1
0
Fork 0
mirror of https://github.com/moby/moby.git synced 2022-11-09 12:21:53 -05:00
moby--moby/daemon/oci_linux.go

1042 lines
30 KiB
Go
Raw Normal View History

package daemon // import "github.com/docker/docker/daemon"
import (
"context"
"fmt"
"io"
"io/ioutil"
"os"
"os/exec"
"path/filepath"
"sort"
"strconv"
"strings"
"github.com/containerd/containerd/containers"
coci "github.com/containerd/containerd/oci"
containertypes "github.com/docker/docker/api/types/container"
"github.com/docker/docker/container"
daemonconfig "github.com/docker/docker/daemon/config"
"github.com/docker/docker/oci"
"github.com/docker/docker/oci/caps"
"github.com/docker/docker/pkg/idtools"
"github.com/docker/docker/pkg/mount"
"github.com/docker/docker/pkg/stringid"
"github.com/docker/docker/rootless/specconv"
volumemounts "github.com/docker/docker/volume/mounts"
"github.com/opencontainers/runc/libcontainer/apparmor"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/devices"
rsystem "github.com/opencontainers/runc/libcontainer/system"
"github.com/opencontainers/runc/libcontainer/user"
specs "github.com/opencontainers/runtime-spec/specs-go"
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
"golang.org/x/sys/unix"
)
const inContainerInitPath = "/sbin/" + daemonconfig.DefaultInitBinary
// WithRlimits sets the container's rlimits along with merging the daemon's rlimits
func WithRlimits(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
var rlimits []specs.POSIXRlimit
// We want to leave the original HostConfig alone so make a copy here
hostConfig := *c.HostConfig
// Merge with the daemon defaults
daemon.mergeUlimits(&hostConfig)
for _, ul := range hostConfig.Ulimits {
rlimits = append(rlimits, specs.POSIXRlimit{
Type: "RLIMIT_" + strings.ToUpper(ul.Name),
Soft: uint64(ul.Soft),
Hard: uint64(ul.Hard),
})
}
s.Process.Rlimits = rlimits
return nil
}
}
// WithLibnetwork sets the libnetwork hook
func WithLibnetwork(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
if s.Hooks == nil {
s.Hooks = &specs.Hooks{}
}
for _, ns := range s.Linux.Namespaces {
if ns.Type == "network" && ns.Path == "" && !c.Config.NetworkDisabled {
target := filepath.Join("/proc", strconv.Itoa(os.Getpid()), "exe")
shortNetCtlrID := stringid.TruncateID(daemon.netController.ID())
s.Hooks.Prestart = append(s.Hooks.Prestart, specs.Hook{
Path: target,
Args: []string{
"libnetwork-setkey",
"-exec-root=" + daemon.configStore.GetExecRoot(),
c.ID,
shortNetCtlrID,
},
})
}
}
return nil
}
}
// WithRootless sets the spec to the rootless configuration
func WithRootless(daemon *Daemon) coci.SpecOpts {
return func(_ context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
var v2Controllers []string
if daemon.getCgroupDriver() == cgroupSystemdDriver {
if !cgroups.IsCgroup2UnifiedMode() {
return errors.New("rootless systemd driver doesn't support cgroup v1")
}
rootlesskitParentEUID := os.Getenv("ROOTLESSKIT_PARENT_EUID")
if rootlesskitParentEUID == "" {
return errors.New("$ROOTLESSKIT_PARENT_EUID is not set (requires RootlessKit v0.8.0)")
}
controllersPath := fmt.Sprintf("/sys/fs/cgroup/user.slice/user-%s.slice/cgroup.controllers", rootlesskitParentEUID)
controllersFile, err := ioutil.ReadFile(controllersPath)
if err != nil {
return err
}
v2Controllers = strings.Fields(string(controllersFile))
}
return specconv.ToRootless(s, v2Controllers)
}
}
// WithOOMScore sets the oom score
func WithOOMScore(score *int) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
s.Process.OOMScoreAdj = score
return nil
}
}
// WithSelinux sets the selinux labels
func WithSelinux(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
s.Process.SelinuxLabel = c.GetProcessLabel()
s.Linux.MountLabel = c.MountLabel
return nil
}
}
// WithApparmor sets the apparmor profile
func WithApparmor(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
if apparmor.IsEnabled() {
var appArmorProfile string
if c.AppArmorProfile != "" {
appArmorProfile = c.AppArmorProfile
} else if c.HostConfig.Privileged {
appArmorProfile = unconfinedAppArmorProfile
} else {
appArmorProfile = defaultApparmorProfile
}
if appArmorProfile == defaultApparmorProfile {
// Unattended upgrades and other fun services can unload AppArmor
// profiles inadvertently. Since we cannot store our profile in
// /etc/apparmor.d, nor can we practically add other ways of
// telling the system to keep our profile loaded, in order to make
// sure that we keep the default profile enabled we dynamically
// reload it if necessary.
if err := ensureDefaultAppArmorProfile(); err != nil {
return err
}
}
s.Process.ApparmorProfile = appArmorProfile
}
return nil
}
}
// WithCapabilities sets the container's capabilties
func WithCapabilities(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
capabilities, err := caps.TweakCapabilities(
caps.DefaultCapabilities(),
c.HostConfig.CapAdd,
c.HostConfig.CapDrop,
c.HostConfig.Capabilities,
c.HostConfig.Privileged,
)
if err != nil {
return err
}
return oci.SetCapabilities(s, capabilities)
}
}
func readUserFile(c *container.Container, p string) (io.ReadCloser, error) {
fp, err := c.GetResourcePath(p)
if err != nil {
return nil, err
}
return os.Open(fp)
}
func getUser(c *container.Container, username string) (uint32, uint32, []uint32, error) {
passwdPath, err := user.GetPasswdPath()
if err != nil {
return 0, 0, nil, err
}
groupPath, err := user.GetGroupPath()
if err != nil {
return 0, 0, nil, err
}
passwdFile, err := readUserFile(c, passwdPath)
if err == nil {
defer passwdFile.Close()
}
groupFile, err := readUserFile(c, groupPath)
if err == nil {
defer groupFile.Close()
}
execUser, err := user.GetExecUser(username, nil, passwdFile, groupFile)
if err != nil {
return 0, 0, nil, err
}
// todo: fix this double read by a change to libcontainer/user pkg
groupFile, err = readUserFile(c, groupPath)
if err == nil {
defer groupFile.Close()
}
var addGroups []int
if len(c.HostConfig.GroupAdd) > 0 {
addGroups, err = user.GetAdditionalGroups(c.HostConfig.GroupAdd, groupFile)
if err != nil {
return 0, 0, nil, err
}
}
uid := uint32(execUser.Uid)
gid := uint32(execUser.Gid)
sgids := append(execUser.Sgids, addGroups...)
var additionalGids []uint32
for _, g := range sgids {
additionalGids = append(additionalGids, uint32(g))
}
return uid, gid, additionalGids, nil
}
func setNamespace(s *specs.Spec, ns specs.LinuxNamespace) {
for i, n := range s.Linux.Namespaces {
if n.Type == ns.Type {
s.Linux.Namespaces[i] = ns
return
}
}
s.Linux.Namespaces = append(s.Linux.Namespaces, ns)
}
// WithNamespaces sets the container's namespaces
func WithNamespaces(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
userNS := false
// user
if c.HostConfig.UsernsMode.IsPrivate() {
uidMap := daemon.idMapping.UIDs()
if uidMap != nil {
userNS = true
ns := specs.LinuxNamespace{Type: "user"}
setNamespace(s, ns)
s.Linux.UIDMappings = specMapping(uidMap)
s.Linux.GIDMappings = specMapping(daemon.idMapping.GIDs())
}
}
// network
if !c.Config.NetworkDisabled {
ns := specs.LinuxNamespace{Type: "network"}
parts := strings.SplitN(string(c.HostConfig.NetworkMode), ":", 2)
if parts[0] == "container" {
nc, err := daemon.getNetworkedContainer(c.ID, c.HostConfig.NetworkMode.ConnectedContainer())
if err != nil {
return err
}
ns.Path = fmt.Sprintf("/proc/%d/ns/net", nc.State.GetPID())
if userNS {
// to share a net namespace, they must also share a user namespace
nsUser := specs.LinuxNamespace{Type: "user"}
nsUser.Path = fmt.Sprintf("/proc/%d/ns/user", nc.State.GetPID())
setNamespace(s, nsUser)
}
} else if c.HostConfig.NetworkMode.IsHost() {
ns.Path = c.NetworkSettings.SandboxKey
}
setNamespace(s, ns)
}
// ipc
ipcMode := c.HostConfig.IpcMode
switch {
case ipcMode.IsContainer():
ns := specs.LinuxNamespace{Type: "ipc"}
ic, err := daemon.getIpcContainer(ipcMode.Container())
if err != nil {
return err
}
ns.Path = fmt.Sprintf("/proc/%d/ns/ipc", ic.State.GetPID())
setNamespace(s, ns)
if userNS {
// to share an IPC namespace, they must also share a user namespace
nsUser := specs.LinuxNamespace{Type: "user"}
nsUser.Path = fmt.Sprintf("/proc/%d/ns/user", ic.State.GetPID())
setNamespace(s, nsUser)
}
case ipcMode.IsHost():
oci.RemoveNamespace(s, specs.LinuxNamespaceType("ipc"))
case ipcMode.IsEmpty():
// A container was created by an older version of the daemon.
// The default behavior used to be what is now called "shareable".
fallthrough
case ipcMode.IsPrivate(), ipcMode.IsShareable(), ipcMode.IsNone():
ns := specs.LinuxNamespace{Type: "ipc"}
setNamespace(s, ns)
default:
return fmt.Errorf("Invalid IPC mode: %v", ipcMode)
}
Implement none, private, and shareable ipc modes Since the commit d88fe447df0e8 ("Add support for sharing /dev/shm/ and /dev/mqueue between containers") container's /dev/shm is mounted on the host first, then bind-mounted inside the container. This is done that way in order to be able to share this container's IPC namespace (and the /dev/shm mount point) with another container. Unfortunately, this functionality breaks container checkpoint/restore (even if IPC is not shared). Since /dev/shm is an external mount, its contents is not saved by `criu checkpoint`, and so upon restore any application that tries to access data under /dev/shm is severily disappointed (which usually results in a fatal crash). This commit solves the issue by introducing new IPC modes for containers (in addition to 'host' and 'container:ID'). The new modes are: - 'shareable': enables sharing this container's IPC with others (this used to be the implicit default); - 'private': disables sharing this container's IPC. In 'private' mode, container's /dev/shm is truly mounted inside the container, without any bind-mounting from the host, which solves the issue. While at it, let's also implement 'none' mode. The motivation, as eloquently put by Justin Cormack, is: > I wondered a while back about having a none shm mode, as currently it is > not possible to have a totally unwriteable container as there is always > a /dev/shm writeable mount. It is a bit of a niche case (and clearly > should never be allowed to be daemon default) but it would be trivial to > add now so maybe we should... ...so here's yet yet another mode: - 'none': no /dev/shm mount inside the container (though it still has its own private IPC namespace). Now, to ultimately solve the abovementioned checkpoint/restore issue, we'd need to make 'private' the default mode, but unfortunately it breaks the backward compatibility. So, let's make the default container IPC mode per-daemon configurable (with the built-in default set to 'shareable' for now). The default can be changed either via a daemon CLI option (--default-shm-mode) or a daemon.json configuration file parameter of the same name. Note one can only set either 'shareable' or 'private' IPC modes as a daemon default (i.e. in this context 'host', 'container', or 'none' do not make much sense). Some other changes this patch introduces are: 1. A mount for /dev/shm is added to default OCI Linux spec. 2. IpcMode.Valid() is simplified to remove duplicated code that parsed 'container:ID' form. Note the old version used to check that ID does not contain a semicolon -- this is no longer the case (tests are modified accordingly). The motivation is we should either do a proper check for container ID validity, or don't check it at all (since it is checked in other places anyway). I chose the latter. 3. IpcMode.Container() is modified to not return container ID if the mode value does not start with "container:", unifying the check to be the same as in IpcMode.IsContainer(). 3. IPC mode unit tests (runconfig/hostconfig_test.go) are modified to add checks for newly added values. [v2: addressed review at https://github.com/moby/moby/pull/34087#pullrequestreview-51345997] [v3: addressed review at https://github.com/moby/moby/pull/34087#pullrequestreview-53902833] [v4: addressed the case of upgrading from older daemon, in this case container.HostConfig.IpcMode is unset and this is valid] [v5: document old and new IpcMode values in api/swagger.yaml] [v6: add the 'none' mode, changelog entry to docs/api/version-history.md] Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
2017-06-27 17:58:50 -04:00
// pid
if c.HostConfig.PidMode.IsContainer() {
ns := specs.LinuxNamespace{Type: "pid"}
pc, err := daemon.getPidContainer(c)
if err != nil {
return err
}
ns.Path = fmt.Sprintf("/proc/%d/ns/pid", pc.State.GetPID())
setNamespace(s, ns)
if userNS {
// to share a PID namespace, they must also share a user namespace
nsUser := specs.LinuxNamespace{Type: "user"}
nsUser.Path = fmt.Sprintf("/proc/%d/ns/user", pc.State.GetPID())
setNamespace(s, nsUser)
}
} else if c.HostConfig.PidMode.IsHost() {
oci.RemoveNamespace(s, specs.LinuxNamespaceType("pid"))
} else {
ns := specs.LinuxNamespace{Type: "pid"}
setNamespace(s, ns)
}
// uts
if c.HostConfig.UTSMode.IsHost() {
oci.RemoveNamespace(s, specs.LinuxNamespaceType("uts"))
s.Hostname = ""
}
// cgroup
if !c.HostConfig.CgroupnsMode.IsEmpty() {
cgroupNsMode := c.HostConfig.CgroupnsMode
if !cgroupNsMode.Valid() {
return fmt.Errorf("invalid cgroup namespace mode: %v", cgroupNsMode)
}
// for cgroup v2: unshare cgroupns even for privileged containers
// https://github.com/containers/libpod/pull/4374#issuecomment-549776387
if cgroupNsMode.IsPrivate() && (cgroups.IsCgroup2UnifiedMode() || !c.HostConfig.Privileged) {
nsCgroup := specs.LinuxNamespace{Type: "cgroup"}
setNamespace(s, nsCgroup)
}
}
return nil
}
}
func specMapping(s []idtools.IDMap) []specs.LinuxIDMapping {
var ids []specs.LinuxIDMapping
for _, item := range s {
ids = append(ids, specs.LinuxIDMapping{
HostID: uint32(item.HostID),
ContainerID: uint32(item.ContainerID),
Size: uint32(item.Size),
})
}
return ids
}
// Get the source mount point of directory passed in as argument. Also return
// optional fields.
func getSourceMount(source string) (string, string, error) {
// Ensure any symlinks are resolved.
sourcePath, err := filepath.EvalSymlinks(source)
if err != nil {
return "", "", err
}
mi, err := mount.GetMounts(mount.ParentsFilter(sourcePath))
if err != nil {
return "", "", err
}
if len(mi) < 1 {
return "", "", fmt.Errorf("Can't find mount point of %s", source)
}
// find the longest mount point
var idx, maxlen int
for i := range mi {
if len(mi[i].Mountpoint) > maxlen {
maxlen = len(mi[i].Mountpoint)
idx = i
}
}
return mi[idx].Mountpoint, mi[idx].Optional, nil
}
const (
sharedPropagationOption = "shared:"
slavePropagationOption = "master:"
)
// hasMountinfoOption checks if any of the passed any of the given option values
// are set in the passed in option string.
func hasMountinfoOption(opts string, vals ...string) bool {
for _, opt := range strings.Split(opts, " ") {
for _, val := range vals {
if strings.HasPrefix(opt, val) {
return true
}
}
}
return false
}
// Ensure mount point on which path is mounted, is shared.
func ensureShared(path string) error {
sourceMount, optionalOpts, err := getSourceMount(path)
if err != nil {
return err
}
// Make sure source mount point is shared.
if !hasMountinfoOption(optionalOpts, sharedPropagationOption) {
return errors.Errorf("path %s is mounted on %s but it is not a shared mount", path, sourceMount)
}
return nil
}
// Ensure mount point on which path is mounted, is either shared or slave.
func ensureSharedOrSlave(path string) error {
sourceMount, optionalOpts, err := getSourceMount(path)
if err != nil {
return err
}
if !hasMountinfoOption(optionalOpts, sharedPropagationOption, slavePropagationOption) {
return errors.Errorf("path %s is mounted on %s but it is not a shared or slave mount", path, sourceMount)
}
return nil
}
// Get the set of mount flags that are set on the mount that contains the given
// path and are locked by CL_UNPRIVILEGED. This is necessary to ensure that
// bind-mounting "with options" will not fail with user namespaces, due to
// kernel restrictions that require user namespace mounts to preserve
// CL_UNPRIVILEGED locked flags.
func getUnprivilegedMountFlags(path string) ([]string, error) {
var statfs unix.Statfs_t
if err := unix.Statfs(path, &statfs); err != nil {
return nil, err
}
// The set of keys come from https://github.com/torvalds/linux/blob/v4.13/fs/namespace.c#L1034-L1048.
unprivilegedFlags := map[uint64]string{
unix.MS_RDONLY: "ro",
unix.MS_NODEV: "nodev",
unix.MS_NOEXEC: "noexec",
unix.MS_NOSUID: "nosuid",
unix.MS_NOATIME: "noatime",
unix.MS_RELATIME: "relatime",
unix.MS_NODIRATIME: "nodiratime",
}
var flags []string
for mask, flag := range unprivilegedFlags {
if uint64(statfs.Flags)&mask == mask {
flags = append(flags, flag)
}
}
return flags, nil
}
var (
mountPropagationMap = map[string]int{
"private": mount.PRIVATE,
"rprivate": mount.RPRIVATE,
"shared": mount.SHARED,
"rshared": mount.RSHARED,
"slave": mount.SLAVE,
"rslave": mount.RSLAVE,
}
mountPropagationReverseMap = map[int]string{
mount.PRIVATE: "private",
mount.RPRIVATE: "rprivate",
mount.SHARED: "shared",
mount.RSHARED: "rshared",
mount.SLAVE: "slave",
mount.RSLAVE: "rslave",
}
)
// inSlice tests whether a string is contained in a slice of strings or not.
// Comparison is case sensitive
func inSlice(slice []string, s string) bool {
for _, ss := range slice {
if s == ss {
return true
}
}
return false
}
// WithMounts sets the container's mounts
func WithMounts(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) (err error) {
if err := daemon.setupContainerMountsRoot(c); err != nil {
return err
}
if err := daemon.setupIpcDirs(c); err != nil {
return err
}
defer func() {
if err != nil {
daemon.cleanupSecretDir(c)
}
}()
if err := daemon.setupSecretDir(c); err != nil {
return err
}
ms, err := daemon.setupMounts(c)
if err != nil {
return err
}
if !c.HostConfig.IpcMode.IsPrivate() && !c.HostConfig.IpcMode.IsEmpty() {
ms = append(ms, c.IpcMounts()...)
}
tmpfsMounts, err := c.TmpfsMounts()
if err != nil {
return err
}
ms = append(ms, tmpfsMounts...)
secretMounts, err := c.SecretMounts()
if err != nil {
return err
}
ms = append(ms, secretMounts...)
sort.Sort(mounts(ms))
mounts := ms
userMounts := make(map[string]struct{})
for _, m := range mounts {
userMounts[m.Destination] = struct{}{}
}
// Copy all mounts from spec to defaultMounts, except for
// - mounts overridden by a user supplied mount;
// - all mounts under /dev if a user supplied /dev is present;
// - /dev/shm, in case IpcMode is none.
// While at it, also
// - set size for /dev/shm from shmsize.
defaultMounts := s.Mounts[:0]
_, mountDev := userMounts["/dev"]
for _, m := range s.Mounts {
if _, ok := userMounts[m.Destination]; ok {
// filter out mount overridden by a user supplied mount
continue
}
if mountDev && strings.HasPrefix(m.Destination, "/dev/") {
// filter out everything under /dev if /dev is user-mounted
continue
}
if m.Destination == "/dev/shm" {
if c.HostConfig.IpcMode.IsNone() {
// filter out /dev/shm for "none" IpcMode
continue
}
// set size for /dev/shm mount from spec
sizeOpt := "size=" + strconv.FormatInt(c.HostConfig.ShmSize, 10)
m.Options = append(m.Options, sizeOpt)
}
defaultMounts = append(defaultMounts, m)
}
s.Mounts = defaultMounts
for _, m := range mounts {
if m.Source == "tmpfs" {
data := m.Data
parser := volumemounts.NewParser("linux")
options := []string{"noexec", "nosuid", "nodev", string(parser.DefaultPropagationMode())}
if data != "" {
options = append(options, strings.Split(data, ",")...)
}
merged, err := mount.MergeTmpfsOptions(options)
if err != nil {
return err
}
s.Mounts = append(s.Mounts, specs.Mount{Destination: m.Destination, Source: m.Source, Type: "tmpfs", Options: merged})
continue
}
mt := specs.Mount{Destination: m.Destination, Source: m.Source, Type: "bind"}
// Determine property of RootPropagation based on volume
// properties. If a volume is shared, then keep root propagation
// shared. This should work for slave and private volumes too.
//
// For slave volumes, it can be either [r]shared/[r]slave.
//
// For private volumes any root propagation value should work.
pFlag := mountPropagationMap[m.Propagation]
switch pFlag {
case mount.SHARED, mount.RSHARED:
if err := ensureShared(m.Source); err != nil {
return err
}
rootpg := mountPropagationMap[s.Linux.RootfsPropagation]
if rootpg != mount.SHARED && rootpg != mount.RSHARED {
s.Linux.RootfsPropagation = mountPropagationReverseMap[mount.SHARED]
}
case mount.SLAVE, mount.RSLAVE:
var fallback bool
if err := ensureSharedOrSlave(m.Source); err != nil {
// For backwards compatibility purposes, treat mounts from the daemon root
// as special since we automatically add rslave propagation to these mounts
// when the user did not set anything, so we should fallback to the old
// behavior which is to use private propagation which is normally the
// default.
if !strings.HasPrefix(m.Source, daemon.root) && !strings.HasPrefix(daemon.root, m.Source) {
return err
}
cm, ok := c.MountPoints[m.Destination]
if !ok {
return err
}
if cm.Spec.BindOptions != nil && cm.Spec.BindOptions.Propagation != "" {
// This means the user explicitly set a propagation, do not fallback in that case.
return err
}
fallback = true
logrus.WithField("container", c.ID).WithField("source", m.Source).Warn("Falling back to default propagation for bind source in daemon root")
}
if !fallback {
rootpg := mountPropagationMap[s.Linux.RootfsPropagation]
if rootpg != mount.SHARED && rootpg != mount.RSHARED && rootpg != mount.SLAVE && rootpg != mount.RSLAVE {
s.Linux.RootfsPropagation = mountPropagationReverseMap[mount.RSLAVE]
}
}
}
bindMode := "rbind"
if m.NonRecursive {
bindMode = "bind"
}
opts := []string{bindMode}
if !m.Writable {
opts = append(opts, "ro")
}
if pFlag != 0 {
opts = append(opts, mountPropagationReverseMap[pFlag])
}
// If we are using user namespaces, then we must make sure that we
// don't drop any of the CL_UNPRIVILEGED "locked" flags of the source
// "mount" when we bind-mount. The reason for this is that at the point
// when runc sets up the root filesystem, it is already inside a user
// namespace, and thus cannot change any flags that are locked.
if daemon.configStore.RemappedRoot != "" {
unprivOpts, err := getUnprivilegedMountFlags(m.Source)
if err != nil {
return err
}
opts = append(opts, unprivOpts...)
}
mt.Options = opts
s.Mounts = append(s.Mounts, mt)
}
if s.Root.Readonly {
for i, m := range s.Mounts {
switch m.Destination {
case "/proc", "/dev/pts", "/dev/shm", "/dev/mqueue", "/dev":
continue
}
if _, ok := userMounts[m.Destination]; !ok {
if !inSlice(m.Options, "ro") {
s.Mounts[i].Options = append(s.Mounts[i].Options, "ro")
}
}
}
}
if c.HostConfig.Privileged {
// clear readonly for /sys
for i := range s.Mounts {
if s.Mounts[i].Destination == "/sys" {
clearReadOnly(&s.Mounts[i])
}
}
s.Linux.ReadonlyPaths = nil
s.Linux.MaskedPaths = nil
}
// TODO: until a kernel/mount solution exists for handling remount in a user namespace,
// we must clear the readonly flag for the cgroups mount (@mrunalp concurs)
if uidMap := daemon.idMapping.UIDs(); uidMap != nil || c.HostConfig.Privileged {
for i, m := range s.Mounts {
if m.Type == "cgroup" {
clearReadOnly(&s.Mounts[i])
}
}
}
return nil
}
}
// WithCommonOptions sets common docker options
func WithCommonOptions(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
if c.BaseFS == nil {
return errors.New("populateCommonSpec: BaseFS of container " + c.ID + " is unexpectedly nil")
}
linkedEnv, err := daemon.setupLinkedContainers(c)
if err != nil {
return err
}
s.Root = &specs.Root{
Path: c.BaseFS.Path(),
Readonly: c.HostConfig.ReadonlyRootfs,
}
if err := c.SetupWorkingDirectory(daemon.idMapping.RootPair()); err != nil {
return err
}
cwd := c.Config.WorkingDir
if len(cwd) == 0 {
cwd = "/"
}
s.Process.Args = append([]string{c.Path}, c.Args...)
// only add the custom init if it is specified and the container is running in its
// own private pid namespace. It does not make sense to add if it is running in the
// host namespace or another container's pid namespace where we already have an init
if c.HostConfig.PidMode.IsPrivate() {
if (c.HostConfig.Init != nil && *c.HostConfig.Init) ||
(c.HostConfig.Init == nil && daemon.configStore.Init) {
s.Process.Args = append([]string{inContainerInitPath, "--", c.Path}, c.Args...)
path := daemon.configStore.InitPath
if path == "" {
path, err = exec.LookPath(daemonconfig.DefaultInitBinary)
if err != nil {
return err
}
}
s.Mounts = append(s.Mounts, specs.Mount{
Destination: inContainerInitPath,
Type: "bind",
Source: path,
Options: []string{"bind", "ro"},
})
}
}
s.Process.Cwd = cwd
s.Process.Env = c.CreateDaemonEnvironment(c.Config.Tty, linkedEnv)
s.Process.Terminal = c.Config.Tty
s.Hostname = c.Config.Hostname
setLinuxDomainname(c, s)
return nil
}
}
// WithCgroups sets the container's cgroups
func WithCgroups(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
var cgroupsPath string
scopePrefix := "docker"
parent := "/docker"
useSystemd := UsingSystemd(daemon.configStore)
if useSystemd {
parent = "system.slice"
if daemon.configStore.Rootless {
parent = "user.slice"
}
}
if c.HostConfig.CgroupParent != "" {
parent = c.HostConfig.CgroupParent
} else if daemon.configStore.CgroupParent != "" {
parent = daemon.configStore.CgroupParent
}
if useSystemd {
cgroupsPath = parent + ":" + scopePrefix + ":" + c.ID
logrus.Debugf("createSpec: cgroupsPath: %s", cgroupsPath)
} else {
cgroupsPath = filepath.Join(parent, c.ID)
}
s.Linux.CgroupsPath = cgroupsPath
p := cgroupsPath
if useSystemd {
initPath, err := cgroups.GetInitCgroup("cpu")
if err != nil {
return err
}
_, err = cgroups.GetOwnCgroup("cpu")
if err != nil {
return err
}
p = filepath.Join(initPath, s.Linux.CgroupsPath)
}
// Clean path to guard against things like ../../../BAD
parentPath := filepath.Dir(p)
if !filepath.IsAbs(parentPath) {
parentPath = filepath.Clean("/" + parentPath)
}
if err := daemon.initCgroupsPath(parentPath); err != nil {
return fmt.Errorf("linux init cgroups path: %v", err)
}
return nil
}
}
// WithDevices sets the container's devices
func WithDevices(daemon *Daemon, c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
// Build lists of devices allowed and created within the container.
var devs []specs.LinuxDevice
devPermissions := s.Linux.Resources.Devices
if c.HostConfig.Privileged && !rsystem.RunningInUserNS() {
hostDevices, err := devices.HostDevices()
if err != nil {
return err
}
for _, d := range hostDevices {
devs = append(devs, oci.Device(d))
}
// adding device mappings in privileged containers
for _, deviceMapping := range c.HostConfig.Devices {
// issue a warning that custom cgroup permissions are ignored in privileged mode
if deviceMapping.CgroupPermissions != "rwm" {
logrus.WithField("container", c.ID).Warnf("custom %s permissions for device %s are ignored in privileged mode", deviceMapping.CgroupPermissions, deviceMapping.PathOnHost)
}
// issue a warning that the device path already exists via /dev mounting in privileged mode
if deviceMapping.PathOnHost == deviceMapping.PathInContainer {
logrus.WithField("container", c.ID).Warnf("path in container %s already exists in privileged mode", deviceMapping.PathInContainer)
continue
}
d, _, err := oci.DevicesFromPath(deviceMapping.PathOnHost, deviceMapping.PathInContainer, "rwm")
if err != nil {
return err
}
devs = append(devs, d...)
}
devPermissions = []specs.LinuxDeviceCgroup{
{
Allow: true,
Access: "rwm",
},
}
} else {
for _, deviceMapping := range c.HostConfig.Devices {
d, dPermissions, err := oci.DevicesFromPath(deviceMapping.PathOnHost, deviceMapping.PathInContainer, deviceMapping.CgroupPermissions)
if err != nil {
return err
}
devs = append(devs, d...)
devPermissions = append(devPermissions, dPermissions...)
}
var err error
devPermissions, err = oci.AppendDevicePermissionsFromCgroupRules(devPermissions, c.HostConfig.DeviceCgroupRules)
if err != nil {
return err
}
}
s.Linux.Devices = append(s.Linux.Devices, devs...)
s.Linux.Resources.Devices = devPermissions
for _, req := range c.HostConfig.DeviceRequests {
if err := daemon.handleDevice(req, s); err != nil {
return err
}
}
return nil
}
}
// WithResources applies the container resources
func WithResources(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
r := c.HostConfig.Resources
weightDevices, err := getBlkioWeightDevices(r)
if err != nil {
return err
}
readBpsDevice, err := getBlkioThrottleDevices(r.BlkioDeviceReadBps)
if err != nil {
return err
}
writeBpsDevice, err := getBlkioThrottleDevices(r.BlkioDeviceWriteBps)
if err != nil {
return err
}
readIOpsDevice, err := getBlkioThrottleDevices(r.BlkioDeviceReadIOps)
if err != nil {
return err
}
writeIOpsDevice, err := getBlkioThrottleDevices(r.BlkioDeviceWriteIOps)
if err != nil {
return err
}
memoryRes := getMemoryResources(r)
cpuRes, err := getCPUResources(r)
if err != nil {
return err
}
blkioWeight := r.BlkioWeight
specResources := &specs.LinuxResources{
Memory: memoryRes,
CPU: cpuRes,
BlockIO: &specs.LinuxBlockIO{
Weight: &blkioWeight,
WeightDevice: weightDevices,
ThrottleReadBpsDevice: readBpsDevice,
ThrottleWriteBpsDevice: writeBpsDevice,
ThrottleReadIOPSDevice: readIOpsDevice,
ThrottleWriteIOPSDevice: writeIOpsDevice,
},
Pids: getPidsLimit(r),
}
if s.Linux.Resources != nil && len(s.Linux.Resources.Devices) > 0 {
specResources.Devices = s.Linux.Resources.Devices
}
s.Linux.Resources = specResources
return nil
}
}
// WithSysctls sets the container's sysctls
func WithSysctls(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
// We merge the sysctls injected above with the HostConfig (latter takes
// precedence for backwards-compatibility reasons).
for k, v := range c.HostConfig.Sysctls {
s.Linux.Sysctl[k] = v
}
return nil
}
}
// WithUser sets the container's user
func WithUser(c *container.Container) coci.SpecOpts {
return func(ctx context.Context, _ coci.Client, _ *containers.Container, s *coci.Spec) error {
uid, gid, additionalGids, err := getUser(c, c.Config.User)
if err != nil {
return err
}
s.Process.User.UID = uid
s.Process.User.GID = gid
s.Process.User.AdditionalGids = additionalGids
return nil
}
}
func (daemon *Daemon) createSpec(c *container.Container) (retSpec *specs.Spec, err error) {
var (
opts []coci.SpecOpts
s = oci.DefaultSpec()
)
opts = append(opts,
WithCommonOptions(daemon, c),
WithCgroups(daemon, c),
WithResources(c),
WithSysctls(c),
WithDevices(daemon, c),
WithUser(c),
WithRlimits(daemon, c),
WithNamespaces(daemon, c),
WithCapabilities(c),
WithSeccomp(daemon, c),
WithMounts(daemon, c),
WithLibnetwork(daemon, c),
WithApparmor(c),
WithSelinux(c),
WithOOMScore(&c.HostConfig.OomScoreAdj),
)
if c.NoNewPrivileges {
opts = append(opts, coci.WithNoNewPrivileges)
}
// Set the masked and readonly paths with regard to the host config options if they are set.
if c.HostConfig.MaskedPaths != nil {
opts = append(opts, coci.WithMaskedPaths(c.HostConfig.MaskedPaths))
}
if c.HostConfig.ReadonlyPaths != nil {
opts = append(opts, coci.WithReadonlyPaths(c.HostConfig.ReadonlyPaths))
}
if daemon.configStore.Rootless {
opts = append(opts, WithRootless(daemon))
}
return &s, coci.ApplyOpts(context.Background(), nil, &containers.Container{
ID: c.ID,
}, &s, opts...)
}
func clearReadOnly(m *specs.Mount) {
var opt []string
for _, o := range m.Options {
if o != "ro" {
opt = append(opt, o)
}
}
m.Options = opt
}
// mergeUlimits merge the Ulimits from HostConfig with daemon defaults, and update HostConfig
func (daemon *Daemon) mergeUlimits(c *containertypes.HostConfig) {
ulimits := c.Ulimits
// Merge ulimits with daemon defaults
ulIdx := make(map[string]struct{})
for _, ul := range ulimits {
ulIdx[ul.Name] = struct{}{}
}
for name, ul := range daemon.configStore.Ulimits {
if _, exists := ulIdx[name]; !exists {
ulimits = append(ulimits, ul)
}
}
c.Ulimits = ulimits
}