On Wed, 2017-04-19 at 00:02 +0000, Bart Van Assche wrote: > On Tue, 2017-04-18 at 16:56 -0700, James Bottomley wrote: > > Actually, I think 3/3 is wrong. You need to start the queue as > > soon as we see it's blocked and the device has gone into DEL. This > > is safe because DEL ensures nothing ever gets from the generic > > request function to queuecommand, so the device driver never sees > > anything. > > Hello James, > > I agree that restarting the queue earlier is safe but I do not agree > that patch 3/3 is wrong. Can you explain why you think that patch 3/3 > is wrong? The shutdown occurs in driver removal, which precedes the queue start in 3/3. To avoid hangs and timeouts, we need the queue to be started before the driver shutdown tries to inject commands. They're not fatal, but they will likely show up as annoying messages (and time delays) which we can avoid by starting the queue the moment we go into DEL. James
