On Fri, 23 Mar 2018 14:52:25 +0100 Michal Hocko <mhocko@xxxxxxxxxx> wrote: > On Fri 23-03-18 09:47:53, Steven Rostedt wrote: > > > > The one solution is to pull the tracing file > > include/trace/events/vmscan.h into mm/ and have a local header to store > > the reclaim_stat structure that both vmscan.h and vmscan.c can > > reference. > > I guess we can live with the public definition as well. > Yes, that would be easier. Thanks! Then if Andrey's patches are not pulled, then this patch will be good to go. -- Steve
