linux-kernel - Re: [PATCH] oom: split out forced OOM killer

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]

Message-ID: <5575E5E6.20908@gmail.com>
Date:	Mon, 08 Jun 2015 14:58:46 -0400
From:	Austin S Hemmelgarn <ahferroin7@...il.com>
To:	David Rientjes <rientjes@...gle.com>
CC:	Michal Hocko <mhocko@...e.cz>,
	Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] oom: split out forced OOM killer

On 2015-06-08 13:59, David Rientjes wrote:
> On Fri, 5 Jun 2015, Austin S Hemmelgarn wrote:
>
>>> I'm not sure what the benefit of this is, and it's adding more code.
>>> Having multiple pathways and requirements, such as constrained_alloc(), to
>>> oom kill a process isn't any clearer, in my opinion.  It also isn't
>>> intended to be optimized since the oom killer called from the page
>>> allocator and from sysrq aren't fastpaths.  To me, this seems like only a
>>> source code level change and doesn't make anything more clear but rather
>>> adds more code and obfuscates the entry path.
>>
>> At the very least, it does make the semantics of sysrq-f much nicer for admins
>> (especially the bit where it ignores the panic_on_oom setting, if the admin
>> wants the system to panic, he'll use sysrq-c).  There have been times I've had
>> to hit sysrq-f multiple times to get to actually kill anything, and this looks
>> to me like it would eliminate that rather annoying issue as well.
>>
>
> Are you saying there's a functional change with this patch/
>
I believe so (haven't actually read the patch itself, just the 
changelog), although it is only a change for certain configurations to a 
very specific and (I hope infrequently) used piece of functionality. 
Like I said above, if I wanted to crash my system, I'd be using sysrq-c; 
and if I'm using sysrq-f, I want _some_ task to die _now_.


Download attachment "smime.p7s" of type "application/pkcs7-signature" (2967 bytes)