On Tue, Jul 04, 2017 at 10:55:06AM +0300, Sagi Grimberg wrote:
> unlike blk_mq_stop_hw_queues and blk_mq_start_stopped_hw_queues
> quiescing/unquiescing respects the submission path rcu grace.
> Also make sure to kick the requeue list when appropriate.
>
> Reviewed-By: James Smart <james.smart@xxxxxxxxxxxx>
> Signed-off-by: Sagi Grimberg <sagi@xxxxxxxxxxx>
> ---
> drivers/nvme/host/fc.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
> index 50cc3b2b0e11..d63263f0b530 100644
> --- a/drivers/nvme/host/fc.c
> +++ b/drivers/nvme/host/fc.c
> @@ -2320,8 +2320,10 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
> if (ret)
> goto out_delete_hw_queue;
>
> - if (ctrl->ctrl.state != NVME_CTRL_NEW)
> - blk_mq_start_stopped_hw_queues(ctrl->ctrl.admin_q, true);
> + if (ctrl->ctrl.state != NVME_CTRL_NEW) {
> + blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
> + blk_mq_kick_requeue_list(ctrl->ctrl.admin_q);
Same question as that in patch 1 about blk_mq_kick_requeue_list().
--
Ming
