linux-kernel - Re: [RFC PATCH] memcg, oom: throttle dump_header for memcg ooms without eligible tasks

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Tue, 16 Oct 2018 20:05:47 +0900
From:   Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>
To:     Michal Hocko <mhocko@...nel.org>
Cc:     Johannes Weiner <hannes@...xchg.org>, linux-mm@...ck.org,
        syzkaller-bugs@...glegroups.com, guro@...com,
        kirill.shutemov@...ux.intel.com, linux-kernel@...r.kernel.org,
        rientjes@...gle.com, yang.s@...baba-inc.com,
        Andrew Morton <akpm@...ux-foundation.org>,
        Sergey Senozhatsky <sergey.senozhatsky.work@...il.com>,
        Petr Mladek <pmladek@...e.com>,
        Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        Steven Rostedt <rostedt@...dmis.org>
Subject: Re: [RFC PATCH] memcg, oom: throttle dump_header for memcg ooms
 without eligible tasks

On 2018/10/16 18:20, Michal Hocko wrote:
>> Anyway, I'm OK if we apply _BOTH_ your patch and my patch. Or I'm OK with simplified
>> one shown below (because you don't like per memcg limit).
> 
> My patch is adding a rate-limit! I really fail to see why we need yet
> another one on top of it. This is just ridiculous. I can see reasons to
> tune that rate limit but adding 2 different mechanisms is just wrong.
> 
> If your NAK to unify the ratelimit for dump_header for all paths
> still holds then I do not care too much to push it forward. But I find
> thiis way of the review feedback counter productive.
> 

Your patch is _NOT_ adding a rate-limit for

  "%s invoked oom-killer: gfp_mask=%#x(%pGg), nodemask=%*pbl, order=%d, oom_score_adj=%hd\n"
  "Out of memory and no killable processes...\n"

lines!

[   97.519229] Out of memory and no killable processes...
[   97.522060] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.525507] Out of memory and no killable processes...
[   97.528817] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.532345] Out of memory and no killable processes...
[   97.534813] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.538270] Out of memory and no killable processes...
[   97.541449] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.546268] Out of memory and no killable processes...
[   97.548823] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.552399] Out of memory and no killable processes...
[   97.554939] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.558616] Out of memory and no killable processes...
[   97.562257] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.565998] Out of memory and no killable processes...
[   97.568642] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.572169] Out of memory and no killable processes...
[   97.575200] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.579357] Out of memory and no killable processes...
[   97.581912] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.585414] Out of memory and no killable processes...
[   97.589191] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.593586] Out of memory and no killable processes...
[   97.596527] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.600118] Out of memory and no killable processes...
[   97.603237] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.606837] Out of memory and no killable processes...
[   97.611550] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.615244] Out of memory and no killable processes...
[   97.617859] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.621634] Out of memory and no killable processes...
[   97.624884] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.629256] Out of memory and no killable processes...
[   97.631885] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.635367] Out of memory and no killable processes...
[   97.638033] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.641827] Out of memory and no killable processes...
[   97.641993] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.648453] Out of memory and no killable processes...
[   97.651481] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.655082] Out of memory and no killable processes...
[   97.657941] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.663036] Out of memory and no killable processes...
[   97.665890] a.out invoked oom-killer: gfp_mask=0x6000c0(GFP_KERNEL), nodemask=(null), order=0, oom_score_adj=-1000
[   97.669473] Out of memory and no killable processes...

We don't need to print these lines every few milliseconds. Even if an exceptional case,
this is a DoS for console users. Printing once (or a few times) per a minute will be
enough. Otherwise, users cannot see what they are typing and what are printed.