2019-06-01 10:08:55 +02:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2006-10-02 02:18:06 -07:00
|
|
|
/*
|
|
|
|
|
* Copyright (C) 2006 IBM Corporation
|
|
|
|
|
*
|
|
|
|
|
* Author: Serge Hallyn <serue@us.ibm.com>
|
|
|
|
|
*
|
2006-10-02 02:18:19 -07:00
|
|
|
* Jun 2006 - namespaces support
|
|
|
|
|
* OpenVZ, SWsoft Inc.
|
|
|
|
|
* Pavel Emelianov <xemul@openvz.org>
|
2006-10-02 02:18:06 -07:00
|
|
|
*/
|
|
|
|
|
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 17:04:11 +09:00
|
|
|
#include <linux/slab.h>
|
2011-05-23 14:51:41 -04:00
|
|
|
#include <linux/export.h>
|
2006-10-02 02:18:06 -07:00
|
|
|
#include <linux/nsproxy.h>
|
nsproxy: Add FOR_EACH_NS_TYPE() X-macro and CLONE_NS_ALL
Introduce the FOR_EACH_NS_TYPE(X) macro as the single source of truth
for the set of (struct type, CLONE_NEW* flag) pairs that define Linux
namespace types.
Currently, the list of CLONE_NEW* flags is duplicated inline in
multiple call sites and would need another copy in each new consumer.
This makes it easy to miss one when a new namespace type is added.
Derive two things from the X-macro:
- CLONE_NS_ALL: Bitmask of all known CLONE_NEW* flags, usable as a
validity mask or iteration bound.
- ns_common_type(): Rewritten to use the X-macro via a leading-comma
_Generic pattern, so the struct-to-flag mapping stays in sync with the
flag set automatically.
Replace the inline flag enumerations in copy_namespaces(),
unshare_nsproxy_namespaces(), check_setns_flags(), and
ksys_unshare() with CLONE_NS_ALL.
When a new namespace type is added, only FOR_EACH_NS_TYPE needs to
be updated; CLONE_NS_ALL, ns_common_type(), and all the call sites
pick up the change automatically.
Cc: Christian Brauner <brauner@kernel.org>
Cc: Günther Noack <gnoack@google.com>
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Link: https://patch.msgid.link/20260312100444.2609563-4-mic@digikod.net
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
2026-03-12 11:04:36 +01:00
|
|
|
#include <linux/ns/ns_common_types.h>
|
2006-10-02 02:18:07 -07:00
|
|
|
#include <linux/init_task.h>
|
2006-12-08 02:37:56 -08:00
|
|
|
#include <linux/mnt_namespace.h>
|
2006-10-02 02:18:14 -07:00
|
|
|
#include <linux/utsname.h>
|
2006-12-08 02:37:59 -08:00
|
|
|
#include <linux/pid_namespace.h>
|
2007-09-26 22:04:26 -07:00
|
|
|
#include <net/net_namespace.h>
|
2008-02-08 04:18:22 -08:00
|
|
|
#include <linux/ipc_namespace.h>
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
#include <linux/time_namespace.h>
|
2020-05-05 16:04:30 +02:00
|
|
|
#include <linux/fs_struct.h>
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
#include <linux/proc_fs.h>
|
2013-04-12 01:50:06 +01:00
|
|
|
#include <linux/proc_ns.h>
|
2010-03-07 17:48:52 -08:00
|
|
|
#include <linux/file.h>
|
|
|
|
|
#include <linux/syscalls.h>
|
2016-01-29 02:54:06 -06:00
|
|
|
#include <linux/cgroup.h>
|
2017-03-08 02:11:36 +05:30
|
|
|
#include <linux/perf_event.h>
|
ns: add active reference count
The namespace tree is, among other things, currently used to support
file handles for namespaces. When a namespace is created it is placed on
the namespace trees and when it is destroyed it is removed from the
namespace trees.
While a namespace is on the namespace trees with a valid reference count
it is possible to reopen it through a namespace file handle. This is all
fine but has some issues that should be addressed.
On current kernels a namespace is visible to userspace in the
following cases:
(1) The namespace is in use by a task.
(2) The namespace is persisted through a VFS object (namespace file
descriptor or bind-mount).
Note that (2) only cares about direct persistence of the namespace
itself not indirectly via e.g., file->f_cred file references or
similar.
(3) The namespace is a hierarchical namespace type and is the parent of
a single or multiple child namespaces.
Case (3) is interesting because it is possible that a parent namespace
might not fulfill any of (1) or (2), i.e., is invisible to userspace but
it may still be resurrected through the NS_GET_PARENT ioctl().
Currently namespace file handles allow much broader access to namespaces
than what is currently possible via (1)-(3). The reason is that
namespaces may remain pinned for completely internal reasons yet are
inaccessible to userspace.
For example, a user namespace my remain pinned by get_cred() calls to
stash the opener's credentials into file->f_cred. As it stands file
handles allow to resurrect such a users namespace even though this
should not be possible via (1)-(3). This is a fundamental uapi change
that we shouldn't do if we don't have to.
Consider the following insane case: Various architectures support the
CONFIG_MMU_LAZY_TLB_REFCOUNT option which uses lazy TLB destruction.
When this option is set a userspace task's struct mm_struct may be used
for kernel threads such as the idle task and will only be destroyed once
the cpu's runqueue switches back to another task. But because of ptrace()
permission checks struct mm_struct stashes the user namespace of the
task that struct mm_struct originally belonged to. The kernel thread
will take a reference on the struct mm_struct and thus pin it.
So on an idle system user namespaces can be persisted for arbitrary
amounts of time which also means that they can be resurrected using
namespace file handles. That makes no sense whatsoever. The problem is
of course excarabted on large systems with a huge number of cpus.
To handle this nicely we introduce an active reference count which
tracks (1)-(3). This is easy to do as all of these things are already
managed centrally. Only (1)-(3) will count towards the active reference
count and only namespaces which are active may be opened via namespace
file handles.
The problem is that namespaces may be resurrected. Which means that they
can become temporarily inactive and will be reactived some time later.
Currently the only example of this is the SIOGCSKNS socket ioctl. The
SIOCGSKNS ioctl allows to open a network namespace file descriptor based
on a socket file descriptor.
If a socket is tied to a network namespace that subsequently becomes
inactive but that socket is persisted by another process in another
network namespace (e.g., via SCM_RIGHTS of pidfd_getfd()) then the
SIOCGSKNS ioctl will resurrect this network namespace.
So calls to open_related_ns() and open_namespace() will end up
resurrecting the corresponding namespace tree.
Note that the active reference count does not regulate the lifetime of
the namespace itself. This is still done by the normal reference count.
The active reference count can only be elevated if the regular reference
count is elevated.
The active reference count also doesn't regulate the presence of a
namespace on the namespace trees. It only regulates its visiblity to
namespace file handles (and in later patches to listns()).
A namespace remains on the namespace trees from creation until its
actual destruction. This will allow the kernel to always reach any
namespace trivially and it will also enable subsystems like bpf to walk
the namespace lists on the system for tracing or general introspection
purposes.
Note that different namespaces have different visibility lifetimes on
current kernels. While most namespace are immediately released when the
last task using them exits, the user- and pid namespace are persisted
and thus both remain accessible via /proc/<pid>/ns/<ns_type>.
The user namespace lifetime is aliged with struct cred and is only
released through exit_creds(). However, it becomes inaccessible to
userspace once the last task using it is reaped, i.e., when
release_task() is called and all proc entries are flushed. Similarly,
the pid namespace is also visible until the last task using it has been
reaped and the associated pid numbers are freed.
The active reference counts of the user- and pid namespace are
decremented once the task is reaped.
Link: https://patch.msgid.link/20251029-work-namespace-nstree-listns-v4-11-2e6f823ebdc0@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-10-29 13:20:24 +01:00
|
|
|
#include <linux/nstree.h>
|
2006-10-02 02:18:07 -07:00
|
|
|
|
2007-07-15 23:41:07 -07:00
|
|
|
static struct kmem_cache *nsproxy_cachep;
|
|
|
|
|
|
2010-03-10 15:23:10 -08:00
|
|
|
struct nsproxy init_nsproxy = {
|
2023-08-17 21:13:32 -07:00
|
|
|
.count = REFCOUNT_INIT(1),
|
2013-08-22 11:39:16 -07:00
|
|
|
.uts_ns = &init_uts_ns,
|
2010-03-10 15:23:10 -08:00
|
|
|
#if defined(CONFIG_POSIX_MQUEUE) || defined(CONFIG_SYSVIPC)
|
2013-08-22 11:39:16 -07:00
|
|
|
.ipc_ns = &init_ipc_ns,
|
2010-03-10 15:23:10 -08:00
|
|
|
#endif
|
2013-08-22 11:39:16 -07:00
|
|
|
.mnt_ns = NULL,
|
|
|
|
|
.pid_ns_for_children = &init_pid_ns,
|
2010-03-10 15:23:10 -08:00
|
|
|
#ifdef CONFIG_NET
|
2013-08-22 11:39:16 -07:00
|
|
|
.net_ns = &init_net,
|
2010-03-10 15:23:10 -08:00
|
|
|
#endif
|
2016-01-29 02:54:06 -06:00
|
|
|
#ifdef CONFIG_CGROUPS
|
|
|
|
|
.cgroup_ns = &init_cgroup_ns,
|
|
|
|
|
#endif
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
#ifdef CONFIG_TIME_NS
|
|
|
|
|
.time_ns = &init_time_ns,
|
|
|
|
|
.time_ns_for_children = &init_time_ns,
|
|
|
|
|
#endif
|
2010-03-10 15:23:10 -08:00
|
|
|
};
|
2006-10-02 02:18:06 -07:00
|
|
|
|
2009-06-17 16:27:56 -07:00
|
|
|
static inline struct nsproxy *create_nsproxy(void)
|
2006-10-02 02:18:06 -07:00
|
|
|
{
|
2009-06-17 16:27:56 -07:00
|
|
|
struct nsproxy *nsproxy;
|
2006-10-02 02:18:06 -07:00
|
|
|
|
2009-06-17 16:27:56 -07:00
|
|
|
nsproxy = kmem_cache_alloc(nsproxy_cachep, GFP_KERNEL);
|
|
|
|
|
if (nsproxy)
|
2023-08-17 21:13:32 -07:00
|
|
|
refcount_set(&nsproxy->count, 1);
|
2009-06-17 16:27:56 -07:00
|
|
|
return nsproxy;
|
2006-10-02 02:18:06 -07:00
|
|
|
}
|
|
|
|
|
|
2025-11-11 22:29:44 +01:00
|
|
|
static inline void nsproxy_free(struct nsproxy *ns)
|
|
|
|
|
{
|
|
|
|
|
put_mnt_ns(ns->mnt_ns);
|
|
|
|
|
put_uts_ns(ns->uts_ns);
|
|
|
|
|
put_ipc_ns(ns->ipc_ns);
|
|
|
|
|
put_pid_ns(ns->pid_ns_for_children);
|
|
|
|
|
put_time_ns(ns->time_ns);
|
|
|
|
|
put_time_ns(ns->time_ns_for_children);
|
|
|
|
|
put_cgroup_ns(ns->cgroup_ns);
|
|
|
|
|
put_net(ns->net_ns);
|
|
|
|
|
kmem_cache_free(nsproxy_cachep, ns);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void deactivate_nsproxy(struct nsproxy *ns)
|
|
|
|
|
{
|
|
|
|
|
nsproxy_ns_active_put(ns);
|
|
|
|
|
nsproxy_free(ns);
|
|
|
|
|
}
|
|
|
|
|
|
2006-10-02 02:18:06 -07:00
|
|
|
/*
|
2007-05-08 00:25:21 -07:00
|
|
|
* Create new nsproxy and all of its the associated namespaces.
|
|
|
|
|
* Return the newly created nsproxy. Do not attach this to the task,
|
|
|
|
|
* leave it to the caller to do proper locking and attach it to task.
|
2006-10-02 02:18:06 -07:00
|
|
|
*/
|
2025-09-01 15:09:51 +02:00
|
|
|
static struct nsproxy *create_new_namespaces(u64 flags,
|
2012-07-26 04:02:49 -07:00
|
|
|
struct task_struct *tsk, struct user_namespace *user_ns,
|
|
|
|
|
struct fs_struct *new_fs)
|
2006-10-02 02:18:06 -07:00
|
|
|
{
|
2007-05-08 00:25:21 -07:00
|
|
|
struct nsproxy *new_nsp;
|
2007-07-15 23:41:06 -07:00
|
|
|
int err;
|
2006-10-02 02:18:06 -07:00
|
|
|
|
2009-06-17 16:27:56 -07:00
|
|
|
new_nsp = create_nsproxy();
|
2007-05-08 00:25:21 -07:00
|
|
|
if (!new_nsp)
|
|
|
|
|
return ERR_PTR(-ENOMEM);
|
2006-10-02 02:18:08 -07:00
|
|
|
|
2026-03-06 17:28:37 +01:00
|
|
|
new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns,
|
|
|
|
|
user_ns, new_fs);
|
2007-07-15 23:41:06 -07:00
|
|
|
if (IS_ERR(new_nsp->mnt_ns)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->mnt_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
goto out_ns;
|
2007-07-15 23:41:06 -07:00
|
|
|
}
|
2007-05-08 00:25:21 -07:00
|
|
|
|
2012-07-26 04:02:49 -07:00
|
|
|
new_nsp->uts_ns = copy_utsname(flags, user_ns, tsk->nsproxy->uts_ns);
|
2007-07-15 23:41:06 -07:00
|
|
|
if (IS_ERR(new_nsp->uts_ns)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->uts_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
goto out_uts;
|
2007-07-15 23:41:06 -07:00
|
|
|
}
|
2007-05-08 00:25:21 -07:00
|
|
|
|
2012-07-26 04:02:49 -07:00
|
|
|
new_nsp->ipc_ns = copy_ipcs(flags, user_ns, tsk->nsproxy->ipc_ns);
|
2007-07-15 23:41:06 -07:00
|
|
|
if (IS_ERR(new_nsp->ipc_ns)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->ipc_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
goto out_ipc;
|
2007-07-15 23:41:06 -07:00
|
|
|
}
|
2007-05-08 00:25:21 -07:00
|
|
|
|
2013-08-22 11:39:16 -07:00
|
|
|
new_nsp->pid_ns_for_children =
|
|
|
|
|
copy_pid_ns(flags, user_ns, tsk->nsproxy->pid_ns_for_children);
|
|
|
|
|
if (IS_ERR(new_nsp->pid_ns_for_children)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->pid_ns_for_children);
|
2007-05-08 00:25:21 -07:00
|
|
|
goto out_pid;
|
2007-07-15 23:41:06 -07:00
|
|
|
}
|
2007-05-08 00:25:21 -07:00
|
|
|
|
2016-01-29 02:54:06 -06:00
|
|
|
new_nsp->cgroup_ns = copy_cgroup_ns(flags, user_ns,
|
|
|
|
|
tsk->nsproxy->cgroup_ns);
|
|
|
|
|
if (IS_ERR(new_nsp->cgroup_ns)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->cgroup_ns);
|
|
|
|
|
goto out_cgroup;
|
|
|
|
|
}
|
|
|
|
|
|
2012-07-26 04:02:49 -07:00
|
|
|
new_nsp->net_ns = copy_net_ns(flags, user_ns, tsk->nsproxy->net_ns);
|
2007-09-26 22:04:26 -07:00
|
|
|
if (IS_ERR(new_nsp->net_ns)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->net_ns);
|
|
|
|
|
goto out_net;
|
|
|
|
|
}
|
|
|
|
|
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
new_nsp->time_ns_for_children = copy_time_ns(flags, user_ns,
|
|
|
|
|
tsk->nsproxy->time_ns_for_children);
|
|
|
|
|
if (IS_ERR(new_nsp->time_ns_for_children)) {
|
|
|
|
|
err = PTR_ERR(new_nsp->time_ns_for_children);
|
|
|
|
|
goto out_time;
|
|
|
|
|
}
|
|
|
|
|
new_nsp->time_ns = get_time_ns(tsk->nsproxy->time_ns);
|
|
|
|
|
|
2007-05-08 00:25:21 -07:00
|
|
|
return new_nsp;
|
|
|
|
|
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
out_time:
|
|
|
|
|
put_net(new_nsp->net_ns);
|
2007-09-26 22:04:26 -07:00
|
|
|
out_net:
|
2016-01-29 02:54:06 -06:00
|
|
|
put_cgroup_ns(new_nsp->cgroup_ns);
|
|
|
|
|
out_cgroup:
|
2025-05-08 14:49:29 -04:00
|
|
|
put_pid_ns(new_nsp->pid_ns_for_children);
|
2007-05-08 00:25:21 -07:00
|
|
|
out_pid:
|
2025-05-08 14:49:29 -04:00
|
|
|
put_ipc_ns(new_nsp->ipc_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
out_ipc:
|
2025-05-08 14:49:29 -04:00
|
|
|
put_uts_ns(new_nsp->uts_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
out_uts:
|
2025-05-08 14:49:29 -04:00
|
|
|
put_mnt_ns(new_nsp->mnt_ns);
|
2007-05-08 00:25:21 -07:00
|
|
|
out_ns:
|
2007-07-15 23:41:07 -07:00
|
|
|
kmem_cache_free(nsproxy_cachep, new_nsp);
|
2007-07-15 23:41:06 -07:00
|
|
|
return ERR_PTR(err);
|
2006-10-02 02:18:06 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* called from clone. This now handles copy for nsproxy and all
|
|
|
|
|
* namespaces therein.
|
|
|
|
|
*/
|
2025-09-01 15:09:51 +02:00
|
|
|
int copy_namespaces(u64 flags, struct task_struct *tsk)
|
2006-10-02 02:18:06 -07:00
|
|
|
{
|
|
|
|
|
struct nsproxy *old_ns = tsk->nsproxy;
|
2012-07-26 00:50:47 -07:00
|
|
|
struct user_namespace *user_ns = task_cred_xxx(tsk, user_ns);
|
2006-10-02 02:18:08 -07:00
|
|
|
struct nsproxy *new_ns;
|
2006-10-02 02:18:06 -07:00
|
|
|
|
nsproxy: Add FOR_EACH_NS_TYPE() X-macro and CLONE_NS_ALL
Introduce the FOR_EACH_NS_TYPE(X) macro as the single source of truth
for the set of (struct type, CLONE_NEW* flag) pairs that define Linux
namespace types.
Currently, the list of CLONE_NEW* flags is duplicated inline in
multiple call sites and would need another copy in each new consumer.
This makes it easy to miss one when a new namespace type is added.
Derive two things from the X-macro:
- CLONE_NS_ALL: Bitmask of all known CLONE_NEW* flags, usable as a
validity mask or iteration bound.
- ns_common_type(): Rewritten to use the X-macro via a leading-comma
_Generic pattern, so the struct-to-flag mapping stays in sync with the
flag set automatically.
Replace the inline flag enumerations in copy_namespaces(),
unshare_nsproxy_namespaces(), check_setns_flags(), and
ksys_unshare() with CLONE_NS_ALL.
When a new namespace type is added, only FOR_EACH_NS_TYPE needs to
be updated; CLONE_NS_ALL, ns_common_type(), and all the call sites
pick up the change automatically.
Cc: Christian Brauner <brauner@kernel.org>
Cc: Günther Noack <gnoack@google.com>
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Link: https://patch.msgid.link/20260312100444.2609563-4-mic@digikod.net
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
2026-03-12 11:04:36 +01:00
|
|
|
if (likely(!(flags & (CLONE_NS_ALL & ~CLONE_NEWUSER)))) {
|
2022-09-20 17:31:19 -07:00
|
|
|
if ((flags & CLONE_VM) ||
|
|
|
|
|
likely(old_ns->time_ns_for_children == old_ns->time_ns)) {
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
get_nsproxy(old_ns);
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
} else if (!ns_capable(user_ns, CAP_SYS_ADMIN))
|
2013-03-09 16:15:23 -08:00
|
|
|
return -EPERM;
|
|
|
|
|
|
2008-04-29 01:01:00 -07:00
|
|
|
/*
|
|
|
|
|
* CLONE_NEWIPC must detach from the undolist: after switching
|
|
|
|
|
* to a new ipc namespace, the semaphore arrays from the old
|
|
|
|
|
* namespace are unreachable. In clone parlance, CLONE_SYSVSEM
|
|
|
|
|
* means share undolist with parent, so we must forbid using
|
|
|
|
|
* it along with CLONE_NEWIPC.
|
|
|
|
|
*/
|
2013-02-27 15:32:09 -03:00
|
|
|
if ((flags & (CLONE_NEWIPC | CLONE_SYSVSEM)) ==
|
2020-11-16 02:00:54 +08:00
|
|
|
(CLONE_NEWIPC | CLONE_SYSVSEM))
|
2013-03-09 16:15:23 -08:00
|
|
|
return -EINVAL;
|
2008-04-29 01:01:00 -07:00
|
|
|
|
2013-02-21 16:44:21 -08:00
|
|
|
new_ns = create_new_namespaces(flags, tsk, user_ns, tsk->fs);
|
2013-03-09 16:15:23 -08:00
|
|
|
if (IS_ERR(new_ns))
|
|
|
|
|
return PTR_ERR(new_ns);
|
2006-10-02 02:18:08 -07:00
|
|
|
|
2022-09-20 17:31:19 -07:00
|
|
|
if ((flags & CLONE_VM) == 0)
|
|
|
|
|
timens_on_fork(new_ns, tsk);
|
ns: Introduce Time Namespace
Time Namespace isolates clock values.
The kernel provides access to several clocks CLOCK_REALTIME,
CLOCK_MONOTONIC, CLOCK_BOOTTIME, etc.
CLOCK_REALTIME
System-wide clock that measures real (i.e., wall-clock) time.
CLOCK_MONOTONIC
Clock that cannot be set and represents monotonic time since
some unspecified starting point.
CLOCK_BOOTTIME
Identical to CLOCK_MONOTONIC, except it also includes any time
that the system is suspended.
For many users, the time namespace means the ability to changes date and
time in a container (CLOCK_REALTIME). Providing per namespace notions of
CLOCK_REALTIME would be complex with a massive overhead, but has a dubious
value.
But in the context of checkpoint/restore functionality, monotonic and
boottime clocks become interesting. Both clocks are monotonic with
unspecified starting points. These clocks are widely used to measure time
slices and set timers. After restoring or migrating processes, it has to be
guaranteed that they never go backward. In an ideal case, the behavior of
these clocks should be the same as for a case when a whole system is
suspended. All this means that it is required to set CLOCK_MONOTONIC and
CLOCK_BOOTTIME clocks, which can be achieved by adding per-namespace
offsets for clocks.
A time namespace is similar to a pid namespace in the way how it is
created: unshare(CLONE_NEWTIME) system call creates a new time namespace,
but doesn't set it to the current process. Then all children of the process
will be born in the new time namespace, or a process can use the setns()
system call to join a namespace.
This scheme allows setting clock offsets for a namespace, before any
processes appear in it.
All available clone flags have been used, so CLONE_NEWTIME uses the highest
bit of CSIGNAL. It means that it can be used only with the unshare() and
the clone3() system calls.
[ tglx: Adjusted paragraph about clone3() to reality and massaged the
changelog a bit. ]
Co-developed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://criu.org/Time_namespace
Link: https://lists.openvz.org/pipermail/criu/2018-June/041504.html
Link: https://lore.kernel.org/r/20191112012724.250792-4-dima@arista.com
2019-11-12 01:26:52 +00:00
|
|
|
|
ns: add active reference count
The namespace tree is, among other things, currently used to support
file handles for namespaces. When a namespace is created it is placed on
the namespace trees and when it is destroyed it is removed from the
namespace trees.
While a namespace is on the namespace trees with a valid reference count
it is possible to reopen it through a namespace file handle. This is all
fine but has some issues that should be addressed.
On current kernels a namespace is visible to userspace in the
following cases:
(1) The namespace is in use by a task.
(2) The namespace is persisted through a VFS object (namespace file
descriptor or bind-mount).
Note that (2) only cares about direct persistence of the namespace
itself not indirectly via e.g., file->f_cred file references or
similar.
(3) The namespace is a hierarchical namespace type and is the parent of
a single or multiple child namespaces.
Case (3) is interesting because it is possible that a parent namespace
might not fulfill any of (1) or (2), i.e., is invisible to userspace but
it may still be resurrected through the NS_GET_PARENT ioctl().
Currently namespace file handles allow much broader access to namespaces
than what is currently possible via (1)-(3). The reason is that
namespaces may remain pinned for completely internal reasons yet are
inaccessible to userspace.
For example, a user namespace my remain pinned by get_cred() calls to
stash the opener's credentials into file->f_cred. As it stands file
handles allow to resurrect such a users namespace even though this
should not be possible via (1)-(3). This is a fundamental uapi change
that we shouldn't do if we don't have to.
Consider the following insane case: Various architectures support the
CONFIG_MMU_LAZY_TLB_REFCOUNT option which uses lazy TLB destruction.
When this option is set a userspace task's struct mm_struct may be used
for kernel threads such as the idle task and will only be destroyed once
the cpu's runqueue switches back to another task. But because of ptrace()
permission checks struct mm_struct stashes the user namespace of the
task that struct mm_struct originally belonged to. The kernel thread
will take a reference on the struct mm_struct and thus pin it.
So on an idle system user namespaces can be persisted for arbitrary
amounts of time which also means that they can be resurrected using
namespace file handles. That makes no sense whatsoever. The problem is
of course excarabted on large systems with a huge number of cpus.
To handle this nicely we introduce an active reference count which
tracks (1)-(3). This is easy to do as all of these things are already
managed centrally. Only (1)-(3) will count towards the active reference
count and only namespaces which are active may be opened via namespace
file handles.
The problem is that namespaces may be resurrected. Which means that they
can become temporarily inactive and will be reactived some time later.
Currently the only example of this is the SIOGCSKNS socket ioctl. The
SIOCGSKNS ioctl allows to open a network namespace file descriptor based
on a socket file descriptor.
If a socket is tied to a network namespace that subsequently becomes
inactive but that socket is persisted by another process in another
network namespace (e.g., via SCM_RIGHTS of pidfd_getfd()) then the
SIOCGSKNS ioctl will resurrect this network namespace.
So calls to open_related_ns() and open_namespace() will end up
resurrecting the corresponding namespace tree.
Note that the active reference count does not regulate the lifetime of
the namespace itself. This is still done by the normal reference count.
The active reference count can only be elevated if the regular reference
count is elevated.
The active reference count also doesn't regulate the presence of a
namespace on the namespace trees. It only regulates its visiblity to
namespace file handles (and in later patches to listns()).
A namespace remains on the namespace trees from creation until its
actual destruction. This will allow the kernel to always reach any
namespace trivially and it will also enable subsystems like bpf to walk
the namespace lists on the system for tracing or general introspection
purposes.
Note that different namespaces have different visibility lifetimes on
current kernels. While most namespace are immediately released when the
last task using them exits, the user- and pid namespace are persisted
and thus both remain accessible via /proc/<pid>/ns/<ns_type>.
The user namespace lifetime is aliged with struct cred and is only
released through exit_creds(). However, it becomes inaccessible to
userspace once the last task using it is reaped, i.e., when
release_task() is called and all proc entries are flushed. Similarly,
the pid namespace is also visible until the last task using it has been
reaped and the associated pid numbers are freed.
The active reference counts of the user- and pid namespace are
decremented once the task is reaped.
Link: https://patch.msgid.link/20251029-work-namespace-nstree-listns-v4-11-2e6f823ebdc0@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-10-29 13:20:24 +01:00
|
|
|
nsproxy_ns_active_get(new_ns);
|
2006-10-02 02:18:08 -07:00
|
|
|
tsk->nsproxy = new_ns;
|
2013-03-09 16:15:23 -08:00
|
|
|
return 0;
|
2006-10-02 02:18:06 -07:00
|
|
|
}
|
|
|
|
|
|
2007-05-08 00:25:21 -07:00
|
|
|
/*
|
|
|
|
|
* Called from unshare. Unshare all the namespaces part of nsproxy.
|
2007-06-23 17:16:25 -07:00
|
|
|
* On success, returns the new nsproxy.
|
2007-05-08 00:25:21 -07:00
|
|
|
*/
|
|
|
|
|
int unshare_nsproxy_namespaces(unsigned long unshare_flags,
|
2012-07-26 05:15:35 -07:00
|
|
|
struct nsproxy **new_nsp, struct cred *new_cred, struct fs_struct *new_fs)
|
2007-05-08 00:25:21 -07:00
|
|
|
{
|
2012-07-26 04:02:49 -07:00
|
|
|
struct user_namespace *user_ns;
|
2026-03-06 17:28:37 +01:00
|
|
|
u64 flags = unshare_flags;
|
2007-05-08 00:25:21 -07:00
|
|
|
int err = 0;
|
|
|
|
|
|
vfs-7.1-rc1.mount.v2
Please consider pulling these changes from the signed vfs-7.1-rc1.mount.v2 tag.
Thanks!
Christian
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCad3vFgAKCRCRxhvAZXjc
onXwAQDwEGvpMUUiuI/JWFqCA5vY5LXXr/36wdcs0iUL1uy9IgEAyOdnYhYkcaX1
3lm87f6OmYkhlq6enJbco7uT4CUzlQA=
=1Ls8
-----END PGP SIGNATURE-----
Merge tag 'vfs-7.1-rc1.mount.v2' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs mount updates from Christian Brauner:
- Add FSMOUNT_NAMESPACE flag to fsmount() that creates a new mount
namespace with the newly created filesystem attached to a copy of the
real rootfs. This returns a namespace file descriptor instead of an
O_PATH mount fd, similar to how OPEN_TREE_NAMESPACE works for
open_tree().
This allows creating a new filesystem and immediately placing it in a
new mount namespace in a single operation, which is useful for
container runtimes and other namespace-based isolation mechanisms.
This accompanies OPEN_TREE_NAMESPACE and avoids a needless detour via
OPEN_TREE_NAMESPACE to get the same effect. Will be especially useful
when you mount an actual filesystem to be used as the container
rootfs.
- Currently, creating a new mount namespace always copies the entire
mount tree from the caller's namespace. For containers and sandboxes
that intend to build their mount table from scratch this is wasteful:
they inherit a potentially large mount tree only to immediately tear
it down.
This series adds support for creating a mount namespace that contains
only a clone of the root mount, with none of the child mounts. Two
new flags are introduced:
- CLONE_EMPTY_MNTNS (0x400000000) for clone3(), using the 64-bit flag space
- UNSHARE_EMPTY_MNTNS (0x00100000) for unshare()
Both flags imply CLONE_NEWNS. The resulting namespace contains a
single nullfs root mount with an immutable empty directory. The
intended workflow is to then mount a real filesystem (e.g., tmpfs)
over the root and build the mount table from there.
- Allow MOVE_MOUNT_BENEATH to target the caller's rootfs, allowing to
switch out the rootfs without pivot_root(2).
The traditional approach to switching the rootfs involves
pivot_root(2) or a chroot_fs_refs()-based mechanism that atomically
updates fs->root for all tasks sharing the same fs_struct. This has
consequences for fork(), unshare(CLONE_FS), and setns().
This series instead decomposes root-switching into individually
atomic, locally-scoped steps:
fd_tree = open_tree(-EBADF, "/newroot", OPEN_TREE_CLONE | OPEN_TREE_CLOEXEC);
fchdir(fd_tree);
move_mount(fd_tree, "", AT_FDCWD, "/", MOVE_MOUNT_BENEATH | MOVE_MOUNT_F_EMPTY_PATH);
chroot(".");
umount2(".", MNT_DETACH);
Since each step only modifies the caller's own state, the
fork/unshare/setns races are eliminated by design.
A key step to making this possible is to remove the locked mount
restriction. Originally MOVE_MOUNT_BENEATH doesn't support mounting
beneath a mount that is locked. The locked mount protects the
underlying mount from being revealed. This is a core mechanism of
unshare(CLONE_NEWUSER | CLONE_NEWNS). The mounts in the new mount
namespace become locked. That effectively makes the new mount table
useless as the caller cannot ever get rid of any of the mounts no
matter how useless they are.
We can lift this restriction though. We simply transfer the locked
property from the top mount to the mount beneath. This works because
what we care about is to protect the underlying mount aka the parent.
The mount mounted between the parent and the top mount takes over the
job of protecting the parent mount from the top mount mount. This
leaves us free to remove the locked property from the top mount which
can consequently be unmounted:
unshare(CLONE_NEWUSER | CLONE_NEWNS)
and we inherit a clone of procfs on /proc then currently we cannot
unmount it as:
umount -l /proc
will fail with EINVAL because the procfs mount is locked.
After this series we can now do:
mount --beneath -t tmpfs tmpfs /proc
umount -l /proc
after which a tmpfs mount has been placed beneath the procfs mount.
The tmpfs mount has become locked and the procfs mount has become
unlocked.
This means you can safely modify an inherited mount table after
unprivileged namespace creation.
Afterwards we simply make it possible to move a mount beneath the
rootfs allowing to upgrade the rootfs.
Removing the locked restriction makes this very useful for containers
created with unshare(CLONE_NEWUSER | CLONE_NEWNS) to reshuffle an
inherited mount table safely and MOVE_MOUNT_BENEATH makes it possible
to switch out the rootfs instead of using the costly pivot_root(2).
* tag 'vfs-7.1-rc1.mount.v2' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
selftests/namespaces: remove unused utils.h include from listns_efault_test
selftests/fsmount_ns: add missing TARGETS and fix cap test
selftests/empty_mntns: fix wrong CLONE_EMPTY_MNTNS hex value in comment
selftests/empty_mntns: fix statmount_alloc() signature mismatch
selftests/statmount: remove duplicate wait_for_pid()
mount: always duplicate mount
selftests/filesystems: add MOVE_MOUNT_BENEATH rootfs tests
move_mount: allow MOVE_MOUNT_BENEATH on the rootfs
move_mount: transfer MNT_LOCKED
selftests/filesystems: add clone3 tests for empty mount namespaces
selftests/filesystems: add tests for empty mount namespaces
namespace: allow creating empty mount namespaces
selftests: add FSMOUNT_NAMESPACE tests
selftests/statmount: add statmount_alloc() helper
tools: update mount.h header
mount: add FSMOUNT_NAMESPACE
mount: simplify __do_loopback()
mount: start iterating from start of rbtree
2026-04-14 19:59:25 -07:00
|
|
|
if (!(flags & (CLONE_NS_ALL & ~CLONE_NEWUSER)))
|
2007-05-08 00:25:21 -07:00
|
|
|
return 0;
|
|
|
|
|
|
2012-07-26 05:15:35 -07:00
|
|
|
user_ns = new_cred ? new_cred->user_ns : current_user_ns();
|
|
|
|
|
if (!ns_capable(user_ns, CAP_SYS_ADMIN))
|
2007-05-08 00:25:21 -07:00
|
|
|
return -EPERM;
|
|
|
|
|
|
2026-03-06 17:28:37 +01:00
|
|
|
/*
|
|
|
|
|
* Convert the 32-bit UNSHARE_EMPTY_MNTNS (which aliases
|
|
|
|
|
* CLONE_PARENT_SETTID) to the unique 64-bit CLONE_EMPTY_MNTNS.
|
|
|
|
|
*/
|
|
|
|
|
if (flags & UNSHARE_EMPTY_MNTNS) {
|
|
|
|
|
flags &= ~(u64)UNSHARE_EMPTY_MNTNS;
|
|
|
|
|
flags |= CLONE_EMPTY_MNTNS;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
*new_nsp = create_new_namespaces(flags, current, user_ns,
|
2012-07-26 05:15:35 -07:00
|
|
|
new_fs ? new_fs : current->fs);
|
2007-10-18 23:39:45 -07:00
|
|
|
if (IS_ERR(*new_nsp)) {
|
2007-05-08 00:25:21 -07:00
|
|
|
err = PTR_ERR(*new_nsp);
|
2007-10-18 23:39:45 -07:00
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
out:
|
2007-05-08 00:25:21 -07:00
|
|
|
return err;
|
|
|
|
|
}
|
2007-07-15 23:41:07 -07:00
|
|
|
|
2007-10-18 23:39:54 -07:00
|
|
|
void switch_task_namespaces(struct task_struct *p, struct nsproxy *new)
|
|
|
|
|
{
|
|
|
|
|
struct nsproxy *ns;
|
|
|
|
|
|
|
|
|
|
might_sleep();
|
|
|
|
|
|
ns: add active reference count
The namespace tree is, among other things, currently used to support
file handles for namespaces. When a namespace is created it is placed on
the namespace trees and when it is destroyed it is removed from the
namespace trees.
While a namespace is on the namespace trees with a valid reference count
it is possible to reopen it through a namespace file handle. This is all
fine but has some issues that should be addressed.
On current kernels a namespace is visible to userspace in the
following cases:
(1) The namespace is in use by a task.
(2) The namespace is persisted through a VFS object (namespace file
descriptor or bind-mount).
Note that (2) only cares about direct persistence of the namespace
itself not indirectly via e.g., file->f_cred file references or
similar.
(3) The namespace is a hierarchical namespace type and is the parent of
a single or multiple child namespaces.
Case (3) is interesting because it is possible that a parent namespace
might not fulfill any of (1) or (2), i.e., is invisible to userspace but
it may still be resurrected through the NS_GET_PARENT ioctl().
Currently namespace file handles allow much broader access to namespaces
than what is currently possible via (1)-(3). The reason is that
namespaces may remain pinned for completely internal reasons yet are
inaccessible to userspace.
For example, a user namespace my remain pinned by get_cred() calls to
stash the opener's credentials into file->f_cred. As it stands file
handles allow to resurrect such a users namespace even though this
should not be possible via (1)-(3). This is a fundamental uapi change
that we shouldn't do if we don't have to.
Consider the following insane case: Various architectures support the
CONFIG_MMU_LAZY_TLB_REFCOUNT option which uses lazy TLB destruction.
When this option is set a userspace task's struct mm_struct may be used
for kernel threads such as the idle task and will only be destroyed once
the cpu's runqueue switches back to another task. But because of ptrace()
permission checks struct mm_struct stashes the user namespace of the
task that struct mm_struct originally belonged to. The kernel thread
will take a reference on the struct mm_struct and thus pin it.
So on an idle system user namespaces can be persisted for arbitrary
amounts of time which also means that they can be resurrected using
namespace file handles. That makes no sense whatsoever. The problem is
of course excarabted on large systems with a huge number of cpus.
To handle this nicely we introduce an active reference count which
tracks (1)-(3). This is easy to do as all of these things are already
managed centrally. Only (1)-(3) will count towards the active reference
count and only namespaces which are active may be opened via namespace
file handles.
The problem is that namespaces may be resurrected. Which means that they
can become temporarily inactive and will be reactived some time later.
Currently the only example of this is the SIOGCSKNS socket ioctl. The
SIOCGSKNS ioctl allows to open a network namespace file descriptor based
on a socket file descriptor.
If a socket is tied to a network namespace that subsequently becomes
inactive but that socket is persisted by another process in another
network namespace (e.g., via SCM_RIGHTS of pidfd_getfd()) then the
SIOCGSKNS ioctl will resurrect this network namespace.
So calls to open_related_ns() and open_namespace() will end up
resurrecting the corresponding namespace tree.
Note that the active reference count does not regulate the lifetime of
the namespace itself. This is still done by the normal reference count.
The active reference count can only be elevated if the regular reference
count is elevated.
The active reference count also doesn't regulate the presence of a
namespace on the namespace trees. It only regulates its visiblity to
namespace file handles (and in later patches to listns()).
A namespace remains on the namespace trees from creation until its
actual destruction. This will allow the kernel to always reach any
namespace trivially and it will also enable subsystems like bpf to walk
the namespace lists on the system for tracing or general introspection
purposes.
Note that different namespaces have different visibility lifetimes on
current kernels. While most namespace are immediately released when the
last task using them exits, the user- and pid namespace are persisted
and thus both remain accessible via /proc/<pid>/ns/<ns_type>.
The user namespace lifetime is aliged with struct cred and is only
released through exit_creds(). However, it becomes inaccessible to
userspace once the last task using it is reaped, i.e., when
release_task() is called and all proc entries are flushed. Similarly,
the pid namespace is also visible until the last task using it has been
reaped and the associated pid numbers are freed.
The active reference counts of the user- and pid namespace are
decremented once the task is reaped.
Link: https://patch.msgid.link/20251029-work-namespace-nstree-listns-v4-11-2e6f823ebdc0@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-10-29 13:20:24 +01:00
|
|
|
if (new)
|
|
|
|
|
nsproxy_ns_active_get(new);
|
|
|
|
|
|
2014-02-03 19:13:49 -08:00
|
|
|
task_lock(p);
|
2007-10-18 23:39:54 -07:00
|
|
|
ns = p->nsproxy;
|
2014-02-03 19:13:49 -08:00
|
|
|
p->nsproxy = new;
|
|
|
|
|
task_unlock(p);
|
2007-10-18 23:39:54 -07:00
|
|
|
|
2020-11-16 02:00:54 +08:00
|
|
|
if (ns)
|
|
|
|
|
put_nsproxy(ns);
|
2007-10-18 23:39:54 -07:00
|
|
|
}
|
|
|
|
|
|
2025-10-29 13:20:23 +01:00
|
|
|
void exit_nsproxy_namespaces(struct task_struct *p)
|
2007-10-18 23:39:54 -07:00
|
|
|
{
|
|
|
|
|
switch_task_namespaces(p, NULL);
|
|
|
|
|
}
|
|
|
|
|
|
ns: add active reference count
The namespace tree is, among other things, currently used to support
file handles for namespaces. When a namespace is created it is placed on
the namespace trees and when it is destroyed it is removed from the
namespace trees.
While a namespace is on the namespace trees with a valid reference count
it is possible to reopen it through a namespace file handle. This is all
fine but has some issues that should be addressed.
On current kernels a namespace is visible to userspace in the
following cases:
(1) The namespace is in use by a task.
(2) The namespace is persisted through a VFS object (namespace file
descriptor or bind-mount).
Note that (2) only cares about direct persistence of the namespace
itself not indirectly via e.g., file->f_cred file references or
similar.
(3) The namespace is a hierarchical namespace type and is the parent of
a single or multiple child namespaces.
Case (3) is interesting because it is possible that a parent namespace
might not fulfill any of (1) or (2), i.e., is invisible to userspace but
it may still be resurrected through the NS_GET_PARENT ioctl().
Currently namespace file handles allow much broader access to namespaces
than what is currently possible via (1)-(3). The reason is that
namespaces may remain pinned for completely internal reasons yet are
inaccessible to userspace.
For example, a user namespace my remain pinned by get_cred() calls to
stash the opener's credentials into file->f_cred. As it stands file
handles allow to resurrect such a users namespace even though this
should not be possible via (1)-(3). This is a fundamental uapi change
that we shouldn't do if we don't have to.
Consider the following insane case: Various architectures support the
CONFIG_MMU_LAZY_TLB_REFCOUNT option which uses lazy TLB destruction.
When this option is set a userspace task's struct mm_struct may be used
for kernel threads such as the idle task and will only be destroyed once
the cpu's runqueue switches back to another task. But because of ptrace()
permission checks struct mm_struct stashes the user namespace of the
task that struct mm_struct originally belonged to. The kernel thread
will take a reference on the struct mm_struct and thus pin it.
So on an idle system user namespaces can be persisted for arbitrary
amounts of time which also means that they can be resurrected using
namespace file handles. That makes no sense whatsoever. The problem is
of course excarabted on large systems with a huge number of cpus.
To handle this nicely we introduce an active reference count which
tracks (1)-(3). This is easy to do as all of these things are already
managed centrally. Only (1)-(3) will count towards the active reference
count and only namespaces which are active may be opened via namespace
file handles.
The problem is that namespaces may be resurrected. Which means that they
can become temporarily inactive and will be reactived some time later.
Currently the only example of this is the SIOGCSKNS socket ioctl. The
SIOCGSKNS ioctl allows to open a network namespace file descriptor based
on a socket file descriptor.
If a socket is tied to a network namespace that subsequently becomes
inactive but that socket is persisted by another process in another
network namespace (e.g., via SCM_RIGHTS of pidfd_getfd()) then the
SIOCGSKNS ioctl will resurrect this network namespace.
So calls to open_related_ns() and open_namespace() will end up
resurrecting the corresponding namespace tree.
Note that the active reference count does not regulate the lifetime of
the namespace itself. This is still done by the normal reference count.
The active reference count can only be elevated if the regular reference
count is elevated.
The active reference count also doesn't regulate the presence of a
namespace on the namespace trees. It only regulates its visiblity to
namespace file handles (and in later patches to listns()).
A namespace remains on the namespace trees from creation until its
actual destruction. This will allow the kernel to always reach any
namespace trivially and it will also enable subsystems like bpf to walk
the namespace lists on the system for tracing or general introspection
purposes.
Note that different namespaces have different visibility lifetimes on
current kernels. While most namespace are immediately released when the
last task using them exits, the user- and pid namespace are persisted
and thus both remain accessible via /proc/<pid>/ns/<ns_type>.
The user namespace lifetime is aliged with struct cred and is only
released through exit_creds(). However, it becomes inaccessible to
userspace once the last task using it is reaped, i.e., when
release_task() is called and all proc entries are flushed. Similarly,
the pid namespace is also visible until the last task using it has been
reaped and the associated pid numbers are freed.
The active reference counts of the user- and pid namespace are
decremented once the task is reaped.
Link: https://patch.msgid.link/20251029-work-namespace-nstree-listns-v4-11-2e6f823ebdc0@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-10-29 13:20:24 +01:00
|
|
|
void switch_cred_namespaces(const struct cred *old, const struct cred *new)
|
|
|
|
|
{
|
|
|
|
|
ns_ref_active_get(new->user_ns);
|
|
|
|
|
ns_ref_active_put(old->user_ns);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void get_cred_namespaces(struct task_struct *tsk)
|
|
|
|
|
{
|
|
|
|
|
ns_ref_active_get(tsk->real_cred->user_ns);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void exit_cred_namespaces(struct task_struct *tsk)
|
|
|
|
|
{
|
|
|
|
|
ns_ref_active_put(tsk->real_cred->user_ns);
|
|
|
|
|
}
|
|
|
|
|
|
2022-09-20 17:31:19 -07:00
|
|
|
int exec_task_namespaces(void)
|
|
|
|
|
{
|
|
|
|
|
struct task_struct *tsk = current;
|
|
|
|
|
struct nsproxy *new;
|
|
|
|
|
|
|
|
|
|
if (tsk->nsproxy->time_ns_for_children == tsk->nsproxy->time_ns)
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
new = create_new_namespaces(0, tsk, current_user_ns(), tsk->fs);
|
|
|
|
|
if (IS_ERR(new))
|
|
|
|
|
return PTR_ERR(new);
|
|
|
|
|
|
|
|
|
|
timens_on_fork(new, tsk);
|
|
|
|
|
switch_task_namespaces(tsk, new);
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
static int check_setns_flags(unsigned long flags)
|
|
|
|
|
{
|
nsproxy: Add FOR_EACH_NS_TYPE() X-macro and CLONE_NS_ALL
Introduce the FOR_EACH_NS_TYPE(X) macro as the single source of truth
for the set of (struct type, CLONE_NEW* flag) pairs that define Linux
namespace types.
Currently, the list of CLONE_NEW* flags is duplicated inline in
multiple call sites and would need another copy in each new consumer.
This makes it easy to miss one when a new namespace type is added.
Derive two things from the X-macro:
- CLONE_NS_ALL: Bitmask of all known CLONE_NEW* flags, usable as a
validity mask or iteration bound.
- ns_common_type(): Rewritten to use the X-macro via a leading-comma
_Generic pattern, so the struct-to-flag mapping stays in sync with the
flag set automatically.
Replace the inline flag enumerations in copy_namespaces(),
unshare_nsproxy_namespaces(), check_setns_flags(), and
ksys_unshare() with CLONE_NS_ALL.
When a new namespace type is added, only FOR_EACH_NS_TYPE needs to
be updated; CLONE_NS_ALL, ns_common_type(), and all the call sites
pick up the change automatically.
Cc: Christian Brauner <brauner@kernel.org>
Cc: Günther Noack <gnoack@google.com>
Signed-off-by: Mickaël Salaün <mic@digikod.net>
Link: https://patch.msgid.link/20260312100444.2609563-4-mic@digikod.net
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
2026-03-12 11:04:36 +01:00
|
|
|
if (!flags || (flags & ~CLONE_NS_ALL))
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
|
|
#ifndef CONFIG_USER_NS
|
|
|
|
|
if (flags & CLONE_NEWUSER)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef CONFIG_PID_NS
|
|
|
|
|
if (flags & CLONE_NEWPID)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef CONFIG_UTS_NS
|
|
|
|
|
if (flags & CLONE_NEWUTS)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef CONFIG_IPC_NS
|
|
|
|
|
if (flags & CLONE_NEWIPC)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef CONFIG_CGROUPS
|
|
|
|
|
if (flags & CLONE_NEWCGROUP)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
|
|
|
|
#ifndef CONFIG_NET_NS
|
|
|
|
|
if (flags & CLONE_NEWNET)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
2020-07-06 17:49:11 +02:00
|
|
|
#ifndef CONFIG_TIME_NS
|
|
|
|
|
if (flags & CLONE_NEWTIME)
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
#endif
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2020-05-05 16:04:30 +02:00
|
|
|
static void put_nsset(struct nsset *nsset)
|
|
|
|
|
{
|
|
|
|
|
unsigned flags = nsset->flags;
|
|
|
|
|
|
|
|
|
|
if (flags & CLONE_NEWUSER)
|
|
|
|
|
put_cred(nsset_cred(nsset));
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
/*
|
|
|
|
|
* We only created a temporary copy if we attached to more than just
|
|
|
|
|
* the mount namespace.
|
|
|
|
|
*/
|
|
|
|
|
if (nsset->fs && (flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS))
|
|
|
|
|
free_fs_struct(nsset->fs);
|
2020-05-05 16:04:30 +02:00
|
|
|
if (nsset->nsproxy)
|
2025-11-11 22:29:44 +01:00
|
|
|
nsproxy_free(nsset->nsproxy);
|
2020-05-05 16:04:30 +02:00
|
|
|
}
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
static int prepare_nsset(unsigned flags, struct nsset *nsset)
|
2020-05-05 16:04:30 +02:00
|
|
|
{
|
|
|
|
|
struct task_struct *me = current;
|
|
|
|
|
|
|
|
|
|
nsset->nsproxy = create_new_namespaces(0, me, current_user_ns(), me->fs);
|
|
|
|
|
if (IS_ERR(nsset->nsproxy))
|
|
|
|
|
return PTR_ERR(nsset->nsproxy);
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
if (flags & CLONE_NEWUSER)
|
2020-05-05 16:04:30 +02:00
|
|
|
nsset->cred = prepare_creds();
|
|
|
|
|
else
|
|
|
|
|
nsset->cred = current_cred();
|
|
|
|
|
if (!nsset->cred)
|
|
|
|
|
goto out;
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
/* Only create a temporary copy of fs_struct if we really need to. */
|
|
|
|
|
if (flags == CLONE_NEWNS) {
|
2020-05-05 16:04:30 +02:00
|
|
|
nsset->fs = me->fs;
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
} else if (flags & CLONE_NEWNS) {
|
|
|
|
|
nsset->fs = copy_fs_struct(me->fs);
|
|
|
|
|
if (!nsset->fs)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
2020-05-05 16:04:30 +02:00
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
nsset->flags = flags;
|
2020-05-05 16:04:30 +02:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
out:
|
|
|
|
|
put_nsset(nsset);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
static inline int validate_ns(struct nsset *nsset, struct ns_common *ns)
|
|
|
|
|
{
|
|
|
|
|
return ns->ops->install(nsset, ns);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* This is the inverse operation to unshare().
|
|
|
|
|
* Ordering is equivalent to the standard ordering used everywhere else
|
|
|
|
|
* during unshare and process creation. The switch to the new set of
|
|
|
|
|
* namespaces occurs at the point of no return after installation of
|
|
|
|
|
* all requested namespaces was successful in commit_nsset().
|
|
|
|
|
*/
|
|
|
|
|
static int validate_nsset(struct nsset *nsset, struct pid *pid)
|
|
|
|
|
{
|
|
|
|
|
int ret = 0;
|
|
|
|
|
unsigned flags = nsset->flags;
|
|
|
|
|
struct user_namespace *user_ns = NULL;
|
|
|
|
|
struct pid_namespace *pid_ns = NULL;
|
|
|
|
|
struct nsproxy *nsp;
|
|
|
|
|
struct task_struct *tsk;
|
|
|
|
|
|
|
|
|
|
/* Take a "snapshot" of the target task's namespaces. */
|
|
|
|
|
rcu_read_lock();
|
|
|
|
|
tsk = pid_task(pid, PIDTYPE_PID);
|
|
|
|
|
if (!tsk) {
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
return -ESRCH;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!ptrace_may_access(tsk, PTRACE_MODE_READ_REALCREDS)) {
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
return -EPERM;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
task_lock(tsk);
|
|
|
|
|
nsp = tsk->nsproxy;
|
|
|
|
|
if (nsp)
|
|
|
|
|
get_nsproxy(nsp);
|
|
|
|
|
task_unlock(tsk);
|
|
|
|
|
if (!nsp) {
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
return -ESRCH;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_PID_NS
|
|
|
|
|
if (flags & CLONE_NEWPID) {
|
|
|
|
|
pid_ns = task_active_pid_ns(tsk);
|
|
|
|
|
if (unlikely(!pid_ns)) {
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
ret = -ESRCH;
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
get_pid_ns(pid_ns);
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_USER_NS
|
|
|
|
|
if (flags & CLONE_NEWUSER)
|
|
|
|
|
user_ns = get_user_ns(__task_cred(tsk)->user_ns);
|
|
|
|
|
#endif
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Install requested namespaces. The caller will have
|
|
|
|
|
* verified earlier that the requested namespaces are
|
|
|
|
|
* supported on this kernel. We don't report errors here
|
|
|
|
|
* if a namespace is requested that isn't supported.
|
|
|
|
|
*/
|
|
|
|
|
#ifdef CONFIG_USER_NS
|
|
|
|
|
if (flags & CLONE_NEWUSER) {
|
|
|
|
|
ret = validate_ns(nsset, &user_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
if (flags & CLONE_NEWNS) {
|
|
|
|
|
ret = validate_ns(nsset, from_mnt_ns(nsp->mnt_ns));
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_UTS_NS
|
|
|
|
|
if (flags & CLONE_NEWUTS) {
|
|
|
|
|
ret = validate_ns(nsset, &nsp->uts_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_IPC_NS
|
|
|
|
|
if (flags & CLONE_NEWIPC) {
|
|
|
|
|
ret = validate_ns(nsset, &nsp->ipc_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_PID_NS
|
|
|
|
|
if (flags & CLONE_NEWPID) {
|
|
|
|
|
ret = validate_ns(nsset, &pid_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_CGROUPS
|
|
|
|
|
if (flags & CLONE_NEWCGROUP) {
|
|
|
|
|
ret = validate_ns(nsset, &nsp->cgroup_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_NET_NS
|
|
|
|
|
if (flags & CLONE_NEWNET) {
|
|
|
|
|
ret = validate_ns(nsset, &nsp->net_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
2020-07-06 17:49:11 +02:00
|
|
|
#ifdef CONFIG_TIME_NS
|
|
|
|
|
if (flags & CLONE_NEWTIME) {
|
|
|
|
|
ret = validate_ns(nsset, &nsp->time_ns->ns);
|
|
|
|
|
if (ret)
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
out:
|
|
|
|
|
if (pid_ns)
|
|
|
|
|
put_pid_ns(pid_ns);
|
|
|
|
|
if (nsp)
|
|
|
|
|
put_nsproxy(nsp);
|
|
|
|
|
put_user_ns(user_ns);
|
|
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
|
}
|
|
|
|
|
|
2020-05-05 16:04:30 +02:00
|
|
|
/*
|
|
|
|
|
* This is the point of no return. There are just a few namespaces
|
|
|
|
|
* that do some actual work here and it's sufficiently minimal that
|
|
|
|
|
* a separate ns_common operation seems unnecessary for now.
|
|
|
|
|
* Unshare is doing the same thing. If we'll end up needing to do
|
|
|
|
|
* more in a given namespace or a helper here is ultimately not
|
|
|
|
|
* exported anymore a simple commit handler for each namespace
|
|
|
|
|
* should be added to ns_common.
|
|
|
|
|
*/
|
|
|
|
|
static void commit_nsset(struct nsset *nsset)
|
|
|
|
|
{
|
|
|
|
|
unsigned flags = nsset->flags;
|
|
|
|
|
struct task_struct *me = current;
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_USER_NS
|
|
|
|
|
if (flags & CLONE_NEWUSER) {
|
|
|
|
|
/* transfer ownership */
|
|
|
|
|
commit_creds(nsset_cred(nsset));
|
|
|
|
|
nsset->cred = NULL;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
/* We only need to commit if we have used a temporary fs_struct. */
|
|
|
|
|
if ((flags & CLONE_NEWNS) && (flags & ~CLONE_NEWNS)) {
|
|
|
|
|
set_fs_root(me->fs, &nsset->fs->root);
|
|
|
|
|
set_fs_pwd(me->fs, &nsset->fs->pwd);
|
|
|
|
|
}
|
|
|
|
|
|
2020-05-05 16:04:30 +02:00
|
|
|
#ifdef CONFIG_IPC_NS
|
|
|
|
|
if (flags & CLONE_NEWIPC)
|
|
|
|
|
exit_sem(me);
|
|
|
|
|
#endif
|
|
|
|
|
|
2020-07-06 17:49:11 +02:00
|
|
|
#ifdef CONFIG_TIME_NS
|
|
|
|
|
if (flags & CLONE_NEWTIME)
|
|
|
|
|
timens_commit(me, nsset->nsproxy->time_ns);
|
|
|
|
|
#endif
|
|
|
|
|
|
2020-05-05 16:04:30 +02:00
|
|
|
/* transfer ownership */
|
|
|
|
|
switch_task_namespaces(me, nsset->nsproxy);
|
|
|
|
|
nsset->nsproxy = NULL;
|
|
|
|
|
}
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
SYSCALL_DEFINE2(setns, int, fd, int, flags)
|
2010-03-07 17:48:52 -08:00
|
|
|
{
|
2024-07-19 20:17:58 -04:00
|
|
|
CLASS(fd, f)(fd);
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
struct ns_common *ns = NULL;
|
2020-05-05 16:04:30 +02:00
|
|
|
struct nsset nsset = {};
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
int err = 0;
|
2010-03-07 17:48:52 -08:00
|
|
|
|
2024-07-19 20:17:58 -04:00
|
|
|
if (fd_empty(f))
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
return -EBADF;
|
2010-03-07 17:48:52 -08:00
|
|
|
|
2024-05-31 14:12:01 -04:00
|
|
|
if (proc_ns_file(fd_file(f))) {
|
|
|
|
|
ns = get_proc_ns(file_inode(fd_file(f)));
|
2025-09-24 13:33:59 +02:00
|
|
|
if (flags && (ns->ns_type != flags))
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
err = -EINVAL;
|
2025-09-24 13:33:59 +02:00
|
|
|
flags = ns->ns_type;
|
2024-05-31 14:12:01 -04:00
|
|
|
} else if (!IS_ERR(pidfd_pid(fd_file(f)))) {
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
err = check_setns_flags(flags);
|
|
|
|
|
} else {
|
2020-06-17 00:33:12 +02:00
|
|
|
err = -EINVAL;
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
}
|
2020-05-05 16:04:30 +02:00
|
|
|
if (err)
|
2010-03-07 17:48:52 -08:00
|
|
|
goto out;
|
|
|
|
|
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
err = prepare_nsset(flags, &nsset);
|
|
|
|
|
if (err)
|
|
|
|
|
goto out;
|
|
|
|
|
|
2024-05-31 14:12:01 -04:00
|
|
|
if (proc_ns_file(fd_file(f)))
|
nsproxy: attach to namespaces via pidfds
For quite a while we have been thinking about using pidfds to attach to
namespaces. This patchset has existed for about a year already but we've
wanted to wait to see how the general api would be received and adopted.
Now that more and more programs in userspace have started using pidfds
for process management it's time to send this one out.
This patch makes it possible to use pidfds to attach to the namespaces
of another process, i.e. they can be passed as the first argument to the
setns() syscall. When only a single namespace type is specified the
semantics are equivalent to passing an nsfd. That means
setns(nsfd, CLONE_NEWNET) equals setns(pidfd, CLONE_NEWNET). However,
when a pidfd is passed, multiple namespace flags can be specified in the
second setns() argument and setns() will attach the caller to all the
specified namespaces all at once or to none of them. Specifying 0 is not
valid together with a pidfd.
Here are just two obvious examples:
setns(pidfd, CLONE_NEWPID | CLONE_NEWNS | CLONE_NEWNET);
setns(pidfd, CLONE_NEWUSER);
Allowing to also attach subsets of namespaces supports various use-cases
where callers setns to a subset of namespaces to retain privilege, perform
an action and then re-attach another subset of namespaces.
If the need arises, as Eric suggested, we can extend this patchset to
assume even more context than just attaching all namespaces. His suggestion
specifically was about assuming the process' root directory when
setns(pidfd, 0) or setns(pidfd, SETNS_PIDFD) is specified. For now, just
keep it flexible in terms of supporting subsets of namespaces but let's
wait until we have users asking for even more context to be assumed. At
that point we can add an extension.
The obvious example where this is useful is a standard container
manager interacting with a running container: pushing and pulling files
or directories, injecting mounts, attaching/execing any kind of process,
managing network devices all these operations require attaching to all
or at least multiple namespaces at the same time. Given that nowadays
most containers are spawned with all namespaces enabled we're currently
looking at at least 14 syscalls, 7 to open the /proc/<pid>/ns/<ns>
nsfds, another 7 to actually perform the namespace switch. With time
namespaces we're looking at about 16 syscalls.
(We could amortize the first 7 or 8 syscalls for opening the nsfds by
stashing them in each container's monitor process but that would mean
we need to send around those file descriptors through unix sockets
everytime we want to interact with the container or keep on-disk
state. Even in scenarios where a caller wants to join a particular
namespace in a particular order callers still profit from batching
other namespaces. That mostly applies to the user namespace but
all container runtimes I found join the user namespace first no matter
if it privileges or deprivileges the container similar to how unshare
behaves.)
With pidfds this becomes a single syscall no matter how many namespaces
are supposed to be attached to.
A decently designed, large-scale container manager usually isn't the
parent of any of the containers it spawns so the containers don't die
when it crashes or needs to update or reinitialize. This means that
for the manager to interact with containers through pids is inherently
racy especially on systems where the maximum pid number is not
significicantly bumped. This is even more problematic since we often spawn
and manage thousands or ten-thousands of containers. Interacting with a
container through a pid thus can become risky quite quickly. Especially
since we allow for an administrator to enable advanced features such as
syscall interception where we're performing syscalls in lieu of the
container. In all of those cases we use pidfds if they are available and
we pass them around as stable references. Using them to setns() to the
target process' namespaces is as reliable as using nsfds. Either the
target process is already dead and we get ESRCH or we manage to attach
to its namespaces but we can't accidently attach to another process'
namespaces. So pidfds lend themselves to be used with this api.
The other main advantage is that with this change the pidfd becomes the
only relevant token for most container interactions and it's the only
token we need to create and send around.
Apart from significiantly reducing the number of syscalls from double
digit to single digit which is a decent reason post-spectre/meltdown
this also allows to switch to a set of namespaces atomically, i.e.
either attaching to all the specified namespaces succeeds or we fail. If
we fail we haven't changed a single namespace. There are currently three
namespaces that can fail (other than for ENOMEM which really is not
very interesting since we then have other problems anyway) for
non-trivial reasons, user, mount, and pid namespaces. We can fail to
attach to a pid namespace if it is not our current active pid namespace
or a descendant of it. We can fail to attach to a user namespace because
we are multi-threaded or because our current mount namespace shares
filesystem state with other tasks, or because we're trying to setns()
to the same user namespace, i.e. the target task has the same user
namespace as we do. We can fail to attach to a mount namespace because
it shares filesystem state with other tasks or because we fail to lookup
the new root for the new mount namespace. In most non-pathological
scenarios these issues can be somewhat mitigated. But there are cases where
we're half-attached to some namespace and failing to attach to another one.
I've talked about some of these problem during the hallway track (something
only the pre-COVID-19 generation will remember) of Plumbers in Los Angeles
in 2018(?). Even if all these issues could be avoided with super careful
userspace coding it would be nicer to have this done in-kernel. Pidfds seem
to lend themselves nicely for this.
The other neat thing about this is that setns() becomes an actual
counterpart to the namespace bits of unshare().
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Serge Hallyn <serge@hallyn.com>
Cc: Jann Horn <jannh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/r/20200505140432.181565-3-christian.brauner@ubuntu.com
2020-05-05 16:04:31 +02:00
|
|
|
err = validate_ns(&nsset, ns);
|
|
|
|
|
else
|
2024-05-31 14:12:01 -04:00
|
|
|
err = validate_nsset(&nsset, pidfd_pid(fd_file(f)));
|
2020-05-05 16:04:30 +02:00
|
|
|
if (!err) {
|
|
|
|
|
commit_nsset(&nsset);
|
|
|
|
|
perf_event_namespaces(current);
|
2010-03-07 17:48:52 -08:00
|
|
|
}
|
2020-05-05 16:04:30 +02:00
|
|
|
put_nsset(&nsset);
|
2010-03-07 17:48:52 -08:00
|
|
|
out:
|
|
|
|
|
return err;
|
|
|
|
|
}
|
|
|
|
|
|
2011-06-28 15:41:10 -04:00
|
|
|
int __init nsproxy_cache_init(void)
|
2007-07-15 23:41:07 -07:00
|
|
|
{
|
2021-09-02 14:55:27 -07:00
|
|
|
nsproxy_cachep = KMEM_CACHE(nsproxy, SLAB_PANIC|SLAB_ACCOUNT);
|
2007-07-15 23:41:07 -07:00
|
|
|
return 0;
|
|
|
|
|
}
|