On Tue, 15 May 2012, Konstantin Khlebnikov wrote: > Hugh Dickins wrote: > > > > I had been hoping to get this stage, where I think we're still in > > agreement (except perhaps on the ordering of function arguments!), > > into 3.5 as a basis for later discussion. > > Yeah, my version differs mostly in function's names and ordering of > arguments. > I use 'long' for last argument in mem_cgroup_update_lru_size(), > and call it once in isolate_lru_pages(), rather than for each isolated page. That sounds very sensible, now that lumpy isn't switching lruvecs: can be done alongside the similarly deferred __mod_zone_page_state()s. > You have single mem_cgroup_page_lruvec() variant, and this is biggest > difference > between our versions. So, Ok, nothing important at this stage. > > Acked-by: Konstantin Khlebnikov <khlebnikov@xxxxxxxxxx> Thanks a lot, I appreciate you going back to delve in, despite being put off by appearances earlier. Yes, I'm only dealing with the trivial passing down of lruvec instead of zone here, where I felt we'd be sure to more or less agree. It seemed the right follow-on to your lruvec work in vmscan.c, giving us both a good base in 3.5-rc1 on which to try out our more interesting bits. Hugh -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>