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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Wed, 24 Feb 2021 16:06:44 -0800
From:   Vipin Sharma <>
To:     Michal Koutný <>
Subject: Re: [RFC 0/2] cgroup: New misc cgroup controller

On Tue, Feb 23, 2021 at 07:24:33PM +0100, Michal Koutný wrote:
> Hello.
> On Thu, Feb 18, 2021 at 11:55:47AM -0800, Vipin Sharma <> wrote:
> > This patch is creating a new misc cgroup controller for allocation and
> > tracking of resources which are not abstract like other cgroup
> > controllers.
> Please don't refer to this as "allocation" anywhere, that has a specific
> meaning (see Resource Distribution Models in
> Documentation/admin-gruide/cgroup-v2.rst).

Yeah, it should be "Limits". I will update the text.

> > This controller was initially proposed as encryption_id but after
> > the feedbacks, it is now changed to misc cgroup.
> >
> Interesting generalization. I wonder what else could fit under this as
> well. (It resembles pids controller on the cover.)
> > Please provide any feedback for this RFC or if it is good for
> > merging then I can send a patch for merging.
> A new controller is added exposed with v1 attributes. I'm not convinced
> it is desirable to change the frozen v1 controllers' API? (And therefore
> promote it as well.)

This is a very trivial controller. I believe once it becomes public it
can be helpful down the line to the v1 users, who cannot use v2 yet, for
some simple resource accounting, like us, we have the need for ASIDs
accounting in v1 until we move to v2. If the community has strong
objection then I can remove v1 support.

> Beware, bikeshedding. The name is very non-descriptive, potentially
> suggesting catch-all semantics. It'd deserve a further thought. My idea
> would be limit(s) or counter controller.

Limit and counter are kind of suggesting the underlying technique for
accounting instead of a generic name to denote resource types managed by
this controller. One can argue that if top level names are similar to
Resource Destribution Model then may be it makes sense to combine
resources with similar accounting technique under one controller.

I am looking at misc as a controller which is for resources not having
proper home in other controllers or not big enough to deserve their own

I agree with you coming up with a name which check all boxes of
requirements won't be possible. We have discussed name sev cgroup,
kvm cgroup, encryption_id cgroup before reaching to an agreement on misc
cgroup. I am open to other names if they are better suited for this
controller and makes more sense.


Powered by blists - more mailing lists