On Tue, 21 Nov 2017, Mike Galbraith wrote: > On Tue, 2017-11-21 at 11:11 -0500, Mikulas Patocka wrote: > > > > So, drop the spinlock unplugging and leave only mutex unplugging, > > reproduce the deadlock and send the stacktraces. > > Nah, I reproduced it five years ago. Is any of that relevant today? > Damned if I know. Your report was the first I've noticed since. > > -Mike If we don't have any reason why it is needed to unplug block requests when a spinlock is taken - so let's not do this. If you add code to the spinlock path and you don't know why, it is cargo cult programming. Mikulas
