On Wed, Nov 11, 2020 at 09:42:17AM -0500, Qian Cai wrote: > On Wed, 2020-11-11 at 17:27 +0800, Ming Lei wrote: > > Can this issue disappear by applying the following change? > > This makes the system boot again as well. OK, actually it isn't necessary to register one new lock key for each hctx(blk_flush_queue) instance, and the current way is really over-kill because there can be lots of hw queues in one system. The original lockdep warning can be avoided by setting one nvme_loop specific lock class simply. If nvme_loop is backed against another nvme_loop, we still can avoid the warning by killing the direct end io chain, or assign another lock class. Will prepare one formal patch tomorrow. Thanks, Ming
