From: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
To: mhocko@suse.com
Cc: aarcange@redhat.com, akpm@linux-foundation.org,
linux-mm@kvack.org, hannes@cmpxchg.org
Subject: Re: [PATCH] mm,page_alloc: Use min watermark for last second allocation attempt.
Date: Sat, 25 Nov 2017 20:07:24 +0900 [thread overview]
Message-ID: <201711252007.HDF15235.SFQOMHLFOVtFJO@I-love.SAKURA.ne.jp> (raw)
In-Reply-To: <20171120093851.gs3zqpmmyacxplor@dhcp22.suse.cz>
Michal Hocko wrote:
> On Fri 17-11-17 19:38:01, Tetsuo Handa wrote:
> [...]
> > [ 1792.835056] Out of memory: Kill process 14294 (idle-priority) score 876 or sacrifice child
> > [ 1792.836073] Killed process 14458 (normal-priority) total-vm:4176kB, anon-rss:88kB, file-rss:0kB, shmem-rss:0kB
>
> Wen you are in a situation when you are killing 88kB process then you
> are most probably going to suffer more oom kills anyway. Optimizing for
> this case is thus questionable at best. You would need to come up with
> a reasonable explanation why the livelock as described by Andrea is not
> possible with the current MM reclaim retry implementation. I am not
> saying the patch is wrong but your justification _is_ wrong.
What I wanted you to check is the fact that there was about 1.5 seconds of time
window and free: was 948KB above min: watermark (which is larger than 88KB) and
total free was 1560KB above min: (which means that making free memory was in progress
rather than suffer more OOM kills).
[ 1792.835056] Out of memory: Kill process 14294 (idle-priority) score 876 or sacrifice child
[ 1792.836073] Killed process 14458 (normal-priority) total-vm:4176kB, anon-rss:88kB, file-rss:0kB, shmem-rss:0kB
[ 1792.837757] oom_reaper: reaped process 14458 (normal-priority), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
[ 1794.366070] systemd-journal invoked oom-killer: gfp_mask=0x14200ca(GFP_HIGHUSER_MOVABLE), nodemask=(null), order=0, oom_score_adj=0
(...snipped...)
[ 1794.366295] Normal free:18448kB min:17500kB low:21872kB high:26244kB active_anon:650740kB inactive_anon:18976kB active_file:220kB inactive_file:436kB unevictable:0kB writepending:0kB present:1048576kB managed:966968kB mlocked:0kB kernel_stack:19568kB pagetables:36132kB bounce:0kB free_pcp:2736kB local_pcp:680kB free_cma:0kB
(...snipped...)
[ 1794.366342] Normal: 557*4kB (UM) 476*8kB (UMH) 126*16kB (UMH) 84*32kB (UM) 16*64kB (UM) 9*128kB (UM) 4*256kB (UM) 4*512kB (UM) 3*1024kB (M) 0*2048kB 0*4096kB = 19060kB
(...snipped...)
[ 1794.366368] Out of memory: Kill process 14294 (idle-priority) score 876 or sacrifice child
[ 1794.367372] Killed process 14459 (normal-priority) total-vm:4176kB, anon-rss:88kB, file-rss:0kB, shmem-rss:0kB
[ 1794.369143] oom_reaper: reaped process 14459 (normal-priority), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
But since I can't find better justification for this patch, I decided to send remaining
patches first: http://lkml.kernel.org/r/1511607169-5084-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp .
--
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: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
prev parent reply other threads:[~2017-11-25 11:07 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-17 10:38 Tetsuo Handa
2017-11-20 9:38 ` Michal Hocko
2017-11-25 11:07 ` Tetsuo Handa [this message]
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=201711252007.HDF15235.SFQOMHLFOVtFJO@I-love.SAKURA.ne.jp \
--to=penguin-kernel@i-love.sakura.ne.jp \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=hannes@cmpxchg.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.com \
/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