On 2019-09-24 11:35:16 [-0500], Scott Wood wrote: > > OK, sounds like stop_one_cpu_nowait() is the way to go then. so I applied the last three patches from the migrate-disable() series and it looks good. Nothing blew up in my testing. There were no objects to the stop_one_cpu_nowait() suggestion and I think that it is important not to lose the state at end. Could you please repost the patches? > -Scott Sebastian
