confile: backport parts of network parsing

This backports part of our robust network parser.

Closes #2271.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: enable "force" for "cgroup-full"

This enables cgroup-full:{mixed,ro,rw}:force and reworks the mount logic.
When cgroup-full was specified we used to bind-mount the cgroups from the host.
That is pretty weird thing to do given that you can simply mount them directly
without going through bind-mounts.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: do_destroy_container()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_destroy_container_on_signal()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: post_start()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: start()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_abort()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_init()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_init_handler()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_poll()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: signal_handler()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: lxc_check_inherited()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: preserve_ns()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: close_ns()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: print_top_failing_dir()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: use correct prefix for includes

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

{commands,start}: remove element from list first

First remove the client from the list then close the fd. Otherwise we open
ourselves to a race where another codepath might be writing to a bad file
descriptor.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

tree-wide: remove locking around openpty()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: remove duplicate lxc_monitor_send_state()

Closes #2177.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

lxccontainer: use wait_for_pid()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: __cg_legacy_setup_limits()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_legacy_set_data()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: convert_devpath()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_set()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_get()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_attach()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: __cg_unified_attach()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: build_full_cgpath_from_monitorpath()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_escape()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: recursive_count_nrtasks()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: recursive_count_nrtasks()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_mount()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: mount_cgroup_full()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_chown()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_enter()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_create()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: remove_path_for_hierarchy()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: create_path_for_hierarchy()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_unified_create_cgroup()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: recursive_destroy()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cgfsng_init()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_unified_get_current_cgroup()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_is_pure_unified()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_hybrid_init()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: lxc_cgfsng_print_basecg_debuginfo()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: lxc_cgfsng_print_hierarchies()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: trim()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: must_append_string()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_hybrid_get_current_cgroup()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: controller_in_clist()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: copy_to_eol()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_hybrid_get_mountpoint()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_hybrid_get_controllers()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: all_controllers_found()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: controller_found()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: controller_list_is_dup()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: controller_lists_intersect()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_legacy_handle_cpuset_hierarchy()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: copy_parent_file()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_legacy_filter_and_set_cpus()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: get_max_cpus()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: lxc_cpumask_to_cpulist()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: lxc_cpumask()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: get_hierarchy()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: must_append_controller()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: string_in_list()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: append_null_to_list()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: add me to authors

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: move cg_legacy_must_prefix_named()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: cg_legacy_must_prefix_named()

s/must_prefix_named/cg_legacy_must_prefix_named/

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: free_string_list()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: fully document remaining variables

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: fully document struct cgfsng_handler_data

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: fully document struct hierarchy

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: order includes

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

confile: add "force" to cgroup:{mixed,ro,rw}

This lets users specify

        lxc.mount.auto = cgroup:mixed:force
or
        lxc.mount.auto = cgroup:ro:force
or
        lxc.mount.auto = cgroup:rw:force

When cgroup namespaces are supported LXC will not mount cgroups for the
container since it assumes that the init system will mount cgroups itself if it
wants to. This assumption already broke when users wanted to run containers
without CAP_SYS_ADMIN. For example, systemd based containers wouldn't start
since systemd needs to mount cgroups (named systemd hierarchy for legacy
cgroups and the unified hierarchy for unified cgroups) to track processes. This
problem was solved by detecting whether the container had CAP_SYS_ADMIN. If it
didn't we performed the cgroup mounts for it.
However, there are more cases when we should be able to mount cgroups for the
container when cgroup namespaces are supported:
- init systems not mounting cgroups themselves:
  A init system that doesn't mount cgroups would not have cgroups available
  especially when combined with custom LSM profiles to prevent cgroup
  {u}mount()ing inside containers.
- application containers:
  Application containers will usually not mount by cgroups themselves.
- read-only cgroups:
  It is useful to be able to mount cgroups read-only to e.g. prevent
  changing cgroup limits from inside the container while at the same time
  allowing the applications to perform introspection on their own cgroups. This
  again is mostly useful for application containers. System containers running
  systemd will usually not work correctly when cgroups are mounted read-only.
To be fair, all of those use-cases could be covered by custom hooks or
lxc.mount.entry entries but exposing it through lxc.mount.auto takes care of
setting correct mount options and adding the necessary logic to e.g. mount
filesystem read-only correctly.

Currently we only extend this to cgroup:{mixed,ro,rw} but technically there's
no reason not to enable the same behavior for cgroup-full:{mixed,ro,rw} as
well. If someone requests this we can simply treat it as a bug and add "force"
for cgroup-full.

Replaces #2136.

Signed-off-by: Shukui Yang <yangshukui@huawei.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

README: add coverity

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

console: ensure that fd is marked EBADF

If the handler closes the file descriptor for the peer or master fd it is
crucial that we mark it as -EBADF. This will prevent lxc_console_delete()
from calling close() on an already closed file descriptor again. I've
observed the double close in the attach code.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

start: don't call close on invalid file descriptor

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

utils: fix lxc_p{close,open}()

If a file descriptor fd is opened by fdopen() and associated with a stream f
will **not** have been dup()ed. This means that fclose(f) will also close the
fd. So never call close(fd) after fdopen(fd) succeeded.
This fixes a double close() Stéphane and I observed when debugging on aarch64
and armf.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

tests: remove lxc-test-ubuntu

This is really taking a long time for not a lot of benefit.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

CODING_STYLE: add section for str{n}cmp()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgroups: use correct mask for chmod()

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

CODING_STYLE: add CODING_STYLE.md

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

CONTRIBUTING: update

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

cgfsng: simplifications and fixes

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

lxclock: remove pthread_atfork_handlers

They shouldn't be needed anymore.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

conf: support mount propagation

Closes #810.

Signed-off-by: Yifeng Tan <tanyifeng1@huawei.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

Fix lxc-console hang

The variable 'descr' is mistakenly covered with 'descr_console'.

Signed-off-by: LiFeng <lifeng68@huawei.com>

Modify .gitignore

add:
src/lxc/cmd/lxc-checkconfig
src/lxc/cmd/lxc-update-config

Signed-off-by: LiFeng <lifeng68@huawei.com>

console: they are really not necessary

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

fix userns helper error handling

In both of these cases if there is actually an error, we won't close the
pipe and the api call will hang. Instead, let's be sure to close the pipe
before waiting, so that it doesn't hang.

Signed-off-by: Tycho Andersen <tycho@tycho.ws>

coverity: #1425971

Dereference after null check

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

coverity: #1426734

Argument cannot be negative

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

coverity: #1429139

Resource leak

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

Restore most cases of am_guest_unpriv

The only cases where we really need to be privileged with respect
to the host is when we are trying to mknod, and in some cases
to do with a physical network device.  This patch leaves the
detection of the network device cases as a TODO.

This should fix the currently broken case of starting a privileged
container with at least one veth nic, nested inside an unprivileged
container.

Cc: Tycho Andersen <tycho@tycho.ws>
Signed-off-by: Serge Hallyn <shallyn@cisco.com>

define am_guest_unpriv

Sometimes we want to know whether we are privileged wrt our
namespaces, and sometimes we want to know whether we are priv
wrt init_user_ns.

Signed-off-by: Serge Hallyn <shallyn@cisco.com>

cgfsng: add required remount flags

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=9566d6742852c527bf5af38af5cbb878dad75705

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>

lsm: fix missing @ in function documentation

Signed-off-by: Marcos Paulo de Souza <marcos.souza.org@gmail.com>

selinux: simplify check for default label

Signed-off-by: Marcos Paulo de Souza <marcos.souza.org@gmail.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>