All of lore.kernel.org
 help / color / mirror / Atom feed
From: ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org (Eric W. Biederman)
To: Colin Walters <walters-gPq2gbYjIk8dnm+yROfE0A@public.gmane.org>
Cc: Kees Cook <keescook-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>,
	netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	Linux Containers
	<containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org>,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	Andy Lutomirski <luto-kltTT9wpgjJwATOyAt5JVQ@public.gmane.org>,
	Seth Forshee
	<seth.forshee-Z7WLFzj8eWMS+FvcfC7Uqw@public.gmane.org>,
	Nikolay Borisov <kernel-6AxghH7DbtA@public.gmane.org>,
	linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	Jann Horn <jann-XZ1E9jl8jIdeoWH0uzbU5w@public.gmane.org>
Subject: Re: [PATCH v2 00/10] userns: sysctl limits for namespaces
Date: Fri, 22 Jul 2016 13:45:07 -0500	[thread overview]
Message-ID: <87poq5y0jw.fsf@x220.int.ebiederm.org> (raw)
In-Reply-To: <1469194399.3817016.673814953.7581706C-2RFepEojUI2N1INw9kWLP6GC3tUn3ZHUQQ4Iyu8u01E@public.gmane.org> (Colin Walters's message of "Fri, 22 Jul 2016 09:33:19 -0400")

Colin Walters <walters-gPq2gbYjIk8dnm+yROfE0A@public.gmane.org> writes:

> On Thu, Jul 21, 2016, at 12:39 PM, Eric W. Biederman wrote:
>> 
>> This patchset addresses two use cases:
>> - Implement a sane upper bound on the number of namespaces.
>> - Provide a way for sandboxes to limit the attack surface from
>>   namespaces.
>
> Perhaps this is obvious, but since you didn't quite explicitly state it;
> do you see this as obsoleting the existing downstream patches
> mentioned in:
> https://lwn.net/Articles/673597/
> It seems conceptually similar to Kees' original approach, right?

Similar yes, and I expect it fills the need.  My primary difference is
that I believe this approach makes sense from a perspective of assuming
that user namespaces or other namespaces are not any buggier than any
other piece of kernel code and that people will use them.

I don't see these limits making sense from a perspective that user
namespaces are flawed and distro kernels should not have enabled them in
the first place.  That was my perception right or wrong of Kees patches
and the related patches that landed in Ubuntu and Debian.

With Kees approach I could not see how to handle the case where some
applications on the system wanted user namespaces and others don't.
Which made it very nasty for future evolution and more deployment of
user namespaces.  Being per user namespace these limits can be used to
sandbox applications without affecting the rest of the system.

> The high level makes sense to me...most interesting is
> per-userns sysctls.  I'll note most current container managers
> mount /proc/sys read-only, and Docker specifically drops
> CAP_SYS_RESOURCE by default, so they'd likely need to learn
> how to undo that if one wanted to support recursive container usage.
> We'd probably need to evaluate the safety of having /proc/sys
> writable generally.  (Also it's rather common to filter out CLONE_NEWUSER
> via seccomp, but that's easy to undo)

Just using a user namespace replaces most of those precautions.

> But that's the flip side - if we're aiming primarily for an upstreamable
> way to *limit* namespace usage, it seems sane to me.

Yes.  The primary target is to stop applications that have gone buggy
and allocated a crazy number of namespaces.  The secondary target
is to allow sandboxes to disable creation of additional namespaces.
Just set the limit to 0 and drop caps, or similarly set the limit
to 1 and create another fresh set of nested namespaces.

Eric

WARNING: multiple messages have this Message-ID (diff)
From: ebiederm@xmission.com (Eric W. Biederman)
To: Colin Walters <walters@verbum.org>
Cc: Linux Containers <containers@lists.linux-foundation.org>,
	Andy Lutomirski <luto@amacapital.net>, Jann Horn <jann@thejh.net>,
	Kees Cook <keescook@chromium.org>,
	Nikolay Borisov <kernel@kyup.com>,
	"Serge E. Hallyn" <serge@hallyn.com>,
	Seth Forshee <seth.forshee@canonical.com>,
	linux-fsdevel@vger.kernel.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-api@vger.kernel.org
Subject: Re: [PATCH v2 00/10] userns: sysctl limits for namespaces
Date: Fri, 22 Jul 2016 13:45:07 -0500	[thread overview]
Message-ID: <87poq5y0jw.fsf@x220.int.ebiederm.org> (raw)
In-Reply-To: <1469194399.3817016.673814953.7581706C@webmail.messagingengine.com> (Colin Walters's message of "Fri, 22 Jul 2016 09:33:19 -0400")

Colin Walters <walters@verbum.org> writes:

> On Thu, Jul 21, 2016, at 12:39 PM, Eric W. Biederman wrote:
>> 
>> This patchset addresses two use cases:
>> - Implement a sane upper bound on the number of namespaces.
>> - Provide a way for sandboxes to limit the attack surface from
>>   namespaces.
>
> Perhaps this is obvious, but since you didn't quite explicitly state it;
> do you see this as obsoleting the existing downstream patches
> mentioned in:
> https://lwn.net/Articles/673597/
> It seems conceptually similar to Kees' original approach, right?

Similar yes, and I expect it fills the need.  My primary difference is
that I believe this approach makes sense from a perspective of assuming
that user namespaces or other namespaces are not any buggier than any
other piece of kernel code and that people will use them.

I don't see these limits making sense from a perspective that user
namespaces are flawed and distro kernels should not have enabled them in
the first place.  That was my perception right or wrong of Kees patches
and the related patches that landed in Ubuntu and Debian.

With Kees approach I could not see how to handle the case where some
applications on the system wanted user namespaces and others don't.
Which made it very nasty for future evolution and more deployment of
user namespaces.  Being per user namespace these limits can be used to
sandbox applications without affecting the rest of the system.

> The high level makes sense to me...most interesting is
> per-userns sysctls.  I'll note most current container managers
> mount /proc/sys read-only, and Docker specifically drops
> CAP_SYS_RESOURCE by default, so they'd likely need to learn
> how to undo that if one wanted to support recursive container usage.
> We'd probably need to evaluate the safety of having /proc/sys
> writable generally.  (Also it's rather common to filter out CLONE_NEWUSER
> via seccomp, but that's easy to undo)

Just using a user namespace replaces most of those precautions.

> But that's the flip side - if we're aiming primarily for an upstreamable
> way to *limit* namespace usage, it seems sane to me.

Yes.  The primary target is to stop applications that have gone buggy
and allocated a crazy number of namespaces.  The secondary target
is to allow sandboxes to disable creation of additional namespaces.
Just set the limit to 0 and drop caps, or similarly set the limit
to 1 and create another fresh set of nested namespaces.

Eric

  parent reply	other threads:[~2016-07-22 18:45 UTC|newest]

Thread overview: 113+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-20  1:13 [PATCH 0/10] userns: sysctl limits for namespaces Eric W. Biederman
     [not found] ` <8737n5dscy.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-20  1:21   ` [PATCH 01/10] sysctl: Stop implicitly passing current into sysctl_table_root.lookup Eric W. Biederman
     [not found]     ` <20160720012138.7894-1-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-20  1:21       ` [PATCH 02/10] userns: Add per user namespace sysctls Eric W. Biederman
     [not found]         ` <20160720012138.7894-2-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-20  5:12           ` Serge E. Hallyn
     [not found]             ` <20160720051226.GC16885-7LNsyQBKDXoIagZqoN9o3w@public.gmane.org>
2016-07-20 15:55               ` Eric W. Biederman
     [not found]                 ` <87shv49ud0.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-20 16:43                   ` Serge E. Hallyn
     [not found]                     ` <20160720164347.GA24378-7LNsyQBKDXoIagZqoN9o3w@public.gmane.org>
2016-07-21 14:41                       ` Eric W. Biederman
     [not found]                         ` <87invz9hpz.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-21 15:07                           ` Serge E. Hallyn
2016-07-20  6:28           ` Nikolay Borisov
     [not found]             ` <578F1A00.6030701-6AxghH7DbtA@public.gmane.org>
2016-07-21 15:37               ` Eric W. Biederman
2016-07-20  1:21       ` [PATCH 03/10] userns: Add a limit on the number of user namespaces Eric W. Biederman
     [not found]         ` <20160720012138.7894-3-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-20  7:03           ` Nikolay Borisov
     [not found]             ` <578F2233.2040708-6AxghH7DbtA@public.gmane.org>
2016-07-21 15:23               ` Eric W. Biederman
2016-07-20  1:21       ` [PATCH 04/10] userns: Generalize the user namespace count into ucount Eric W. Biederman
2016-07-20  1:21       ` [PATCH 05/10] pidns: Add a limit on the number of pid namespaces Eric W. Biederman
2016-07-20  1:21       ` [PATCH 06/10] utsns: Add a limit on the number of uts namespaces Eric W. Biederman
2016-07-20  1:21       ` [PATCH 07/10] ipcns: Add a limit on the number of ipc namespaces Eric W. Biederman
2016-07-20  1:21       ` [PATCH 08/10] cgroupns: Add a limit on the number of cgroup namespaces Eric W. Biederman
2016-07-20  1:21       ` [PATCH 09/10] netns: Add a limit on the number of net namespaces Eric W. Biederman
2016-07-20  1:21       ` [PATCH 10/10] mntns: Add a limit on the number of mount namespaces Eric W. Biederman
2016-07-20  1:58       ` [PATCH 01/10] sysctl: Stop implicitly passing current into sysctl_table_root.lookup Eric W. Biederman
2016-07-20  5:06       ` Serge E. Hallyn
2016-07-20  4:02   ` [PATCH 0/10] userns: sysctl limits for namespaces Kees Cook
     [not found]     ` <CAGXu5jKWjHKxtPSGuogjR+XE+SbNWiUp2bOxugpD+hUex9QeZA-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-07-21 14:49       ` Eric W. Biederman
2016-07-21 16:58       ` Eric W. Biederman
     [not found]         ` <871t2n53o5.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-21 18:46           ` Kees Cook
2016-07-21 16:39   ` [PATCH v2 00/10] " Eric W. Biederman
2016-07-21 16:39     ` Eric W. Biederman
2016-07-22 13:33     ` Colin Walters
     [not found]       ` <1469194399.3817016.673814953.7581706C-2RFepEojUI2N1INw9kWLP6GC3tUn3ZHUQQ4Iyu8u01E@public.gmane.org>
2016-07-22 18:45         ` Eric W. Biederman [this message]
2016-07-22 18:45           ` Eric W. Biederman
2016-07-22 21:46           ` Kees Cook
2016-07-23  2:11             ` Eric W. Biederman
2016-07-23  2:11               ` Eric W. Biederman
2016-07-23  2:11               ` Eric W. Biederman
     [not found]             ` <CAGXu5j+X7eWggkwpBpABsFe4hqK5LN1mYJ2TH91qj3iSe6rtcQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-07-23  2:11               ` Eric W. Biederman
     [not found]           ` <87poq5y0jw.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-22 21:46             ` Kees Cook
     [not found]     ` <87d1m754jc.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-21 16:40       ` [PATCH v2 01/10] sysctl: Stop implicitly passing current into sysctl_table_root.lookup Eric W. Biederman
2016-07-21 16:40         ` Eric W. Biederman
2016-07-21 16:40         ` [PATCH v2 04/10] userns: Generalize the user namespace count into ucount Eric W. Biederman
     [not found]           ` <20160721164014.17534-4-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:09             ` Serge E. Hallyn
2016-07-25 23:09           ` Serge E. Hallyn
2016-07-25 23:09             ` Serge E. Hallyn
2016-07-21 16:40         ` [PATCH v2 06/10] utsns: Add a limit on the number of uts namespaces Eric W. Biederman
     [not found]           ` <20160721164014.17534-6-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:09             ` Serge E. Hallyn
2016-07-25 23:09               ` Serge E. Hallyn
     [not found]         ` <20160721164014.17534-1-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-21 16:40           ` [PATCH v2 02/10] userns: Add per user namespace sysctls Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
     [not found]             ` <20160721164014.17534-2-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-26  0:02               ` Eric W. Biederman
2016-07-26  0:02                 ` Eric W. Biederman
     [not found]                 ` <878twp6zd2.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-26  0:24                   ` David Miller
2016-07-26  0:24                     ` David Miller
     [not found]                     ` <20160725.172406.352408511647766870.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2016-07-26  0:44                       ` Eric W. Biederman
2016-07-26  0:44                         ` Eric W. Biederman
     [not found]                         ` <87k2g95it9.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-26  2:58                           ` David Miller
2016-07-26  2:58                             ` David Miller
     [not found]                             ` <20160725.195836.1881375122073635224.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
2016-07-26  4:00                               ` Eric W. Biederman
2016-07-26  4:00                             ` Eric W. Biederman
2016-07-26  4:00                               ` Eric W. Biederman
2016-07-21 16:40           ` [PATCH v2 03/10] userns: Add a limit on the number of user namespaces Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
     [not found]             ` <20160721164014.17534-3-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:05               ` Serge E. Hallyn
2016-07-25 23:05             ` Serge E. Hallyn
2016-07-21 16:40           ` [PATCH v2 04/10] userns: Generalize the user namespace count into ucount Eric W. Biederman
2016-07-21 16:40           ` [PATCH v2 05/10] pidns: Add a limit on the number of pid namespaces Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
     [not found]             ` <20160721164014.17534-5-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:09               ` Serge E. Hallyn
2016-07-25 23:09                 ` Serge E. Hallyn
2016-07-21 16:40           ` [PATCH v2 06/10] utsns: Add a limit on the number of uts namespaces Eric W. Biederman
2016-07-21 16:40           ` [PATCH v2 07/10] ipcns: Add a limit on the number of ipc namespaces Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
2016-07-25 23:10             ` Serge E. Hallyn
     [not found]             ` <20160721164014.17534-7-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:10               ` Serge E. Hallyn
2016-07-21 16:40           ` [PATCH v2 08/10] cgroupns: Add a limit on the number of cgroup namespaces Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
2016-07-25 23:12             ` Serge E. Hallyn
2016-07-25 23:12               ` Serge E. Hallyn
     [not found]             ` <20160721164014.17534-8-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:12               ` Serge E. Hallyn
2016-07-21 16:40           ` [PATCH v2 09/10] netns: Add a limit on the number of net namespaces Eric W. Biederman
2016-07-21 16:40             ` Eric W. Biederman
2016-07-25 23:13             ` Serge E. Hallyn
     [not found]             ` <20160721164014.17534-9-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:13               ` Serge E. Hallyn
2016-07-26  6:01               ` Andrei Vagin
2016-07-26  6:01                 ` Andrei Vagin
     [not found]                 ` <CANaxB-zfkvu9XYkuYB47tYFGLeVnWBhpgEUaSrGD4+-O8ThcNw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-07-26 20:00                   ` Eric W. Biederman
2016-07-26 20:00                     ` Eric W. Biederman
2016-07-21 16:40           ` [PATCH v2 10/10] mntns: Add a limit on the number of mount namespaces Eric W. Biederman
2016-07-21 16:40         ` Eric W. Biederman
     [not found]           ` <20160721164014.17534-10-ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
2016-07-25 23:15             ` Serge E. Hallyn
2016-07-25 23:15               ` Serge E. Hallyn
2016-07-22 13:33       ` [PATCH v2 00/10] userns: sysctl limits for namespaces Colin Walters
2016-07-26 10:27       ` Michael Kerrisk (man-pages)
2016-07-26 10:27         ` Michael Kerrisk (man-pages)
     [not found]         ` <94b608ae-1d06-5c41-cbd5-94e663a2163a-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2016-07-26 15:14           ` Eric W. Biederman
2016-07-26 15:14             ` Eric W. Biederman
2016-07-26 15:14             ` Eric W. Biederman
2016-07-26 10:30       ` Michael Kerrisk (man-pages)
2016-07-26 10:30         ` Michael Kerrisk (man-pages)
     [not found]         ` <6be70177-a81d-7ed8-d2c9-a596d4d6a165-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2016-07-26 15:06           ` Eric W. Biederman
2016-07-26 15:06         ` Eric W. Biederman
2016-07-26 15:06           ` Eric W. Biederman
2016-07-26 15:06           ` Eric W. Biederman
2016-07-26 16:52           ` Kees Cook
2016-07-26 16:52             ` Kees Cook
     [not found]             ` <CAGXu5j+ZtZAFpW0ajnnBTRZk17ez6XYF=meEJY_rF7wwgyZB5Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-07-26 17:29               ` Michael Kerrisk (man-pages)
2016-07-26 17:29             ` Michael Kerrisk (man-pages)
2016-07-26 17:29               ` Michael Kerrisk (man-pages)
     [not found]               ` <CAKgNAkhCvb-WkBbwebbBw6cfhg0FJ-SSbvK=aS1WK+BQ55eO5Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-07-26 20:44                 ` Kees Cook
2016-07-26 20:44               ` Kees Cook
     [not found]           ` <87shuwtp4c.fsf-JOvCrm2gF+uungPnsOpG7nhyD016LWXt@public.gmane.org>
2016-07-26 16:52             ` Kees Cook
2016-08-08 21:16       ` Eric W. Biederman
2016-08-08 21:16         ` Eric W. Biederman
2016-08-08 21:16         ` Eric W. Biederman

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=87poq5y0jw.fsf@x220.int.ebiederm.org \
    --to=ebiederm-as9lmozglivwk0htik3j/w@public.gmane.org \
    --cc=containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
    --cc=jann-XZ1E9jl8jIdeoWH0uzbU5w@public.gmane.org \
    --cc=keescook-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org \
    --cc=kernel-6AxghH7DbtA@public.gmane.org \
    --cc=linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=luto-kltTT9wpgjJwATOyAt5JVQ@public.gmane.org \
    --cc=netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=seth.forshee-Z7WLFzj8eWMS+FvcfC7Uqw@public.gmane.org \
    --cc=walters-gPq2gbYjIk8dnm+yROfE0A@public.gmane.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.