On Mon, Mar 22, 2021 at 08:37:26AM +0100, Christoph Hellwig wrote: > When we reset/teardown a controller, we must freeze and quiesce the > namespaces request queues to make sure that we safely stop inflight I/O > submissions. Freeze is mandatory because if our hctx map changed between > reconnects, blk_mq_update_nr_hw_queues will immediately attempt to freeze > the queue, and if it still has pending submissions (that are still > quiesced) it will hang. > > However, by freezing the namespaces request queues, and only unfreezing > them when we successfully reconnect, inflight submissions that are > running concurrently can now block grabbing the nshead srcu until either > we successfully reconnect or ctrl_loss_tmo expired (or the user > explicitly disconnected). > > This caused a deadlock when a different controller (different path on the > same subsystem) became live (i.e. optimized/non-optimized). This is > because nvme_mpath_set_live needs to synchronize the nshead srcu before > requeueing I/O in order to make sure that current_path is visible to > future (re-)submisions. However the srcu lock is taken by a blocked > submission on a frozen request queue, and we have a deadlock. > > In order to fix this use the blk_mq_submit_bio_direct API to submit the > bio to the low-level driver, which does not block on the queue free > but instead allows nvme-multipath to pick another path or queue up the > bio. Looks good. Reviewed-by: Keith Busch <kbusch@xxxxxxxxxx>
