Hi Christoph,

On Mon, Jul 18, 2022 at 08:29:28AM +0200, Christoph Hellwig wrote:

This was just a rather broken way to paper over a core block layer bug
that is now fixed.

This reverts commit cebbe577cb17ed9b04b50d9e6802a8bacffbadca.

This change will break START_DEV/STOP_DEV, which is supposed to run
multiple cycles after the device is added, especially this way can
help to implement error recovery from userside, such as one ubq_daemon
is crashed/hang, the device can be recovered by sending STOP_DEV/START_DEV
commands again after new ubq_daemon is setup.

So here we do need separated request_queue/disk, and the model is
similar with scsi's, in which disk rebind needs to be supported
and GD_OWNS_QUEUE can't be set.


Thanks,
Ming