From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx153.postini.com [74.125.245.153]) by kanga.kvack.org (Postfix) with SMTP id D06546B007B for ; Tue, 19 Jun 2012 23:29:08 -0400 (EDT) Received: by lbjn8 with SMTP id n8so11085lbj.14 for ; Tue, 19 Jun 2012 20:29:06 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <4FE0C514.4040807@jp.fujitsu.com> References: <1340038051-29502-1-git-send-email-yinghan@google.com> <1340038051-29502-4-git-send-email-yinghan@google.com> <4FE0C514.4040807@jp.fujitsu.com> Date: Tue, 19 Jun 2012 20:29:06 -0700 Message-ID: Subject: Re: [PATCH V5 4/5] mm, vmscan: fix do_try_to_free_pages() livelock From: Ying Han Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org List-ID: To: KOSAKI Motohiro Cc: mhocko@suse.cz, hannes@cmpxchg.org, mel@csn.ul.ie, kamezawa.hiroyu@jp.fujitsu.com, riel@redhat.com, dhillf@gmail.com, hughd@google.com, dan.magenheimer@oracle.com, akpm@linux-foundation.org, linux-mm@kvack.org On Tue, Jun 19, 2012 at 11:29 AM, KOSAKI Motohiro wrote: > On 6/18/2012 12:47 PM, Ying Han wrote: >> Currently, do_try_to_free_pages() can enter livelock. Because of, >> now vmscan has two conflicted policies. >> >> 1) kswapd sleep when it couldn't reclaim any page even though >> =A0 =A0reach priority 0. This is because to avoid kswapd() infinite >> =A0 =A0loop. That said, kswapd assume direct reclaim makes enough >> =A0 =A0free pages either regular page reclaim or oom-killer. >> =A0 =A0This logic makes kswapd -> direct-reclaim dependency. >> 2) direct reclaim continue to reclaim without oom-killer until >> =A0 =A0kswapd turn on zone->all_unreclaimble. This is because >> =A0 =A0to avoid too early oom-kill. >> =A0 =A0This logic makes direct-reclaim -> kswapd dependency. >> >> In worst case, direct-reclaim may continue to page reclaim forever >> when kswapd is slept and any other thread don't wakeup kswapd. >> >> We can't turn on zone->all_unreclaimable because this is racy. >> direct reclaim path don't take any lock. Thus this patch removes >> zone->all_unreclaimable field completely and recalculates every >> time. >> >> Note: we can't take the idea that direct-reclaim see zone->pages_scanned >> directly and kswapd continue to use zone->all_unreclaimable. Because, >> it is racy. commit 929bea7c71 (vmscan: all_unreclaimable() use >> zone->all_unreclaimable as a name) describes the detail. >> >> Reported-by: Aaditya Kumar >> Reported-by: Ying Han >> Signed-off-by: KOSAKI Motohiro >> Acked-by: Rik van Riel > > Please drop this. I've got some review comment about this patch and > i need respin. but thank you for paying attention this. Thanks for the heads up. Are you working on the new version of it, since I included this patch in my softlimit reclaim patchset as a replacement of one patch i had. --Ying > > > -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org