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
Add a simple struct nsset. It holds all necessary pieces to switch to a new
set of namespaces without leaving a task in a half-switched state which we
will make use of in the next patch. This patch switches the existing setns
logic over without causing a change in setns() behavior. This brings
setns() closer to how unshare() works(). The prepare_ns() function is
responsible to prepare all necessary information. This has two reasons.
First it minimizes dependencies between individual namespaces, i.e. all
install handler can expect that all fields are properly initialized
independent in what order they are called in. Second, this makes the code
easier to maintain and easier to follow if it needs to be changed.
The prepare_ns() helper will only be switched over to use a flags argument
in the next patch. Here it will still use nstype as a simple integer
argument which was argued would be clearer. I'm not particularly
opinionated about this if it really helps or not. The struct nsset itself
already contains the flags field since its name already indicates that it
can contain information required by different namespaces. None of this
should have functional consequences.
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/20200505140432.181565-2-christian.brauner@ubuntu.com
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
Based on 1 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license as published by
the free software foundation version 2 of the license
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 315 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Armijn Hemel <armijn@tjaldur.nl>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190531190115.503150771@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce the ability to create new cgroup namespace. The newly created
cgroup namespace remembers the cgroup of the process at the point
of creation of the cgroup namespace (referred as cgroupns-root).
The main purpose of cgroup namespace is to virtualize the contents
of /proc/self/cgroup file. Processes inside a cgroup namespace
are only able to see paths relative to their namespace root
(unless they are moved outside of their cgroupns-root, at which point
they will see a relative path from their cgroupns-root).
For a correctly setup container this enables container-tools
(like libcontainer, lxc, lmctfy, etc.) to create completely virtualized
containers without leaking system level cgroup hierarchy to the task.
This patch only implements the 'unshare' part of the cgroupns.
Signed-off-by: Aditya Kali <adityakali@google.com>
Signed-off-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
a) make get_proc_ns() return a pointer to struct ns_common
b) mirror ns_ops in dentry->d_fsdata of ns dentries, so that
is_mnt_ns_file() could get away with fewer dereferences.
That way struct proc_ns becomes invisible outside of fs/proc/*.c
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The synchronous syncrhonize_rcu in switch_task_namespaces makes setns
a sufficiently expensive system call that people have complained.
Upon inspect nsproxy no longer needs rcu protection for remote reads.
remote reads are rare. So optimize for same process reads and write
by switching using rask_lock instead.
This yields a simpler to understand lock, and a faster setns system call.
In particular this fixes a performance regression observed
by Rafael David Tinoco <rafael.tinoco@canonical.com>.
This is effectively a revert of Pavel Emelyanov's commit
cf7b708c8d Make access to task's nsproxy lighter
from 2007. The race this originialy fixed no longer exists as
do_notify_parent uses task_active_pid_ns(parent) instead of
parent->nsproxy.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Pull namespace changes from Eric Biederman:
"This is an assorted mishmash of small cleanups, enhancements and bug
fixes.
The major theme is user namespace mount restrictions. nsown_capable
is killed as it encourages not thinking about details that need to be
considered. A very hard to hit pid namespace exiting bug was finally
tracked and fixed. A couple of cleanups to the basic namespace
infrastructure.
Finally there is an enhancement that makes per user namespace
capabilities usable as capabilities, and an enhancement that allows
the per userns root to nice other processes in the user namespace"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
userns: Kill nsown_capable it makes the wrong thing easy
capabilities: allow nice if we are privileged
pidns: Don't have unshare(CLONE_NEWPID) imply CLONE_THREAD
userns: Allow PR_CAPBSET_DROP in a user namespace.
namespaces: Simplify copy_namespaces so it is clear what is going on.
pidns: Fix hang in zap_pid_ns_processes by sending a potentially extra wakeup
sysfs: Restrict mounting sysfs
userns: Better restrictions on when proc and sysfs can be mounted
vfs: Don't copy mount bind mounts of /proc/<pid>/ns/mnt between namespaces
kernel/nsproxy.c: Improving a snippet of code.
proc: Restrict mounting the proc filesystem
vfs: Lock in place mounts from more privileged users
Remove the test for the impossible case where tsk->nsproxy == NULL. Fork
will never be called with tsk->nsproxy == NULL.
Only call get_nsproxy when we don't need to generate a new_nsproxy,
and mark the case where we don't generate a new nsproxy as likely.
Remove the code to drop an unnecessarily acquired nsproxy value.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
nsproxy.pid_ns is *not* the task's pid namespace. The name should clarify
that.
This makes it more obvious that setns on a pid namespace is weird --
it won't change the pid namespace shown in procfs.
Signed-off-by: Andy Lutomirski <luto@amacapital.net>
Reviewed-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It seems GCC generates a better code in that way, so I changed that statement.
Btw, they have the same semantic, so I'm sending this patch due to performance issues.
Acked-by: Serge E. Hallyn <serge.hallyn@ubuntu.com>
Signed-off-by: Raphael S.Carvalho <raphael.scarv@gmail.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Pull vfs pile (part one) from Al Viro:
"Assorted stuff - cleaning namei.c up a bit, fixing ->d_name/->d_parent
locking violations, etc.
The most visible changes here are death of FS_REVAL_DOT (replaced with
"has ->d_weak_revalidate()") and a new helper getting from struct file
to inode. Some bits of preparation to xattr method interface changes.
Misc patches by various people sent this cycle *and* ocfs2 fixes from
several cycles ago that should've been upstream right then.
PS: the next vfs pile will be xattr stuff."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (46 commits)
saner proc_get_inode() calling conventions
proc: avoid extra pde_put() in proc_fill_super()
fs: change return values from -EACCES to -EPERM
fs/exec.c: make bprm_mm_init() static
ocfs2/dlm: use GFP_ATOMIC inside a spin_lock
ocfs2: fix possible use-after-free with AIO
ocfs2: Fix oops in ocfs2_fast_symlink_readpage() code path
get_empty_filp()/alloc_file() leave both ->f_pos and ->f_version zero
target: writev() on single-element vector is pointless
export kernel_write(), convert open-coded instances
fs: encode_fh: return FILEID_INVALID if invalid fid_type
kill f_vfsmnt
vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op
nfsd: handle vfs_getattr errors in acl protocol
switch vfs_getattr() to struct path
default SET_PERSONALITY() in linux/elf.h
ceph: prepopulate inodes only when request is aborted
d_hash_and_lookup(): export, switch open-coded instances
9p: switch v9fs_set_create_acl() to inode+fid, do it before d_instantiate()
9p: split dropping the acls from v9fs_set_create_acl()
...
- Add CLONE_THREAD to the unshare flags if CLONE_NEWUSER is selected
As changing user namespaces is only valid if all there is only
a single thread.
- Restore the code to add CLONE_VM if CLONE_THREAD is selected and
the code to addCLONE_SIGHAND if CLONE_VM is selected.
Making the constraints in the code clear.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Modify create_new_namespaces to explicitly take a user namespace
parameter, instead of implicitly through the task_struct.
This allows an implementation of unshare(CLONE_NEWUSER) where
the new user namespace is not stored onto the current task_struct
until after all of the namespaces are created.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
- Push the permission check from the core setns syscall into
the setns install methods where the user namespace of the
target namespace can be determined, and used in a ns_capable
call.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
If an unprivileged user has the appropriate capabilities in their
current user namespace allow the creation of new namespaces.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
This will allow for support for unprivileged mounts in a new user namespace.
Acked-by: "Serge E. Hallyn" <serge@hallyn.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Unsharing of the pid namespace unlike unsharing of other namespaces
does not take affect immediately. Instead it affects the children
created with fork and clone. The first of these children becomes the init
process of the new pid namespace, the rest become oddball children
of pid 0. From the point of view of the new pid namespace the process
that created it is pid 0, as it's pid does not map.
A couple of different semantics were considered but this one was
settled on because it is easy to implement and it is usable from
pam modules. The core reasons for the existence of unshare.
I took a survey of the callers of pam modules and the following
appears to be a representative sample of their logic.
{
setup stuff include pam
child = fork();
if (!child) {
setuid()
exec /bin/bash
}
waitpid(child);
pam and other cleanup
}
As you can see there is a fork to create the unprivileged user
space process. Which means that the unprivileged user space
process will appear as pid 1 in the new pid namespace. Further
most login processes do not cope with extraneous children which
means shifting the duty of reaping extraneous child process to
the creator of those extraneous children makes the system more
comprehensible.
The practical reason for this set of pid namespace semantics is
that it is simple to implement and verify they work correctly.
Whereas an implementation that requres changing the struct
pid on a process comes with a lot more races and pain. Not
the least of which is that glibc caches getpid().
These semantics are implemented by having two notions
of the pid namespace of a proces. There is task_active_pid_ns
which is the pid namspace the process was created with
and the pid namespace that all pids are presented to
that process in. The task_active_pid_ns is stored
in the struct pid of the task.
Then there is the pid namespace that will be used for children
that pid namespace is stored in task->nsproxy->pid_ns.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
