ static void ublk_stop_dev(struct ublk_device *ub)
                ublk_unquiesce_dev(ub);
        }
        del_gendisk(ub->ub_disk);
+
+       /* Sync with ublk_abort_queue() by holding the lock */
+       spin_lock(&ub->lock);
+       disk = ub->ub_disk;
        ub->dev_info.state = UBLK_S_DEV_DEAD;
        ub->dev_info.ublksrv_pid = -1;
-       put_disk(ub->ub_disk);
        ub->ub_disk = NULL;
+       spin_unlock(&ub->lock);
+       put_disk(disk);
  unlock:
        mutex_unlock(&ub->mutex);
        ublk_cancel_dev(ub);