On Mon, 2 Aug 2010 13:43:12 -0700 Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > On Fri, 30 Jul 2010 20:02:13 +0900 (JST) > KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx> wrote: > > > > On Fri, 30 Jul 2010 09:12:26 +0900 (JST) KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx> wrote: > > > > > > > > On Sat, 17 Jul 2010 12:16:33 -0700 (PDT) David Rientjes <rientjes@xxxxxxxxxx> wrote: > > > > > > > > > > > This a complete rewrite of the oom killer's badness() heuristic > > > > > > > > > > Any comments here, or are we ready to proceed? > > > > > > > > > > Gimme those acked-bys, reviewed-bys and tested-bys, please! > > > > > > > > If he continue to resend all of rewrite patch, I continue to refuse them. > > > > I explained it multi times. > > > > > > There are about 1000 emails on this topic. Please briefly explain it again. > > > > Major homework are > > > > - make patch series instead unreviewable all in one patch. > > Sometimes that's not very practical and the splitup isn't necessarily a > lot easier to understand and review. > > It's still possible to review the end result - just read the patched code. > > > - kill oom_score_adj > > Unclear why? > One reason I poitned out is that this new parameter is hard to use for admins and library writers. old oom_adj was defined as an parameter works as (memory usage of app)/oom_adj. new oom_score_adj was define as (memory usage of app * oom_score_adj)/ system_memory Then, an applications' oom_score on a host is quite different from on the other host. This operation is very new rather than a simple interface updates. This opinion was rejected. Anyway, I believe the value other than OOM_DISABLE is useless, I have no concerns. I'll use memcg if I want to control this kind of things. Because I know the new calculation logic works better at default, I welcome this patch itself in general. Thanks, -Kame -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>