| Message ID | 1499154912-10420-4-git-send-email-sagi@grimberg.me (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, Jul 04, 2017 at 10:55:07AM +0300, Sagi Grimberg wrote: > Before we iterate over the tags, we need to make sure that > no inflight requests are being queued, so use quiesce > helper for that. > > Signed-off-by: Sagi Grimberg <sagi@grimberg.me> > --- > drivers/nvme/target/loop.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c > index 3d51341e62ee..891002ee005f 100644 > --- a/drivers/nvme/target/loop.c > +++ b/drivers/nvme/target/loop.c > @@ -434,7 +434,7 @@ static void nvme_loop_shutdown_ctrl(struct nvme_loop_ctrl *ctrl) > if (ctrl->ctrl.state == NVME_CTRL_LIVE) > nvme_shutdown_ctrl(&ctrl->ctrl); > > - blk_mq_stop_hw_queues(ctrl->ctrl.admin_q); > + blk_mq_quiesce_queue(ctrl->ctrl.admin_q); > blk_mq_tagset_busy_iter(&ctrl->admin_tag_set, > nvme_cancel_request, &ctrl->ctrl); > nvme_loop_destroy_admin_queue(ctrl); Is the queue killed before calling nvme_loop_destroy_admin_queue()? If not, simply quiescing and not unquiescing will hang in blk_cleanup_queue(). So I suggest to put blk_mq_unquiesce_queue() after blk_mq_tagset_busy_iter().
>> diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c >> index 3d51341e62ee..891002ee005f 100644 >> --- a/drivers/nvme/target/loop.c >> +++ b/drivers/nvme/target/loop.c >> @@ -434,7 +434,7 @@ static void nvme_loop_shutdown_ctrl(struct nvme_loop_ctrl *ctrl) >> if (ctrl->ctrl.state == NVME_CTRL_LIVE) >> nvme_shutdown_ctrl(&ctrl->ctrl); >> >> - blk_mq_stop_hw_queues(ctrl->ctrl.admin_q); >> + blk_mq_quiesce_queue(ctrl->ctrl.admin_q); >> blk_mq_tagset_busy_iter(&ctrl->admin_tag_set, >> nvme_cancel_request, &ctrl->ctrl); >> nvme_loop_destroy_admin_queue(ctrl); > > Is the queue killed before calling nvme_loop_destroy_admin_queue()? No, its not. > If not, simply quiescing and not unquiescing will hang in > blk_cleanup_queue(). Why should it? blk_cleanup_queue does not hang, even under IO load... > So I suggest to put blk_mq_unquiesce_queue() after blk_mq_tagset_busy_iter(). I can do that, but I don't understand why...
diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c index 3d51341e62ee..891002ee005f 100644 --- a/drivers/nvme/target/loop.c +++ b/drivers/nvme/target/loop.c @@ -434,7 +434,7 @@ static void nvme_loop_shutdown_ctrl(struct nvme_loop_ctrl *ctrl) if (ctrl->ctrl.state == NVME_CTRL_LIVE) nvme_shutdown_ctrl(&ctrl->ctrl); - blk_mq_stop_hw_queues(ctrl->ctrl.admin_q); + blk_mq_quiesce_queue(ctrl->ctrl.admin_q); blk_mq_tagset_busy_iter(&ctrl->admin_tag_set, nvme_cancel_request, &ctrl->ctrl); nvme_loop_destroy_admin_queue(ctrl);
Before we iterate over the tags, we need to make sure that no inflight requests are being queued, so use quiesce helper for that. Signed-off-by: Sagi Grimberg <sagi@grimberg.me> --- drivers/nvme/target/loop.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)