Inside nvme_kill_queues(), we have to start hw queues for draining requests in sw queues, .dispatch list and requeue list, so use blk_mq_start_hw_queues() instead of blk_mq_start_stopped_hw_queues() which only run queues if queues are stopped, but the queues may have been started already, for example nvme_start_queues() is called in reset work function. blk_mq_start_hw_queues() run hw queues in current context, instead of running asynchronously like before. Given nvme_kill_queues() is run from either remove context or reset worker context, both are fine to run hw queue directly. And the mutex of namespaces_mutex isn't a problem too becasue nvme_start_freeze() runs hw queue in this way already. Cc: stable@xxxxxxxxxxxxxxx Reported-by: Zhang Yi <yizhan@xxxxxxxxxx> Reviewed-by: Keith Busch <keith.busch@xxxxxxxxx> Reviewed-by: Christoph Hellwig <hch@xxxxxx> Reviewed-by: Johannes Thumshirn <jthumshirn@xxxxxxx> Signed-off-by: Ming Lei <ming.lei@xxxxxxxxxx> --- drivers/nvme/host/core.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index d5e0906262ea..40d5e4a9e8d7 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -2437,7 +2437,13 @@ void nvme_kill_queues(struct nvme_ctrl *ctrl) revalidate_disk(ns->disk); blk_set_queue_dying(ns->queue); blk_mq_abort_requeue_list(ns->queue); - blk_mq_start_stopped_hw_queues(ns->queue, true); + + /* + * Forcibly start all queues to avoid having stuck requests. + * Note that we must ensure the queues are not stopped + * when the final removal happens. + */ + blk_mq_start_hw_queues(ns->queue); } mutex_unlock(&ctrl->namespaces_mutex); } -- 2.9.4