On Fri 03-11-17 23:08:35, Tetsuo Handa wrote: > Michal Hocko wrote: > > On Fri 03-11-17 22:46:29, Tetsuo Handa wrote: > > [...] > > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c > > > index c274960..547e9cb 100644 > > > --- a/mm/page_alloc.c > > > +++ b/mm/page_alloc.c > > > @@ -3312,11 +3312,10 @@ void warn_alloc(gfp_t gfp_mask, nodemask_t > > > *nodemask, const char *fmt, ...) > > > } > > > > > > /* > > > - * Go through the zonelist yet one more time, keep very high watermark > > > - * here, this is only to catch a parallel oom killing, we must fail if > > > - * we're still under heavy pressure. But make sure that this reclaim > > > - * attempt shall not depend on __GFP_DIRECT_RECLAIM && !__GFP_NORETRY > > > - * allocation which will never fail due to oom_lock already held. > > > + * This allocation attempt must not depend on __GFP_DIRECT_RECLAIM && > > > + * !__GFP_NORETRY allocation which will never fail due to oom_lock > > > + * already held. And since this allocation attempt does not sleep, > > > + * there is no reason we must use high watermark here. > > > */ > > > page = get_page_from_freelist((gfp_mask | __GFP_HARDWALL) & > > > ~__GFP_DIRECT_RECLAIM, order, > > > > Which patch does this depend on? > > This patch is preparation for "mm,oom: Move last second allocation to inside > the OOM killer." patch in order to use changelog close to what you suggested. > That is, I will move this comment and get_page_from_freelist() together to > alloc_pages_before_oomkill(), after we recorded why using ALLOC_WMARK_HIGH.
Is it really worth a separate patch, though? Aren't you overcomplicating things again? -- Michal Hocko SUSE Labs
