linux-kernel - Re: memcg creates an unkillable task in 3.2-rc2

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:	Mon, 29 Jul 2013 11:52:11 -0700
From:	ebiederm@...ssion.com (Eric W. Biederman)
To:	Johannes Weiner <hannes@...xchg.org>
Cc:	Tejun Heo <tj@...nel.org>, Michal Hocko <mhocko@...e.cz>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	cgroups@...r.kernel.org, containers@...ts.linux-foundation.org,
	linux-kernel@...r.kernel.org, kent.overstreet@...il.com,
	Li Zefan <lizefan@...wei.com>,
	Glauber Costa <glommer@...il.com>
Subject: Re: memcg creates an unkillable task in 3.2-rc2

Johannes Weiner <hannes@...xchg.org> writes:

> On Mon, Jul 29, 2013 at 10:03:35AM -0700, Eric W. Biederman wrote:
>> Yes.  From the looks of the looks of it the cgroup implementation is
>> rather badly borked right now, and definitely not up to the standards of
>> the other core pieces of the kernel.  One of the reasons I was rather
>> apalled when systemd started using them.  Until the code actually works
>> reliably and the races are removed most people's systems would be much
>> better off with cgroups compiled out.
>> 
>> A single unified hierarchy is a really nasty idea for the same set of
>> reasons. You have to recompile to disable a controller to see if it that
>> controller's bugs are what are causing problems on your production
>> system.  Compiles or even just a reboot is a very heavy hammer to ask
>> people to use when they are triaging a problem.
>
> That's not how it works.  You can always select which controllers you
> want to mount during runtime.  Unified hierarchy only means that there
> is one cgroup tree for all mounted controllers, rather than every
> controller having its own separate cgroup tree.
>
> If you are like most users and currently mount all controllers in the
> same directory so that their cgroup trees overlap and appear to be a
> single tree, nothing changes for you.

My practical need is that I need the ability modify which controllers I
am using on a per group basis.  So that I can make corrall new processes
in a different set of controllers than currently existing processes.

I might just be missing something but I don't see how to do that with
all of the controllers mounted to the same filesystem.

So while I currently have a single mount of cgroupfs, and don't yet see
a need for orthogonal classification of processes.  To keep bugs and
craziness under control I expect I will be implementing a mount of
cgroupfs per controller within a month.

But except for the need to limit the scope of bugs this is all getting
rather badly off topic.

Eric
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/