From: Ming Lei <ming.lei@xxxxxxxxxx> commit 1b95e817916069ec45a7f259d088fd1c091a8cc6 upstream. Error recovery can be interrupted by controller removal, then the controller is left as quiesced, and IO hang can be caused. Fix the issue by unquiescing controller unconditionally when removing namespaces. This way is reasonable and safe given forward progress can be made when removing namespaces. Reviewed-by: Keith Busch <kbusch@xxxxxxxxxx> Reviewed-by: Sagi Grimberg <sagi@xxxxxxxxxxx> Reported-by: Chunguang Xu <brookxu.cn@xxxxxxxxx> Closes: https://lore.kernel.org/linux-nvme/cover.1685350577.git.chunguang.xu@xxxxxxxxxx/ Cc: stable@xxxxxxxxxxxxxxx Signed-off-by: Ming Lei <ming.lei@xxxxxxxxxx> Signed-off-by: Keith Busch <kbusch@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- drivers/nvme/host/core.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -4728,6 +4728,12 @@ void nvme_remove_namespaces(struct nvme_ */ nvme_mpath_clear_ctrl_paths(ctrl); + /* + * Unquiesce io queues so any pending IO won't hang, especially + * those submitted from scan work + */ + nvme_unquiesce_io_queues(ctrl); + /* prevent racing with ns scanning */ flush_work(&ctrl->scan_work); @@ -4737,10 +4743,8 @@ void nvme_remove_namespaces(struct nvme_ * removing the namespaces' disks; fail all the queues now to avoid * potentially having to clean up the failed sync later. */ - if (ctrl->state == NVME_CTRL_DEAD) { + if (ctrl->state == NVME_CTRL_DEAD) nvme_mark_namespaces_dead(ctrl); - nvme_unquiesce_io_queues(ctrl); - } /* this is a no-op when called from the controller reset handler */ nvme_change_ctrl_state(ctrl, NVME_CTRL_DELETING_NOIO);
