From: Michal Hocko <mhocko@kernel.org>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Hugh Dickins <hughd@google.com>,
Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>,
Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Johannes Weiner <hannes@cmpxchg.org>,
Mel Gorman <mgorman@suse.de>,
David Rientjes <rientjes@google.com>,
Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>,
Hillf Danton <hillf.zj@alibaba-inc.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
Joonsoo Kim <js1304@gmail.com>
Subject: Re: [PATCH] mm, oom: protect !costly allocations some more
Date: Tue, 8 Mar 2016 13:22:41 +0100 [thread overview]
Message-ID: <20160308122241.GD13542@dhcp22.suse.cz> (raw)
In-Reply-To: <56DEB394.40602@suse.cz>
On Tue 08-03-16 12:12:20, Vlastimil Babka wrote:
> On 03/08/2016 11:10 AM, Michal Hocko wrote:
> > On Tue 08-03-16 10:52:15, Vlastimil Babka wrote:
> >> On 03/08/2016 10:46 AM, Michal Hocko wrote:
> > [...]
> >>>>> @@ -3294,6 +3289,18 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
> >>>>> did_some_progress > 0, no_progress_loops))
> >>>>> goto retry;
> >>>>>
> >>>>> + /*
> >>>>> + * !costly allocations are really important and we have to make sure
> >>>>> + * the compaction wasn't deferred or didn't bail out early due to locks
> >>>>> + * contention before we go OOM.
> >>>>> + */
> >>>>> + if (order && order <= PAGE_ALLOC_COSTLY_ORDER) {
> >>>>> + if (compact_result <= COMPACT_CONTINUE)
> >>>>
> >>>> Same here.
> >>>> I was going to say that this didn't have effect on Sergey's test, but
> >>>> turns out it did :)
> >>>
> >>> This should work as expected because compact_result is unsigned long
> >>> and so this is the unsigned arithmetic. I can make
> >>> #define COMPACT_NONE -1UL
> >>>
> >>> to make the intention more obvious if you prefer, though.
> >>
> >> Well, what wasn't obvious to me is actually that here (unlike in the
> >> test above) it was actually intended that COMPACT_NONE doesn't result in
> >> a retry. But it makes sense, otherwise we would retry endlessly if
> >> reclaim couldn't form a higher-order page, right.
> >
> > Yeah, that was the whole point. An alternative would be moving the test
> > into should_compact_retry(order, compact_result, contended_compaction)
> > which would be CONFIG_COMPACTION specific so we can get rid of the
> > COMPACT_NONE altogether. Something like the following. We would lose the
> > always initialized compact_result but this would matter only for
> > order==0 and we check for that. Even gcc doesn't complain.
>
> Yeah I like this version better, you can add my Acked-By.
OK, patch updated and I will post it as a reply to the original email.
> Thanks.
>
> > A more important question is whether the criteria I have chosen are
> > reasonable and reasonably independent on the particular implementation
> > of the compaction. I still cannot convince myself about the convergence
> > here. Is it possible that the compaction would keep returning
> > compact_result <= COMPACT_CONTINUE while not making any progress at all?
>
> Theoretically, if reclaim/compaction suitability decisions and
> allocation attempts didn't match the watermark checks, including the
> alloc_flags and classzone_idx parameters. Possible scenarios:
>
> - reclaim thinks compaction has enough to proceed, but compaction thinks
> otherwise and returns COMPACT_SKIPPED
> - compaction thinks it succeeded and returns COMPACT_PARTIAL, but
> allocation attempt fails
> - and perhaps some other combinations
But that might happen right now as well so it wouldn't be a regression,
right?
--
Michal Hocko
SUSE Labs
next prev parent reply other threads:[~2016-03-08 12:22 UTC|newest]
Thread overview: 150+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-15 18:19 [PATCH 0/3] OOM detection rework v4 Michal Hocko
2015-12-15 18:19 ` [PATCH 1/3] mm, oom: rework oom detection Michal Hocko
2016-01-14 22:58 ` David Rientjes
2016-01-16 1:07 ` Tetsuo Handa
2016-01-19 22:48 ` David Rientjes
2016-01-20 11:13 ` Tetsuo Handa
2016-01-20 13:13 ` Michal Hocko
2016-04-04 8:23 ` Vladimir Davydov
2016-04-04 9:42 ` Michal Hocko
2015-12-15 18:19 ` [PATCH 2/3] mm: throttle on IO only when there are too many dirty and writeback pages Michal Hocko
2016-03-17 11:35 ` Tetsuo Handa
2016-03-17 12:01 ` Michal Hocko
2015-12-15 18:19 ` [PATCH 3/3] mm: use watermak checks for __GFP_REPEAT high order allocations Michal Hocko
2015-12-16 23:35 ` [PATCH 0/3] OOM detection rework v4 Andrew Morton
2015-12-18 12:12 ` Michal Hocko
2015-12-16 23:58 ` Andrew Morton
2015-12-18 13:15 ` Michal Hocko
2015-12-18 16:35 ` Johannes Weiner
2015-12-24 12:41 ` Tetsuo Handa
2015-12-28 12:08 ` Tetsuo Handa
2015-12-28 14:13 ` Tetsuo Handa
2016-01-06 12:44 ` Vlastimil Babka
2016-01-08 12:37 ` Michal Hocko
2015-12-29 16:32 ` Michal Hocko
2015-12-30 15:05 ` Tetsuo Handa
2016-01-02 15:47 ` Tetsuo Handa
2016-01-20 12:24 ` Michal Hocko
2016-01-27 23:18 ` David Rientjes
2016-01-28 21:19 ` Michal Hocko
2015-12-29 16:27 ` Michal Hocko
2016-01-28 20:40 ` [PATCH 4/3] mm, oom: drop the last allocation attempt before out_of_memory Michal Hocko
2016-01-28 21:36 ` Johannes Weiner
2016-01-28 23:19 ` David Rientjes
2016-01-28 23:51 ` Johannes Weiner
2016-01-29 10:39 ` Tetsuo Handa
2016-01-29 15:32 ` Michal Hocko
2016-01-30 12:18 ` Tetsuo Handa
2016-01-29 15:23 ` Michal Hocko
2016-01-29 15:24 ` Michal Hocko
2016-01-28 21:19 ` [PATCH 5/3] mm, vmscan: make zone_reclaimable_pages more precise Michal Hocko
2016-01-28 23:20 ` David Rientjes
2016-01-29 3:41 ` Hillf Danton
2016-01-29 10:35 ` Tetsuo Handa
2016-01-29 15:17 ` Michal Hocko
2016-01-29 21:30 ` Tetsuo Handa
2016-02-03 13:27 ` [PATCH 0/3] OOM detection rework v4 Michal Hocko
2016-02-03 22:58 ` David Rientjes
2016-02-04 12:57 ` Michal Hocko
2016-02-04 13:10 ` Tetsuo Handa
2016-02-04 13:39 ` Michal Hocko
2016-02-04 14:24 ` Michal Hocko
2016-02-07 4:09 ` Tetsuo Handa
2016-02-15 20:06 ` Michal Hocko
2016-02-16 13:10 ` Tetsuo Handa
2016-02-16 15:19 ` Michal Hocko
2016-02-25 3:47 ` Hugh Dickins
2016-02-25 6:48 ` Sergey Senozhatsky
2016-02-25 9:17 ` Hillf Danton
2016-02-25 9:27 ` Michal Hocko
2016-02-25 9:48 ` Hillf Danton
2016-02-25 11:02 ` Sergey Senozhatsky
2016-02-25 9:23 ` Michal Hocko
2016-02-26 6:32 ` Hugh Dickins
2016-02-26 7:54 ` Hillf Danton
2016-02-26 9:24 ` Michal Hocko
2016-02-26 10:27 ` Hillf Danton
2016-02-26 13:49 ` Michal Hocko
2016-02-26 9:33 ` Michal Hocko
2016-02-29 21:02 ` Michal Hocko
2016-03-02 2:19 ` Joonsoo Kim
2016-03-02 9:50 ` Michal Hocko
2016-03-02 13:32 ` Joonsoo Kim
2016-03-02 14:06 ` Michal Hocko
2016-03-02 14:34 ` Joonsoo Kim
2016-03-03 9:26 ` Michal Hocko
2016-03-03 10:29 ` Tetsuo Handa
2016-03-03 14:10 ` Joonsoo Kim
2016-03-03 15:25 ` Michal Hocko
2016-03-04 5:23 ` Joonsoo Kim
2016-03-04 15:15 ` Michal Hocko
2016-03-04 17:39 ` Michal Hocko
2016-03-07 5:23 ` Joonsoo Kim
2016-03-03 15:50 ` Vlastimil Babka
2016-03-03 16:26 ` Michal Hocko
2016-03-04 7:10 ` Joonsoo Kim
2016-03-02 15:01 ` Minchan Kim
2016-03-07 16:08 ` [PATCH] mm, oom: protect !costly allocations some more (was: Re: [PATCH 0/3] OOM detection rework v4) Michal Hocko
2016-03-08 3:51 ` Sergey Senozhatsky
2016-03-08 9:08 ` Michal Hocko
2016-03-08 9:24 ` Sergey Senozhatsky
2016-03-08 9:24 ` [PATCH] mm, oom: protect !costly allocations some more Vlastimil Babka
2016-03-08 9:32 ` Sergey Senozhatsky
2016-03-08 9:46 ` Michal Hocko
2016-03-08 9:52 ` Vlastimil Babka
2016-03-08 10:10 ` Michal Hocko
2016-03-08 11:12 ` Vlastimil Babka
2016-03-08 12:22 ` Michal Hocko [this message]
2016-03-08 12:29 ` Vlastimil Babka
2016-03-08 9:58 ` [PATCH] mm, oom: protect !costly allocations some more (was: Re: [PATCH 0/3] OOM detection rework v4) Sergey Senozhatsky
2016-03-08 13:57 ` Michal Hocko
2016-03-08 10:36 ` Hugh Dickins
2016-03-08 13:42 ` [PATCH 0/2] oom rework: high order enahncements Michal Hocko
2016-03-08 13:42 ` [PATCH 1/3] mm, compaction: change COMPACT_ constants into enum Michal Hocko
2016-03-08 14:19 ` Vlastimil Babka
2016-03-09 3:55 ` Hillf Danton
2016-03-08 13:42 ` [PATCH 2/3] mm, compaction: cover all compaction mode in compact_zone Michal Hocko
2016-03-08 14:22 ` Vlastimil Babka
2016-03-09 3:57 ` Hillf Danton
2016-03-08 13:42 ` [PATCH 3/3] mm, oom: protect !costly allocations some more Michal Hocko
2016-03-08 14:34 ` Vlastimil Babka
2016-03-08 14:48 ` Michal Hocko
2016-03-08 15:03 ` Vlastimil Babka
2016-03-09 11:11 ` Michal Hocko
2016-03-09 14:07 ` Vlastimil Babka
2016-03-11 12:17 ` Hugh Dickins
2016-03-11 13:06 ` Michal Hocko
2016-03-11 19:08 ` Hugh Dickins
2016-03-14 16:21 ` Michal Hocko
2016-03-08 15:19 ` [PATCH] mm, oom: protect !costly allocations some more (was: Re: [PATCH 0/3] OOM detection rework v4) Joonsoo Kim
2016-03-08 16:05 ` Michal Hocko
2016-03-08 17:03 ` Joonsoo Kim
2016-03-09 10:41 ` Michal Hocko
2016-03-11 14:53 ` Joonsoo Kim
2016-03-11 15:20 ` Michal Hocko
2016-02-29 20:35 ` [PATCH 0/3] OOM detection rework v4 Michal Hocko
2016-03-01 7:29 ` Hugh Dickins
2016-03-01 13:38 ` Michal Hocko
2016-03-01 14:40 ` Michal Hocko
2016-03-01 18:14 ` Vlastimil Babka
2016-03-02 2:55 ` Joonsoo Kim
2016-03-02 12:37 ` Michal Hocko
2016-03-02 14:06 ` Joonsoo Kim
[not found] ` <20160302122410.GD26686@dhcp22.suse.cz>
2016-03-02 13:22 ` Vlastimil Babka
2016-03-02 2:28 ` Joonsoo Kim
2016-03-02 12:39 ` Michal Hocko
2016-03-03 9:54 ` Hugh Dickins
2016-03-03 12:32 ` Michal Hocko
2016-03-03 20:57 ` Hugh Dickins
2016-03-04 7:41 ` Vlastimil Babka
2016-03-04 7:53 ` Joonsoo Kim
2016-03-04 12:28 ` Michal Hocko
2016-03-11 10:45 ` Tetsuo Handa
2016-03-11 13:08 ` Michal Hocko
2016-03-11 13:32 ` Tetsuo Handa
2016-03-11 15:28 ` Michal Hocko
2016-03-11 16:49 ` Tetsuo Handa
2016-03-11 17:00 ` Michal Hocko
2016-03-11 17:20 ` Tetsuo Handa
2016-03-12 4:08 ` Tetsuo Handa
2016-03-13 14:41 ` Tetsuo Handa
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160308122241.GD13542@dhcp22.suse.cz \
--to=mhocko@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=hannes@cmpxchg.org \
--cc=hillf.zj@alibaba-inc.com \
--cc=hughd@google.com \
--cc=js1304@gmail.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=penguin-kernel@i-love.sakura.ne.jp \
--cc=rientjes@google.com \
--cc=sergey.senozhatsky.work@gmail.com \
--cc=torvalds@linux-foundation.org \
--cc=vbabka@suse.cz \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).