On Fri, Jun 17, 2022 at 10:11:25PM +0800, Yu Kuai wrote: > > What's worse, io hung is possible in theory because wakeups might be > missed. For example, 2 * wake_batch tags are put, while only wake_batch > threads are worken: Not just in theory, io hanging sounds like the problem reported here: https://bugzilla.kernel.org/show_bug.cgi?id=215679 I only knew that it had nothing to do with the nvme driver, though, but they closed the bug after it stopped happening with different hardware.
