vbatts/containerd
1
0
Fork 0
Code Pull requests Releases Activity

Vendor dependencies

Browse source 
Signed-off-by: Alexander Morozov <lk4d4@docker.com>
This commit is contained in:
Alexander Morozov 2016-01-05 14:56:27 -08:00
parent 9f1a65e941
commit e17cf1ad4f
474 changed files with 71500 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

152
vendor/src/github.com/opencontainers/runc/libcontainer/README.md vendored Normal file
View file

@ -0,0 +1,152 @@
Libcontainer provides a native Go implementation for creating containers
with namespaces, cgroups, capabilities, and filesystem access controls.
It allows you to manage the lifecycle of the container performing additional operations
after the container is created.
#### Container
A container is a self contained execution environment that shares the kernel of the
host system and which is (optionally) isolated from other containers in the system.
#### Using libcontainer
To create a container you first have to initialize an instance of a factory
that will handle the creation and initialization for a container.
Because containers are spawned in a two step process you will need to provide
arguments to a binary that will be executed as the init process for the container.
To use the current binary that is spawning the containers and acting as the parent
you can use `os.Args[0]` and we have a command called `init` setup.
```go
root, err := libcontainer.New("/var/lib/container", libcontainer.InitArgs(os.Args[0], "init"))
if err != nil {
log.Fatal(err)
}
```
Once you have an instance of the factory created we can create a configuration
struct describing how the container is to be created. A sample would look similar to this:
```go
config := &configs.Config{
Rootfs: rootfs,
Capabilities: []string{
"CAP_CHOWN",
"CAP_DAC_OVERRIDE",
"CAP_FSETID",
"CAP_FOWNER",
"CAP_MKNOD",
"CAP_NET_RAW",
"CAP_SETGID",
"CAP_SETUID",
"CAP_SETFCAP",
"CAP_SETPCAP",
"CAP_NET_BIND_SERVICE",
"CAP_SYS_CHROOT",
"CAP_KILL",
"CAP_AUDIT_WRITE",
},
Namespaces: configs.Namespaces([]configs.Namespace{
{Type: configs.NEWNS},
{Type: configs.NEWUTS},
{Type: configs.NEWIPC},
{Type: configs.NEWPID},
{Type: configs.NEWNET},
}),
Cgroups: &configs.Cgroup{
Name: "test-container",
Parent: "system",
AllowAllDevices: false,
AllowedDevices: configs.DefaultAllowedDevices,
},
Devices: configs.DefaultAutoCreatedDevices,
Hostname: "testing",
Networks: []*configs.Network{
{
Type: "loopback",
Address: "127.0.0.1/0",
Gateway: "localhost",
},
},
Rlimits: []configs.Rlimit{
{
Type: syscall.RLIMIT_NOFILE,
Hard: uint64(1024),
Soft: uint64(1024),
},
},
}
```
Once you have the configuration populated you can create a container:
```go
container, err := root.Create("container-id", config)
```
To spawn bash as the initial process inside the container and have the
processes pid returned in order to wait, signal, or kill the process:
```go
process := &libcontainer.Process{
Args: []string{"/bin/bash"},
Env: []string{"PATH=/bin"},
User: "daemon",
Stdin: os.Stdin,
Stdout: os.Stdout,
Stderr: os.Stderr,
}
err := container.Start(process)
if err != nil {
log.Fatal(err)
}
// wait for the process to finish.
status, err := process.Wait()
if err != nil {
log.Fatal(err)
}
// destroy the container.
container.Destroy()
```
Additional ways to interact with a running container are:
```go
// return all the pids for all processes running inside the container.
processes, err := container.Processes()
// get detailed cpu, memory, io, and network statistics for the container and
// it's processes.
stats, err := container.Stats()
// pause all processes inside the container.
container.Pause()
// resume all paused processes.
container.Resume()
```
#### Checkpoint & Restore
libcontainer now integrates [CRIU](http://criu.org/) for checkpointing and restoring containers.
This let's you save the state of a process running inside a container to disk, and then restore
that state into a new process, on the same machine or on another machine.
`criu` version 1.5.2 or higher is required to use checkpoint and restore.
If you don't already have `criu` installed, you can build it from source, following the
[online instructions](http://criu.org/Installation). `criu` is also installed in the docker image
generated when building libcontainer with docker.
## Copyright and license
Code and documentation copyright 2014 Docker, inc. Code released under the Apache 2.0 license.
Docs released under Creative commons.

334
vendor/src/github.com/opencontainers/runc/libcontainer/SPEC.md vendored Normal file
View file

@ -0,0 +1,334 @@
## Container Specification - v1
This is the standard configuration for version 1 containers. It includes
namespaces, standard filesystem setup, a default Linux capability set, and
information about resource reservations. It also has information about any
populated environment settings for the processes running inside a container.
Along with the configuration of how a container is created the standard also
discusses actions that can be performed on a container to manage and inspect
information about the processes running inside.
The v1 profile is meant to be able to accommodate the majority of applications
with a strong security configuration.
### System Requirements and Compatibility
Minimum requirements:
* Kernel version - 3.10 recommended 2.6.2x minimum(with backported patches)
* Mounted cgroups with each subsystem in its own hierarchy
### Namespaces
| Flag | Enabled |
| ------------ | ------- |
| CLONE_NEWPID | 1 |
| CLONE_NEWUTS | 1 |
| CLONE_NEWIPC | 1 |
| CLONE_NEWNET | 1 |
| CLONE_NEWNS | 1 |
| CLONE_NEWUSER | 1 |
Namespaces are created for the container via the `clone` syscall.
### Filesystem
A root filesystem must be provided to a container for execution. The container
will use this root filesystem (rootfs) to jail and spawn processes inside where
the binaries and system libraries are local to that directory. Any binaries
to be executed must be contained within this rootfs.
Mounts that happen inside the container are automatically cleaned up when the
container exits as the mount namespace is destroyed and the kernel will
unmount all the mounts that were setup within that namespace.
For a container to execute properly there are certain filesystems that
are required to be mounted within the rootfs that the runtime will setup.
| Path | Type | Flags | Data |
| ----------- | ------ | -------------------------------------- | ---------------------------------------- |
| /proc | proc | MS_NOEXEC,MS_NOSUID,MS_NODEV | |
| /dev | tmpfs | MS_NOEXEC,MS_STRICTATIME | mode=755 |
| /dev/shm | tmpfs | MS_NOEXEC,MS_NOSUID,MS_NODEV | mode=1777,size=65536k |
| /dev/mqueue | mqueue | MS_NOEXEC,MS_NOSUID,MS_NODEV | |
| /dev/pts | devpts | MS_NOEXEC,MS_NOSUID | newinstance,ptmxmode=0666,mode=620,gid=5 |
| /sys | sysfs | MS_NOEXEC,MS_NOSUID,MS_NODEV,MS_RDONLY | |
After a container's filesystems are mounted within the newly created
mount namespace `/dev` will need to be populated with a set of device nodes.
It is expected that a rootfs does not need to have any device nodes specified
for `/dev` within the rootfs as the container will setup the correct devices
that are required for executing a container's process.
| Path | Mode | Access |
| ------------ | ---- | ---------- |
| /dev/null | 0666 | rwm |
| /dev/zero | 0666 | rwm |
| /dev/full | 0666 | rwm |
| /dev/tty | 0666 | rwm |
| /dev/random | 0666 | rwm |
| /dev/urandom | 0666 | rwm |
| /dev/fuse | 0666 | rwm |
**ptmx**
`/dev/ptmx` will need to be a symlink to the host's `/dev/ptmx` within
the container.
The use of a pseudo TTY is optional within a container and it should support both.
If a pseudo is provided to the container `/dev/console` will need to be
setup by binding the console in `/dev/` after it has been populated and mounted
in tmpfs.
| Source | Destination | UID GID | Mode | Type |
| --------------- | ------------ | ------- | ---- | ---- |
| *pty host path* | /dev/console | 0 0 | 0600 | bind |
After `/dev/null` has been setup we check for any external links between
the container's io, STDIN, STDOUT, STDERR. If the container's io is pointing
to `/dev/null` outside the container we close and `dup2` the the `/dev/null`
that is local to the container's rootfs.
After the container has `/proc` mounted a few standard symlinks are setup
within `/dev/` for the io.
| Source | Destination |
| --------------- | ----------- |
| /proc/self/fd | /dev/fd |
| /proc/self/fd/0 | /dev/stdin |
| /proc/self/fd/1 | /dev/stdout |
| /proc/self/fd/2 | /dev/stderr |
A `pivot_root` is used to change the root for the process, effectively
jailing the process inside the rootfs.
```c
put_old = mkdir(...);
pivot_root(rootfs, put_old);
chdir("/");
unmount(put_old, MS_DETACH);
rmdir(put_old);
```
For container's running with a rootfs inside `ramfs` a `MS_MOVE` combined
with a `chroot` is required as `pivot_root` is not supported in `ramfs`.
```c
mount(rootfs, "/", NULL, MS_MOVE, NULL);
chroot(".");
chdir("/");
```
The `umask` is set back to `0022` after the filesystem setup has been completed.
### Resources
Cgroups are used to handle resource allocation for containers. This includes
system resources like cpu, memory, and device access.
| Subsystem | Enabled |
| ---------- | ------- |
| devices | 1 |
| memory | 1 |
| cpu | 1 |
| cpuacct | 1 |
| cpuset | 1 |
| blkio | 1 |
| perf_event | 1 |
| freezer | 1 |
| hugetlb | 1 |
All cgroup subsystem are joined so that statistics can be collected from
each of the subsystems. Freezer does not expose any stats but is joined
so that containers can be paused and resumed.
The parent process of the container's init must place the init pid inside
the correct cgroups before the initialization begins. This is done so
that no processes or threads escape the cgroups. This sync is
done via a pipe ( specified in the runtime section below ) that the container's
init process will block waiting for the parent to finish setup.
### Security
The standard set of Linux capabilities that are set in a container
provide a good default for security and flexibility for the applications.
| Capability | Enabled |
| -------------------- | ------- |
| CAP_NET_RAW | 1 |
| CAP_NET_BIND_SERVICE | 1 |
| CAP_AUDIT_READ | 1 |
| CAP_AUDIT_WRITE | 1 |
| CAP_DAC_OVERRIDE | 1 |
| CAP_SETFCAP | 1 |
| CAP_SETPCAP | 1 |
| CAP_SETGID | 1 |
| CAP_SETUID | 1 |
| CAP_MKNOD | 1 |
| CAP_CHOWN | 1 |
| CAP_FOWNER | 1 |
| CAP_FSETID | 1 |
| CAP_KILL | 1 |
| CAP_SYS_CHROOT | 1 |
| CAP_NET_BROADCAST | 0 |
| CAP_SYS_MODULE | 0 |
| CAP_SYS_RAWIO | 0 |
| CAP_SYS_PACCT | 0 |
| CAP_SYS_ADMIN | 0 |
| CAP_SYS_NICE | 0 |
| CAP_SYS_RESOURCE | 0 |
| CAP_SYS_TIME | 0 |
| CAP_SYS_TTY_CONFIG | 0 |
| CAP_AUDIT_CONTROL | 0 |
| CAP_MAC_OVERRIDE | 0 |
| CAP_MAC_ADMIN | 0 |
| CAP_NET_ADMIN | 0 |
| CAP_SYSLOG | 0 |
| CAP_DAC_READ_SEARCH | 0 |
| CAP_LINUX_IMMUTABLE | 0 |
| CAP_IPC_LOCK | 0 |
| CAP_IPC_OWNER | 0 |
| CAP_SYS_PTRACE | 0 |
| CAP_SYS_BOOT | 0 |
| CAP_LEASE | 0 |
| CAP_WAKE_ALARM | 0 |
| CAP_BLOCK_SUSPE | 0 |
Additional security layers like [apparmor](https://wiki.ubuntu.com/AppArmor)
and [selinux](http://selinuxproject.org/page/Main_Page) can be used with
the containers. A container should support setting an apparmor profile or
selinux process and mount labels if provided in the configuration.
Standard apparmor profile:
```c
#include <tunables/global>
profile <profile_name> flags=(attach_disconnected,mediate_deleted) {
#include <abstractions/base>
network,
capability,
file,
umount,
deny @{PROC}/sys/fs/** wklx,
deny @{PROC}/sysrq-trigger rwklx,
deny @{PROC}/mem rwklx,
deny @{PROC}/kmem rwklx,
deny @{PROC}/sys/kernel/[^s][^h][^m]* wklx,
deny @{PROC}/sys/kernel/*/** wklx,
deny mount,
deny /sys/[^f]*/** wklx,
deny /sys/f[^s]*/** wklx,
deny /sys/fs/[^c]*/** wklx,
deny /sys/fs/c[^g]*/** wklx,
deny /sys/fs/cg[^r]*/** wklx,
deny /sys/firmware/efi/efivars/** rwklx,
deny /sys/kernel/security/** rwklx,
}
```
*TODO: seccomp work is being done to find a good default config*
### Runtime and Init Process
During container creation the parent process needs to talk to the container's init
process and have a form of synchronization. This is accomplished by creating
a pipe that is passed to the container's init. When the init process first spawns
it will block on its side of the pipe until the parent closes its side. This
allows the parent to have time to set the new process inside a cgroup hierarchy
and/or write any uid/gid mappings required for user namespaces.
The pipe is passed to the init process via FD 3.
The application consuming libcontainer should be compiled statically. libcontainer
does not define any init process and the arguments provided are used to `exec` the
process inside the application. There should be no long running init within the
container spec.
If a pseudo tty is provided to a container it will open and `dup2` the console
as the container's STDIN, STDOUT, STDERR as well as mounting the console
as `/dev/console`.
An extra set of mounts are provided to a container and setup for use. A container's
rootfs can contain some non portable files inside that can cause side effects during
execution of a process. These files are usually created and populated with the container
specific information via the runtime.
**Extra runtime files:**
* /etc/hosts
* /etc/resolv.conf
* /etc/hostname
* /etc/localtime
#### Defaults
There are a few defaults that can be overridden by users, but in their omission
these apply to processes within a container.
| Type | Value |
| ------------------- | ------------------------------ |
| Parent Death Signal | SIGKILL |
| UID | 0 |
| GID | 0 |
| GROUPS | 0, NULL |
| CWD | "/" |
| $HOME | Current user's home dir or "/" |
| Readonly rootfs | false |
| Pseudo TTY | false |
## Actions
After a container is created there is a standard set of actions that can
be done to the container. These actions are part of the public API for
a container.
| Action | Description |
| -------------- | ------------------------------------------------------------------ |
| Get processes | Return all the pids for processes running inside a container |
| Get Stats | Return resource statistics for the container as a whole |
| Wait | Wait waits on the container's init process ( pid 1 ) |
| Wait Process | Wait on any of the container's processes returning the exit status |
| Destroy | Kill the container's init process and remove any filesystem state |
| Signal | Send a signal to the container's init process |
| Signal Process | Send a signal to any of the container's processes |
| Pause | Pause all processes inside the container |
| Resume | Resume all processes inside the container if paused |
| Exec | Execute a new process inside of the container ( requires setns ) |
| Set | Setup configs of the container after it's created |
### Execute a new process inside of a running container.
User can execute a new process inside of a running container. Any binaries to be
executed must be accessible within the container's rootfs.
The started process will run inside the container's rootfs. Any changes
made by the process to the container's filesystem will persist after the
process finished executing.
The started process will join all the container's existing namespaces. When the
container is paused, the process will also be paused and will resume when
the container is unpaused. The started process will only run when the container's
primary process (PID 1) is running, and will not be restarted when the container
is restarted.
#### Planned additions
The started process will have its own cgroups nested inside the container's
cgroups. This is used for process tracking and optionally resource allocation
handling for the new process. Freezer cgroup is required, the rest of the cgroups
are optional. The process executor must place its pid inside the correct
cgroups before starting the process. This is done so that no child processes or
threads can escape the cgroups.
When the process is stopped, the process executor will try (in a best-effort way)
to stop all its children and remove the sub-cgroups.

38
vendor/src/github.com/opencontainers/runc/libcontainer/apparmor/apparmor.go vendored Normal file
View file

@ -0,0 +1,38 @@
// +build apparmor,linux
package apparmor
// #cgo LDFLAGS: -lapparmor
// #include <sys/apparmor.h>
// #include <stdlib.h>
import "C"
import (
"io/ioutil"
"os"
"unsafe"
)
// IsEnabled returns true if apparmor is enabled for the host.
func IsEnabled() bool {
if _, err := os.Stat("/sys/kernel/security/apparmor"); err == nil && os.Getenv("container") == "" {
if _, err = os.Stat("/sbin/apparmor_parser"); err == nil {
buf, err := ioutil.ReadFile("/sys/module/apparmor/parameters/enabled")
return err == nil && len(buf) > 1 && buf[0] == 'Y'
}
}
return false
}
// ApplyProfile will apply the profile with the specified name to the process after
// the next exec.
func ApplyProfile(name string) error {
if name == "" {
return nil
}
cName := C.CString(name)
defer C.free(unsafe.Pointer(cName))
if _, err := C.aa_change_onexec(cName); err != nil {
return err
}
return nil
}

20
vendor/src/github.com/opencontainers/runc/libcontainer/apparmor/apparmor_disabled.go vendored Normal file
View file

@ -0,0 +1,20 @@
// +build !apparmor !linux
package apparmor
import (
"errors"
)
var ErrApparmorNotEnabled = errors.New("apparmor: config provided but apparmor not supported")
func IsEnabled() bool {
return false
}
func ApplyProfile(name string) error {
if name != "" {
return ErrApparmorNotEnabled
}
return nil
}

69
vendor/src/github.com/opencontainers/runc/libcontainer/capabilities_linux.go vendored Normal file
View file

@ -0,0 +1,69 @@
// +build linux
package libcontainer
import (
"fmt"
"os"
"strings"
"github.com/syndtr/gocapability/capability"
)
const allCapabilityTypes = capability.CAPS | capability.BOUNDS
var capabilityMap map[string]capability.Cap
func init() {
capabilityMap = make(map[string]capability.Cap)
last := capability.CAP_LAST_CAP
// workaround for RHEL6 which has no /proc/sys/kernel/cap_last_cap
if last == capability.Cap(63) {
last = capability.CAP_BLOCK_SUSPEND
}
for _, cap := range capability.List() {
if cap > last {
continue
}
capKey := fmt.Sprintf("CAP_%s", strings.ToUpper(cap.String()))
capabilityMap[capKey] = cap
}
}
func newCapWhitelist(caps []string) (*whitelist, error) {
l := []capability.Cap{}
for _, c := range caps {
v, ok := capabilityMap[c]
if !ok {
return nil, fmt.Errorf("unknown capability %q", c)
}
l = append(l, v)
}
pid, err := capability.NewPid(os.Getpid())
if err != nil {
return nil, err
}
return &whitelist{
keep: l,
pid: pid,
}, nil
}
type whitelist struct {
pid capability.Capabilities
keep []capability.Cap
}
// dropBoundingSet drops the capability bounding set to those specified in the whitelist.
func (w *whitelist) dropBoundingSet() error {
w.pid.Clear(capability.BOUNDS)
w.pid.Set(capability.BOUNDS, w.keep...)
return w.pid.Apply(capability.BOUNDS)
}
// drop drops all capabilities for the current process except those specified in the whitelist.
func (w *whitelist) drop() error {
w.pid.Clear(allCapabilityTypes)
w.pid.Set(allCapabilityTypes, w.keep...)
return w.pid.Apply(allCapabilityTypes)
}

61
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/cgroups.go vendored Normal file
View file

@ -0,0 +1,61 @@
// +build linux
package cgroups
import (
"fmt"
"github.com/opencontainers/runc/libcontainer/configs"
)
type Manager interface {
// Apply cgroup configuration to the process with the specified pid
Apply(pid int) error
// Returns the PIDs inside the cgroup set
GetPids() ([]int, error)
// Returns statistics for the cgroup set
GetStats() (*Stats, error)
// Toggles the freezer cgroup according with specified state
Freeze(state configs.FreezerState) error
// Destroys the cgroup set
Destroy() error
// NewCgroupManager() and LoadCgroupManager() require following attributes:
// Paths map[string]string
// Cgroups *cgroups.Cgroup
// Paths maps cgroup subsystem to path at which it is mounted.
// Cgroups specifies specific cgroup settings for the various subsystems
// Returns cgroup paths to save in a state file and to be able to
// restore the object later.
GetPaths() map[string]string
// Set the cgroup as configured.
Set(container *configs.Config) error
}
type NotFoundError struct {
Subsystem string
}
func (e *NotFoundError) Error() string {
return fmt.Sprintf("mountpoint for %s not found", e.Subsystem)
}
func NewNotFoundError(sub string) error {
return &NotFoundError{
Subsystem: sub,
}
}
func IsNotFound(err error) bool {
if err == nil {
return false
}
_, ok := err.(*NotFoundError)
return ok
}

3
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/cgroups_unsupported.go vendored Normal file
View file

@ -0,0 +1,3 @@
// +build !linux
package cgroups

350
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/apply_raw.go vendored Normal file
View file

@ -0,0 +1,350 @@
// +build linux
package fs
import (
"errors"
"fmt"
"io"
"io/ioutil"
"os"
"path/filepath"
"strconv"
"sync"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
var (
subsystems = subsystemSet{
&CpusetGroup{},
&DevicesGroup{},
&MemoryGroup{},
&CpuGroup{},
&CpuacctGroup{},
&BlkioGroup{},
&HugetlbGroup{},
&NetClsGroup{},
&NetPrioGroup{},
&PerfEventGroup{},
&FreezerGroup{},
}
CgroupProcesses = "cgroup.procs"
HugePageSizes, _ = cgroups.GetHugePageSize()
)
var errSubsystemDoesNotExist = errors.New("cgroup: subsystem does not exist")
type subsystemSet []subsystem
func (s subsystemSet) Get(name string) (subsystem, error) {
for _, ss := range s {
if ss.Name() == name {
return ss, nil
}
}
return nil, errSubsystemDoesNotExist
}
type subsystem interface {
// Name returns the name of the subsystem.
Name() string
// Returns the stats, as 'stats', corresponding to the cgroup under 'path'.
GetStats(path string, stats *cgroups.Stats) error
// Removes the cgroup represented by 'cgroupData'.
Remove(*cgroupData) error
// Creates and joins the cgroup represented by 'cgroupData'.
Apply(*cgroupData) error
// Set the cgroup represented by cgroup.
Set(path string, cgroup *configs.Cgroup) error
}
type Manager struct {
mu sync.Mutex
Cgroups *configs.Cgroup
Paths map[string]string
}
// The absolute path to the root of the cgroup hierarchies.
var cgroupRootLock sync.Mutex
var cgroupRoot string
// Gets the cgroupRoot.
func getCgroupRoot() (string, error) {
cgroupRootLock.Lock()
defer cgroupRootLock.Unlock()
if cgroupRoot != "" {
return cgroupRoot, nil
}
root, err := cgroups.FindCgroupMountpointDir()
if err != nil {
return "", err
}
if _, err := os.Stat(root); err != nil {
return "", err
}
cgroupRoot = root
return cgroupRoot, nil
}
type cgroupData struct {
root string
parent string
name string
config *configs.Cgroup
pid int
}
func (m *Manager) Apply(pid int) (err error) {
if m.Cgroups == nil {
return nil
}
var c = m.Cgroups
d, err := getCgroupData(m.Cgroups, pid)
if err != nil {
return err
}
paths := make(map[string]string)
defer func() {
if err != nil {
cgroups.RemovePaths(paths)
}
}()
for _, sys := range subsystems {
if err := sys.Apply(d); err != nil {
return err
}
// TODO: Apply should, ideally, be reentrant or be broken up into a separate
// create and join phase so that the cgroup hierarchy for a container can be
// created then join consists of writing the process pids to cgroup.procs
p, err := d.path(sys.Name())
if err != nil {
if cgroups.IsNotFound(err) {
continue
}
return err
}
paths[sys.Name()] = p
}
m.Paths = paths
if paths["cpu"] != "" {
if err := CheckCpushares(paths["cpu"], c.Resources.CpuShares); err != nil {
return err
}
}
return nil
}
func (m *Manager) Destroy() error {
m.mu.Lock()
defer m.mu.Unlock()
if err := cgroups.RemovePaths(m.Paths); err != nil {
return err
}
m.Paths = make(map[string]string)
return nil
}
func (m *Manager) GetPaths() map[string]string {
m.mu.Lock()
paths := m.Paths
m.mu.Unlock()
return paths
}
func (m *Manager) GetStats() (*cgroups.Stats, error) {
m.mu.Lock()
defer m.mu.Unlock()
stats := cgroups.NewStats()
for name, path := range m.Paths {
sys, err := subsystems.Get(name)
if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
continue
}
if err := sys.GetStats(path, stats); err != nil {
return nil, err
}
}
return stats, nil
}
func (m *Manager) Set(container *configs.Config) error {
for name, path := range m.Paths {
sys, err := subsystems.Get(name)
if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
continue
}
if err := sys.Set(path, container.Cgroups); err != nil {
return err
}
}
return nil
}
// Freeze toggles the container's freezer cgroup depending on the state
// provided
func (m *Manager) Freeze(state configs.FreezerState) error {
d, err := getCgroupData(m.Cgroups, 0)
if err != nil {
return err
}
dir, err := d.path("freezer")
if err != nil {
return err
}
prevState := m.Cgroups.Resources.Freezer
m.Cgroups.Resources.Freezer = state
freezer, err := subsystems.Get("freezer")
if err != nil {
return err
}
err = freezer.Set(dir, m.Cgroups)
if err != nil {
m.Cgroups.Resources.Freezer = prevState
return err
}
return nil
}
func (m *Manager) GetPids() ([]int, error) {
d, err := getCgroupData(m.Cgroups, 0)
if err != nil {
return nil, err
}
dir, err := d.path("devices")
if err != nil {
return nil, err
}
return cgroups.GetPids(dir)
}
func getCgroupData(c *configs.Cgroup, pid int) (*cgroupData, error) {
root, err := getCgroupRoot()
if err != nil {
return nil, err
}
return &cgroupData{
root: root,
parent: c.Parent,
name: c.Name,
config: c,
pid: pid,
}, nil
}
func (raw *cgroupData) parentPath(subsystem, mountpoint, root string) (string, error) {
// Use GetThisCgroupDir instead of GetInitCgroupDir, because the creating
// process could in container and shared pid namespace with host, and
// /proc/1/cgroup could point to whole other world of cgroups.
initPath, err := cgroups.GetThisCgroupDir(subsystem)
if err != nil {
return "", err
}
// This is needed for nested containers, because in /proc/self/cgroup we
// see pathes from host, which don't exist in container.
relDir, err := filepath.Rel(root, initPath)
if err != nil {
return "", err
}
return filepath.Join(mountpoint, relDir), nil
}
func (raw *cgroupData) path(subsystem string) (string, error) {
mnt, root, err := cgroups.FindCgroupMountpointAndRoot(subsystem)
// If we didn't mount the subsystem, there is no point we make the path.
if err != nil {
return "", err
}
cgPath := filepath.Join(raw.parent, raw.name)
// If the cgroup name/path is absolute do not look relative to the cgroup of the init process.
if filepath.IsAbs(cgPath) {
// Sometimes subsystems can be mounted togethger as 'cpu,cpuacct'.
return filepath.Join(raw.root, filepath.Base(mnt), cgPath), nil
}
parentPath, err := raw.parentPath(subsystem, mnt, root)
if err != nil {
return "", err
}
return filepath.Join(parentPath, cgPath), nil
}
func (raw *cgroupData) join(subsystem string) (string, error) {
path, err := raw.path(subsystem)
if err != nil {
return "", err
}
if err := os.MkdirAll(path, 0755); err != nil {
return "", err
}
if err := writeFile(path, CgroupProcesses, strconv.Itoa(raw.pid)); err != nil {
return "", err
}
return path, nil
}
func writeFile(dir, file, data string) error {
// Normally dir should not be empty, one case is that cgroup subsystem
// is not mounted, we will get empty dir, and we want it fail here.
if dir == "" {
return fmt.Errorf("no such directory for %s.", file)
}
return ioutil.WriteFile(filepath.Join(dir, file), []byte(data), 0700)
}
func readFile(dir, file string) (string, error) {
data, err := ioutil.ReadFile(filepath.Join(dir, file))
return string(data), err
}
func removePath(p string, err error) error {
if err != nil {
return err
}
if p != "" {
return os.RemoveAll(p)
}
return nil
}
func CheckCpushares(path string, c int64) error {
var cpuShares int64
if c == 0 {
return nil
}
fd, err := os.Open(filepath.Join(path, "cpu.shares"))
if err != nil {
return err
}
defer fd.Close()
_, err = fmt.Fscanf(fd, "%d", &cpuShares)
if err != nil && err != io.EOF {
return err
}
if c > cpuShares {
return fmt.Errorf("The maximum allowed cpu-shares is %d", cpuShares)
} else if c < cpuShares {
return fmt.Errorf("The minimum allowed cpu-shares is %d", cpuShares)
}
return nil
}

242
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/blkio.go vendored Normal file
View file

@ -0,0 +1,242 @@
// +build linux
package fs
import (
"bufio"
"fmt"
"os"
"path/filepath"
"strconv"
"strings"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type BlkioGroup struct {
}
func (s *BlkioGroup) Name() string {
return "blkio"
}
func (s *BlkioGroup) Apply(d *cgroupData) error {
dir, err := d.join("blkio")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *BlkioGroup) Set(path string, cgroup *configs.Cgroup) error {
if cgroup.Resources.BlkioWeight != 0 {
if err := writeFile(path, "blkio.weight", strconv.FormatUint(uint64(cgroup.Resources.BlkioWeight), 10)); err != nil {
return err
}
}
if cgroup.Resources.BlkioLeafWeight != 0 {
if err := writeFile(path, "blkio.leaf_weight", strconv.FormatUint(uint64(cgroup.Resources.BlkioLeafWeight), 10)); err != nil {
return err
}
}
for _, wd := range cgroup.Resources.BlkioWeightDevice {
if err := writeFile(path, "blkio.weight_device", wd.WeightString()); err != nil {
return err
}
if err := writeFile(path, "blkio.leaf_weight_device", wd.LeafWeightString()); err != nil {
return err
}
}
for _, td := range cgroup.Resources.BlkioThrottleReadBpsDevice {
if err := writeFile(path, "blkio.throttle.read_bps_device", td.String()); err != nil {
return err
}
}
for _, td := range cgroup.Resources.BlkioThrottleWriteBpsDevice {
if err := writeFile(path, "blkio.throttle.write_bps_device", td.String()); err != nil {
return err
}
}
for _, td := range cgroup.Resources.BlkioThrottleReadIOPSDevice {
if err := writeFile(path, "blkio.throttle.read_iops_device", td.String()); err != nil {
return err
}
}
for _, td := range cgroup.Resources.BlkioThrottleWriteIOPSDevice {
if err := writeFile(path, "blkio.throttle.write_iops_device", td.String()); err != nil {
return err
}
}
return nil
}
func (s *BlkioGroup) Remove(d *cgroupData) error {
return removePath(d.path("blkio"))
}
/*
examples:
blkio.sectors
8:0 6792
blkio.io_service_bytes
8:0 Read 1282048
8:0 Write 2195456
8:0 Sync 2195456
8:0 Async 1282048
8:0 Total 3477504
Total 3477504
blkio.io_serviced
8:0 Read 124
8:0 Write 104
8:0 Sync 104
8:0 Async 124
8:0 Total 228
Total 228
blkio.io_queued
8:0 Read 0
8:0 Write 0
8:0 Sync 0
8:0 Async 0
8:0 Total 0
Total 0
*/
func splitBlkioStatLine(r rune) bool {
return r == ' ' || r == ':'
}
func getBlkioStat(path string) ([]cgroups.BlkioStatEntry, error) {
var blkioStats []cgroups.BlkioStatEntry
f, err := os.Open(path)
if err != nil {
if os.IsNotExist(err) {
return blkioStats, nil
}
return nil, err
}
defer f.Close()
sc := bufio.NewScanner(f)
for sc.Scan() {
// format: dev type amount
fields := strings.FieldsFunc(sc.Text(), splitBlkioStatLine)
if len(fields) < 3 {
if len(fields) == 2 && fields[0] == "Total" {
// skip total line
continue
} else {
return nil, fmt.Errorf("Invalid line found while parsing %s: %s", path, sc.Text())
}
}
v, err := strconv.ParseUint(fields[0], 10, 64)
if err != nil {
return nil, err
}
major := v
v, err = strconv.ParseUint(fields[1], 10, 64)
if err != nil {
return nil, err
}
minor := v
op := ""
valueField := 2
if len(fields) == 4 {
op = fields[2]
valueField = 3
}
v, err = strconv.ParseUint(fields[valueField], 10, 64)
if err != nil {
return nil, err
}
blkioStats = append(blkioStats, cgroups.BlkioStatEntry{Major: major, Minor: minor, Op: op, Value: v})
}
return blkioStats, nil
}
func (s *BlkioGroup) GetStats(path string, stats *cgroups.Stats) error {
// Try to read CFQ stats available on all CFQ enabled kernels first
if blkioStats, err := getBlkioStat(filepath.Join(path, "blkio.io_serviced_recursive")); err == nil && blkioStats != nil {
return getCFQStats(path, stats)
}
return getStats(path, stats) // Use generic stats as fallback
}
func getCFQStats(path string, stats *cgroups.Stats) error {
var blkioStats []cgroups.BlkioStatEntry
var err error
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.sectors_recursive")); err != nil {
return err
}
stats.BlkioStats.SectorsRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_service_bytes_recursive")); err != nil {
return err
}
stats.BlkioStats.IoServiceBytesRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_serviced_recursive")); err != nil {
return err
}
stats.BlkioStats.IoServicedRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_queued_recursive")); err != nil {
return err
}
stats.BlkioStats.IoQueuedRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_service_time_recursive")); err != nil {
return err
}
stats.BlkioStats.IoServiceTimeRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_wait_time_recursive")); err != nil {
return err
}
stats.BlkioStats.IoWaitTimeRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_merged_recursive")); err != nil {
return err
}
stats.BlkioStats.IoMergedRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.time_recursive")); err != nil {
return err
}
stats.BlkioStats.IoTimeRecursive = blkioStats
return nil
}
func getStats(path string, stats *cgroups.Stats) error {
var blkioStats []cgroups.BlkioStatEntry
var err error
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.throttle.io_service_bytes")); err != nil {
return err
}
stats.BlkioStats.IoServiceBytesRecursive = blkioStats
if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.throttle.io_serviced")); err != nil {
return err
}
stats.BlkioStats.IoServicedRecursive = blkioStats
return nil
}

99
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpu.go vendored Normal file
View file

@ -0,0 +1,99 @@
// +build linux
package fs
import (
"bufio"
"os"
"path/filepath"
"strconv"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type CpuGroup struct {
}
func (s *CpuGroup) Name() string {
return "cpu"
}
func (s *CpuGroup) Apply(d *cgroupData) error {
// We always want to join the cpu group, to allow fair cpu scheduling
// on a container basis
dir, err := d.join("cpu")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *CpuGroup) Set(path string, cgroup *configs.Cgroup) error {
if cgroup.Resources.CpuShares != 0 {
if err := writeFile(path, "cpu.shares", strconv.FormatInt(cgroup.Resources.CpuShares, 10)); err != nil {
return err
}
}
if cgroup.Resources.CpuPeriod != 0 {
if err := writeFile(path, "cpu.cfs_period_us", strconv.FormatInt(cgroup.Resources.CpuPeriod, 10)); err != nil {
return err
}
}
if cgroup.Resources.CpuQuota != 0 {
if err := writeFile(path, "cpu.cfs_quota_us", strconv.FormatInt(cgroup.Resources.CpuQuota, 10)); err != nil {
return err
}
}
if cgroup.Resources.CpuRtPeriod != 0 {
if err := writeFile(path, "cpu.rt_period_us", strconv.FormatInt(cgroup.Resources.CpuRtPeriod, 10)); err != nil {
return err
}
}
if cgroup.Resources.CpuRtRuntime != 0 {
if err := writeFile(path, "cpu.rt_runtime_us", strconv.FormatInt(cgroup.Resources.CpuRtRuntime, 10)); err != nil {
return err
}
}
return nil
}
func (s *CpuGroup) Remove(d *cgroupData) error {
return removePath(d.path("cpu"))
}
func (s *CpuGroup) GetStats(path string, stats *cgroups.Stats) error {
f, err := os.Open(filepath.Join(path, "cpu.stat"))
if err != nil {
if os.IsNotExist(err) {
return nil
}
return err
}
defer f.Close()
sc := bufio.NewScanner(f)
for sc.Scan() {
t, v, err := getCgroupParamKeyValue(sc.Text())
if err != nil {
return err
}
switch t {
case "nr_periods":
stats.CpuStats.ThrottlingData.Periods = v
case "nr_throttled":
stats.CpuStats.ThrottlingData.ThrottledPeriods = v
case "throttled_time":
stats.CpuStats.ThrottlingData.ThrottledTime = v
}
}
return nil
}

121
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuacct.go vendored Normal file
View file

@ -0,0 +1,121 @@
// +build linux
package fs
import (
"fmt"
"io/ioutil"
"path/filepath"
"strconv"
"strings"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/system"
)
const (
cgroupCpuacctStat = "cpuacct.stat"
nanosecondsInSecond = 1000000000
)
var clockTicks = uint64(system.GetClockTicks())
type CpuacctGroup struct {
}
func (s *CpuacctGroup) Name() string {
return "cpuacct"
}
func (s *CpuacctGroup) Apply(d *cgroupData) error {
// we just want to join this group even though we don't set anything
if _, err := d.join("cpuacct"); err != nil && !cgroups.IsNotFound(err) {
return err
}
return nil
}
func (s *CpuacctGroup) Set(path string, cgroup *configs.Cgroup) error {
return nil
}
func (s *CpuacctGroup) Remove(d *cgroupData) error {
return removePath(d.path("cpuacct"))
}
func (s *CpuacctGroup) GetStats(path string, stats *cgroups.Stats) error {
userModeUsage, kernelModeUsage, err := getCpuUsageBreakdown(path)
if err != nil {
return err
}
totalUsage, err := getCgroupParamUint(path, "cpuacct.usage")
if err != nil {
return err
}
percpuUsage, err := getPercpuUsage(path)
if err != nil {
return err
}
stats.CpuStats.CpuUsage.TotalUsage = totalUsage
stats.CpuStats.CpuUsage.PercpuUsage = percpuUsage
stats.CpuStats.CpuUsage.UsageInUsermode = userModeUsage
stats.CpuStats.CpuUsage.UsageInKernelmode = kernelModeUsage
return nil
}
// Returns user and kernel usage breakdown in nanoseconds.
func getCpuUsageBreakdown(path string) (uint64, uint64, error) {
userModeUsage := uint64(0)
kernelModeUsage := uint64(0)
const (
userField = "user"
systemField = "system"
)
// Expected format:
// user <usage in ticks>
// system <usage in ticks>
data, err := ioutil.ReadFile(filepath.Join(path, cgroupCpuacctStat))
if err != nil {
return 0, 0, err
}
fields := strings.Fields(string(data))
if len(fields) != 4 {
return 0, 0, fmt.Errorf("failure - %s is expected to have 4 fields", filepath.Join(path, cgroupCpuacctStat))
}
if fields[0] != userField {
return 0, 0, fmt.Errorf("unexpected field %q in %q, expected %q", fields[0], cgroupCpuacctStat, userField)
}
if fields[2] != systemField {
return 0, 0, fmt.Errorf("unexpected field %q in %q, expected %q", fields[2], cgroupCpuacctStat, systemField)
}
if userModeUsage, err = strconv.ParseUint(fields[1], 10, 64); err != nil {
return 0, 0, err
}
if kernelModeUsage, err = strconv.ParseUint(fields[3], 10, 64); err != nil {
return 0, 0, err
}
return (userModeUsage * nanosecondsInSecond) / clockTicks, (kernelModeUsage * nanosecondsInSecond) / clockTicks, nil
}
func getPercpuUsage(path string) ([]uint64, error) {
percpuUsage := []uint64{}
data, err := ioutil.ReadFile(filepath.Join(path, "cpuacct.usage_percpu"))
if err != nil {
return percpuUsage, err
}
for _, value := range strings.Fields(string(data)) {
value, err := strconv.ParseUint(value, 10, 64)
if err != nil {
return percpuUsage, fmt.Errorf("Unable to convert param value to uint64: %s", err)
}
percpuUsage = append(percpuUsage, value)
}
return percpuUsage, nil
}

138
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuset.go vendored Normal file
View file

@ -0,0 +1,138 @@
// +build linux
package fs
import (
"bytes"
"io/ioutil"
"os"
"path/filepath"
"strconv"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type CpusetGroup struct {
}
func (s *CpusetGroup) Name() string {
return "cpuset"
}
func (s *CpusetGroup) Apply(d *cgroupData) error {
dir, err := d.path("cpuset")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
return s.ApplyDir(dir, d.config, d.pid)
}
func (s *CpusetGroup) Set(path string, cgroup *configs.Cgroup) error {
if cgroup.Resources.CpusetCpus != "" {
if err := writeFile(path, "cpuset.cpus", cgroup.Resources.CpusetCpus); err != nil {
return err
}
}
if cgroup.Resources.CpusetMems != "" {
if err := writeFile(path, "cpuset.mems", cgroup.Resources.CpusetMems); err != nil {
return err
}
}
return nil
}
func (s *CpusetGroup) Remove(d *cgroupData) error {
return removePath(d.path("cpuset"))
}
func (s *CpusetGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}
func (s *CpusetGroup) ApplyDir(dir string, cgroup *configs.Cgroup, pid int) error {
// This might happen if we have no cpuset cgroup mounted.
// Just do nothing and don't fail.
if dir == "" {
return nil
}
root, err := getCgroupRoot()
if err != nil {
return err
}
if err := s.ensureParent(dir, root); err != nil {
return err
}
// the default values inherit from parent cgroup are already set in
// s.ensureParent, cover these if we have our own
if err := s.Set(dir, cgroup); err != nil {
return err
}
// because we are not using d.join we need to place the pid into the procs file
// unlike the other subsystems
if err := writeFile(dir, "cgroup.procs", strconv.Itoa(pid)); err != nil {
return err
}
return nil
}
func (s *CpusetGroup) getSubsystemSettings(parent string) (cpus []byte, mems []byte, err error) {
if cpus, err = ioutil.ReadFile(filepath.Join(parent, "cpuset.cpus")); err != nil {
return
}
if mems, err = ioutil.ReadFile(filepath.Join(parent, "cpuset.mems")); err != nil {
return
}
return cpus, mems, nil
}
// ensureParent makes sure that the parent directory of current is created
// and populated with the proper cpus and mems files copied from
// it's parent.
func (s *CpusetGroup) ensureParent(current, root string) error {
parent := filepath.Dir(current)
if filepath.Clean(parent) == root {
return nil
}
if err := s.ensureParent(parent, root); err != nil {
return err
}
if err := os.MkdirAll(current, 0755); err != nil {
return err
}
return s.copyIfNeeded(current, parent)
}
// copyIfNeeded copies the cpuset.cpus and cpuset.mems from the parent
// directory to the current directory if the file's contents are 0
func (s *CpusetGroup) copyIfNeeded(current, parent string) error {
var (
err error
currentCpus, currentMems []byte
parentCpus, parentMems []byte
)
if currentCpus, currentMems, err = s.getSubsystemSettings(current); err != nil {
return err
}
if parentCpus, parentMems, err = s.getSubsystemSettings(parent); err != nil {
return err
}
if s.isEmpty(currentCpus) {
if err := writeFile(current, "cpuset.cpus", string(parentCpus)); err != nil {
return err
}
}
if s.isEmpty(currentMems) {
if err := writeFile(current, "cpuset.mems", string(parentMems)); err != nil {
return err
}
}
return nil
}
func (s *CpusetGroup) isEmpty(b []byte) bool {
return len(bytes.Trim(b, "\n")) == 0
}

65
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/devices.go vendored Normal file
View file

@ -0,0 +1,65 @@
// +build linux
package fs
import (
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type DevicesGroup struct {
}
func (s *DevicesGroup) Name() string {
return "devices"
}
func (s *DevicesGroup) Apply(d *cgroupData) error {
dir, err := d.join("devices")
if err != nil {
// We will return error even it's `not found` error, devices
// cgroup is hard requirement for container's security.
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *DevicesGroup) Set(path string, cgroup *configs.Cgroup) error {
if !cgroup.Resources.AllowAllDevices {
if err := writeFile(path, "devices.deny", "a"); err != nil {
return err
}
for _, dev := range cgroup.Resources.AllowedDevices {
if err := writeFile(path, "devices.allow", dev.CgroupString()); err != nil {
return err
}
}
return nil
}
if err := writeFile(path, "devices.allow", "a"); err != nil {
return err
}
for _, dev := range cgroup.Resources.DeniedDevices {
if err := writeFile(path, "devices.deny", dev.CgroupString()); err != nil {
return err
}
}
return nil
}
func (s *DevicesGroup) Remove(d *cgroupData) error {
return removePath(d.path("devices"))
}
func (s *DevicesGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

66
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/freezer.go vendored Normal file
View file

@ -0,0 +1,66 @@
// +build linux
package fs
import (
"fmt"
"strings"
"time"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type FreezerGroup struct {
}
func (s *FreezerGroup) Name() string {
return "freezer"
}
func (s *FreezerGroup) Apply(d *cgroupData) error {
dir, err := d.join("freezer")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *FreezerGroup) Set(path string, cgroup *configs.Cgroup) error {
switch cgroup.Resources.Freezer {
case configs.Frozen, configs.Thawed:
if err := writeFile(path, "freezer.state", string(cgroup.Resources.Freezer)); err != nil {
return err
}
for {
state, err := readFile(path, "freezer.state")
if err != nil {
return err
}
if strings.TrimSpace(state) == string(cgroup.Resources.Freezer) {
break
}
time.Sleep(1 * time.Millisecond)
}
case configs.Undefined:
return nil
default:
return fmt.Errorf("Invalid argument '%s' to freezer.state", string(cgroup.Resources.Freezer))
}
return nil
}
func (s *FreezerGroup) Remove(d *cgroupData) error {
return removePath(d.path("freezer"))
}
func (s *FreezerGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

3
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/fs_unsupported.go vendored Normal file
View file

@ -0,0 +1,3 @@
// +build !linux
package fs

76
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/hugetlb.go vendored Normal file
View file

@ -0,0 +1,76 @@
// +build linux
package fs
import (
"fmt"
"strconv"
"strings"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type HugetlbGroup struct {
}
func (s *HugetlbGroup) Name() string {
return "hugetlb"
}
func (s *HugetlbGroup) Apply(d *cgroupData) error {
dir, err := d.join("hugetlb")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *HugetlbGroup) Set(path string, cgroup *configs.Cgroup) error {
for _, hugetlb := range cgroup.Resources.HugetlbLimit {
if err := writeFile(path, strings.Join([]string{"hugetlb", hugetlb.Pagesize, "limit_in_bytes"}, "."), strconv.FormatUint(hugetlb.Limit, 10)); err != nil {
return err
}
}
return nil
}
func (s *HugetlbGroup) Remove(d *cgroupData) error {
return removePath(d.path("hugetlb"))
}
func (s *HugetlbGroup) GetStats(path string, stats *cgroups.Stats) error {
hugetlbStats := cgroups.HugetlbStats{}
for _, pageSize := range HugePageSizes {
usage := strings.Join([]string{"hugetlb", pageSize, "usage_in_bytes"}, ".")
value, err := getCgroupParamUint(path, usage)
if err != nil {
return fmt.Errorf("failed to parse %s - %v", usage, err)
}
hugetlbStats.Usage = value
maxUsage := strings.Join([]string{"hugetlb", pageSize, "max_usage_in_bytes"}, ".")
value, err = getCgroupParamUint(path, maxUsage)
if err != nil {
return fmt.Errorf("failed to parse %s - %v", maxUsage, err)
}
hugetlbStats.MaxUsage = value
failcnt := strings.Join([]string{"hugetlb", pageSize, "failcnt"}, ".")
value, err = getCgroupParamUint(path, failcnt)
if err != nil {
return fmt.Errorf("failed to parse %s - %v", failcnt, err)
}
hugetlbStats.Failcnt = value
stats.HugetlbStats[pageSize] = hugetlbStats
}
return nil
}

187
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/memory.go vendored Normal file
View file

@ -0,0 +1,187 @@
// +build linux
package fs
import (
"bufio"
"fmt"
"os"
"path/filepath"
"strconv"
"strings"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type MemoryGroup struct {
}
func (s *MemoryGroup) Name() string {
return "memory"
}
func (s *MemoryGroup) Apply(d *cgroupData) (err error) {
path, err := d.path("memory")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if memoryAssigned(d.config) {
if path != "" {
if err := os.MkdirAll(path, 0755); err != nil {
return err
}
}
if err := s.Set(path, d.config); err != nil {
return err
}
}
defer func() {
if err != nil {
os.RemoveAll(path)
}
}()
// We need to join memory cgroup after set memory limits, because
// kmem.limit_in_bytes can only be set when the cgroup is empty.
_, err = d.join("memory")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
return nil
}
func (s *MemoryGroup) Set(path string, cgroup *configs.Cgroup) error {
if cgroup.Resources.Memory != 0 {
if err := writeFile(path, "memory.limit_in_bytes", strconv.FormatInt(cgroup.Resources.Memory, 10)); err != nil {
return err
}
}
if cgroup.Resources.MemoryReservation != 0 {
if err := writeFile(path, "memory.soft_limit_in_bytes", strconv.FormatInt(cgroup.Resources.MemoryReservation, 10)); err != nil {
return err
}
}
if cgroup.Resources.MemorySwap > 0 {
if err := writeFile(path, "memory.memsw.limit_in_bytes", strconv.FormatInt(cgroup.Resources.MemorySwap, 10)); err != nil {
return err
}
}
if cgroup.Resources.KernelMemory > 0 {
if err := writeFile(path, "memory.kmem.limit_in_bytes", strconv.FormatInt(cgroup.Resources.KernelMemory, 10)); err != nil {
return err
}
}
if cgroup.Resources.OomKillDisable {
if err := writeFile(path, "memory.oom_control", "1"); err != nil {
return err
}
}
if cgroup.Resources.MemorySwappiness >= 0 && cgroup.Resources.MemorySwappiness <= 100 {
if err := writeFile(path, "memory.swappiness", strconv.FormatInt(cgroup.Resources.MemorySwappiness, 10)); err != nil {
return err
}
} else if cgroup.Resources.MemorySwappiness == -1 {
return nil
} else {
return fmt.Errorf("invalid value:%d. valid memory swappiness range is 0-100", cgroup.Resources.MemorySwappiness)
}
return nil
}
func (s *MemoryGroup) Remove(d *cgroupData) error {
return removePath(d.path("memory"))
}
func (s *MemoryGroup) GetStats(path string, stats *cgroups.Stats) error {
// Set stats from memory.stat.
statsFile, err := os.Open(filepath.Join(path, "memory.stat"))
if err != nil {
if os.IsNotExist(err) {
return nil
}
return err
}
defer statsFile.Close()
sc := bufio.NewScanner(statsFile)
for sc.Scan() {
t, v, err := getCgroupParamKeyValue(sc.Text())
if err != nil {
return fmt.Errorf("failed to parse memory.stat (%q) - %v", sc.Text(), err)
}
stats.MemoryStats.Stats[t] = v
}
stats.MemoryStats.Cache = stats.MemoryStats.Stats["cache"]
memoryUsage, err := getMemoryData(path, "")
if err != nil {
return err
}
stats.MemoryStats.Usage = memoryUsage
swapUsage, err := getMemoryData(path, "memsw")
if err != nil {
return err
}
stats.MemoryStats.SwapUsage = swapUsage
kernelUsage, err := getMemoryData(path, "kmem")
if err != nil {
return err
}
stats.MemoryStats.KernelUsage = kernelUsage
return nil
}
func memoryAssigned(cgroup *configs.Cgroup) bool {
return cgroup.Resources.Memory != 0 ||
cgroup.Resources.MemoryReservation != 0 ||
cgroup.Resources.MemorySwap > 0 ||
cgroup.Resources.KernelMemory > 0 ||
cgroup.Resources.OomKillDisable ||
cgroup.Resources.MemorySwappiness != -1
}
func getMemoryData(path, name string) (cgroups.MemoryData, error) {
memoryData := cgroups.MemoryData{}
moduleName := "memory"
if name != "" {
moduleName = strings.Join([]string{"memory", name}, ".")
}
usage := strings.Join([]string{moduleName, "usage_in_bytes"}, ".")
maxUsage := strings.Join([]string{moduleName, "max_usage_in_bytes"}, ".")
failcnt := strings.Join([]string{moduleName, "failcnt"}, ".")
value, err := getCgroupParamUint(path, usage)
if err != nil {
if moduleName != "memory" && os.IsNotExist(err) {
return cgroups.MemoryData{}, nil
}
return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", usage, err)
}
memoryData.Usage = value
value, err = getCgroupParamUint(path, maxUsage)
if err != nil {
if moduleName != "memory" && os.IsNotExist(err) {
return cgroups.MemoryData{}, nil
}
return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", maxUsage, err)
}
memoryData.MaxUsage = value
value, err = getCgroupParamUint(path, failcnt)
if err != nil {
if moduleName != "memory" && os.IsNotExist(err) {
return cgroups.MemoryData{}, nil
}
return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", failcnt, err)
}
memoryData.Failcnt = value
return memoryData, nil
}

32
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/name.go vendored Normal file
View file

@ -0,0 +1,32 @@
// +build linux
package fs
import (
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type NameGroup struct {
GroupName string
}
func (s *NameGroup) Name() string {
return s.GroupName
}
func (s *NameGroup) Apply(d *cgroupData) error {
return nil
}
func (s *NameGroup) Set(path string, cgroup *configs.Cgroup) error {
return nil
}
func (s *NameGroup) Remove(d *cgroupData) error {
return nil
}
func (s *NameGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

46
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_cls.go vendored Normal file
View file

@ -0,0 +1,46 @@
// +build linux
package fs
import (
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type NetClsGroup struct {
}
func (s *NetClsGroup) Name() string {
return "net_cls"
}
func (s *NetClsGroup) Apply(d *cgroupData) error {
dir, err := d.join("net_cls")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *NetClsGroup) Set(path string, cgroup *configs.Cgroup) error {
if cgroup.Resources.NetClsClassid != "" {
if err := writeFile(path, "net_cls.classid", cgroup.Resources.NetClsClassid); err != nil {
return err
}
}
return nil
}
func (s *NetClsGroup) Remove(d *cgroupData) error {
return removePath(d.path("net_cls"))
}
func (s *NetClsGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

46
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_prio.go vendored Normal file
View file

@ -0,0 +1,46 @@
// +build linux
package fs
import (
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type NetPrioGroup struct {
}
func (s *NetPrioGroup) Name() string {
return "net_prio"
}
func (s *NetPrioGroup) Apply(d *cgroupData) error {
dir, err := d.join("net_prio")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := s.Set(dir, d.config); err != nil {
return err
}
return nil
}
func (s *NetPrioGroup) Set(path string, cgroup *configs.Cgroup) error {
for _, prioMap := range cgroup.Resources.NetPrioIfpriomap {
if err := writeFile(path, "net_prio.ifpriomap", prioMap.CgroupString()); err != nil {
return err
}
}
return nil
}
func (s *NetPrioGroup) Remove(d *cgroupData) error {
return removePath(d.path("net_prio"))
}
func (s *NetPrioGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

35
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/perf_event.go vendored Normal file
View file

@ -0,0 +1,35 @@
// +build linux
package fs
import (
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type PerfEventGroup struct {
}
func (s *PerfEventGroup) Name() string {
return "perf_event"
}
func (s *PerfEventGroup) Apply(d *cgroupData) error {
// we just want to join this group even though we don't set anything
if _, err := d.join("perf_event"); err != nil && !cgroups.IsNotFound(err) {
return err
}
return nil
}
func (s *PerfEventGroup) Set(path string, cgroup *configs.Cgroup) error {
return nil
}
func (s *PerfEventGroup) Remove(d *cgroupData) error {
return removePath(d.path("perf_event"))
}
func (s *PerfEventGroup) GetStats(path string, stats *cgroups.Stats) error {
return nil
}

79
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/fs/utils.go vendored Normal file
View file

@ -0,0 +1,79 @@
// +build linux
package fs
import (
"errors"
"fmt"
"io/ioutil"
"path/filepath"
"strconv"
"strings"
)
var (
ErrNotSupportStat = errors.New("stats are not supported for subsystem")
ErrNotValidFormat = errors.New("line is not a valid key value format")
)
// Saturates negative values at zero and returns a uint64.
// Due to kernel bugs, some of the memory cgroup stats can be negative.
func parseUint(s string, base, bitSize int) (uint64, error) {
value, err := strconv.ParseUint(s, base, bitSize)
if err != nil {
intValue, intErr := strconv.ParseInt(s, base, bitSize)
// 1. Handle negative values greater than MinInt64 (and)
// 2. Handle negative values lesser than MinInt64
if intErr == nil && intValue < 0 {
return 0, nil
} else if intErr != nil && intErr.(*strconv.NumError).Err == strconv.ErrRange && intValue < 0 {
return 0, nil
}
return value, err
}
return value, nil
}
// Parses a cgroup param and returns as name, value
// i.e. "io_service_bytes 1234" will return as io_service_bytes, 1234
func getCgroupParamKeyValue(t string) (string, uint64, error) {
parts := strings.Fields(t)
switch len(parts) {
case 2:
value, err := parseUint(parts[1], 10, 64)
if err != nil {
return "", 0, fmt.Errorf("unable to convert param value (%q) to uint64: %v", parts[1], err)
}
return parts[0], value, nil
default:
return "", 0, ErrNotValidFormat
}
}
// Gets a single uint64 value from the specified cgroup file.
func getCgroupParamUint(cgroupPath, cgroupFile string) (uint64, error) {
fileName := filepath.Join(cgroupPath, cgroupFile)
contents, err := ioutil.ReadFile(fileName)
if err != nil {
return 0, err
}
res, err := parseUint(strings.TrimSpace(string(contents)), 10, 64)
if err != nil {
return res, fmt.Errorf("unable to parse %q as a uint from Cgroup file %q", string(contents), fileName)
}
return res, nil
}
// Gets a string value from the specified cgroup file
func getCgroupParamString(cgroupPath, cgroupFile string) (string, error) {
contents, err := ioutil.ReadFile(filepath.Join(cgroupPath, cgroupFile))
if err != nil {
return "", err
}
return strings.TrimSpace(string(contents)), nil
}

92
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/stats.go vendored Normal file
View file

@ -0,0 +1,92 @@
// +build linux
package cgroups
type ThrottlingData struct {
// Number of periods with throttling active
Periods uint64 `json:"periods,omitempty"`
// Number of periods when the container hit its throttling limit.
ThrottledPeriods uint64 `json:"throttled_periods,omitempty"`
// Aggregate time the container was throttled for in nanoseconds.
ThrottledTime uint64 `json:"throttled_time,omitempty"`
}
// All CPU stats are aggregate since container inception.
type CpuUsage struct {
// Total CPU time consumed.
// Units: nanoseconds.
TotalUsage uint64 `json:"total_usage,omitempty"`
// Total CPU time consumed per core.
// Units: nanoseconds.
PercpuUsage []uint64 `json:"percpu_usage,omitempty"`
// Time spent by tasks of the cgroup in kernel mode.
// Units: nanoseconds.
UsageInKernelmode uint64 `json:"usage_in_kernelmode"`
// Time spent by tasks of the cgroup in user mode.
// Units: nanoseconds.
UsageInUsermode uint64 `json:"usage_in_usermode"`
}
type CpuStats struct {
CpuUsage CpuUsage `json:"cpu_usage,omitempty"`
ThrottlingData ThrottlingData `json:"throttling_data,omitempty"`
}
type MemoryData struct {
Usage uint64 `json:"usage,omitempty"`
MaxUsage uint64 `json:"max_usage,omitempty"`
Failcnt uint64 `json:"failcnt"`
}
type MemoryStats struct {
// memory used for cache
Cache uint64 `json:"cache,omitempty"`
// usage of memory
Usage MemoryData `json:"usage,omitempty"`
// usage of memory + swap
SwapUsage MemoryData `json:"swap_usage,omitempty"`
// usafe of kernel memory
KernelUsage MemoryData `json:"kernel_usage,omitempty"`
Stats map[string]uint64 `json:"stats,omitempty"`
}
type BlkioStatEntry struct {
Major uint64 `json:"major,omitempty"`
Minor uint64 `json:"minor,omitempty"`
Op string `json:"op,omitempty"`
Value uint64 `json:"value,omitempty"`
}
type BlkioStats struct {
// number of bytes tranferred to and from the block device
IoServiceBytesRecursive []BlkioStatEntry `json:"io_service_bytes_recursive,omitempty"`
IoServicedRecursive []BlkioStatEntry `json:"io_serviced_recursive,omitempty"`
IoQueuedRecursive []BlkioStatEntry `json:"io_queue_recursive,omitempty"`
IoServiceTimeRecursive []BlkioStatEntry `json:"io_service_time_recursive,omitempty"`
IoWaitTimeRecursive []BlkioStatEntry `json:"io_wait_time_recursive,omitempty"`
IoMergedRecursive []BlkioStatEntry `json:"io_merged_recursive,omitempty"`
IoTimeRecursive []BlkioStatEntry `json:"io_time_recursive,omitempty"`
SectorsRecursive []BlkioStatEntry `json:"sectors_recursive,omitempty"`
}
type HugetlbStats struct {
// current res_counter usage for hugetlb
Usage uint64 `json:"usage,omitempty"`
// maximum usage ever recorded.
MaxUsage uint64 `json:"max_usage,omitempty"`
// number of times htgetlb usage allocation failure.
Failcnt uint64 `json:"failcnt"`
}
type Stats struct {
CpuStats CpuStats `json:"cpu_stats,omitempty"`
MemoryStats MemoryStats `json:"memory_stats,omitempty"`
BlkioStats BlkioStats `json:"blkio_stats,omitempty"`
// the map is in the format "size of hugepage: stats of the hugepage"
HugetlbStats map[string]HugetlbStats `json:"hugetlb_stats,omitempty"`
}
func NewStats() *Stats {
memoryStats := MemoryStats{Stats: make(map[string]uint64)}
hugetlbStats := make(map[string]HugetlbStats)
return &Stats{MemoryStats: memoryStats, HugetlbStats: hugetlbStats}
}

51
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_nosystemd.go vendored Normal file
View file

@ -0,0 +1,51 @@
// +build !linux
package systemd
import (
"fmt"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
)
type Manager struct {
Cgroups *configs.Cgroup
Paths map[string]string
}
func UseSystemd() bool {
return false
}
func (m *Manager) Apply(pid int) error {
return fmt.Errorf("Systemd not supported")
}
func (m *Manager) GetPids() ([]int, error) {
return nil, fmt.Errorf("Systemd not supported")
}
func (m *Manager) Destroy() error {
return fmt.Errorf("Systemd not supported")
}
func (m *Manager) GetPaths() map[string]string {
return nil
}
func (m *Manager) GetStats() (*cgroups.Stats, error) {
return nil, fmt.Errorf("Systemd not supported")
}
func (m *Manager) Set(container *configs.Config) error {
return nil, fmt.Errorf("Systemd not supported")
}
func (m *Manager) Freeze(state configs.FreezerState) error {
return fmt.Errorf("Systemd not supported")
}
func Freeze(c *configs.Cgroup, state configs.FreezerState) error {
return fmt.Errorf("Systemd not supported")
}

644
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_systemd.go vendored Normal file
View file

@ -0,0 +1,644 @@
// +build linux
package systemd
import (
"errors"
"fmt"
"io/ioutil"
"os"
"path/filepath"
"strconv"
"strings"
"sync"
"time"
systemdDbus "github.com/coreos/go-systemd/dbus"
systemdUtil "github.com/coreos/go-systemd/util"
"github.com/godbus/dbus"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/cgroups/fs"
"github.com/opencontainers/runc/libcontainer/configs"
)
type Manager struct {
mu sync.Mutex
Cgroups *configs.Cgroup
Paths map[string]string
}
type subsystem interface {
// Name returns the name of the subsystem.
Name() string
// Returns the stats, as 'stats', corresponding to the cgroup under 'path'.
GetStats(path string, stats *cgroups.Stats) error
// Set the cgroup represented by cgroup.
Set(path string, cgroup *configs.Cgroup) error
}
var errSubsystemDoesNotExist = errors.New("cgroup: subsystem does not exist")
type subsystemSet []subsystem
func (s subsystemSet) Get(name string) (subsystem, error) {
for _, ss := range s {
if ss.Name() == name {
return ss, nil
}
}
return nil, errSubsystemDoesNotExist
}
var subsystems = subsystemSet{
&fs.CpusetGroup{},
&fs.DevicesGroup{},
&fs.MemoryGroup{},
&fs.CpuGroup{},
&fs.CpuacctGroup{},
&fs.BlkioGroup{},
&fs.HugetlbGroup{},
&fs.PerfEventGroup{},
&fs.FreezerGroup{},
&fs.NetPrioGroup{},
&fs.NetClsGroup{},
&fs.NameGroup{GroupName: "name=systemd"},
}
const (
testScopeWait = 4
)
var (
connLock sync.Mutex
theConn *systemdDbus.Conn
hasStartTransientUnit bool
hasTransientDefaultDependencies bool
)
func newProp(name string, units interface{}) systemdDbus.Property {
return systemdDbus.Property{
Name: name,
Value: dbus.MakeVariant(units),
}
}
func UseSystemd() bool {
if !systemdUtil.IsRunningSystemd() {
return false
}
connLock.Lock()
defer connLock.Unlock()
if theConn == nil {
var err error
theConn, err = systemdDbus.New()
if err != nil {
return false
}
// Assume we have StartTransientUnit
hasStartTransientUnit = true
// But if we get UnknownMethod error we don't
if _, err := theConn.StartTransientUnit("test.scope", "invalid", nil, nil); err != nil {
if dbusError, ok := err.(dbus.Error); ok {
if dbusError.Name == "org.freedesktop.DBus.Error.UnknownMethod" {
hasStartTransientUnit = false
return hasStartTransientUnit
}
}
}
// Ensure the scope name we use doesn't exist. Use the Pid to
// avoid collisions between multiple libcontainer users on a
// single host.
scope := fmt.Sprintf("libcontainer-%d-systemd-test-default-dependencies.scope", os.Getpid())
testScopeExists := true
for i := 0; i <= testScopeWait; i++ {
if _, err := theConn.StopUnit(scope, "replace", nil); err != nil {
if dbusError, ok := err.(dbus.Error); ok {
if strings.Contains(dbusError.Name, "org.freedesktop.systemd1.NoSuchUnit") {
testScopeExists = false
break
}
}
}
time.Sleep(time.Millisecond)
}
// Bail out if we can't kill this scope without testing for DefaultDependencies
if testScopeExists {
return hasStartTransientUnit
}
// Assume StartTransientUnit on a scope allows DefaultDependencies
hasTransientDefaultDependencies = true
ddf := newProp("DefaultDependencies", false)
if _, err := theConn.StartTransientUnit(scope, "replace", []systemdDbus.Property{ddf}, nil); err != nil {
if dbusError, ok := err.(dbus.Error); ok {
if strings.Contains(dbusError.Name, "org.freedesktop.DBus.Error.PropertyReadOnly") {
hasTransientDefaultDependencies = false
}
}
}
// Not critical because of the stop unit logic above.
theConn.StopUnit(scope, "replace", nil)
}
return hasStartTransientUnit
}
func getIfaceForUnit(unitName string) string {
if strings.HasSuffix(unitName, ".scope") {
return "Scope"
}
if strings.HasSuffix(unitName, ".service") {
return "Service"
}
return "Unit"
}
func (m *Manager) Apply(pid int) error {
var (
c = m.Cgroups
unitName = getUnitName(c)
slice = "system.slice"
properties []systemdDbus.Property
)
if c.Parent != "" {
slice = c.Parent
}
properties = append(properties,
systemdDbus.PropSlice(slice),
systemdDbus.PropDescription("docker container "+c.Name),
newProp("PIDs", []uint32{uint32(pid)}),
)
// Always enable accounting, this gets us the same behaviour as the fs implementation,
// plus the kernel has some problems with joining the memory cgroup at a later time.
properties = append(properties,
newProp("MemoryAccounting", true),
newProp("CPUAccounting", true),
newProp("BlockIOAccounting", true))
if hasTransientDefaultDependencies {
properties = append(properties,
newProp("DefaultDependencies", false))
}
if c.Resources.Memory != 0 {
properties = append(properties,
newProp("MemoryLimit", uint64(c.Resources.Memory)))
}
if c.Resources.CpuShares != 0 {
properties = append(properties,
newProp("CPUShares", uint64(c.Resources.CpuShares)))
}
if c.Resources.BlkioWeight != 0 {
properties = append(properties,
newProp("BlockIOWeight", uint64(c.Resources.BlkioWeight)))
}
// We need to set kernel memory before processes join cgroup because
// kmem.limit_in_bytes can only be set when the cgroup is empty.
// And swap memory limit needs to be set after memory limit, only
// memory limit is handled by systemd, so it's kind of ugly here.
if c.Resources.KernelMemory > 0 {
if err := setKernelMemory(c); err != nil {
return err
}
}
if _, err := theConn.StartTransientUnit(unitName, "replace", properties, nil); err != nil {
return err
}
if err := joinDevices(c, pid); err != nil {
return err
}
// TODO: CpuQuota and CpuPeriod not available in systemd
// we need to manually join the cpu.cfs_quota_us and cpu.cfs_period_us
if err := joinCpu(c, pid); err != nil {
return err
}
// TODO: MemoryReservation and MemorySwap not available in systemd
if err := joinMemory(c, pid); err != nil {
return err
}
// we need to manually join the freezer, net_cls, net_prio and cpuset cgroup in systemd
// because it does not currently support it via the dbus api.
if err := joinFreezer(c, pid); err != nil {
return err
}
if err := joinNetPrio(c, pid); err != nil {
return err
}
if err := joinNetCls(c, pid); err != nil {
return err
}
if err := joinCpuset(c, pid); err != nil {
return err
}
if err := joinHugetlb(c, pid); err != nil {
return err
}
if err := joinPerfEvent(c, pid); err != nil {
return err
}
// FIXME: Systemd does have `BlockIODeviceWeight` property, but we got problem
// using that (at least on systemd 208, see https://github.com/opencontainers/runc/libcontainer/pull/354),
// so use fs work around for now.
if err := joinBlkio(c, pid); err != nil {
return err
}
paths := make(map[string]string)
for _, s := range subsystems {
subsystemPath, err := getSubsystemPath(m.Cgroups, s.Name())
if err != nil {
// Don't fail if a cgroup hierarchy was not found, just skip this subsystem
if cgroups.IsNotFound(err) {
continue
}
return err
}
paths[s.Name()] = subsystemPath
}
m.Paths = paths
if paths["cpu"] != "" {
if err := fs.CheckCpushares(paths["cpu"], c.Resources.CpuShares); err != nil {
return err
}
}
return nil
}
func (m *Manager) Destroy() error {
m.mu.Lock()
defer m.mu.Unlock()
theConn.StopUnit(getUnitName(m.Cgroups), "replace", nil)
if err := cgroups.RemovePaths(m.Paths); err != nil {
return err
}
m.Paths = make(map[string]string)
return nil
}
func (m *Manager) GetPaths() map[string]string {
m.mu.Lock()
paths := m.Paths
m.mu.Unlock()
return paths
}
func writeFile(dir, file, data string) error {
// Normally dir should not be empty, one case is that cgroup subsystem
// is not mounted, we will get empty dir, and we want it fail here.
if dir == "" {
return fmt.Errorf("no such directory for %s.", file)
}
return ioutil.WriteFile(filepath.Join(dir, file), []byte(data), 0700)
}
func join(c *configs.Cgroup, subsystem string, pid int) (string, error) {
path, err := getSubsystemPath(c, subsystem)
if err != nil {
return "", err
}
if err := os.MkdirAll(path, 0755); err != nil {
return "", err
}
if err := writeFile(path, "cgroup.procs", strconv.Itoa(pid)); err != nil {
return "", err
}
return path, nil
}
func joinCpu(c *configs.Cgroup, pid int) error {
path, err := getSubsystemPath(c, "cpu")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if c.Resources.CpuQuota != 0 {
if err = writeFile(path, "cpu.cfs_quota_us", strconv.FormatInt(c.Resources.CpuQuota, 10)); err != nil {
return err
}
}
if c.Resources.CpuPeriod != 0 {
if err = writeFile(path, "cpu.cfs_period_us", strconv.FormatInt(c.Resources.CpuPeriod, 10)); err != nil {
return err
}
}
if c.Resources.CpuRtPeriod != 0 {
if err = writeFile(path, "cpu.rt_period_us", strconv.FormatInt(c.Resources.CpuRtPeriod, 10)); err != nil {
return err
}
}
if c.Resources.CpuRtRuntime != 0 {
if err = writeFile(path, "cpu.rt_runtime_us", strconv.FormatInt(c.Resources.CpuRtRuntime, 10)); err != nil {
return err
}
}
return nil
}
func joinFreezer(c *configs.Cgroup, pid int) error {
path, err := join(c, "freezer", pid)
if err != nil && !cgroups.IsNotFound(err) {
return err
}
freezer, err := subsystems.Get("freezer")
if err != nil {
return err
}
return freezer.Set(path, c)
}
func joinNetPrio(c *configs.Cgroup, pid int) error {
path, err := join(c, "net_prio", pid)
if err != nil && !cgroups.IsNotFound(err) {
return err
}
netPrio, err := subsystems.Get("net_prio")
if err != nil {
return err
}
return netPrio.Set(path, c)
}
func joinNetCls(c *configs.Cgroup, pid int) error {
path, err := join(c, "net_cls", pid)
if err != nil && !cgroups.IsNotFound(err) {
return err
}
netcls, err := subsystems.Get("net_cls")
if err != nil {
return err
}
return netcls.Set(path, c)
}
func getSubsystemPath(c *configs.Cgroup, subsystem string) (string, error) {
mountpoint, err := cgroups.FindCgroupMountpoint(subsystem)
if err != nil {
return "", err
}
initPath, err := cgroups.GetInitCgroupDir(subsystem)
if err != nil {
return "", err
}
slice := "system.slice"
if c.Parent != "" {
slice = c.Parent
}
return filepath.Join(mountpoint, initPath, slice, getUnitName(c)), nil
}
func (m *Manager) Freeze(state configs.FreezerState) error {
path, err := getSubsystemPath(m.Cgroups, "freezer")
if err != nil {
return err
}
prevState := m.Cgroups.Resources.Freezer
m.Cgroups.Resources.Freezer = state
freezer, err := subsystems.Get("freezer")
if err != nil {
return err
}
err = freezer.Set(path, m.Cgroups)
if err != nil {
m.Cgroups.Resources.Freezer = prevState
return err
}
return nil
}
func (m *Manager) GetPids() ([]int, error) {
path, err := getSubsystemPath(m.Cgroups, "devices")
if err != nil {
return nil, err
}
return cgroups.GetPids(path)
}
func (m *Manager) GetStats() (*cgroups.Stats, error) {
m.mu.Lock()
defer m.mu.Unlock()
stats := cgroups.NewStats()
for name, path := range m.Paths {
sys, err := subsystems.Get(name)
if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
continue
}
if err := sys.GetStats(path, stats); err != nil {
return nil, err
}
}
return stats, nil
}
func (m *Manager) Set(container *configs.Config) error {
for name, path := range m.Paths {
sys, err := subsystems.Get(name)
if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
continue
}
if err := sys.Set(path, container.Cgroups); err != nil {
return err
}
}
return nil
}
func getUnitName(c *configs.Cgroup) string {
return fmt.Sprintf("%s-%s.scope", c.ScopePrefix, c.Name)
}
// Atm we can't use the systemd device support because of two missing things:
// * Support for wildcards to allow mknod on any device
// * Support for wildcards to allow /dev/pts support
//
// The second is available in more recent systemd as "char-pts", but not in e.g. v208 which is
// in wide use. When both these are available we will be able to switch, but need to keep the old
// implementation for backwards compat.
//
// Note: we can't use systemd to set up the initial limits, and then change the cgroup
// because systemd will re-write the device settings if it needs to re-apply the cgroup context.
// This happens at least for v208 when any sibling unit is started.
func joinDevices(c *configs.Cgroup, pid int) error {
path, err := join(c, "devices", pid)
// Even if it's `not found` error, we'll return err because devices cgroup
// is hard requirement for container security.
if err != nil {
return err
}
devices, err := subsystems.Get("devices")
if err != nil {
return err
}
return devices.Set(path, c)
}
func setKernelMemory(c *configs.Cgroup) error {
path, err := getSubsystemPath(c, "memory")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
if err := os.MkdirAll(path, 0755); err != nil {
return err
}
if c.Resources.KernelMemory > 0 {
err = writeFile(path, "memory.kmem.limit_in_bytes", strconv.FormatInt(c.Resources.KernelMemory, 10))
if err != nil {
return err
}
}
return nil
}
func joinMemory(c *configs.Cgroup, pid int) error {
path, err := getSubsystemPath(c, "memory")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
// -1 disables memoryswap
if c.Resources.MemorySwap > 0 {
err = writeFile(path, "memory.memsw.limit_in_bytes", strconv.FormatInt(c.Resources.MemorySwap, 10))
if err != nil {
return err
}
}
if c.Resources.MemoryReservation > 0 {
err = writeFile(path, "memory.soft_limit_in_bytes", strconv.FormatInt(c.Resources.MemoryReservation, 10))
if err != nil {
return err
}
}
if c.Resources.OomKillDisable {
if err := writeFile(path, "memory.oom_control", "1"); err != nil {
return err
}
}
if c.Resources.MemorySwappiness >= 0 && c.Resources.MemorySwappiness <= 100 {
err = writeFile(path, "memory.swappiness", strconv.FormatInt(c.Resources.MemorySwappiness, 10))
if err != nil {
return err
}
} else if c.Resources.MemorySwappiness == -1 {
return nil
} else {
return fmt.Errorf("invalid value:%d. valid memory swappiness range is 0-100", c.Resources.MemorySwappiness)
}
return nil
}
// systemd does not atm set up the cpuset controller, so we must manually
// join it. Additionally that is a very finicky controller where each
// level must have a full setup as the default for a new directory is "no cpus"
func joinCpuset(c *configs.Cgroup, pid int) error {
path, err := getSubsystemPath(c, "cpuset")
if err != nil && !cgroups.IsNotFound(err) {
return err
}
s := &fs.CpusetGroup{}
return s.ApplyDir(path, c, pid)
}
// `BlockIODeviceWeight` property of systemd does not work properly, and systemd
// expects device path instead of major minor numbers, which is also confusing
// for users. So we use fs work around for now.
func joinBlkio(c *configs.Cgroup, pid int) error {
path, err := getSubsystemPath(c, "blkio")
if err != nil {
return err
}
// systemd doesn't directly support this in the dbus properties
if c.Resources.BlkioLeafWeight != 0 {
if err := writeFile(path, "blkio.leaf_weight", strconv.FormatUint(uint64(c.Resources.BlkioLeafWeight), 10)); err != nil {
return err
}
}
for _, wd := range c.Resources.BlkioWeightDevice {
if err := writeFile(path, "blkio.weight_device", wd.WeightString()); err != nil {
return err
}
if err := writeFile(path, "blkio.leaf_weight_device", wd.LeafWeightString()); err != nil {
return err
}
}
for _, td := range c.Resources.BlkioThrottleReadBpsDevice {
if err := writeFile(path, "blkio.throttle.read_bps_device", td.String()); err != nil {
return err
}
}
for _, td := range c.Resources.BlkioThrottleWriteBpsDevice {
if err := writeFile(path, "blkio.throttle.write_bps_device", td.String()); err != nil {
return err
}
}
for _, td := range c.Resources.BlkioThrottleReadIOPSDevice {
if err := writeFile(path, "blkio.throttle.read_iops_device", td.String()); err != nil {
return err
}
}
for _, td := range c.Resources.BlkioThrottleWriteIOPSDevice {
if err := writeFile(path, "blkio.throttle.write_iops_device", td.String()); err != nil {
return err
}
}
return nil
}
func joinHugetlb(c *configs.Cgroup, pid int) error {
path, err := join(c, "hugetlb", pid)
if err != nil && !cgroups.IsNotFound(err) {
return err
}
hugetlb, err := subsystems.Get("hugetlb")
if err != nil {
return err
}
return hugetlb.Set(path, c)
}
func joinPerfEvent(c *configs.Cgroup, pid int) error {
path, err := join(c, "perf_event", pid)
if err != nil && !cgroups.IsNotFound(err) {
return err
}
perfEvent, err := subsystems.Get("perf_event")
if err != nil {
return err
}
return perfEvent.Set(path, c)
}

347
vendor/src/github.com/opencontainers/runc/libcontainer/cgroups/utils.go vendored Normal file
View file

@ -0,0 +1,347 @@
// +build linux
package cgroups
import (
"bufio"
"fmt"
"io/ioutil"
"os"
"path/filepath"
"strconv"
"strings"
"time"
"github.com/docker/docker/pkg/mount"
"github.com/docker/go-units"
)
const cgroupNamePrefix = "name="
// https://www.kernel.org/doc/Documentation/cgroups/cgroups.txt
func FindCgroupMountpoint(subsystem string) (string, error) {
// We are not using mount.GetMounts() because it's super-inefficient,
// parsing it directly sped up x10 times because of not using Sscanf.
// It was one of two major performance drawbacks in container start.
f, err := os.Open("/proc/self/mountinfo")
if err != nil {
return "", err
}
defer f.Close()
scanner := bufio.NewScanner(f)
for scanner.Scan() {
txt := scanner.Text()
fields := strings.Split(txt, " ")
for _, opt := range strings.Split(fields[len(fields)-1], ",") {
if opt == subsystem {
return fields[4], nil
}
}
}
if err := scanner.Err(); err != nil {
return "", err
}
return "", NewNotFoundError(subsystem)
}
func FindCgroupMountpointAndRoot(subsystem string) (string, string, error) {
f, err := os.Open("/proc/self/mountinfo")
if err != nil {
return "", "", err
}
defer f.Close()
scanner := bufio.NewScanner(f)
for scanner.Scan() {
txt := scanner.Text()
fields := strings.Split(txt, " ")
for _, opt := range strings.Split(fields[len(fields)-1], ",") {
if opt == subsystem {
return fields[4], fields[3], nil
}
}
}
if err := scanner.Err(); err != nil {
return "", "", err
}
return "", "", NewNotFoundError(subsystem)
}
func FindCgroupMountpointDir() (string, error) {
f, err := os.Open("/proc/self/mountinfo")
if err != nil {
return "", err
}
defer f.Close()
scanner := bufio.NewScanner(f)
for scanner.Scan() {
text := scanner.Text()
fields := strings.Split(text, " ")
// Safe as mountinfo encodes mountpoints with spaces as \040.
index := strings.Index(text, " - ")
postSeparatorFields := strings.Fields(text[index+3:])
if len(postSeparatorFields) < 3 {
return "", fmt.Errorf("Error found less than 3 fields post '-' in %q", text)
}
if postSeparatorFields[0] == "cgroup" {
return filepath.Dir(fields[4]), nil
}
}
if err := scanner.Err(); err != nil {
return "", err
}
return "", NewNotFoundError("cgroup")
}
type Mount struct {
Mountpoint string
Root string
Subsystems []string
}
func (m Mount) GetThisCgroupDir(cgroups map[string]string) (string, error) {
if len(m.Subsystems) == 0 {
return "", fmt.Errorf("no subsystem for mount")
}
return getControllerPath(m.Subsystems[0], cgroups)
}
func GetCgroupMounts() ([]Mount, error) {
mounts, err := mount.GetMounts()
if err != nil {
return nil, err
}
all, err := GetAllSubsystems()
if err != nil {
return nil, err
}
allMap := make(map[string]bool)
for _, s := range all {
allMap[s] = true
}
res := []Mount{}
for _, mount := range mounts {
if mount.Fstype == "cgroup" {
m := Mount{Mountpoint: mount.Mountpoint, Root: mount.Root}
for _, opt := range strings.Split(mount.VfsOpts, ",") {
if strings.HasPrefix(opt, cgroupNamePrefix) {
m.Subsystems = append(m.Subsystems, opt[len(cgroupNamePrefix):])
}
if allMap[opt] {
m.Subsystems = append(m.Subsystems, opt)
}
}
res = append(res, m)
}
}
return res, nil
}
// Returns all the cgroup subsystems supported by the kernel
func GetAllSubsystems() ([]string, error) {
f, err := os.Open("/proc/cgroups")
if err != nil {
return nil, err
}
defer f.Close()
subsystems := []string{}
s := bufio.NewScanner(f)
for s.Scan() {
if err := s.Err(); err != nil {
return nil, err
}
text := s.Text()
if text[0] != '#' {
parts := strings.Fields(text)
if len(parts) >= 4 && parts[3] != "0" {
subsystems = append(subsystems, parts[0])
}
}
}
return subsystems, nil
}
// Returns the relative path to the cgroup docker is running in.
func GetThisCgroupDir(subsystem string) (string, error) {
cgroups, err := ParseCgroupFile("/proc/self/cgroup")
if err != nil {
return "", err
}
return getControllerPath(subsystem, cgroups)
}
func GetInitCgroupDir(subsystem string) (string, error) {
cgroups, err := ParseCgroupFile("/proc/1/cgroup")
if err != nil {
return "", err
}
return getControllerPath(subsystem, cgroups)
}
func readProcsFile(dir string) ([]int, error) {
f, err := os.Open(filepath.Join(dir, "cgroup.procs"))
if err != nil {
return nil, err
}
defer f.Close()
var (
s = bufio.NewScanner(f)
out = []int{}
)
for s.Scan() {
if t := s.Text(); t != "" {
pid, err := strconv.Atoi(t)
if err != nil {
return nil, err
}
out = append(out, pid)
}
}
return out, nil
}
func ParseCgroupFile(path string) (map[string]string, error) {
f, err := os.Open(path)
if err != nil {
return nil, err
}
defer f.Close()
s := bufio.NewScanner(f)
cgroups := make(map[string]string)
for s.Scan() {
if err := s.Err(); err != nil {
return nil, err
}
text := s.Text()
parts := strings.Split(text, ":")
for _, subs := range strings.Split(parts[1], ",") {
cgroups[subs] = parts[2]
}
}
return cgroups, nil
}
func getControllerPath(subsystem string, cgroups map[string]string) (string, error) {
if p, ok := cgroups[subsystem]; ok {
return p, nil
}
if p, ok := cgroups[cgroupNamePrefix+subsystem]; ok {
return p, nil
}
return "", NewNotFoundError(subsystem)
}
func PathExists(path string) bool {
if _, err := os.Stat(path); err != nil {
return false
}
return true
}
func EnterPid(cgroupPaths map[string]string, pid int) error {
for _, path := range cgroupPaths {
if PathExists(path) {
if err := ioutil.WriteFile(filepath.Join(path, "cgroup.procs"),
[]byte(strconv.Itoa(pid)), 0700); err != nil {
return err
}
}
}
return nil
}
// RemovePaths iterates over the provided paths removing them.
// We trying to remove all paths five times with increasing delay between tries.
// If after all there are not removed cgroups - appropriate error will be
// returned.
func RemovePaths(paths map[string]string) (err error) {
delay := 10 * time.Millisecond
for i := 0; i < 5; i++ {
if i != 0 {
time.Sleep(delay)
delay *= 2
}
for s, p := range paths {
os.RemoveAll(p)
// TODO: here probably should be logging
_, err := os.Stat(p)
// We need this strange way of checking cgroups existence because
// RemoveAll almost always returns error, even on already removed
// cgroups
if os.IsNotExist(err) {
delete(paths, s)
}
}
if len(paths) == 0 {
return nil
}
}
return fmt.Errorf("Failed to remove paths: %s", paths)
}
func GetHugePageSize() ([]string, error) {
var pageSizes []string
sizeList := []string{"B", "kB", "MB", "GB", "TB", "PB"}
files, err := ioutil.ReadDir("/sys/kernel/mm/hugepages")
if err != nil {
return pageSizes, err
}
for _, st := range files {
nameArray := strings.Split(st.Name(), "-")
pageSize, err := units.RAMInBytes(nameArray[1])
if err != nil {
return []string{}, err
}
sizeString := units.CustomSize("%g%s", float64(pageSize), 1024.0, sizeList)
pageSizes = append(pageSizes, sizeString)
}
return pageSizes, nil
}
// GetPids returns all pids, that were added to cgroup at path and to all its
// subcgroups.
func GetPids(path string) ([]int, error) {
var pids []int
// collect pids from all sub-cgroups
err := filepath.Walk(path, func(p string, info os.FileInfo, iErr error) error {
dir, file := filepath.Split(p)
if file != "cgroup.procs" {
return nil
}
if iErr != nil {
return iErr
}
cPids, err := readProcsFile(dir)
if err != nil {
return err
}
pids = append(pids, cPids...)
return nil
})
return pids, err
}

10
vendor/src/github.com/opencontainers/runc/libcontainer/compat_1.5_linux.go vendored Normal file
View file

@ -0,0 +1,10 @@
// +build linux,!go1.5
package libcontainer
import "syscall"
// GidMappingsEnableSetgroups was added in Go 1.5, so do nothing when building
// with earlier versions
func enableSetgroups(sys *syscall.SysProcAttr) {
}

61
vendor/src/github.com/opencontainers/runc/libcontainer/configs/blkio_device.go vendored Normal file
View file

@ -0,0 +1,61 @@
package configs
import "fmt"
// blockIODevice holds major:minor format supported in blkio cgroup
type blockIODevice struct {
// Major is the device's major number
Major int64 `json:"major"`
// Minor is the device's minor number
Minor int64 `json:"minor"`
}
// WeightDevice struct holds a `major:minor weight`|`major:minor leaf_weight` pair
type WeightDevice struct {
blockIODevice
// Weight is the bandwidth rate for the device, range is from 10 to 1000
Weight uint16 `json:"weight"`
// LeafWeight is the bandwidth rate for the device while competing with the cgroup's child cgroups, range is from 10 to 1000, cfq scheduler only
LeafWeight uint16 `json:"leafWeight"`
}
// NewWeightDevice returns a configured WeightDevice pointer
func NewWeightDevice(major, minor int64, weight, leafWeight uint16) *WeightDevice {
wd := &WeightDevice{}
wd.Major = major
wd.Minor = minor
wd.Weight = weight
wd.LeafWeight = leafWeight
return wd
}
// WeightString formats the struct to be writable to the cgroup specific file
func (wd *WeightDevice) WeightString() string {
return fmt.Sprintf("%d:%d %d", wd.Major, wd.Minor, wd.Weight)
}
// LeafWeightString formats the struct to be writable to the cgroup specific file
func (wd *WeightDevice) LeafWeightString() string {
return fmt.Sprintf("%d:%d %d", wd.Major, wd.Minor, wd.LeafWeight)
}
// ThrottleDevice struct holds a `major:minor rate_per_second` pair
type ThrottleDevice struct {
blockIODevice
// Rate is the IO rate limit per cgroup per device
Rate uint64 `json:"rate"`
}
// NewThrottleDevice returns a configured ThrottleDevice pointer
func NewThrottleDevice(major, minor int64, rate uint64) *ThrottleDevice {
td := &ThrottleDevice{}
td.Major = major
td.Minor = minor
td.Rate = rate
return td
}
// String formats the struct to be writable to the cgroup specific file
func (td *ThrottleDevice) String() string {
return fmt.Sprintf("%d:%d %d", td.Major, td.Minor, td.Rate)
}

105
vendor/src/github.com/opencontainers/runc/libcontainer/configs/cgroup_unix.go vendored Normal file
View file

@ -0,0 +1,105 @@
// +build linux freebsd
package configs
type FreezerState string
const (
Undefined FreezerState = ""
Frozen FreezerState = "FROZEN"
Thawed FreezerState = "THAWED"
)
type Cgroup struct {
Name string `json:"name"`
// name of parent cgroup or slice
Parent string `json:"parent"`
// ScopePrefix decribes prefix for the scope name
ScopePrefix string `json:"scope_prefix"`
// Resources contains various cgroups settings to apply
Resources *Resources `json:"resources"`
}
type Resources struct {
// If this is true allow access to any kind of device within the container. If false, allow access only to devices explicitly listed in the allowed_devices list.
AllowAllDevices bool `json:"allow_all_devices"`
AllowedDevices []*Device `json:"allowed_devices"`
DeniedDevices []*Device `json:"denied_devices"`
// Memory limit (in bytes)
Memory int64 `json:"memory"`
// Memory reservation or soft_limit (in bytes)
MemoryReservation int64 `json:"memory_reservation"`
// Total memory usage (memory + swap); set `-1' to disable swap
MemorySwap int64 `json:"memory_swap"`
// Kernel memory limit (in bytes)
KernelMemory int64 `json:"kernel_memory"`
// CPU shares (relative weight vs. other containers)
CpuShares int64 `json:"cpu_shares"`
// CPU hardcap limit (in usecs). Allowed cpu time in a given period.
CpuQuota int64 `json:"cpu_quota"`
// CPU period to be used for hardcapping (in usecs). 0 to use system default.
CpuPeriod int64 `json:"cpu_period"`
// How many time CPU will use in realtime scheduling (in usecs).
CpuRtRuntime int64 `json:"cpu_quota"`
// CPU period to be used for realtime scheduling (in usecs).
CpuRtPeriod int64 `json:"cpu_period"`
// CPU to use
CpusetCpus string `json:"cpuset_cpus"`
// MEM to use
CpusetMems string `json:"cpuset_mems"`
// Specifies per cgroup weight, range is from 10 to 1000.
BlkioWeight uint16 `json:"blkio_weight"`
// Specifies tasks' weight in the given cgroup while competing with the cgroup's child cgroups, range is from 10 to 1000, cfq scheduler only
BlkioLeafWeight uint16 `json:"blkio_leaf_weight"`
// Weight per cgroup per device, can override BlkioWeight.
BlkioWeightDevice []*WeightDevice `json:"blkio_weight_device"`
// IO read rate limit per cgroup per device, bytes per second.
BlkioThrottleReadBpsDevice []*ThrottleDevice `json:"blkio_throttle_read_bps_device"`
// IO write rate limit per cgroup per divice, bytes per second.
BlkioThrottleWriteBpsDevice []*ThrottleDevice `json:"blkio_throttle_write_bps_device"`
// IO read rate limit per cgroup per device, IO per second.
BlkioThrottleReadIOPSDevice []*ThrottleDevice `json:"blkio_throttle_read_iops_device"`
// IO write rate limit per cgroup per device, IO per second.
BlkioThrottleWriteIOPSDevice []*ThrottleDevice `json:"blkio_throttle_write_iops_device"`
// set the freeze value for the process
Freezer FreezerState `json:"freezer"`
// Hugetlb limit (in bytes)
HugetlbLimit []*HugepageLimit `json:"hugetlb_limit"`
// Whether to disable OOM Killer
OomKillDisable bool `json:"oom_kill_disable"`
// Tuning swappiness behaviour per cgroup
MemorySwappiness int64 `json:"memory_swappiness"`
// Set priority of network traffic for container
NetPrioIfpriomap []*IfPrioMap `json:"net_prio_ifpriomap"`
// Set class identifier for container's network packets
NetClsClassid string `json:"net_cls_classid"`
}

6
vendor/src/github.com/opencontainers/runc/libcontainer/configs/cgroup_unsupported.go vendored Normal file
View file

@ -0,0 +1,6 @@
// +build !windows,!linux,!freebsd
package configs
type Cgroup struct {
}

6
vendor/src/github.com/opencontainers/runc/libcontainer/configs/cgroup_windows.go vendored Normal file
View file

@ -0,0 +1,6 @@
package configs
// TODO Windows: This can ultimately be entirely factored out on Windows as
// cgroups are a Unix-specific construct.
type Cgroup struct {
}

252
vendor/src/github.com/opencontainers/runc/libcontainer/configs/config.go vendored Normal file
View file

@ -0,0 +1,252 @@
package configs
import (
"bytes"
"encoding/json"
"os/exec"
)
type Rlimit struct {
Type int `json:"type"`
Hard uint64 `json:"hard"`
Soft uint64 `json:"soft"`
}
// IDMap represents UID/GID Mappings for User Namespaces.
type IDMap struct {
ContainerID int `json:"container_id"`
HostID int `json:"host_id"`
Size int `json:"size"`
}
// Seccomp represents syscall restrictions
// By default, only the native architecture of the kernel is allowed to be used
// for syscalls. Additional architectures can be added by specifying them in
// Architectures.
type Seccomp struct {
DefaultAction Action `json:"default_action"`
Architectures []string `json:"architectures"`
Syscalls []*Syscall `json:"syscalls"`
}
// An action to be taken upon rule match in Seccomp
type Action int
const (
Kill Action = iota + 1
Errno
Trap
Allow
Trace
)
// A comparison operator to be used when matching syscall arguments in Seccomp
type Operator int
const (
EqualTo Operator = iota + 1
NotEqualTo
GreaterThan
GreaterThanOrEqualTo
LessThan
LessThanOrEqualTo
MaskEqualTo
)
// A rule to match a specific syscall argument in Seccomp
type Arg struct {
Index uint `json:"index"`
Value uint64 `json:"value"`
ValueTwo uint64 `json:"value_two"`
Op Operator `json:"op"`
}
// An rule to match a syscall in Seccomp
type Syscall struct {
Name string `json:"name"`
Action Action `json:"action"`
Args []*Arg `json:"args"`
}
// TODO Windows. Many of these fields should be factored out into those parts
// which are common across platforms, and those which are platform specific.
// Config defines configuration options for executing a process inside a contained environment.
type Config struct {
// NoPivotRoot will use MS_MOVE and a chroot to jail the process into the container's rootfs
// This is a common option when the container is running in ramdisk
NoPivotRoot bool `json:"no_pivot_root"`
// ParentDeathSignal specifies the signal that is sent to the container's process in the case
// that the parent process dies.
ParentDeathSignal int `json:"parent_death_signal"`
// PivotDir allows a custom directory inside the container's root filesystem to be used as pivot, when NoPivotRoot is not set.
// When a custom PivotDir not set, a temporary dir inside the root filesystem will be used. The pivot dir needs to be writeable.
// This is required when using read only root filesystems. In these cases, a read/writeable path can be (bind) mounted somewhere inside the root filesystem to act as pivot.
PivotDir string `json:"pivot_dir"`
// Path to a directory containing the container's root filesystem.
Rootfs string `json:"rootfs"`
// Readonlyfs will remount the container's rootfs as readonly where only externally mounted
// bind mounts are writtable.
Readonlyfs bool `json:"readonlyfs"`
// Specifies the mount propagation flags to be applied to /.
RootPropagation int `json:"rootPropagation"`
// Mounts specify additional source and destination paths that will be mounted inside the container's
// rootfs and mount namespace if specified
Mounts []*Mount `json:"mounts"`
// The device nodes that should be automatically created within the container upon container start. Note, make sure that the node is marked as allowed in the cgroup as well!
Devices []*Device `json:"devices"`
MountLabel string `json:"mount_label"`
// Hostname optionally sets the container's hostname if provided
Hostname string `json:"hostname"`
// Namespaces specifies the container's namespaces that it should setup when cloning the init process
// If a namespace is not provided that namespace is shared from the container's parent process
Namespaces Namespaces `json:"namespaces"`
// Capabilities specify the capabilities to keep when executing the process inside the container
// All capbilities not specified will be dropped from the processes capability mask
Capabilities []string `json:"capabilities"`
// Networks specifies the container's network setup to be created
Networks []*Network `json:"networks"`
// Routes can be specified to create entries in the route table as the container is started
Routes []*Route `json:"routes"`
// Cgroups specifies specific cgroup settings for the various subsystems that the container is
// placed into to limit the resources the container has available
Cgroups *Cgroup `json:"cgroups"`
// AppArmorProfile specifies the profile to apply to the process running in the container and is
// change at the time the process is execed
AppArmorProfile string `json:"apparmor_profile"`
// ProcessLabel specifies the label to apply to the process running in the container. It is
// commonly used by selinux
ProcessLabel string `json:"process_label"`
// Rlimits specifies the resource limits, such as max open files, to set in the container
// If Rlimits are not set, the container will inherit rlimits from the parent process
Rlimits []Rlimit `json:"rlimits"`
// OomScoreAdj specifies the adjustment to be made by the kernel when calculating oom scores
// for a process. Valid values are between the range [-1000, '1000'], where processes with
// higher scores are preferred for being killed.
// More information about kernel oom score calculation here: https://lwn.net/Articles/317814/
OomScoreAdj int `json:"oom_score_adj"`
// AdditionalGroups specifies the gids that should be added to supplementary groups
// in addition to those that the user belongs to.
AdditionalGroups []string `json:"additional_groups"`
// UidMappings is an array of User ID mappings for User Namespaces
UidMappings []IDMap `json:"uid_mappings"`
// GidMappings is an array of Group ID mappings for User Namespaces
GidMappings []IDMap `json:"gid_mappings"`
// MaskPaths specifies paths within the container's rootfs to mask over with a bind
// mount pointing to /dev/null as to prevent reads of the file.
MaskPaths []string `json:"mask_paths"`
// ReadonlyPaths specifies paths within the container's rootfs to remount as read-only
// so that these files prevent any writes.
ReadonlyPaths []string `json:"readonly_paths"`
// Sysctl is a map of properties and their values. It is the equivalent of using
// sysctl -w my.property.name value in Linux.
Sysctl map[string]string `json:"sysctl"`
// Seccomp allows actions to be taken whenever a syscall is made within the container.
// A number of rules are given, each having an action to be taken if a syscall matches it.
// A default action to be taken if no rules match is also given.
Seccomp *Seccomp `json:"seccomp"`
// Hooks are a collection of actions to perform at various container lifecycle events.
// Hooks are not able to be marshaled to json but they are also not needed to.
Hooks *Hooks `json:"-"`
// Version is the version of opencontainer specification that is supported.
Version string `json:"version"`
}
type Hooks struct {
// Prestart commands are executed after the container namespaces are created,
// but before the user supplied command is executed from init.
Prestart []Hook
// Poststart commands are executed after the container init process starts.
Poststart []Hook
// Poststop commands are executed after the container init process exits.
Poststop []Hook
}
// HookState is the payload provided to a hook on execution.
type HookState struct {
Version string `json:"version"`
ID string `json:"id"`
Pid int `json:"pid"`
Root string `json:"root"`
}
type Hook interface {
// Run executes the hook with the provided state.
Run(HookState) error
}
// NewFunctionHooks will call the provided function when the hook is run.
func NewFunctionHook(f func(HookState) error) FuncHook {
return FuncHook{
run: f,
}
}
type FuncHook struct {
run func(HookState) error
}
func (f FuncHook) Run(s HookState) error {
return f.run(s)
}
type Command struct {
Path string `json:"path"`
Args []string `json:"args"`
Env []string `json:"env"`
Dir string `json:"dir"`
}
// NewCommandHooks will execute the provided command when the hook is run.
func NewCommandHook(cmd Command) CommandHook {
return CommandHook{
Command: cmd,
}
}
type CommandHook struct {
Command
}
func (c Command) Run(s HookState) error {
b, err := json.Marshal(s)
if err != nil {
return err
}
cmd := exec.Cmd{
Path: c.Path,
Args: c.Args,
Env: c.Env,
Stdin: bytes.NewReader(b),
}
return cmd.Run()
}

51
vendor/src/github.com/opencontainers/runc/libcontainer/configs/config_unix.go vendored Normal file
View file

@ -0,0 +1,51 @@
// +build freebsd linux
package configs
import "fmt"
// Gets the root uid for the process on host which could be non-zero
// when user namespaces are enabled.
func (c Config) HostUID() (int, error) {
if c.Namespaces.Contains(NEWUSER) {
if c.UidMappings == nil {
return -1, fmt.Errorf("User namespaces enabled, but no user mappings found.")
}
id, found := c.hostIDFromMapping(0, c.UidMappings)
if !found {
return -1, fmt.Errorf("User namespaces enabled, but no root user mapping found.")
}
return id, nil
}
// Return default root uid 0
return 0, nil
}
// Gets the root gid for the process on host which could be non-zero
// when user namespaces are enabled.
func (c Config) HostGID() (int, error) {
if c.Namespaces.Contains(NEWUSER) {
if c.GidMappings == nil {
return -1, fmt.Errorf("User namespaces enabled, but no gid mappings found.")
}
id, found := c.hostIDFromMapping(0, c.GidMappings)
if !found {
return -1, fmt.Errorf("User namespaces enabled, but no root group mapping found.")
}
return id, nil
}
// Return default root gid 0
return 0, nil
}
// Utility function that gets a host ID for a container ID from user namespace map
// if that ID is present in the map.
func (c Config) hostIDFromMapping(containerID int, uMap []IDMap) (int, bool) {
for _, m := range uMap {
if (containerID >= m.ContainerID) && (containerID <= (m.ContainerID + m.Size - 1)) {
hostID := m.HostID + (containerID - m.ContainerID)
return hostID, true
}
}
return -1, false
}

54
vendor/src/github.com/opencontainers/runc/libcontainer/configs/device.go vendored Normal file
View file

@ -0,0 +1,54 @@
package configs
import (
"fmt"
"os"
)
const (
Wildcard = -1
)
// TODO Windows: This can be factored out in the future
type Device struct {
// Device type, block, char, etc.
Type rune `json:"type"`
// Path to the device.
Path string `json:"path"`
// Major is the device's major number.
Major int64 `json:"major"`
// Minor is the device's minor number.
Minor int64 `json:"minor"`
// Cgroup permissions format, rwm.
Permissions string `json:"permissions"`
// FileMode permission bits for the device.
FileMode os.FileMode `json:"file_mode"`
// Uid of the device.
Uid uint32 `json:"uid"`
// Gid of the device.
Gid uint32 `json:"gid"`
}
func (d *Device) CgroupString() string {
return fmt.Sprintf("%c %s:%s %s", d.Type, deviceNumberString(d.Major), deviceNumberString(d.Minor), d.Permissions)
}
func (d *Device) Mkdev() int {
return int((d.Major << 8) | (d.Minor & 0xff) | ((d.Minor & 0xfff00) << 12))
}
// deviceNumberString converts the device number to a string return result.
func deviceNumberString(number int64) string {
if number == Wildcard {
return "*"
}
return fmt.Sprint(number)
}

139
vendor/src/github.com/opencontainers/runc/libcontainer/configs/device_defaults.go vendored Normal file
View file

@ -0,0 +1,139 @@
// +build linux freebsd
package configs
var (
// These are devices that are to be both allowed and created.
DefaultSimpleDevices = []*Device{
// /dev/null and zero
{
Path: "/dev/null",
Type: 'c',
Major: 1,
Minor: 3,
Permissions: "rwm",
FileMode: 0666,
},
{
Path: "/dev/zero",
Type: 'c',
Major: 1,
Minor: 5,
Permissions: "rwm",
FileMode: 0666,
},
{
Path: "/dev/full",
Type: 'c',
Major: 1,
Minor: 7,
Permissions: "rwm",
FileMode: 0666,
},
// consoles and ttys
{
Path: "/dev/tty",
Type: 'c',
Major: 5,
Minor: 0,
Permissions: "rwm",
FileMode: 0666,
},
// /dev/urandom,/dev/random
{
Path: "/dev/urandom",
Type: 'c',
Major: 1,
Minor: 9,
Permissions: "rwm",
FileMode: 0666,
},
{
Path: "/dev/random",
Type: 'c',
Major: 1,
Minor: 8,
Permissions: "rwm",
FileMode: 0666,
},
}
DefaultAllowedDevices = append([]*Device{
// allow mknod for any device
{
Type: 'c',
Major: Wildcard,
Minor: Wildcard,
Permissions: "m",
},
{
Type: 'b',
Major: Wildcard,
Minor: Wildcard,
Permissions: "m",
},
{
Path: "/dev/console",
Type: 'c',
Major: 5,
Minor: 1,
Permissions: "rwm",
},
{
Path: "/dev/tty0",
Type: 'c',
Major: 4,
Minor: 0,
Permissions: "rwm",
},
{
Path: "/dev/tty1",
Type: 'c',
Major: 4,
Minor: 1,
Permissions: "rwm",
},
// /dev/pts/ - pts namespaces are "coming soon"
{
Path: "",
Type: 'c',
Major: 136,
Minor: Wildcard,
Permissions: "rwm",
},
{
Path: "",
Type: 'c',
Major: 5,
Minor: 2,
Permissions: "rwm",
},
// tuntap
{
Path: "",
Type: 'c',
Major: 10,
Minor: 200,
Permissions: "rwm",
},
}, DefaultSimpleDevices...)
DefaultAutoCreatedDevices = append([]*Device{
{
// /dev/fuse is created but not allowed.
// This is to allow java to work. Because java
// Insists on there being a /dev/fuse
// https://github.com/docker/docker/issues/514
// https://github.com/docker/docker/issues/2393
//
Path: "/dev/fuse",
Type: 'c',
Major: 10,
Minor: 229,
Permissions: "rwm",
},
}, DefaultSimpleDevices...)
)

9
vendor/src/github.com/opencontainers/runc/libcontainer/configs/hugepage_limit.go vendored Normal file
View file

@ -0,0 +1,9 @@
package configs
type HugepageLimit struct {
// which type of hugepage to limit.
Pagesize string `json:"page_size"`
// usage limit for hugepage.
Limit uint64 `json:"limit"`
}

14
vendor/src/github.com/opencontainers/runc/libcontainer/configs/interface_priority_map.go vendored Normal file
View file

@ -0,0 +1,14 @@
package configs
import (
"fmt"
)
type IfPrioMap struct {
Interface string `json:"interface"`
Priority int64 `json:"priority"`
}
func (i *IfPrioMap) CgroupString() string {
return fmt.Sprintf("%s %d", i.Interface, i.Priority)
}

30
vendor/src/github.com/opencontainers/runc/libcontainer/configs/mount.go vendored Normal file
View file

@ -0,0 +1,30 @@
package configs
type Mount struct {
// Source path for the mount.
Source string `json:"source"`
// Destination path for the mount inside the container.
Destination string `json:"destination"`
// Device the mount is for.
Device string `json:"device"`
// Mount flags.
Flags int `json:"flags"`
// Propagation Flags
PropagationFlags []int `json:"propagation_flags"`
// Mount data applied to the mount.
Data string `json:"data"`
// Relabel source if set, "z" indicates shared, "Z" indicates unshared.
Relabel string `json:"relabel"`
// Optional Command to be run before Source is mounted.
PremountCmds []Command `json:"premount_cmds"`
// Optional Command to be run after Source is mounted.
PostmountCmds []Command `json:"postmount_cmds"`
}

5
vendor/src/github.com/opencontainers/runc/libcontainer/configs/namespaces.go vendored Normal file
View file

@ -0,0 +1,5 @@
package configs
type NamespaceType string
type Namespaces []Namespace

31
vendor/src/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall.go vendored Normal file
View file

@ -0,0 +1,31 @@
// +build linux
package configs
import "syscall"
func (n *Namespace) Syscall() int {
return namespaceInfo[n.Type]
}
var namespaceInfo = map[NamespaceType]int{
NEWNET: syscall.CLONE_NEWNET,
NEWNS: syscall.CLONE_NEWNS,
NEWUSER: syscall.CLONE_NEWUSER,
NEWIPC: syscall.CLONE_NEWIPC,
NEWUTS: syscall.CLONE_NEWUTS,
NEWPID: syscall.CLONE_NEWPID,
}
// CloneFlags parses the container's Namespaces options to set the correct
// flags on clone, unshare. This functions returns flags only for new namespaces.
func (n *Namespaces) CloneFlags() uintptr {
var flag int
for _, v := range *n {
if v.Path != "" {
continue
}
flag |= namespaceInfo[v.Type]
}
return uintptr(flag)
}

15
vendor/src/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall_unsupported.go vendored Normal file
View file

@ -0,0 +1,15 @@
// +build !linux,!windows
package configs
func (n *Namespace) Syscall() int {
panic("No namespace syscall support")
return 0
}
// CloneFlags parses the container's Namespaces options to set the correct
// flags on clone, unshare. This functions returns flags only for new namespaces.
func (n *Namespaces) CloneFlags() uintptr {
panic("No namespace syscall support")
return uintptr(0)
}

89
vendor/src/github.com/opencontainers/runc/libcontainer/configs/namespaces_unix.go vendored Normal file
View file

@ -0,0 +1,89 @@
// +build linux freebsd
package configs
import "fmt"
const (
NEWNET NamespaceType = "NEWNET"
NEWPID NamespaceType = "NEWPID"
NEWNS NamespaceType = "NEWNS"
NEWUTS NamespaceType = "NEWUTS"
NEWIPC NamespaceType = "NEWIPC"
NEWUSER NamespaceType = "NEWUSER"
)
func NamespaceTypes() []NamespaceType {
return []NamespaceType{
NEWNET,
NEWPID,
NEWNS,
NEWUTS,
NEWIPC,
NEWUSER,
}
}
// Namespace defines configuration for each namespace. It specifies an
// alternate path that is able to be joined via setns.
type Namespace struct {
Type NamespaceType `json:"type"`
Path string `json:"path"`
}
func (n *Namespace) GetPath(pid int) string {
if n.Path != "" {
return n.Path
}
return fmt.Sprintf("/proc/%d/ns/%s", pid, n.file())
}
func (n *Namespace) file() string {
file := ""
switch n.Type {
case NEWNET:
file = "net"
case NEWNS:
file = "mnt"
case NEWPID:
file = "pid"
case NEWIPC:
file = "ipc"
case NEWUSER:
file = "user"
case NEWUTS:
file = "uts"
}
return file
}
func (n *Namespaces) Remove(t NamespaceType) bool {
i := n.index(t)
if i == -1 {
return false
}
*n = append((*n)[:i], (*n)[i+1:]...)
return true
}
func (n *Namespaces) Add(t NamespaceType, path string) {
i := n.index(t)
if i == -1 {
*n = append(*n, Namespace{Type: t, Path: path})
return
}
(*n)[i].Path = path
}
func (n *Namespaces) index(t NamespaceType) int {
for i, ns := range *n {
if ns.Type == t {
return i
}
}
return -1
}
func (n *Namespaces) Contains(t NamespaceType) bool {
return n.index(t) != -1
}

8
vendor/src/github.com/opencontainers/runc/libcontainer/configs/namespaces_unsupported.go vendored Normal file
View file

@ -0,0 +1,8 @@
// +build !linux,!freebsd
package configs
// Namespace defines configuration for each namespace. It specifies an
// alternate path that is able to be joined via setns.
type Namespace struct {
}

72
vendor/src/github.com/opencontainers/runc/libcontainer/configs/network.go vendored Normal file
View file

@ -0,0 +1,72 @@
package configs
// Network defines configuration for a container's networking stack
//
// The network configuration can be omitted from a container causing the
// container to be setup with the host's networking stack
type Network struct {
// Type sets the networks type, commonly veth and loopback
Type string `json:"type"`
// Name of the network interface
Name string `json:"name"`
// The bridge to use.
Bridge string `json:"bridge"`
// MacAddress contains the MAC address to set on the network interface
MacAddress string `json:"mac_address"`
// Address contains the IPv4 and mask to set on the network interface
Address string `json:"address"`
// Gateway sets the gateway address that is used as the default for the interface
Gateway string `json:"gateway"`
// IPv6Address contains the IPv6 and mask to set on the network interface
IPv6Address string `json:"ipv6_address"`
// IPv6Gateway sets the ipv6 gateway address that is used as the default for the interface
IPv6Gateway string `json:"ipv6_gateway"`
// Mtu sets the mtu value for the interface and will be mirrored on both the host and
// container's interfaces if a pair is created, specifically in the case of type veth
// Note: This does not apply to loopback interfaces.
Mtu int `json:"mtu"`
// TxQueueLen sets the tx_queuelen value for the interface and will be mirrored on both the host and
// container's interfaces if a pair is created, specifically in the case of type veth
// Note: This does not apply to loopback interfaces.
TxQueueLen int `json:"txqueuelen"`
// HostInterfaceName is a unique name of a veth pair that resides on in the host interface of the
// container.
HostInterfaceName string `json:"host_interface_name"`
// HairpinMode specifies if hairpin NAT should be enabled on the virtual interface
// bridge port in the case of type veth
// Note: This is unsupported on some systems.
// Note: This does not apply to loopback interfaces.
HairpinMode bool `json:"hairpin_mode"`
}
// Routes can be specified to create entries in the route table as the container is started
//
// All of destination, source, and gateway should be either IPv4 or IPv6.
// One of the three options must be present, and omitted entries will use their
// IP family default for the route table. For IPv4 for example, setting the
// gateway to 1.2.3.4 and the interface to eth0 will set up a standard
// destination of 0.0.0.0(or *) when viewed in the route table.
type Route struct {
// Sets the destination and mask, should be a CIDR. Accepts IPv4 and IPv6
Destination string `json:"destination"`
// Sets the source and mask, should be a CIDR. Accepts IPv4 and IPv6
Source string `json:"source"`
// Sets the gateway. Accepts IPv4 and IPv6
Gateway string `json:"gateway"`
// The device to set this route up for, for example: eth0
InterfaceName string `json:"interface_name"`
}

93
vendor/src/github.com/opencontainers/runc/libcontainer/configs/validate/config.go vendored Normal file
View file

@ -0,0 +1,93 @@
package validate
import (
"fmt"
"os"
"path/filepath"
"github.com/opencontainers/runc/libcontainer/configs"
)
type Validator interface {
Validate(*configs.Config) error
}
func New() Validator {
return &ConfigValidator{}
}
type ConfigValidator struct {
}
func (v *ConfigValidator) Validate(config *configs.Config) error {
if err := v.rootfs(config); err != nil {
return err
}
if err := v.network(config); err != nil {
return err
}
if err := v.hostname(config); err != nil {
return err
}
if err := v.security(config); err != nil {
return err
}
if err := v.usernamespace(config); err != nil {
return err
}
return nil
}
// rootfs validates the the rootfs is an absolute path and is not a symlink
// to the container's root filesystem.
func (v *ConfigValidator) rootfs(config *configs.Config) error {
cleaned, err := filepath.Abs(config.Rootfs)
if err != nil {
return err
}
if cleaned, err = filepath.EvalSymlinks(cleaned); err != nil {
return err
}
if config.Rootfs != cleaned {
return fmt.Errorf("%s is not an absolute path or is a symlink", config.Rootfs)
}
return nil
}
func (v *ConfigValidator) network(config *configs.Config) error {
if !config.Namespaces.Contains(configs.NEWNET) {
if len(config.Networks) > 0 || len(config.Routes) > 0 {
return fmt.Errorf("unable to apply network settings without a private NET namespace")
}
}
return nil
}
func (v *ConfigValidator) hostname(config *configs.Config) error {
if config.Hostname != "" && !config.Namespaces.Contains(configs.NEWUTS) {
return fmt.Errorf("unable to set hostname without a private UTS namespace")
}
return nil
}
func (v *ConfigValidator) security(config *configs.Config) error {
// restrict sys without mount namespace
if (len(config.MaskPaths) > 0 || len(config.ReadonlyPaths) > 0) &&
!config.Namespaces.Contains(configs.NEWNS) {
return fmt.Errorf("unable to restrict sys entries without a private MNT namespace")
}
return nil
}
func (v *ConfigValidator) usernamespace(config *configs.Config) error {
if config.Namespaces.Contains(configs.NEWUSER) {
if _, err := os.Stat("/proc/self/ns/user"); os.IsNotExist(err) {
return fmt.Errorf("USER namespaces aren't enabled in the kernel")
}
} else {
if config.UidMappings != nil || config.GidMappings != nil {
return fmt.Errorf("User namespace mappings specified, but USER namespace isn't enabled in the config")
}
}
return nil
}

15
vendor/src/github.com/opencontainers/runc/libcontainer/console.go vendored Normal file
View file

@ -0,0 +1,15 @@
package libcontainer
import "io"
// Console represents a pseudo TTY.
type Console interface {
io.ReadWriter
io.Closer
// Path returns the filesystem path to the slave side of the pty.
Path() string
// Fd returns the fd for the master of the pty.
Fd() uintptr
}

13
vendor/src/github.com/opencontainers/runc/libcontainer/console_freebsd.go vendored Normal file
View file

@ -0,0 +1,13 @@
// +build freebsd
package libcontainer
import (
"errors"
)
// NewConsole returns an initalized console that can be used within a container by copying bytes
// from the master side to the slave that is attached as the tty for the container's init process.
func NewConsole(uid, gid int) (Console, error) {
return nil, errors.New("libcontainer console is not supported on FreeBSD")
}

145
vendor/src/github.com/opencontainers/runc/libcontainer/console_linux.go vendored Normal file
View file

@ -0,0 +1,145 @@
package libcontainer
import (
"fmt"
"os"
"path/filepath"
"syscall"
"unsafe"
"github.com/opencontainers/runc/libcontainer/label"
)
// NewConsole returns an initalized console that can be used within a container by copying bytes
// from the master side to the slave that is attached as the tty for the container's init process.
func NewConsole(uid, gid int) (Console, error) {
master, err := os.OpenFile("/dev/ptmx", syscall.O_RDWR|syscall.O_NOCTTY|syscall.O_CLOEXEC, 0)
if err != nil {
return nil, err
}
console, err := ptsname(master)
if err != nil {
return nil, err
}
if err := unlockpt(master); err != nil {
return nil, err
}
if err := os.Chmod(console, 0600); err != nil {
return nil, err
}
if err := os.Chown(console, uid, gid); err != nil {
return nil, err
}
return &linuxConsole{
slavePath: console,
master: master,
}, nil
}
// newConsoleFromPath is an internal function returning an initialized console for use inside
// a container's MNT namespace.
func newConsoleFromPath(slavePath string) *linuxConsole {
return &linuxConsole{
slavePath: slavePath,
}
}
// linuxConsole is a linux psuedo TTY for use within a container.
type linuxConsole struct {
master *os.File
slavePath string
}
func (c *linuxConsole) Fd() uintptr {
return c.master.Fd()
}
func (c *linuxConsole) Path() string {
return c.slavePath
}
func (c *linuxConsole) Read(b []byte) (int, error) {
return c.master.Read(b)
}
func (c *linuxConsole) Write(b []byte) (int, error) {
return c.master.Write(b)
}
func (c *linuxConsole) Close() error {
if m := c.master; m != nil {
return m.Close()
}
return nil
}
// mount initializes the console inside the rootfs mounting with the specified mount label
// and applying the correct ownership of the console.
func (c *linuxConsole) mount(rootfs, mountLabel string) error {
oldMask := syscall.Umask(0000)
defer syscall.Umask(oldMask)
if err := label.SetFileLabel(c.slavePath, mountLabel); err != nil {
return err
}
dest := filepath.Join(rootfs, "/dev/console")
f, err := os.Create(dest)
if err != nil && !os.IsExist(err) {
return err
}
if f != nil {
f.Close()
}
return syscall.Mount(c.slavePath, dest, "bind", syscall.MS_BIND, "")
}
// dupStdio opens the slavePath for the console and dups the fds to the current
// processes stdio, fd 0,1,2.
func (c *linuxConsole) dupStdio() error {
slave, err := c.open(syscall.O_RDWR)
if err != nil {
return err
}
fd := int(slave.Fd())
for _, i := range []int{0, 1, 2} {
if err := syscall.Dup3(fd, i, 0); err != nil {
return err
}
}
return nil
}
// open is a clone of os.OpenFile without the O_CLOEXEC used to open the pty slave.
func (c *linuxConsole) open(flag int) (*os.File, error) {
r, e := syscall.Open(c.slavePath, flag, 0)
if e != nil {
return nil, &os.PathError{
Op: "open",
Path: c.slavePath,
Err: e,
}
}
return os.NewFile(uintptr(r), c.slavePath), nil
}
func ioctl(fd uintptr, flag, data uintptr) error {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, fd, flag, data); err != 0 {
return err
}
return nil
}
// unlockpt unlocks the slave pseudoterminal device corresponding to the master pseudoterminal referred to by f.
// unlockpt should be called before opening the slave side of a pty.
func unlockpt(f *os.File) error {
var u int32
return ioctl(f.Fd(), syscall.TIOCSPTLCK, uintptr(unsafe.Pointer(&u)))
}
// ptsname retrieves the name of the first available pts for the given master.
func ptsname(f *os.File) (string, error) {
var n int32
if err := ioctl(f.Fd(), syscall.TIOCGPTN, uintptr(unsafe.Pointer(&n))); err != nil {
return "", err
}
return fmt.Sprintf("/dev/pts/%d", n), nil
}

30
vendor/src/github.com/opencontainers/runc/libcontainer/console_windows.go vendored Normal file
View file

@ -0,0 +1,30 @@
package libcontainer
// NewConsole returns an initalized console that can be used within a container
func NewConsole(uid, gid int) (Console, error) {
return &windowsConsole{}, nil
}
// windowsConsole is a Windows psuedo TTY for use within a container.
type windowsConsole struct {
}
func (c *windowsConsole) Fd() uintptr {
return 0
}
func (c *windowsConsole) Path() string {
return ""
}
func (c *windowsConsole) Read(b []byte) (int, error) {
return 0, nil
}
func (c *windowsConsole) Write(b []byte) (int, error) {
return 0, nil
}
func (c *windowsConsole) Close() error {
return nil
}

140
vendor/src/github.com/opencontainers/runc/libcontainer/container.go vendored Normal file
View file

@ -0,0 +1,140 @@
// Libcontainer provides a native Go implementation for creating containers
// with namespaces, cgroups, capabilities, and filesystem access controls.
// It allows you to manage the lifecycle of the container performing additional operations
// after the container is created.
package libcontainer
import (
"os"
"github.com/opencontainers/runc/libcontainer/configs"
)
// The status of a container.
type Status int
const (
// The container exists and is running.
Running Status = iota + 1
// The container exists, it is in the process of being paused.
Pausing
// The container exists, but all its processes are paused.
Paused
// The container exists, but its state is saved on disk
Checkpointed
// The container does not exist.
Destroyed
)
func (s Status) String() string {
switch s {
case Running:
return "running"
case Pausing:
return "pausing"
case Paused:
return "paused"
case Checkpointed:
return "checkpointed"
case Destroyed:
return "destroyed"
default:
return "undefined"
}
}
// BaseState represents the platform agnostic pieces relating to a
// running container's state
type BaseState struct {
// ID is the container ID.
ID string `json:"id"`
// InitProcessPid is the init process id in the parent namespace.
InitProcessPid int `json:"init_process_pid"`
// InitProcessStartTime is the init process start time.
InitProcessStartTime string `json:"init_process_start"`
// Config is the container's configuration.
Config configs.Config `json:"config"`
}
// A libcontainer container object.
//
// Each container is thread-safe within the same process. Since a container can
// be destroyed by a separate process, any function may return that the container
// was not found. BaseContainer includes methods that are platform agnostic.
type BaseContainer interface {
// Returns the ID of the container
ID() string
// Returns the current status of the container.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// Systemerror - System error.
Status() (Status, error)
// State returns the current container's state information.
//
// errors:
// Systemerror - System error.
State() (*State, error)
// Returns the current config of the container.
Config() configs.Config
// Returns the PIDs inside this container. The PIDs are in the namespace of the calling process.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// Systemerror - System error.
//
// Some of the returned PIDs may no longer refer to processes in the Container, unless
// the Container state is PAUSED in which case every PID in the slice is valid.
Processes() ([]int, error)
// Returns statistics for the container.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// Systemerror - System error.
Stats() (*Stats, error)
// Set resources of container as configured
//
// We can use this to change resources when containers are running.
//
// errors:
// Systemerror - System error.
Set(config configs.Config) error
// Start a process inside the container. Returns error if process fails to
// start. You can track process lifecycle with passed Process structure.
//
// errors:
// ContainerDestroyed - Container no longer exists,
// ConfigInvalid - config is invalid,
// ContainerPaused - Container is paused,
// Systemerror - System error.
Start(process *Process) (err error)
// Destroys the container after killing all running processes.
//
// Any event registrations are removed before the container is destroyed.
// No error is returned if the container is already destroyed.
//
// errors:
// Systemerror - System error.
Destroy() error
// Signal sends the provided signal code to the container's initial process.
//
// errors:
// Systemerror - System error.
Signal(s os.Signal) error
}

1064
vendor/src/github.com/opencontainers/runc/libcontainer/container_linux.go vendored Normal file
View file

File diff suppressed because it is too large Load diff

13
vendor/src/github.com/opencontainers/runc/libcontainer/container_nouserns_linux.go vendored Normal file
View file

@ -0,0 +1,13 @@
// +build !go1.4
package libcontainer
import (
"fmt"
"syscall"
)
// not available before go 1.4
func (c *linuxContainer) addUidGidMappings(sys *syscall.SysProcAttr) error {
return fmt.Errorf("User namespace is not supported in golang < 1.4")
}

26
vendor/src/github.com/opencontainers/runc/libcontainer/container_userns_linux.go vendored Normal file
View file

@ -0,0 +1,26 @@
// +build go1.4
package libcontainer
import "syscall"
// Converts IDMap to SysProcIDMap array and adds it to SysProcAttr.
func (c *linuxContainer) addUidGidMappings(sys *syscall.SysProcAttr) error {
if c.config.UidMappings != nil {
sys.UidMappings = make([]syscall.SysProcIDMap, len(c.config.UidMappings))
for i, um := range c.config.UidMappings {
sys.UidMappings[i].ContainerID = um.ContainerID
sys.UidMappings[i].HostID = um.HostID
sys.UidMappings[i].Size = um.Size
}
}
if c.config.GidMappings != nil {
sys.GidMappings = make([]syscall.SysProcIDMap, len(c.config.GidMappings))
for i, gm := range c.config.GidMappings {
sys.GidMappings[i].ContainerID = gm.ContainerID
sys.GidMappings[i].HostID = gm.HostID
sys.GidMappings[i].Size = gm.Size
}
}
return nil
}

20
vendor/src/github.com/opencontainers/runc/libcontainer/container_windows.go vendored Normal file
View file

@ -0,0 +1,20 @@
package libcontainer
// State represents a running container's state
type State struct {
BaseState
// Platform specific fields below here
}
// A libcontainer container object.
//
// Each container is thread-safe within the same process. Since a container can
// be destroyed by a separate process, any function may return that the container
// was not found.
type Container interface {
BaseContainer
// Methods below here are platform specific
}

36
vendor/src/github.com/opencontainers/runc/libcontainer/criu_opts_unix.go vendored Normal file
View file

@ -0,0 +1,36 @@
// +build linux freebsd
package libcontainer
// cgroup restoring strategy provided by criu
type cg_mode uint32
const (
CRIU_CG_MODE_SOFT cg_mode = 3 + iota // restore cgroup properties if only dir created by criu
CRIU_CG_MODE_FULL // always restore all cgroups and their properties
CRIU_CG_MODE_STRICT // restore all, requiring them to not present in the system
CRIU_CG_MODE_DEFAULT // the same as CRIU_CG_MODE_SOFT
)
type CriuPageServerInfo struct {
Address string // IP address of CRIU page server
Port int32 // port number of CRIU page server
}
type VethPairName struct {
ContainerInterfaceName string
HostInterfaceName string
}
type CriuOpts struct {
ImagesDirectory string // directory for storing image files
WorkDirectory string // directory to cd and write logs/pidfiles/stats to
LeaveRunning bool // leave container in running state after checkpoint
TcpEstablished bool // checkpoint/restore established TCP connections
ExternalUnixConnections bool // allow external unix connections
ShellJob bool // allow to dump and restore shell jobs
FileLocks bool // handle file locks, for safety
PageServer CriuPageServerInfo // allow to dump to criu page server
VethPairs []VethPairName // pass the veth to criu when restore
ManageCgroupsMode cg_mode // dump or restore cgroup mode
}

6
vendor/src/github.com/opencontainers/runc/libcontainer/criu_opts_windows.go vendored Normal file
View file

@ -0,0 +1,6 @@
package libcontainer
// TODO Windows: This can ultimately be entirely factored out as criu is
// a Unix concept not relevant on Windows.
type CriuOpts struct {
}

2
vendor/src/github.com/opencontainers/runc/libcontainer/criurpc/Makefile vendored Normal file
View file

@ -0,0 +1,2 @@
gen: criurpc.proto
protoc --go_out=. criurpc.proto

689
vendor/src/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.pb.go vendored Normal file
View file

@ -0,0 +1,689 @@
// Code generated by protoc-gen-go.
// source: criurpc.proto
// DO NOT EDIT!
/*
Package criurpc is a generated protocol buffer package.
It is generated from these files:
criurpc.proto
It has these top-level messages:
CriuPageServerInfo
CriuVethPair
ExtMountMap
InheritFd
CgroupRoot
UnixSk
CriuOpts
CriuDumpResp
CriuRestoreResp
CriuNotify
CriuReq
CriuResp
*/
package criurpc
import proto "github.com/golang/protobuf/proto"
import math "math"
// Reference imports to suppress errors if they are not otherwise used.
var _ = proto.Marshal
var _ = math.Inf
type CriuReqType int32
const (
CriuReqType_EMPTY CriuReqType = 0
CriuReqType_DUMP CriuReqType = 1
CriuReqType_RESTORE CriuReqType = 2
CriuReqType_CHECK CriuReqType = 3
CriuReqType_PRE_DUMP CriuReqType = 4
CriuReqType_PAGE_SERVER CriuReqType = 5
CriuReqType_NOTIFY CriuReqType = 6
CriuReqType_CPUINFO_DUMP CriuReqType = 7
CriuReqType_CPUINFO_CHECK CriuReqType = 8
)
var CriuReqType_name = map[int32]string{
0: "EMPTY",
1: "DUMP",
2: "RESTORE",
3: "CHECK",
4: "PRE_DUMP",
5: "PAGE_SERVER",
6: "NOTIFY",
7: "CPUINFO_DUMP",
8: "CPUINFO_CHECK",
}
var CriuReqType_value = map[string]int32{
"EMPTY": 0,
"DUMP": 1,
"RESTORE": 2,
"CHECK": 3,
"PRE_DUMP": 4,
"PAGE_SERVER": 5,
"NOTIFY": 6,
"CPUINFO_DUMP": 7,
"CPUINFO_CHECK": 8,
}
func (x CriuReqType) Enum() *CriuReqType {
p := new(CriuReqType)
*p = x
return p
}
func (x CriuReqType) String() string {
return proto.EnumName(CriuReqType_name, int32(x))
}
func (x *CriuReqType) UnmarshalJSON(data []byte) error {
value, err := proto.UnmarshalJSONEnum(CriuReqType_value, data, "CriuReqType")
if err != nil {
return err
}
*x = CriuReqType(value)
return nil
}
type CriuPageServerInfo struct {
Address *string `protobuf:"bytes,1,opt,name=address" json:"address,omitempty"`
Port *int32 `protobuf:"varint,2,opt,name=port" json:"port,omitempty"`
Pid *int32 `protobuf:"varint,3,opt,name=pid" json:"pid,omitempty"`
Fd *int32 `protobuf:"varint,4,opt,name=fd" json:"fd,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuPageServerInfo) Reset() { *m = CriuPageServerInfo{} }
func (m *CriuPageServerInfo) String() string { return proto.CompactTextString(m) }
func (*CriuPageServerInfo) ProtoMessage() {}
func (m *CriuPageServerInfo) GetAddress() string {
if m != nil && m.Address != nil {
return *m.Address
}
return ""
}
func (m *CriuPageServerInfo) GetPort() int32 {
if m != nil && m.Port != nil {
return *m.Port
}
return 0
}
func (m *CriuPageServerInfo) GetPid() int32 {
if m != nil && m.Pid != nil {
return *m.Pid
}
return 0
}
func (m *CriuPageServerInfo) GetFd() int32 {
if m != nil && m.Fd != nil {
return *m.Fd
}
return 0
}
type CriuVethPair struct {
IfIn *string `protobuf:"bytes,1,req,name=if_in" json:"if_in,omitempty"`
IfOut *string `protobuf:"bytes,2,req,name=if_out" json:"if_out,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuVethPair) Reset() { *m = CriuVethPair{} }
func (m *CriuVethPair) String() string { return proto.CompactTextString(m) }
func (*CriuVethPair) ProtoMessage() {}
func (m *CriuVethPair) GetIfIn() string {
if m != nil && m.IfIn != nil {
return *m.IfIn
}
return ""
}
func (m *CriuVethPair) GetIfOut() string {
if m != nil && m.IfOut != nil {
return *m.IfOut
}
return ""
}
type ExtMountMap struct {
Key *string `protobuf:"bytes,1,req,name=key" json:"key,omitempty"`
Val *string `protobuf:"bytes,2,req,name=val" json:"val,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *ExtMountMap) Reset() { *m = ExtMountMap{} }
func (m *ExtMountMap) String() string { return proto.CompactTextString(m) }
func (*ExtMountMap) ProtoMessage() {}
func (m *ExtMountMap) GetKey() string {
if m != nil && m.Key != nil {
return *m.Key
}
return ""
}
func (m *ExtMountMap) GetVal() string {
if m != nil && m.Val != nil {
return *m.Val
}
return ""
}
type InheritFd struct {
Key *string `protobuf:"bytes,1,req,name=key" json:"key,omitempty"`
Fd *int32 `protobuf:"varint,2,req,name=fd" json:"fd,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *InheritFd) Reset() { *m = InheritFd{} }
func (m *InheritFd) String() string { return proto.CompactTextString(m) }
func (*InheritFd) ProtoMessage() {}
func (m *InheritFd) GetKey() string {
if m != nil && m.Key != nil {
return *m.Key
}
return ""
}
func (m *InheritFd) GetFd() int32 {
if m != nil && m.Fd != nil {
return *m.Fd
}
return 0
}
type CgroupRoot struct {
Ctrl *string `protobuf:"bytes,1,opt,name=ctrl" json:"ctrl,omitempty"`
Path *string `protobuf:"bytes,2,req,name=path" json:"path,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CgroupRoot) Reset() { *m = CgroupRoot{} }
func (m *CgroupRoot) String() string { return proto.CompactTextString(m) }
func (*CgroupRoot) ProtoMessage() {}
func (m *CgroupRoot) GetCtrl() string {
if m != nil && m.Ctrl != nil {
return *m.Ctrl
}
return ""
}
func (m *CgroupRoot) GetPath() string {
if m != nil && m.Path != nil {
return *m.Path
}
return ""
}
type UnixSk struct {
Inode *uint32 `protobuf:"varint,1,req,name=inode" json:"inode,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *UnixSk) Reset() { *m = UnixSk{} }
func (m *UnixSk) String() string { return proto.CompactTextString(m) }
func (*UnixSk) ProtoMessage() {}
func (m *UnixSk) GetInode() uint32 {
if m != nil && m.Inode != nil {
return *m.Inode
}
return 0
}
type CriuOpts struct {
ImagesDirFd *int32 `protobuf:"varint,1,req,name=images_dir_fd" json:"images_dir_fd,omitempty"`
Pid *int32 `protobuf:"varint,2,opt,name=pid" json:"pid,omitempty"`
LeaveRunning *bool `protobuf:"varint,3,opt,name=leave_running" json:"leave_running,omitempty"`
ExtUnixSk *bool `protobuf:"varint,4,opt,name=ext_unix_sk" json:"ext_unix_sk,omitempty"`
TcpEstablished *bool `protobuf:"varint,5,opt,name=tcp_established" json:"tcp_established,omitempty"`
EvasiveDevices *bool `protobuf:"varint,6,opt,name=evasive_devices" json:"evasive_devices,omitempty"`
ShellJob *bool `protobuf:"varint,7,opt,name=shell_job" json:"shell_job,omitempty"`
FileLocks *bool `protobuf:"varint,8,opt,name=file_locks" json:"file_locks,omitempty"`
LogLevel *int32 `protobuf:"varint,9,opt,name=log_level,def=2" json:"log_level,omitempty"`
LogFile *string `protobuf:"bytes,10,opt,name=log_file" json:"log_file,omitempty"`
Ps *CriuPageServerInfo `protobuf:"bytes,11,opt,name=ps" json:"ps,omitempty"`
NotifyScripts *bool `protobuf:"varint,12,opt,name=notify_scripts" json:"notify_scripts,omitempty"`
Root *string `protobuf:"bytes,13,opt,name=root" json:"root,omitempty"`
ParentImg *string `protobuf:"bytes,14,opt,name=parent_img" json:"parent_img,omitempty"`
TrackMem *bool `protobuf:"varint,15,opt,name=track_mem" json:"track_mem,omitempty"`
AutoDedup *bool `protobuf:"varint,16,opt,name=auto_dedup" json:"auto_dedup,omitempty"`
WorkDirFd *int32 `protobuf:"varint,17,opt,name=work_dir_fd" json:"work_dir_fd,omitempty"`
LinkRemap *bool `protobuf:"varint,18,opt,name=link_remap" json:"link_remap,omitempty"`
Veths []*CriuVethPair `protobuf:"bytes,19,rep,name=veths" json:"veths,omitempty"`
CpuCap *uint32 `protobuf:"varint,20,opt,name=cpu_cap,def=4294967295" json:"cpu_cap,omitempty"`
ForceIrmap *bool `protobuf:"varint,21,opt,name=force_irmap" json:"force_irmap,omitempty"`
ExecCmd []string `protobuf:"bytes,22,rep,name=exec_cmd" json:"exec_cmd,omitempty"`
ExtMnt []*ExtMountMap `protobuf:"bytes,23,rep,name=ext_mnt" json:"ext_mnt,omitempty"`
ManageCgroups *bool `protobuf:"varint,24,opt,name=manage_cgroups" json:"manage_cgroups,omitempty"`
CgRoot []*CgroupRoot `protobuf:"bytes,25,rep,name=cg_root" json:"cg_root,omitempty"`
RstSibling *bool `protobuf:"varint,26,opt,name=rst_sibling" json:"rst_sibling,omitempty"`
InheritFd []*InheritFd `protobuf:"bytes,27,rep,name=inherit_fd" json:"inherit_fd,omitempty"`
AutoExtMnt *bool `protobuf:"varint,28,opt,name=auto_ext_mnt" json:"auto_ext_mnt,omitempty"`
ExtSharing *bool `protobuf:"varint,29,opt,name=ext_sharing" json:"ext_sharing,omitempty"`
ExtMasters *bool `protobuf:"varint,30,opt,name=ext_masters" json:"ext_masters,omitempty"`
SkipMnt []string `protobuf:"bytes,31,rep,name=skip_mnt" json:"skip_mnt,omitempty"`
EnableFs []string `protobuf:"bytes,32,rep,name=enable_fs" json:"enable_fs,omitempty"`
UnixSkIno []*UnixSk `protobuf:"bytes,33,rep,name=unix_sk_ino" json:"unix_sk_ino,omitempty"`
ManageCgroupsMode *uint32 `protobuf:"varint,34,opt,name=manage_cgroups_mode" json:"manage_cgroups_mode,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuOpts) Reset() { *m = CriuOpts{} }
func (m *CriuOpts) String() string { return proto.CompactTextString(m) }
func (*CriuOpts) ProtoMessage() {}
const Default_CriuOpts_LogLevel int32 = 2
const Default_CriuOpts_CpuCap uint32 = 4294967295
func (m *CriuOpts) GetImagesDirFd() int32 {
if m != nil && m.ImagesDirFd != nil {
return *m.ImagesDirFd
}
return 0
}
func (m *CriuOpts) GetPid() int32 {
if m != nil && m.Pid != nil {
return *m.Pid
}
return 0
}
func (m *CriuOpts) GetLeaveRunning() bool {
if m != nil && m.LeaveRunning != nil {
return *m.LeaveRunning
}
return false
}
func (m *CriuOpts) GetExtUnixSk() bool {
if m != nil && m.ExtUnixSk != nil {
return *m.ExtUnixSk
}
return false
}
func (m *CriuOpts) GetTcpEstablished() bool {
if m != nil && m.TcpEstablished != nil {
return *m.TcpEstablished
}
return false
}
func (m *CriuOpts) GetEvasiveDevices() bool {
if m != nil && m.EvasiveDevices != nil {
return *m.EvasiveDevices
}
return false
}
func (m *CriuOpts) GetShellJob() bool {
if m != nil && m.ShellJob != nil {
return *m.ShellJob
}
return false
}
func (m *CriuOpts) GetFileLocks() bool {
if m != nil && m.FileLocks != nil {
return *m.FileLocks
}
return false
}
func (m *CriuOpts) GetLogLevel() int32 {
if m != nil && m.LogLevel != nil {
return *m.LogLevel
}
return Default_CriuOpts_LogLevel
}
func (m *CriuOpts) GetLogFile() string {
if m != nil && m.LogFile != nil {
return *m.LogFile
}
return ""
}
func (m *CriuOpts) GetPs() *CriuPageServerInfo {
if m != nil {
return m.Ps
}
return nil
}
func (m *CriuOpts) GetNotifyScripts() bool {
if m != nil && m.NotifyScripts != nil {
return *m.NotifyScripts
}
return false
}
func (m *CriuOpts) GetRoot() string {
if m != nil && m.Root != nil {
return *m.Root
}
return ""
}
func (m *CriuOpts) GetParentImg() string {
if m != nil && m.ParentImg != nil {
return *m.ParentImg
}
return ""
}
func (m *CriuOpts) GetTrackMem() bool {
if m != nil && m.TrackMem != nil {
return *m.TrackMem
}
return false
}
func (m *CriuOpts) GetAutoDedup() bool {
if m != nil && m.AutoDedup != nil {
return *m.AutoDedup
}
return false
}
func (m *CriuOpts) GetWorkDirFd() int32 {
if m != nil && m.WorkDirFd != nil {
return *m.WorkDirFd
}
return 0
}
func (m *CriuOpts) GetLinkRemap() bool {
if m != nil && m.LinkRemap != nil {
return *m.LinkRemap
}
return false
}
func (m *CriuOpts) GetVeths() []*CriuVethPair {
if m != nil {
return m.Veths
}
return nil
}
func (m *CriuOpts) GetCpuCap() uint32 {
if m != nil && m.CpuCap != nil {
return *m.CpuCap
}
return Default_CriuOpts_CpuCap
}
func (m *CriuOpts) GetForceIrmap() bool {
if m != nil && m.ForceIrmap != nil {
return *m.ForceIrmap
}
return false
}
func (m *CriuOpts) GetExecCmd() []string {
if m != nil {
return m.ExecCmd
}
return nil
}
func (m *CriuOpts) GetExtMnt() []*ExtMountMap {
if m != nil {
return m.ExtMnt
}
return nil
}
func (m *CriuOpts) GetManageCgroups() bool {
if m != nil && m.ManageCgroups != nil {
return *m.ManageCgroups
}
return false
}
func (m *CriuOpts) GetCgRoot() []*CgroupRoot {
if m != nil {
return m.CgRoot
}
return nil
}
func (m *CriuOpts) GetRstSibling() bool {
if m != nil && m.RstSibling != nil {
return *m.RstSibling
}
return false
}
func (m *CriuOpts) GetInheritFd() []*InheritFd {
if m != nil {
return m.InheritFd
}
return nil
}
func (m *CriuOpts) GetAutoExtMnt() bool {
if m != nil && m.AutoExtMnt != nil {
return *m.AutoExtMnt
}
return false
}
func (m *CriuOpts) GetExtSharing() bool {
if m != nil && m.ExtSharing != nil {
return *m.ExtSharing
}
return false
}
func (m *CriuOpts) GetExtMasters() bool {
if m != nil && m.ExtMasters != nil {
return *m.ExtMasters
}
return false
}
func (m *CriuOpts) GetSkipMnt() []string {
if m != nil {
return m.SkipMnt
}
return nil
}
func (m *CriuOpts) GetEnableFs() []string {
if m != nil {
return m.EnableFs
}
return nil
}
func (m *CriuOpts) GetUnixSkIno() []*UnixSk {
if m != nil {
return m.UnixSkIno
}
return nil
}
func (m *CriuOpts) GetManageCgroupsMode() uint32 {
if m != nil && m.ManageCgroupsMode != nil {
return *m.ManageCgroupsMode
}
return 0
}
type CriuDumpResp struct {
Restored *bool `protobuf:"varint,1,opt,name=restored" json:"restored,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuDumpResp) Reset() { *m = CriuDumpResp{} }
func (m *CriuDumpResp) String() string { return proto.CompactTextString(m) }
func (*CriuDumpResp) ProtoMessage() {}
func (m *CriuDumpResp) GetRestored() bool {
if m != nil && m.Restored != nil {
return *m.Restored
}
return false
}
type CriuRestoreResp struct {
Pid *int32 `protobuf:"varint,1,req,name=pid" json:"pid,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuRestoreResp) Reset() { *m = CriuRestoreResp{} }
func (m *CriuRestoreResp) String() string { return proto.CompactTextString(m) }
func (*CriuRestoreResp) ProtoMessage() {}
func (m *CriuRestoreResp) GetPid() int32 {
if m != nil && m.Pid != nil {
return *m.Pid
}
return 0
}
type CriuNotify struct {
Script *string `protobuf:"bytes,1,opt,name=script" json:"script,omitempty"`
Pid *int32 `protobuf:"varint,2,opt,name=pid" json:"pid,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuNotify) Reset() { *m = CriuNotify{} }
func (m *CriuNotify) String() string { return proto.CompactTextString(m) }
func (*CriuNotify) ProtoMessage() {}
func (m *CriuNotify) GetScript() string {
if m != nil && m.Script != nil {
return *m.Script
}
return ""
}
func (m *CriuNotify) GetPid() int32 {
if m != nil && m.Pid != nil {
return *m.Pid
}
return 0
}
type CriuReq struct {
Type *CriuReqType `protobuf:"varint,1,req,name=type,enum=CriuReqType" json:"type,omitempty"`
Opts *CriuOpts `protobuf:"bytes,2,opt,name=opts" json:"opts,omitempty"`
NotifySuccess *bool `protobuf:"varint,3,opt,name=notify_success" json:"notify_success,omitempty"`
//
// When set service won't close the connection but
// will wait for more req-s to appear. Works not
// for all request types.
KeepOpen *bool `protobuf:"varint,4,opt,name=keep_open" json:"keep_open,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuReq) Reset() { *m = CriuReq{} }
func (m *CriuReq) String() string { return proto.CompactTextString(m) }
func (*CriuReq) ProtoMessage() {}
func (m *CriuReq) GetType() CriuReqType {
if m != nil && m.Type != nil {
return *m.Type
}
return CriuReqType_EMPTY
}
func (m *CriuReq) GetOpts() *CriuOpts {
if m != nil {
return m.Opts
}
return nil
}
func (m *CriuReq) GetNotifySuccess() bool {
if m != nil && m.NotifySuccess != nil {
return *m.NotifySuccess
}
return false
}
func (m *CriuReq) GetKeepOpen() bool {
if m != nil && m.KeepOpen != nil {
return *m.KeepOpen
}
return false
}
type CriuResp struct {
Type *CriuReqType `protobuf:"varint,1,req,name=type,enum=CriuReqType" json:"type,omitempty"`
Success *bool `protobuf:"varint,2,req,name=success" json:"success,omitempty"`
Dump *CriuDumpResp `protobuf:"bytes,3,opt,name=dump" json:"dump,omitempty"`
Restore *CriuRestoreResp `protobuf:"bytes,4,opt,name=restore" json:"restore,omitempty"`
Notify *CriuNotify `protobuf:"bytes,5,opt,name=notify" json:"notify,omitempty"`
Ps *CriuPageServerInfo `protobuf:"bytes,6,opt,name=ps" json:"ps,omitempty"`
CrErrno *int32 `protobuf:"varint,7,opt,name=cr_errno" json:"cr_errno,omitempty"`
XXX_unrecognized []byte `json:"-"`
}
func (m *CriuResp) Reset() { *m = CriuResp{} }
func (m *CriuResp) String() string { return proto.CompactTextString(m) }
func (*CriuResp) ProtoMessage() {}
func (m *CriuResp) GetType() CriuReqType {
if m != nil && m.Type != nil {
return *m.Type
}
return CriuReqType_EMPTY
}
func (m *CriuResp) GetSuccess() bool {
if m != nil && m.Success != nil {
return *m.Success
}
return false
}
func (m *CriuResp) GetDump() *CriuDumpResp {
if m != nil {
return m.Dump
}
return nil
}
func (m *CriuResp) GetRestore() *CriuRestoreResp {
if m != nil {
return m.Restore
}
return nil
}
func (m *CriuResp) GetNotify() *CriuNotify {
if m != nil {
return m.Notify
}
return nil
}
func (m *CriuResp) GetPs() *CriuPageServerInfo {
if m != nil {
return m.Ps
}
return nil
}
func (m *CriuResp) GetCrErrno() int32 {
if m != nil && m.CrErrno != nil {
return *m.CrErrno
}
return 0
}
func init() {
proto.RegisterEnum("CriuReqType", CriuReqType_name, CriuReqType_value)
}

142
vendor/src/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.proto vendored Normal file
View file

@ -0,0 +1,142 @@
message criu_page_server_info {
optional string address = 1;
optional int32 port = 2;
optional int32 pid = 3;
optional int32 fd = 4;
}
message criu_veth_pair {
required string if_in = 1;
required string if_out = 2;
};
message ext_mount_map {
required string key = 1;
required string val = 2;
};
message inherit_fd {
required string key = 1;
required int32 fd = 2;
};
message cgroup_root {
optional string ctrl = 1;
required string path = 2;
};
message unix_sk {
required uint32 inode = 1;
};
message criu_opts {
required int32 images_dir_fd = 1;
optional int32 pid = 2; /* if not set on dump, will dump requesting process */
optional bool leave_running = 3;
optional bool ext_unix_sk = 4;
optional bool tcp_established = 5;
optional bool evasive_devices = 6;
optional bool shell_job = 7;
optional bool file_locks = 8;
optional int32 log_level = 9 [default = 2];
optional string log_file = 10; /* No subdirs are allowed. Consider using work-dir */
optional criu_page_server_info ps = 11;
optional bool notify_scripts = 12;
optional string root = 13;
optional string parent_img = 14;
optional bool track_mem = 15;
optional bool auto_dedup = 16;
optional int32 work_dir_fd = 17;
optional bool link_remap = 18;
repeated criu_veth_pair veths = 19;
optional uint32 cpu_cap = 20 [default = 0xffffffff];
optional bool force_irmap = 21;
repeated string exec_cmd = 22;
repeated ext_mount_map ext_mnt = 23;
optional bool manage_cgroups = 24; /* backward compatibility */
repeated cgroup_root cg_root = 25;
optional bool rst_sibling = 26; /* swrk only */
repeated inherit_fd inherit_fd = 27; /* swrk only */
optional bool auto_ext_mnt = 28;
optional bool ext_sharing = 29;
optional bool ext_masters = 30;
repeated string skip_mnt = 31;
repeated string enable_fs = 32;
repeated unix_sk unix_sk_ino = 33;
optional uint32 manage_cgroups_mode = 34;
}
message criu_dump_resp {
optional bool restored = 1;
}
message criu_restore_resp {
required int32 pid = 1;
}
message criu_notify {
optional string script = 1;
optional int32 pid = 2;
}
enum criu_req_type {
EMPTY = 0;
DUMP = 1;
RESTORE = 2;
CHECK = 3;
PRE_DUMP = 4;
PAGE_SERVER = 5;
NOTIFY = 6;
CPUINFO_DUMP = 7;
CPUINFO_CHECK = 8;
}
/*
* Request -- each type corresponds to must-be-there
* request arguments of respective type
*/
message criu_req {
required criu_req_type type = 1;
optional criu_opts opts = 2;
optional bool notify_success = 3;
/*
* When set service won't close the connection but
* will wait for more req-s to appear. Works not
* for all request types.
*/
optional bool keep_open = 4;
}
/*
* Responce -- it states whether the request was served
* and additional request-specific informarion
*/
message criu_resp {
required criu_req_type type = 1;
required bool success = 2;
optional criu_dump_resp dump = 3;
optional criu_restore_resp restore = 4;
optional criu_notify notify = 5;
optional criu_page_server_info ps = 6;
optional int32 cr_errno = 7;
}

65
vendor/src/github.com/opencontainers/runc/libcontainer/error.go vendored Normal file
View file

@ -0,0 +1,65 @@
package libcontainer
import "io"
// API error code type.
type ErrorCode int
// API error codes.
const (
// Factory errors
IdInUse ErrorCode = iota
InvalidIdFormat
// Container errors
ContainerNotExists
ContainerPaused
ContainerNotStopped
ContainerNotRunning
// Process errors
ProcessNotExecuted
// Common errors
ConfigInvalid
ConsoleExists
SystemError
)
func (c ErrorCode) String() string {
switch c {
case IdInUse:
return "Id already in use"
case InvalidIdFormat:
return "Invalid format"
case ContainerPaused:
return "Container paused"
case ConfigInvalid:
return "Invalid configuration"
case SystemError:
return "System error"
case ContainerNotExists:
return "Container does not exist"
case ContainerNotStopped:
return "Container is not stopped"
case ContainerNotRunning:
return "Container is not running"
case ConsoleExists:
return "Console exists for process"
default:
return "Unknown error"
}
}
// API Error type.
type Error interface {
error
// Returns a verbose string including the error message
// and a representation of the stack trace suitable for
// printing.
Detail(w io.Writer) error
// Returns the error code for this error.
Code() ErrorCode
}

45
vendor/src/github.com/opencontainers/runc/libcontainer/factory.go vendored Normal file
View file

@ -0,0 +1,45 @@
package libcontainer
import (
"github.com/opencontainers/runc/libcontainer/configs"
)
type Factory interface {
// Creates a new container with the given id and starts the initial process inside it.
// id must be a string containing only letters, digits and underscores and must contain
// between 1 and 1024 characters, inclusive.
//
// The id must not already be in use by an existing container. Containers created using
// a factory with the same path (and file system) must have distinct ids.
//
// Returns the new container with a running process.
//
// errors:
// IdInUse - id is already in use by a container
// InvalidIdFormat - id has incorrect format
// ConfigInvalid - config is invalid
// Systemerror - System error
//
// On error, any partially created container parts are cleaned up (the operation is atomic).
Create(id string, config *configs.Config) (Container, error)
// Load takes an ID for an existing container and returns the container information
// from the state. This presents a read only view of the container.
//
// errors:
// Path does not exist
// Container is stopped
// System error
Load(id string) (Container, error)
// StartInitialization is an internal API to libcontainer used during the reexec of the
// container.
//
// Errors:
// Pipe connection error
// System error
StartInitialization() error
// Type returns info string about factory type (e.g. lxc, libcontainer...)
Type() string
}

274
vendor/src/github.com/opencontainers/runc/libcontainer/factory_linux.go vendored Normal file
View file

@ -0,0 +1,274 @@
// +build linux
package libcontainer
import (
"encoding/json"
"fmt"
"io/ioutil"
"os"
"os/exec"
"path/filepath"
"regexp"
"strconv"
"syscall"
"github.com/docker/docker/pkg/mount"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/cgroups/fs"
"github.com/opencontainers/runc/libcontainer/cgroups/systemd"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/configs/validate"
)
const (
stateFilename = "state.json"
)
var (
idRegex = regexp.MustCompile(`^[\w_-]+$`)
maxIdLen = 1024
)
// InitArgs returns an options func to configure a LinuxFactory with the
// provided init arguments.
func InitArgs(args ...string) func(*LinuxFactory) error {
return func(l *LinuxFactory) error {
name := args[0]
if filepath.Base(name) == name {
if lp, err := exec.LookPath(name); err == nil {
name = lp
}
} else {
abs, err := filepath.Abs(name)
if err != nil {
return err
}
name = abs
}
l.InitPath = "/proc/self/exe"
l.InitArgs = append([]string{name}, args[1:]...)
return nil
}
}
// InitPath returns an options func to configure a LinuxFactory with the
// provided absolute path to the init binary and arguements.
func InitPath(path string, args ...string) func(*LinuxFactory) error {
return func(l *LinuxFactory) error {
l.InitPath = path
l.InitArgs = args
return nil
}
}
// SystemdCgroups is an options func to configure a LinuxFactory to return
// containers that use systemd to create and manage cgroups.
func SystemdCgroups(l *LinuxFactory) error {
l.NewCgroupsManager = func(config *configs.Cgroup, paths map[string]string) cgroups.Manager {
return &systemd.Manager{
Cgroups: config,
Paths: paths,
}
}
return nil
}
// Cgroupfs is an options func to configure a LinuxFactory to return
// containers that use the native cgroups filesystem implementation to
// create and manage cgroups.
func Cgroupfs(l *LinuxFactory) error {
l.NewCgroupsManager = func(config *configs.Cgroup, paths map[string]string) cgroups.Manager {
return &fs.Manager{
Cgroups: config,
Paths: paths,
}
}
return nil
}
// TmpfsRoot is an option func to mount LinuxFactory.Root to tmpfs.
func TmpfsRoot(l *LinuxFactory) error {
mounted, err := mount.Mounted(l.Root)
if err != nil {
return err
}
if !mounted {
if err := syscall.Mount("tmpfs", l.Root, "tmpfs", 0, ""); err != nil {
return err
}
}
return nil
}
// New returns a linux based container factory based in the root directory and
// configures the factory with the provided option funcs.
func New(root string, options ...func(*LinuxFactory) error) (Factory, error) {
if root != "" {
if err := os.MkdirAll(root, 0700); err != nil {
return nil, newGenericError(err, SystemError)
}
}
l := &LinuxFactory{
Root: root,
Validator: validate.New(),
CriuPath: "criu",
}
InitArgs(os.Args[0], "init")(l)
Cgroupfs(l)
for _, opt := range options {
if err := opt(l); err != nil {
return nil, err
}
}
return l, nil
}
// LinuxFactory implements the default factory interface for linux based systems.
type LinuxFactory struct {
// Root directory for the factory to store state.
Root string
// InitPath is the absolute path to the init binary.
InitPath string
// InitArgs are arguments for calling the init responsibilities for spawning
// a container.
InitArgs []string
// CriuPath is the path to the criu binary used for checkpoint and restore of
// containers.
CriuPath string
// Validator provides validation to container configurations.
Validator validate.Validator
// NewCgroupsManager returns an initialized cgroups manager for a single container.
NewCgroupsManager func(config *configs.Cgroup, paths map[string]string) cgroups.Manager
}
func (l *LinuxFactory) Create(id string, config *configs.Config) (Container, error) {
if l.Root == "" {
return nil, newGenericError(fmt.Errorf("invalid root"), ConfigInvalid)
}
if err := l.validateID(id); err != nil {
return nil, err
}
if err := l.Validator.Validate(config); err != nil {
return nil, newGenericError(err, ConfigInvalid)
}
containerRoot := filepath.Join(l.Root, id)
if _, err := os.Stat(containerRoot); err == nil {
return nil, newGenericError(fmt.Errorf("container with id exists: %v", id), IdInUse)
} else if !os.IsNotExist(err) {
return nil, newGenericError(err, SystemError)
}
if err := os.MkdirAll(containerRoot, 0700); err != nil {
return nil, newGenericError(err, SystemError)
}
c := &linuxContainer{
id: id,
root: containerRoot,
config: config,
initPath: l.InitPath,
initArgs: l.InitArgs,
criuPath: l.CriuPath,
cgroupManager: l.NewCgroupsManager(config.Cgroups, nil),
}
c.state = &stoppedState{c: c}
return c, nil
}
func (l *LinuxFactory) Load(id string) (Container, error) {
if l.Root == "" {
return nil, newGenericError(fmt.Errorf("invalid root"), ConfigInvalid)
}
containerRoot := filepath.Join(l.Root, id)
state, err := l.loadState(containerRoot)
if err != nil {
return nil, err
}
r := &nonChildProcess{
processPid: state.InitProcessPid,
processStartTime: state.InitProcessStartTime,
fds: state.ExternalDescriptors,
}
c := &linuxContainer{
initProcess: r,
id: id,
config: &state.Config,
initPath: l.InitPath,
initArgs: l.InitArgs,
criuPath: l.CriuPath,
cgroupManager: l.NewCgroupsManager(state.Config.Cgroups, state.CgroupPaths),
root: containerRoot,
}
c.state = &nullState{c: c}
return c, nil
}
func (l *LinuxFactory) Type() string {
return "libcontainer"
}
// StartInitialization loads a container by opening the pipe fd from the parent to read the configuration and state
// This is a low level implementation detail of the reexec and should not be consumed externally
func (l *LinuxFactory) StartInitialization() (err error) {
fdStr := os.Getenv("_LIBCONTAINER_INITPIPE")
pipefd, err := strconv.Atoi(fdStr)
if err != nil {
return fmt.Errorf("error converting env var _LIBCONTAINER_INITPIPE(%q) to an int: %s", fdStr, err)
}
var (
pipe = os.NewFile(uintptr(pipefd), "pipe")
it = initType(os.Getenv("_LIBCONTAINER_INITTYPE"))
)
// clear the current process's environment to clean any libcontainer
// specific env vars.
os.Clearenv()
defer func() {
// if we have an error during the initialization of the container's init then send it back to the
// parent process in the form of an initError.
if err != nil {
// ensure that any data sent from the parent is consumed so it doesn't
// receive ECONNRESET when the child writes to the pipe.
ioutil.ReadAll(pipe)
if err := json.NewEncoder(pipe).Encode(newSystemError(err)); err != nil {
panic(err)
}
}
// ensure that this pipe is always closed
pipe.Close()
}()
i, err := newContainerInit(it, pipe)
if err != nil {
return err
}
return i.Init()
}
func (l *LinuxFactory) loadState(root string) (*State, error) {
f, err := os.Open(filepath.Join(root, stateFilename))
if err != nil {
if os.IsNotExist(err) {
return nil, newGenericError(err, ContainerNotExists)
}
return nil, newGenericError(err, SystemError)
}
defer f.Close()
var state *State
if err := json.NewDecoder(f).Decode(&state); err != nil {
return nil, newGenericError(err, SystemError)
}
return state, nil
}
func (l *LinuxFactory) validateID(id string) error {
if !idRegex.MatchString(id) {
return newGenericError(fmt.Errorf("invalid id format: %v", id), InvalidIdFormat)
}
if len(id) > maxIdLen {
return newGenericError(fmt.Errorf("invalid id format: %v", id), InvalidIdFormat)
}
return nil
}

74
vendor/src/github.com/opencontainers/runc/libcontainer/generic_error.go vendored Normal file
View file

@ -0,0 +1,74 @@
package libcontainer
import (
"fmt"
"io"
"text/template"
"time"
"github.com/opencontainers/runc/libcontainer/stacktrace"
)
var errorTemplate = template.Must(template.New("error").Parse(`Timestamp: {{.Timestamp}}
Code: {{.ECode}}
{{if .Message }}
Message: {{.Message}}
{{end}}
Frames:{{range $i, $frame := .Stack.Frames}}
---
{{$i}}: {{$frame.Function}}
Package: {{$frame.Package}}
File: {{$frame.File}}@{{$frame.Line}}{{end}}
`))
func newGenericError(err error, c ErrorCode) Error {
if le, ok := err.(Error); ok {
return le
}
gerr := &genericError{
Timestamp: time.Now(),
Err: err,
ECode: c,
Stack: stacktrace.Capture(1),
}
if err != nil {
gerr.Message = err.Error()
}
return gerr
}
func newSystemError(err error) Error {
if le, ok := err.(Error); ok {
return le
}
gerr := &genericError{
Timestamp: time.Now(),
Err: err,
ECode: SystemError,
Stack: stacktrace.Capture(1),
}
if err != nil {
gerr.Message = err.Error()
}
return gerr
}
type genericError struct {
Timestamp time.Time
ECode ErrorCode
Err error `json:"-"`
Message string
Stack stacktrace.Stacktrace
}
func (e *genericError) Error() string {
return fmt.Sprintf("[%d] %s: %s", e.ECode, e.ECode, e.Message)
}
func (e *genericError) Code() ErrorCode {
return e.ECode
}
func (e *genericError) Detail(w io.Writer) error {
return errorTemplate.Execute(w, e)
}

334
vendor/src/github.com/opencontainers/runc/libcontainer/init_linux.go vendored Normal file
View file

@ -0,0 +1,334 @@
// +build linux
package libcontainer
import (
"encoding/json"
"fmt"
"io/ioutil"
"net"
"os"
"strconv"
"strings"
"syscall"
"github.com/Sirupsen/logrus"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/system"
"github.com/opencontainers/runc/libcontainer/user"
"github.com/opencontainers/runc/libcontainer/utils"
"github.com/vishvananda/netlink"
)
type initType string
const (
initSetns initType = "setns"
initStandard initType = "standard"
)
type pid struct {
Pid int `json:"pid"`
}
// network is an internal struct used to setup container networks.
type network struct {
configs.Network
// TempVethPeerName is a unique temporary veth peer name that was placed into
// the container's namespace.
TempVethPeerName string `json:"temp_veth_peer_name"`
}
// initConfig is used for transferring parameters from Exec() to Init()
type initConfig struct {
Args []string `json:"args"`
Env []string `json:"env"`
Cwd string `json:"cwd"`
Capabilities []string `json:"capabilities"`
User string `json:"user"`
Config *configs.Config `json:"config"`
Console string `json:"console"`
Networks []*network `json:"network"`
PassedFilesCount int `json:"passed_files_count"`
}
type initer interface {
Init() error
}
func newContainerInit(t initType, pipe *os.File) (initer, error) {
var config *initConfig
if err := json.NewDecoder(pipe).Decode(&config); err != nil {
return nil, err
}
if err := populateProcessEnvironment(config.Env); err != nil {
return nil, err
}
switch t {
case initSetns:
return &linuxSetnsInit{
config: config,
}, nil
case initStandard:
return &linuxStandardInit{
parentPid: syscall.Getppid(),
config: config,
}, nil
}
return nil, fmt.Errorf("unknown init type %q", t)
}
// populateProcessEnvironment loads the provided environment variables into the
// current processes's environment.
func populateProcessEnvironment(env []string) error {
for _, pair := range env {
p := strings.SplitN(pair, "=", 2)
if len(p) < 2 {
return fmt.Errorf("invalid environment '%v'", pair)
}
if err := os.Setenv(p[0], p[1]); err != nil {
return err
}
}
return nil
}
// finalizeNamespace drops the caps, sets the correct user
// and working dir, and closes any leaked file descriptors
// before executing the command inside the namespace
func finalizeNamespace(config *initConfig) error {
// Ensure that all unwanted fds we may have accidentally
// inherited are marked close-on-exec so they stay out of the
// container
if err := utils.CloseExecFrom(config.PassedFilesCount + 3); err != nil {
return err
}
capabilities := config.Config.Capabilities
if config.Capabilities != nil {
capabilities = config.Capabilities
}
w, err := newCapWhitelist(capabilities)
if err != nil {
return err
}
// drop capabilities in bounding set before changing user
if err := w.dropBoundingSet(); err != nil {
return err
}
// preserve existing capabilities while we change users
if err := system.SetKeepCaps(); err != nil {
return err
}
if err := setupUser(config); err != nil {
return err
}
if err := system.ClearKeepCaps(); err != nil {
return err
}
// drop all other capabilities
if err := w.drop(); err != nil {
return err
}
if config.Cwd != "" {
if err := syscall.Chdir(config.Cwd); err != nil {
return err
}
}
return nil
}
// joinExistingNamespaces gets all the namespace paths specified for the container and
// does a setns on the namespace fd so that the current process joins the namespace.
func joinExistingNamespaces(namespaces []configs.Namespace) error {
for _, ns := range namespaces {
if ns.Path != "" {
f, err := os.OpenFile(ns.Path, os.O_RDONLY, 0)
if err != nil {
return err
}
err = system.Setns(f.Fd(), uintptr(ns.Syscall()))
f.Close()
if err != nil {
return err
}
}
}
return nil
}
// setupUser changes the groups, gid, and uid for the user inside the container
func setupUser(config *initConfig) error {
// Set up defaults.
defaultExecUser := user.ExecUser{
Uid: syscall.Getuid(),
Gid: syscall.Getgid(),
Home: "/",
}
passwdPath, err := user.GetPasswdPath()
if err != nil {
return err
}
groupPath, err := user.GetGroupPath()
if err != nil {
return err
}
execUser, err := user.GetExecUserPath(config.User, &defaultExecUser, passwdPath, groupPath)
if err != nil {
return err
}
var addGroups []int
if len(config.Config.AdditionalGroups) > 0 {
addGroups, err = user.GetAdditionalGroupsPath(config.Config.AdditionalGroups, groupPath)
if err != nil {
return err
}
}
// before we change to the container's user make sure that the processes STDIO
// is correctly owned by the user that we are switching to.
if err := fixStdioPermissions(execUser); err != nil {
return err
}
suppGroups := append(execUser.Sgids, addGroups...)
if err := syscall.Setgroups(suppGroups); err != nil {
return err
}
if err := system.Setgid(execUser.Gid); err != nil {
return err
}
if err := system.Setuid(execUser.Uid); err != nil {
return err
}
// if we didn't get HOME already, set it based on the user's HOME
if envHome := os.Getenv("HOME"); envHome == "" {
if err := os.Setenv("HOME", execUser.Home); err != nil {
return err
}
}
return nil
}
// fixStdioPermissions fixes the permissions of PID 1's STDIO within the container to the specified user.
// The ownership needs to match because it is created outside of the container and needs to be
// localized.
func fixStdioPermissions(u *user.ExecUser) error {
var null syscall.Stat_t
if err := syscall.Stat("/dev/null", &null); err != nil {
return err
}
for _, fd := range []uintptr{
os.Stdin.Fd(),
os.Stderr.Fd(),
os.Stdout.Fd(),
} {
var s syscall.Stat_t
if err := syscall.Fstat(int(fd), &s); err != nil {
return err
}
// skip chown of /dev/null if it was used as one of the STDIO fds.
if s.Rdev == null.Rdev {
continue
}
if err := syscall.Fchown(int(fd), u.Uid, u.Gid); err != nil {
return err
}
}
return nil
}
// setupNetwork sets up and initializes any network interface inside the container.
func setupNetwork(config *initConfig) error {
for _, config := range config.Networks {
strategy, err := getStrategy(config.Type)
if err != nil {
return err
}
if err := strategy.initialize(config); err != nil {
return err
}
}
return nil
}
func setupRoute(config *configs.Config) error {
for _, config := range config.Routes {
_, dst, err := net.ParseCIDR(config.Destination)
if err != nil {
return err
}
src := net.ParseIP(config.Source)
if src == nil {
return fmt.Errorf("Invalid source for route: %s", config.Source)
}
gw := net.ParseIP(config.Gateway)
if gw == nil {
return fmt.Errorf("Invalid gateway for route: %s", config.Gateway)
}
l, err := netlink.LinkByName(config.InterfaceName)
if err != nil {
return err
}
route := &netlink.Route{
Scope: netlink.SCOPE_UNIVERSE,
Dst: dst,
Src: src,
Gw: gw,
LinkIndex: l.Attrs().Index,
}
if err := netlink.RouteAdd(route); err != nil {
return err
}
}
return nil
}
func setupRlimits(config *configs.Config) error {
for _, rlimit := range config.Rlimits {
l := &syscall.Rlimit{Max: rlimit.Hard, Cur: rlimit.Soft}
if err := syscall.Setrlimit(rlimit.Type, l); err != nil {
return fmt.Errorf("error setting rlimit type %v: %v", rlimit.Type, err)
}
}
return nil
}
func setOomScoreAdj(oomScoreAdj int) error {
path := "/proc/self/oom_score_adj"
return ioutil.WriteFile(path, []byte(strconv.Itoa(oomScoreAdj)), 0700)
}
// killCgroupProcesses freezes then iterates over all the processes inside the
// manager's cgroups sending a SIGKILL to each process then waiting for them to
// exit.
func killCgroupProcesses(m cgroups.Manager) error {
var procs []*os.Process
if err := m.Freeze(configs.Frozen); err != nil {
logrus.Warn(err)
}
pids, err := m.GetPids()
if err != nil {
m.Freeze(configs.Thawed)
return err
}
for _, pid := range pids {
if p, err := os.FindProcess(pid); err == nil {
procs = append(procs, p)
if err := p.Kill(); err != nil {
logrus.Warn(err)
}
}
}
if err := m.Freeze(configs.Thawed); err != nil {
logrus.Warn(err)
}
for _, p := range procs {
if _, err := p.Wait(); err != nil {
logrus.Warn(err)
}
}
return nil
}

76
vendor/src/github.com/opencontainers/runc/libcontainer/label/label.go vendored Normal file
View file

@ -0,0 +1,76 @@
// +build !selinux !linux
package label
// InitLabels returns the process label and file labels to be used within
// the container. A list of options can be passed into this function to alter
// the labels.
func InitLabels(options []string) (string, string, error) {
return "", "", nil
}
func GenLabels(options string) (string, string, error) {
return "", "", nil
}
func FormatMountLabel(src string, mountLabel string) string {
return src
}
func SetProcessLabel(processLabel string) error {
return nil
}
func SetFileLabel(path string, fileLabel string) error {
return nil
}
func SetFileCreateLabel(fileLabel string) error {
return nil
}
func Relabel(path string, fileLabel string, shared bool) error {
return nil
}
func GetPidLabel(pid int) (string, error) {
return "", nil
}
func Init() {
}
func ReserveLabel(label string) error {
return nil
}
func UnreserveLabel(label string) error {
return nil
}
// DupSecOpt takes an process label and returns security options that
// can be used to set duplicate labels on future container processes
func DupSecOpt(src string) []string {
return nil
}
// DisableSecOpt returns a security opt that can disable labeling
// support for future container processes
func DisableSecOpt() []string {
return nil
}
// Validate checks that the label does not include unexpected options
func Validate(label string) error {
return nil
}
// RelabelNeeded checks whether the user requested a relabel
func RelabelNeeded(label string) bool {
return false
}
// IsShared checks that the label includes a "shared" mark
func IsShared(label string) bool {
return false
}

192
vendor/src/github.com/opencontainers/runc/libcontainer/label/label_selinux.go vendored Normal file
View file

@ -0,0 +1,192 @@
// +build selinux,linux
package label
import (
"fmt"
"strings"
"github.com/opencontainers/runc/libcontainer/selinux"
)
// Valid Label Options
var validOptions = map[string]bool{
"disable": true,
"type": true,
"user": true,
"role": true,
"level": true,
}
var ErrIncompatibleLabel = fmt.Errorf("Bad SELinux option z and Z can not be used together")
// InitLabels returns the process label and file labels to be used within
// the container. A list of options can be passed into this function to alter
// the labels. The labels returned will include a random MCS String, that is
// guaranteed to be unique.
func InitLabels(options []string) (string, string, error) {
if !selinux.SelinuxEnabled() {
return "", "", nil
}
processLabel, mountLabel := selinux.GetLxcContexts()
if processLabel != "" {
pcon := selinux.NewContext(processLabel)
mcon := selinux.NewContext(mountLabel)
for _, opt := range options {
if opt == "disable" {
return "", "", nil
}
if i := strings.Index(opt, ":"); i == -1 {
return "", "", fmt.Errorf("Bad label option %q, valid options 'disable' or \n'user, role, level, type' followed by ':' and a value", opt)
}
con := strings.SplitN(opt, ":", 2)
if !validOptions[con[0]] {
return "", "", fmt.Errorf("Bad label option %q, valid options 'disable, user, role, level, type'", con[0])
}
pcon[con[0]] = con[1]
if con[0] == "level" || con[0] == "user" {
mcon[con[0]] = con[1]
}
}
processLabel = pcon.Get()
mountLabel = mcon.Get()
}
return processLabel, mountLabel, nil
}
// DEPRECATED: The GenLabels function is only to be used during the transition to the official API.
func GenLabels(options string) (string, string, error) {
return InitLabels(strings.Fields(options))
}
// FormatMountLabel returns a string to be used by the mount command.
// The format of this string will be used to alter the labeling of the mountpoint.
// The string returned is suitable to be used as the options field of the mount command.
// If you need to have additional mount point options, you can pass them in as
// the first parameter. Second parameter is the label that you wish to apply
// to all content in the mount point.
func FormatMountLabel(src, mountLabel string) string {
if mountLabel != "" {
switch src {
case "":
src = fmt.Sprintf("context=%q", mountLabel)
default:
src = fmt.Sprintf("%s,context=%q", src, mountLabel)
}
}
return src
}
// SetProcessLabel takes a process label and tells the kernel to assign the
// label to the next program executed by the current process.
func SetProcessLabel(processLabel string) error {
if processLabel == "" {
return nil
}
return selinux.Setexeccon(processLabel)
}
// GetProcessLabel returns the process label that the kernel will assign
// to the next program executed by the current process. If "" is returned
// this indicates that the default labeling will happen for the process.
func GetProcessLabel() (string, error) {
return selinux.Getexeccon()
}
// SetFileLabel modifies the "path" label to the specified file label
func SetFileLabel(path string, fileLabel string) error {
if selinux.SelinuxEnabled() && fileLabel != "" {
return selinux.Setfilecon(path, fileLabel)
}
return nil
}
// Tell the kernel the label for all files to be created
func SetFileCreateLabel(fileLabel string) error {
if selinux.SelinuxEnabled() {
return selinux.Setfscreatecon(fileLabel)
}
return nil
}
// Change the label of path to the filelabel string.
// It changes the MCS label to s0 if shared is true.
// This will allow all containers to share the content.
func Relabel(path string, fileLabel string, shared bool) error {
if !selinux.SelinuxEnabled() {
return nil
}
if fileLabel == "" {
return nil
}
exclude_paths := map[string]bool{"/": true, "/usr": true, "/etc": true}
if exclude_paths[path] {
return fmt.Errorf("Relabeling of %s is not allowed", path)
}
if shared {
c := selinux.NewContext(fileLabel)
c["level"] = "s0"
fileLabel = c.Get()
}
return selinux.Chcon(path, fileLabel, true)
}
// GetPidLabel will return the label of the process running with the specified pid
func GetPidLabel(pid int) (string, error) {
return selinux.Getpidcon(pid)
}
// Init initialises the labeling system
func Init() {
selinux.SelinuxEnabled()
}
// ReserveLabel will record the fact that the MCS label has already been used.
// This will prevent InitLabels from using the MCS label in a newly created
// container
func ReserveLabel(label string) error {
selinux.ReserveLabel(label)
return nil
}
// UnreserveLabel will remove the reservation of the MCS label.
// This will allow InitLabels to use the MCS label in a newly created
// containers
func UnreserveLabel(label string) error {
selinux.FreeLxcContexts(label)
return nil
}
// DupSecOpt takes an process label and returns security options that
// can be used to set duplicate labels on future container processes
func DupSecOpt(src string) []string {
return selinux.DupSecOpt(src)
}
// DisableSecOpt returns a security opt that can disable labeling
// support for future container processes
func DisableSecOpt() []string {
return selinux.DisableSecOpt()
}
// Validate checks that the label does not include unexpected options
func Validate(label string) error {
if strings.Contains(label, "z") && strings.Contains(label, "Z") {
return ErrIncompatibleLabel
}
return nil
}
// RelabelNeeded checks whether the user requested a relabel
func RelabelNeeded(label string) bool {
return strings.Contains(label, "z") || strings.Contains(label, "Z")
}
// IsShared checks that the label includes a "shared" mark
func IsShared(label string) bool {
return strings.Contains(label, "z")
}

62
vendor/src/github.com/opencontainers/runc/libcontainer/message_linux.go vendored Normal file
View file

@ -0,0 +1,62 @@
// +build linux
package libcontainer
import (
"syscall"
"github.com/vishvananda/netlink/nl"
)
// list of known message types we want to send to bootstrap program
// The number is randomly chosen to not conflict with known netlink types
const (
InitMsg uint16 = 62000
PidAttr uint16 = 27281
ConsolePathAttr uint16 = 27282
// When syscall.NLA_HDRLEN is in gccgo, take this out.
syscall_NLA_HDRLEN = (syscall.SizeofNlAttr + syscall.NLA_ALIGNTO - 1) & ^(syscall.NLA_ALIGNTO - 1)
)
type Int32msg struct {
Type uint16
Value uint32
}
// int32msg has the following representation
// | nlattr len | nlattr type |
// | uint32 value |
func (msg *Int32msg) Serialize() []byte {
buf := make([]byte, msg.Len())
native := nl.NativeEndian()
native.PutUint16(buf[0:2], uint16(msg.Len()))
native.PutUint16(buf[2:4], msg.Type)
native.PutUint32(buf[4:8], msg.Value)
return buf
}
func (msg *Int32msg) Len() int {
return syscall_NLA_HDRLEN + 4
}
// bytemsg has the following representation
// | nlattr len | nlattr type |
// | value | pad |
type Bytemsg struct {
Type uint16
Value []byte
}
func (msg *Bytemsg) Serialize() []byte {
l := msg.Len()
buf := make([]byte, (l+syscall.NLA_ALIGNTO-1) & ^(syscall.NLA_ALIGNTO-1))
native := nl.NativeEndian()
native.PutUint16(buf[0:2], uint16(l))
native.PutUint16(buf[2:4], msg.Type)
copy(buf[4:], msg.Value)
return buf
}
func (msg *Bytemsg) Len() int {
return syscall_NLA_HDRLEN + len(msg.Value) + 1 // null-terminated
}

259
vendor/src/github.com/opencontainers/runc/libcontainer/network_linux.go vendored Normal file
View file

@ -0,0 +1,259 @@
// +build linux
package libcontainer
import (
"fmt"
"io/ioutil"
"net"
"path/filepath"
"strconv"
"strings"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/utils"
"github.com/vishvananda/netlink"
)
var strategies = map[string]networkStrategy{
"veth": &veth{},
"loopback": &loopback{},
}
// networkStrategy represents a specific network configuration for
// a container's networking stack
type networkStrategy interface {
create(*network, int) error
initialize(*network) error
detach(*configs.Network) error
attach(*configs.Network) error
}
// getStrategy returns the specific network strategy for the
// provided type.
func getStrategy(tpe string) (networkStrategy, error) {
s, exists := strategies[tpe]
if !exists {
return nil, fmt.Errorf("unknown strategy type %q", tpe)
}
return s, nil
}
// Returns the network statistics for the network interfaces represented by the NetworkRuntimeInfo.
func getNetworkInterfaceStats(interfaceName string) (*NetworkInterface, error) {
out := &NetworkInterface{Name: interfaceName}
// This can happen if the network runtime information is missing - possible if the
// container was created by an old version of libcontainer.
if interfaceName == "" {
return out, nil
}
type netStatsPair struct {
// Where to write the output.
Out *uint64
// The network stats file to read.
File string
}
// Ingress for host veth is from the container. Hence tx_bytes stat on the host veth is actually number of bytes received by the container.
netStats := []netStatsPair{
{Out: &out.RxBytes, File: "tx_bytes"},
{Out: &out.RxPackets, File: "tx_packets"},
{Out: &out.RxErrors, File: "tx_errors"},
{Out: &out.RxDropped, File: "tx_dropped"},
{Out: &out.TxBytes, File: "rx_bytes"},
{Out: &out.TxPackets, File: "rx_packets"},
{Out: &out.TxErrors, File: "rx_errors"},
{Out: &out.TxDropped, File: "rx_dropped"},
}
for _, netStat := range netStats {
data, err := readSysfsNetworkStats(interfaceName, netStat.File)
if err != nil {
return nil, err
}
*(netStat.Out) = data
}
return out, nil
}
// Reads the specified statistics available under /sys/class/net/<EthInterface>/statistics
func readSysfsNetworkStats(ethInterface, statsFile string) (uint64, error) {
data, err := ioutil.ReadFile(filepath.Join("/sys/class/net", ethInterface, "statistics", statsFile))
if err != nil {
return 0, err
}
return strconv.ParseUint(strings.TrimSpace(string(data)), 10, 64)
}
// loopback is a network strategy that provides a basic loopback device
type loopback struct {
}
func (l *loopback) create(n *network, nspid int) error {
return nil
}
func (l *loopback) initialize(config *network) error {
return netlink.LinkSetUp(&netlink.Device{LinkAttrs: netlink.LinkAttrs{Name: "lo"}})
}
func (l *loopback) attach(n *configs.Network) (err error) {
return nil
}
func (l *loopback) detach(n *configs.Network) (err error) {
return nil
}
// veth is a network strategy that uses a bridge and creates
// a veth pair, one that is attached to the bridge on the host and the other
// is placed inside the container's namespace
type veth struct {
}
func (v *veth) detach(n *configs.Network) (err error) {
return netlink.LinkSetMaster(&netlink.Device{LinkAttrs: netlink.LinkAttrs{Name: n.HostInterfaceName}}, nil)
}
// attach a container network interface to an external network
func (v *veth) attach(n *configs.Network) (err error) {
brl, err := netlink.LinkByName(n.Bridge)
if err != nil {
return err
}
br, ok := brl.(*netlink.Bridge)
if !ok {
return fmt.Errorf("Wrong device type %T", brl)
}
host, err := netlink.LinkByName(n.HostInterfaceName)
if err != nil {
return err
}
if err := netlink.LinkSetMaster(host, br); err != nil {
return err
}
if err := netlink.LinkSetMTU(host, n.Mtu); err != nil {
return err
}
if n.HairpinMode {
if err := netlink.LinkSetHairpin(host, true); err != nil {
return err
}
}
if err := netlink.LinkSetUp(host); err != nil {
return err
}
return nil
}
func (v *veth) create(n *network, nspid int) (err error) {
tmpName, err := v.generateTempPeerName()
if err != nil {
return err
}
n.TempVethPeerName = tmpName
if n.Bridge == "" {
return fmt.Errorf("bridge is not specified")
}
veth := &netlink.Veth{
LinkAttrs: netlink.LinkAttrs{
Name: n.HostInterfaceName,
TxQLen: n.TxQueueLen,
},
PeerName: n.TempVethPeerName,
}
if err := netlink.LinkAdd(veth); err != nil {
return err
}
defer func() {
if err != nil {
netlink.LinkDel(veth)
}
}()
if err := v.attach(&n.Network); err != nil {
return err
}
child, err := netlink.LinkByName(n.TempVethPeerName)
if err != nil {
return err
}
return netlink.LinkSetNsPid(child, nspid)
}
func (v *veth) generateTempPeerName() (string, error) {
return utils.GenerateRandomName("veth", 7)
}
func (v *veth) initialize(config *network) error {
peer := config.TempVethPeerName
if peer == "" {
return fmt.Errorf("peer is not specified")
}
child, err := netlink.LinkByName(peer)
if err != nil {
return err
}
if err := netlink.LinkSetDown(child); err != nil {
return err
}
if err := netlink.LinkSetName(child, config.Name); err != nil {
return err
}
// get the interface again after we changed the name as the index also changes.
if child, err = netlink.LinkByName(config.Name); err != nil {
return err
}
if config.MacAddress != "" {
mac, err := net.ParseMAC(config.MacAddress)
if err != nil {
return err
}
if err := netlink.LinkSetHardwareAddr(child, mac); err != nil {
return err
}
}
ip, err := netlink.ParseAddr(config.Address)
if err != nil {
return err
}
if err := netlink.AddrAdd(child, ip); err != nil {
return err
}
if config.IPv6Address != "" {
ip6, err := netlink.ParseAddr(config.IPv6Address)
if err != nil {
return err
}
if err := netlink.AddrAdd(child, ip6); err != nil {
return err
}
}
if err := netlink.LinkSetMTU(child, config.Mtu); err != nil {
return err
}
if err := netlink.LinkSetUp(child); err != nil {
return err
}
if config.Gateway != "" {
gw := net.ParseIP(config.Gateway)
if err := netlink.RouteAdd(&netlink.Route{
Scope: netlink.SCOPE_UNIVERSE,
LinkIndex: child.Attrs().Index,
Gw: gw,
}); err != nil {
return err
}
}
if config.IPv6Gateway != "" {
gw := net.ParseIP(config.IPv6Gateway)
if err := netlink.RouteAdd(&netlink.Route{
Scope: netlink.SCOPE_UNIVERSE,
LinkIndex: child.Attrs().Index,
Gw: gw,
}); err != nil {
return err
}
}
return nil
}

63
vendor/src/github.com/opencontainers/runc/libcontainer/notify_linux.go vendored Normal file
View file

@ -0,0 +1,63 @@
// +build linux
package libcontainer
import (
"fmt"
"io/ioutil"
"os"
"path/filepath"
"syscall"
)
const oomCgroupName = "memory"
// notifyOnOOM returns channel on which you can expect event about OOM,
// if process died without OOM this channel will be closed.
// s is current *libcontainer.State for container.
func notifyOnOOM(paths map[string]string) (<-chan struct{}, error) {
dir := paths[oomCgroupName]
if dir == "" {
return nil, fmt.Errorf("There is no path for %q in state", oomCgroupName)
}
oomControl, err := os.Open(filepath.Join(dir, "memory.oom_control"))
if err != nil {
return nil, err
}
fd, _, syserr := syscall.RawSyscall(syscall.SYS_EVENTFD2, 0, syscall.FD_CLOEXEC, 0)
if syserr != 0 {
oomControl.Close()
return nil, syserr
}
eventfd := os.NewFile(fd, "eventfd")
eventControlPath := filepath.Join(dir, "cgroup.event_control")
data := fmt.Sprintf("%d %d", eventfd.Fd(), oomControl.Fd())
if err := ioutil.WriteFile(eventControlPath, []byte(data), 0700); err != nil {
eventfd.Close()
oomControl.Close()
return nil, err
}
ch := make(chan struct{})
go func() {
defer func() {
close(ch)
eventfd.Close()
oomControl.Close()
}()
buf := make([]byte, 8)
for {
if _, err := eventfd.Read(buf); err != nil {
return
}
// When a cgroup is destroyed, an event is sent to eventfd.
// So if the control path is gone, return instead of notifying.
if _, err := os.Lstat(eventControlPath); os.IsNotExist(err) {
return
}
ch <- struct{}{}
}
}()
return ch, nil
}

25
vendor/src/github.com/opencontainers/runc/libcontainer/nsenter/README.md vendored Normal file
View file

@ -0,0 +1,25 @@
## nsenter
The `nsenter` package registers a special init constructor that is called before
the Go runtime has a chance to boot. This provides us the ability to `setns` on
existing namespaces and avoid the issues that the Go runtime has with multiple
threads. This constructor will be called if this package is registered,
imported, in your go application.
The `nsenter` package will `import "C"` and it uses [cgo](https://golang.org/cmd/cgo/)
package. In cgo, if the import of "C" is immediately preceded by a comment, that comment,
called the preamble, is used as a header when compiling the C parts of the package.
So every time we import package `nsenter`, the C code function `nsexec()` would be
called. And package `nsenter` is now only imported in Docker execdriver, so every time
before we call `execdriver.Exec()`, that C code would run.
`nsexec()` will first check the environment variable `_LIBCONTAINER_INITPID`
which will give the process of the container that should be joined. Namespaces fd will
be found from `/proc/[pid]/ns` and set by `setns` syscall.
And then get the pipe number from `_LIBCONTAINER_INITPIPE`, error message could
be transfered through it. If tty is added, `_LIBCONTAINER_CONSOLE_PATH` will
have value and start a console for output.
Finally, `nsexec()` will clone a child process , exit the parent process and let
the Go runtime take over.

12
vendor/src/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go vendored Normal file
View file

@ -0,0 +1,12 @@
// +build linux,!gccgo
package nsenter
/*
#cgo CFLAGS: -Wall
extern void nsexec();
void __attribute__((constructor)) init(void) {
nsexec();
}
*/
import "C"

25
vendor/src/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go vendored Normal file
View file

@ -0,0 +1,25 @@
// +build linux,gccgo
package nsenter
/*
#cgo CFLAGS: -Wall
extern void nsexec();
void __attribute__((constructor)) init(void) {
nsexec();
}
*/
import "C"
// AlwaysFalse is here to stay false
// (and be exported so the compiler doesn't optimize out its reference)
var AlwaysFalse bool
func init() {
if AlwaysFalse {
// by referencing this C init() in a noop test, it will ensure the compiler
// links in the C function.
// https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65134
C.init()
}
}

5
vendor/src/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go vendored Normal file
View file

@ -0,0 +1,5 @@
// +build !linux !cgo
package nsenter
import "C"

260
vendor/src/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c vendored Normal file
View file

@ -0,0 +1,260 @@
#define _GNU_SOURCE
#include <stdlib.h>
#include <unistd.h>
#include <stdio.h>
#include <errno.h>
#include <string.h>
#include <linux/limits.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
#include <fcntl.h>
#include <signal.h>
#include <setjmp.h>
#include <sched.h>
#include <signal.h>
#include <linux/netlink.h>
#include <linux/types.h>
#include <stdint.h>
#include <sys/socket.h>
/* All arguments should be above stack, because it grows down */
struct clone_arg {
/*
* Reserve some space for clone() to locate arguments
* and retcode in this place
*/
char stack[4096] __attribute__ ((aligned(16)));
char stack_ptr[0];
jmp_buf *env;
};
#define pr_perror(fmt, ...) fprintf(stderr, "nsenter: " fmt ": %m\n", ##__VA_ARGS__)
static int child_func(void *_arg)
{
struct clone_arg *arg = (struct clone_arg *)_arg;
longjmp(*arg->env, 1);
}
// Use raw setns syscall for versions of glibc that don't include it (namely glibc-2.12)
#if __GLIBC__ == 2 && __GLIBC_MINOR__ < 14
#define _GNU_SOURCE
#include "syscall.h"
#if defined(__NR_setns) && !defined(SYS_setns)
#define SYS_setns __NR_setns
#endif
#ifdef SYS_setns
int setns(int fd, int nstype)
{
return syscall(SYS_setns, fd, nstype);
}
#endif
#endif
static int clone_parent(jmp_buf * env) __attribute__ ((noinline));
static int clone_parent(jmp_buf * env)
{
struct clone_arg ca;
int child;
ca.env = env;
child = clone(child_func, ca.stack_ptr, CLONE_PARENT | SIGCHLD, &ca);
return child;
}
static uint32_t readint32(char *buf)
{
return *(uint32_t *) buf;
}
// list of known message types we want to send to bootstrap program
// These are defined in libcontainer/message_linux.go
#define INIT_MSG 62000
#define PID_ATTR 27281
#define CONSOLE_PATH_ATTR 27282
void nsexec()
{
char *namespaces[] = { "ipc", "uts", "net", "pid", "mnt", "user" };
const int num = sizeof(namespaces) / sizeof(char *);
jmp_buf env;
char buf[PATH_MAX], *val;
int i, tfd, self_tfd, child, n, len, pipenum, consolefd = -1;
pid_t pid = 0;
// if we dont have INITTYPE or this is the init process, skip the bootstrap process
val = getenv("_LIBCONTAINER_INITTYPE");
if (val == NULL || strcmp(val, "standard") == 0) {
return;
}
if (strcmp(val, "setns") != 0) {
pr_perror("Invalid inittype %s", val);
exit(1);
}
val = getenv("_LIBCONTAINER_INITPIPE");
if (val == NULL) {
pr_perror("Child pipe not found");
exit(1);
}
pipenum = atoi(val);
snprintf(buf, sizeof(buf), "%d", pipenum);
if (strcmp(val, buf)) {
pr_perror("Unable to parse _LIBCONTAINER_INITPIPE");
exit(1);
}
char nlbuf[NLMSG_HDRLEN];
struct nlmsghdr *nh;
if ((n = read(pipenum, nlbuf, NLMSG_HDRLEN)) != NLMSG_HDRLEN) {
pr_perror("Failed to read netlink header, got %d", n);
exit(1);
}
nh = (struct nlmsghdr *)nlbuf;
if (nh->nlmsg_type == NLMSG_ERROR) {
pr_perror("Invalid netlink header message");
exit(1);
}
if (nh->nlmsg_type != INIT_MSG) {
pr_perror("Unexpected netlink message type %d", nh->nlmsg_type);
exit(1);
}
// read the netlink payload
len = NLMSG_PAYLOAD(nh, 0);
char data[len];
if ((n = read(pipenum, data, len)) != len) {
pr_perror("Failed to read netlink payload, got %d", n);
exit(1);
}
int start = 0;
struct nlattr *attr;
while (start < len) {
int payload_len;
attr = (struct nlattr *)((void *)data + start);
start += NLA_HDRLEN;
payload_len = attr->nla_len - NLA_HDRLEN;
switch (attr->nla_type) {
case PID_ATTR:
pid = (pid_t) readint32(data + start);
break;
case CONSOLE_PATH_ATTR:
consolefd = open((char *)data + start, O_RDWR);
if (consolefd < 0) {
pr_perror("Failed to open console %s", (char *)data + start);
exit(1);
}
break;
}
start += NLA_ALIGN(payload_len);
}
// required pid to be passed
if (pid == 0) {
pr_perror("missing pid");
exit(1);
}
/* Check that the specified process exists */
snprintf(buf, PATH_MAX - 1, "/proc/%d/ns", pid);
tfd = open(buf, O_DIRECTORY | O_RDONLY);
if (tfd == -1) {
pr_perror("Failed to open \"%s\"", buf);
exit(1);
}
self_tfd = open("/proc/self/ns", O_DIRECTORY | O_RDONLY);
if (self_tfd == -1) {
pr_perror("Failed to open /proc/self/ns");
exit(1);
}
for (i = 0; i < num; i++) {
struct stat st;
struct stat self_st;
int fd;
/* Symlinks on all namespaces exist for dead processes, but they can't be opened */
if (fstatat(tfd, namespaces[i], &st, 0) == -1) {
// Ignore nonexistent namespaces.
if (errno == ENOENT)
continue;
}
/* Skip namespaces we're already part of */
if (fstatat(self_tfd, namespaces[i], &self_st, 0) != -1 && st.st_ino == self_st.st_ino) {
continue;
}
fd = openat(tfd, namespaces[i], O_RDONLY);
if (fd == -1) {
pr_perror("Failed to open ns file %s for ns %s", buf, namespaces[i]);
exit(1);
}
// Set the namespace.
if (setns(fd, 0) == -1) {
pr_perror("Failed to setns for %s", namespaces[i]);
exit(1);
}
close(fd);
}
close(self_tfd);
close(tfd);
if (setjmp(env) == 1) {
// Child
if (setsid() == -1) {
pr_perror("setsid failed");
exit(1);
}
if (consolefd != -1) {
if (ioctl(consolefd, TIOCSCTTY, 0) == -1) {
pr_perror("ioctl TIOCSCTTY failed");
exit(1);
}
if (dup3(consolefd, STDIN_FILENO, 0) != STDIN_FILENO) {
pr_perror("Failed to dup 0");
exit(1);
}
if (dup3(consolefd, STDOUT_FILENO, 0) != STDOUT_FILENO) {
pr_perror("Failed to dup 1");
exit(1);
}
if (dup3(consolefd, STDERR_FILENO, 0) != STDERR_FILENO) {
pr_perror("Failed to dup 2");
exit(1);
}
}
// Finish executing, let the Go runtime take over.
return;
}
// Parent
// We must fork to actually enter the PID namespace, use CLONE_PARENT
// so the child can have the right parent, and we don't need to forward
// the child's exit code or resend its death signal.
child = clone_parent(&env);
if (child < 0) {
pr_perror("Unable to fork");
exit(1);
}
len = snprintf(buf, sizeof(buf), "{ \"pid\" : %d }\n", child);
if (write(pipenum, buf, len) != len) {
pr_perror("Unable to send a child pid");
kill(child, SIGKILL);
exit(1);
}
exit(0);
}

105
vendor/src/github.com/opencontainers/runc/libcontainer/process.go vendored Normal file
View file

@ -0,0 +1,105 @@
package libcontainer
import (
"fmt"
"io"
"math"
"os"
)
type processOperations interface {
wait() (*os.ProcessState, error)
signal(sig os.Signal) error
pid() int
}
// Process specifies the configuration and IO for a process inside
// a container.
type Process struct {
// The command to be run followed by any arguments.
Args []string
// Env specifies the environment variables for the process.
Env []string
// User will set the uid and gid of the executing process running inside the container
// local to the container's user and group configuration.
User string
// Cwd will change the processes current working directory inside the container's rootfs.
Cwd string
// Stdin is a pointer to a reader which provides the standard input stream.
Stdin io.Reader
// Stdout is a pointer to a writer which receives the standard output stream.
Stdout io.Writer
// Stderr is a pointer to a writer which receives the standard error stream.
Stderr io.Writer
// ExtraFiles specifies additional open files to be inherited by the container
ExtraFiles []*os.File
// consolePath is the path to the console allocated to the container.
consolePath string
// Capabilities specify the capabilities to keep when executing the process inside the container
// All capabilities not specified will be dropped from the processes capability mask
Capabilities []string
ops processOperations
}
// Wait waits for the process to exit.
// Wait releases any resources associated with the Process
func (p Process) Wait() (*os.ProcessState, error) {
if p.ops == nil {
return nil, newGenericError(fmt.Errorf("invalid process"), ProcessNotExecuted)
}
return p.ops.wait()
}
// Pid returns the process ID
func (p Process) Pid() (int, error) {
// math.MinInt32 is returned here, because it's invalid value
// for the kill() system call.
if p.ops == nil {
return math.MinInt32, newGenericError(fmt.Errorf("invalid process"), ProcessNotExecuted)
}
return p.ops.pid(), nil
}
// Signal sends a signal to the Process.
func (p Process) Signal(sig os.Signal) error {
if p.ops == nil {
return newGenericError(fmt.Errorf("invalid process"), ProcessNotExecuted)
}
return p.ops.signal(sig)
}
// IO holds the process's STDIO
type IO struct {
Stdin io.WriteCloser
Stdout io.ReadCloser
Stderr io.ReadCloser
}
// NewConsole creates new console for process and returns it
func (p *Process) NewConsole(rootuid int) (Console, error) {
console, err := NewConsole(rootuid, rootuid)
if err != nil {
return nil, err
}
p.consolePath = console.Path()
return console, nil
}
// ConsoleFromPath sets the process's console with the path provided
func (p *Process) ConsoleFromPath(path string) error {
if p.consolePath != "" {
return newGenericError(fmt.Errorf("console path already exists for process"), ConsoleExists)
}
p.consolePath = path
return nil
}

363
vendor/src/github.com/opencontainers/runc/libcontainer/process_linux.go vendored Normal file
View file

@ -0,0 +1,363 @@
// +build linux
package libcontainer
import (
"encoding/json"
"errors"
"io"
"os"
"os/exec"
"path/filepath"
"strconv"
"syscall"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/system"
)
type parentProcess interface {
// pid returns the pid for the running process.
pid() int
// start starts the process execution.
start() error
// send a SIGKILL to the process and wait for the exit.
terminate() error
// wait waits on the process returning the process state.
wait() (*os.ProcessState, error)
// startTime return's the process start time.
startTime() (string, error)
signal(os.Signal) error
externalDescriptors() []string
setExternalDescriptors(fds []string)
}
type setnsProcess struct {
cmd *exec.Cmd
parentPipe *os.File
childPipe *os.File
cgroupPaths map[string]string
config *initConfig
fds []string
process *Process
bootstrapData io.Reader
}
func (p *setnsProcess) startTime() (string, error) {
return system.GetProcessStartTime(p.pid())
}
func (p *setnsProcess) signal(sig os.Signal) error {
s, ok := sig.(syscall.Signal)
if !ok {
return errors.New("os: unsupported signal type")
}
return syscall.Kill(p.pid(), s)
}
func (p *setnsProcess) start() (err error) {
defer p.parentPipe.Close()
err = p.cmd.Start()
p.childPipe.Close()
if err != nil {
return newSystemError(err)
}
if p.bootstrapData != nil {
if _, err := io.Copy(p.parentPipe, p.bootstrapData); err != nil {
return newSystemError(err)
}
}
if err = p.execSetns(); err != nil {
return newSystemError(err)
}
if len(p.cgroupPaths) > 0 {
if err := cgroups.EnterPid(p.cgroupPaths, p.pid()); err != nil {
return newSystemError(err)
}
}
if err := json.NewEncoder(p.parentPipe).Encode(p.config); err != nil {
return newSystemError(err)
}
if err := syscall.Shutdown(int(p.parentPipe.Fd()), syscall.SHUT_WR); err != nil {
return newSystemError(err)
}
// wait for the child process to fully complete and receive an error message
// if one was encoutered
var ierr *genericError
if err := json.NewDecoder(p.parentPipe).Decode(&ierr); err != nil && err != io.EOF {
return newSystemError(err)
}
if ierr != nil {
p.wait()
return newSystemError(ierr)
}
return nil
}
// execSetns runs the process that executes C code to perform the setns calls
// because setns support requires the C process to fork off a child and perform the setns
// before the go runtime boots, we wait on the process to die and receive the child's pid
// over the provided pipe.
func (p *setnsProcess) execSetns() error {
status, err := p.cmd.Process.Wait()
if err != nil {
p.cmd.Wait()
return newSystemError(err)
}
if !status.Success() {
p.cmd.Wait()
return newSystemError(&exec.ExitError{ProcessState: status})
}
var pid *pid
if err := json.NewDecoder(p.parentPipe).Decode(&pid); err != nil {
p.cmd.Wait()
return newSystemError(err)
}
process, err := os.FindProcess(pid.Pid)
if err != nil {
return err
}
p.cmd.Process = process
p.process.ops = p
return nil
}
// terminate sends a SIGKILL to the forked process for the setns routine then waits to
// avoid the process becomming a zombie.
func (p *setnsProcess) terminate() error {
if p.cmd.Process == nil {
return nil
}
err := p.cmd.Process.Kill()
if _, werr := p.wait(); err == nil {
err = werr
}
return err
}
func (p *setnsProcess) wait() (*os.ProcessState, error) {
err := p.cmd.Wait()
// Return actual ProcessState even on Wait error
return p.cmd.ProcessState, err
}
func (p *setnsProcess) pid() int {
return p.cmd.Process.Pid
}
func (p *setnsProcess) externalDescriptors() []string {
return p.fds
}
func (p *setnsProcess) setExternalDescriptors(newFds []string) {
p.fds = newFds
}
type initProcess struct {
cmd *exec.Cmd
parentPipe *os.File
childPipe *os.File
config *initConfig
manager cgroups.Manager
container *linuxContainer
fds []string
process *Process
}
func (p *initProcess) pid() int {
return p.cmd.Process.Pid
}
func (p *initProcess) externalDescriptors() []string {
return p.fds
}
func (p *initProcess) start() (err error) {
defer p.parentPipe.Close()
err = p.cmd.Start()
p.process.ops = p
p.childPipe.Close()
if err != nil {
p.process.ops = nil
return newSystemError(err)
}
// Save the standard descriptor names before the container process
// can potentially move them (e.g., via dup2()). If we don't do this now,
// we won't know at checkpoint time which file descriptor to look up.
fds, err := getPipeFds(p.pid())
if err != nil {
return newSystemError(err)
}
p.setExternalDescriptors(fds)
// Do this before syncing with child so that no children
// can escape the cgroup
if err := p.manager.Apply(p.pid()); err != nil {
return newSystemError(err)
}
defer func() {
if err != nil {
// TODO: should not be the responsibility to call here
p.manager.Destroy()
}
}()
if p.config.Config.Hooks != nil {
s := configs.HookState{
Version: p.container.config.Version,
ID: p.container.id,
Pid: p.pid(),
Root: p.config.Config.Rootfs,
}
for _, hook := range p.config.Config.Hooks.Prestart {
if err := hook.Run(s); err != nil {
return newSystemError(err)
}
}
}
if err := p.createNetworkInterfaces(); err != nil {
return newSystemError(err)
}
if err := p.sendConfig(); err != nil {
return newSystemError(err)
}
// wait for the child process to fully complete and receive an error message
// if one was encoutered
var ierr *genericError
if err := json.NewDecoder(p.parentPipe).Decode(&ierr); err != nil && err != io.EOF {
return newSystemError(err)
}
if ierr != nil {
return newSystemError(ierr)
}
return nil
}
func (p *initProcess) wait() (*os.ProcessState, error) {
err := p.cmd.Wait()
if err != nil {
return p.cmd.ProcessState, err
}
// we should kill all processes in cgroup when init is died if we use host PID namespace
if p.cmd.SysProcAttr.Cloneflags&syscall.CLONE_NEWPID == 0 {
killCgroupProcesses(p.manager)
}
return p.cmd.ProcessState, nil
}
func (p *initProcess) terminate() error {
if p.cmd.Process == nil {
return nil
}
err := p.cmd.Process.Kill()
if _, werr := p.wait(); err == nil {
err = werr
}
return err
}
func (p *initProcess) startTime() (string, error) {
return system.GetProcessStartTime(p.pid())
}
func (p *initProcess) sendConfig() error {
// send the state to the container's init process then shutdown writes for the parent
if err := json.NewEncoder(p.parentPipe).Encode(p.config); err != nil {
return err
}
// shutdown writes for the parent side of the pipe
return syscall.Shutdown(int(p.parentPipe.Fd()), syscall.SHUT_WR)
}
func (p *initProcess) createNetworkInterfaces() error {
for _, config := range p.config.Config.Networks {
strategy, err := getStrategy(config.Type)
if err != nil {
return err
}
n := &network{
Network: *config,
}
if err := strategy.create(n, p.pid()); err != nil {
return err
}
p.config.Networks = append(p.config.Networks, n)
}
return nil
}
func (p *initProcess) signal(sig os.Signal) error {
s, ok := sig.(syscall.Signal)
if !ok {
return errors.New("os: unsupported signal type")
}
return syscall.Kill(p.pid(), s)
}
func (p *initProcess) setExternalDescriptors(newFds []string) {
p.fds = newFds
}
func getPipeFds(pid int) ([]string, error) {
fds := make([]string, 3)
dirPath := filepath.Join("/proc", strconv.Itoa(pid), "/fd")
for i := 0; i < 3; i++ {
f := filepath.Join(dirPath, strconv.Itoa(i))
target, err := os.Readlink(f)
if err != nil {
return fds, err
}
fds[i] = target
}
return fds, nil
}
// InitializeIO creates pipes for use with the process's STDIO
// and returns the opposite side for each
func (p *Process) InitializeIO(rootuid int) (i *IO, err error) {
var fds []uintptr
i = &IO{}
// cleanup in case of an error
defer func() {
if err != nil {
for _, fd := range fds {
syscall.Close(int(fd))
}
}
}()
// STDIN
r, w, err := os.Pipe()
if err != nil {
return nil, err
}
fds = append(fds, r.Fd(), w.Fd())
p.Stdin, i.Stdin = r, w
// STDOUT
if r, w, err = os.Pipe(); err != nil {
return nil, err
}
fds = append(fds, r.Fd(), w.Fd())
p.Stdout, i.Stdout = w, r
// STDERR
if r, w, err = os.Pipe(); err != nil {
return nil, err
}
fds = append(fds, r.Fd(), w.Fd())
p.Stderr, i.Stderr = w, r
// change ownership of the pipes incase we are in a user namespace
for _, fd := range fds {
if err := syscall.Fchown(int(fd), rootuid, rootuid); err != nil {
return nil, err
}
}
return i, nil
}

122
vendor/src/github.com/opencontainers/runc/libcontainer/restored_process.go vendored Normal file
View file

@ -0,0 +1,122 @@
// +build linux
package libcontainer
import (
"fmt"
"os"
"github.com/opencontainers/runc/libcontainer/system"
)
func newRestoredProcess(pid int, fds []string) (*restoredProcess, error) {
var (
err error
)
proc, err := os.FindProcess(pid)
if err != nil {
return nil, err
}
started, err := system.GetProcessStartTime(pid)
if err != nil {
return nil, err
}
return &restoredProcess{
proc: proc,
processStartTime: started,
fds: fds,
}, nil
}
type restoredProcess struct {
proc *os.Process
processStartTime string
fds []string
}
func (p *restoredProcess) start() error {
return newGenericError(fmt.Errorf("restored process cannot be started"), SystemError)
}
func (p *restoredProcess) pid() int {
return p.proc.Pid
}
func (p *restoredProcess) terminate() error {
err := p.proc.Kill()
if _, werr := p.wait(); err == nil {
err = werr
}
return err
}
func (p *restoredProcess) wait() (*os.ProcessState, error) {
// TODO: how do we wait on the actual process?
// maybe use --exec-cmd in criu
st, err := p.proc.Wait()
if err != nil {
return nil, err
}
return st, nil
}
func (p *restoredProcess) startTime() (string, error) {
return p.processStartTime, nil
}
func (p *restoredProcess) signal(s os.Signal) error {
return p.proc.Signal(s)
}
func (p *restoredProcess) externalDescriptors() []string {
return p.fds
}
func (p *restoredProcess) setExternalDescriptors(newFds []string) {
p.fds = newFds
}
// nonChildProcess represents a process where the calling process is not
// the parent process. This process is created when a factory loads a container from
// a persisted state.
type nonChildProcess struct {
processPid int
processStartTime string
fds []string
}
func (p *nonChildProcess) start() error {
return newGenericError(fmt.Errorf("restored process cannot be started"), SystemError)
}
func (p *nonChildProcess) pid() int {
return p.processPid
}
func (p *nonChildProcess) terminate() error {
return newGenericError(fmt.Errorf("restored process cannot be terminated"), SystemError)
}
func (p *nonChildProcess) wait() (*os.ProcessState, error) {
return nil, newGenericError(fmt.Errorf("restored process cannot be waited on"), SystemError)
}
func (p *nonChildProcess) startTime() (string, error) {
return p.processStartTime, nil
}
func (p *nonChildProcess) signal(s os.Signal) error {
proc, err := os.FindProcess(p.processPid)
if err != nil {
return err
}
return proc.Signal(s)
}
func (p *nonChildProcess) externalDescriptors() []string {
return p.fds
}
func (p *nonChildProcess) setExternalDescriptors(newFds []string) {
p.fds = newFds
}

669
vendor/src/github.com/opencontainers/runc/libcontainer/rootfs_linux.go vendored Normal file
View file

@ -0,0 +1,669 @@
// +build linux
package libcontainer
import (
"fmt"
"io/ioutil"
"os"
"os/exec"
"path"
"path/filepath"
"strings"
"syscall"
"time"
"github.com/docker/docker/pkg/mount"
"github.com/docker/docker/pkg/symlink"
"github.com/opencontainers/runc/libcontainer/cgroups"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/label"
)
const defaultMountFlags = syscall.MS_NOEXEC | syscall.MS_NOSUID | syscall.MS_NODEV
// setupRootfs sets up the devices, mount points, and filesystems for use inside a
// new mount namespace.
func setupRootfs(config *configs.Config, console *linuxConsole) (err error) {
if err := prepareRoot(config); err != nil {
return newSystemError(err)
}
setupDev := len(config.Devices) != 0
for _, m := range config.Mounts {
for _, precmd := range m.PremountCmds {
if err := mountCmd(precmd); err != nil {
return newSystemError(err)
}
}
if err := mountToRootfs(m, config.Rootfs, config.MountLabel); err != nil {
return newSystemError(err)
}
for _, postcmd := range m.PostmountCmds {
if err := mountCmd(postcmd); err != nil {
return newSystemError(err)
}
}
}
if setupDev {
if err := createDevices(config); err != nil {
return newSystemError(err)
}
if err := setupPtmx(config, console); err != nil {
return newSystemError(err)
}
if err := setupDevSymlinks(config.Rootfs); err != nil {
return newSystemError(err)
}
}
if err := syscall.Chdir(config.Rootfs); err != nil {
return newSystemError(err)
}
if config.NoPivotRoot {
err = msMoveRoot(config.Rootfs)
} else {
err = pivotRoot(config.Rootfs, config.PivotDir)
}
if err != nil {
return newSystemError(err)
}
if setupDev {
if err := reOpenDevNull(); err != nil {
return newSystemError(err)
}
}
if config.Readonlyfs {
if err := setReadonly(); err != nil {
return newSystemError(err)
}
}
syscall.Umask(0022)
return nil
}
func mountCmd(cmd configs.Command) error {
command := exec.Command(cmd.Path, cmd.Args[:]...)
command.Env = cmd.Env
command.Dir = cmd.Dir
if out, err := command.CombinedOutput(); err != nil {
return fmt.Errorf("%#v failed: %s: %v", cmd, string(out), err)
}
return nil
}
func mountToRootfs(m *configs.Mount, rootfs, mountLabel string) error {
var (
dest = m.Destination
)
if !strings.HasPrefix(dest, rootfs) {
dest = filepath.Join(rootfs, dest)
}
switch m.Device {
case "proc", "sysfs":
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
// Selinux kernels do not support labeling of /proc or /sys
return mountPropagate(m, rootfs, "")
case "mqueue":
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
if err := mountPropagate(m, rootfs, mountLabel); err != nil {
// older kernels do not support labeling of /dev/mqueue
if err := mountPropagate(m, rootfs, ""); err != nil {
return err
}
}
return label.SetFileLabel(dest, mountLabel)
case "tmpfs":
stat, err := os.Stat(dest)
if err != nil {
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
}
if err := mountPropagate(m, rootfs, mountLabel); err != nil {
return err
}
if stat != nil {
if err = os.Chmod(dest, stat.Mode()); err != nil {
return err
}
}
return nil
case "devpts":
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
return mountPropagate(m, rootfs, mountLabel)
case "securityfs":
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
return mountPropagate(m, rootfs, mountLabel)
case "bind":
stat, err := os.Stat(m.Source)
if err != nil {
// error out if the source of a bind mount does not exist as we will be
// unable to bind anything to it.
return err
}
// ensure that the destination of the bind mount is resolved of symlinks at mount time because
// any previous mounts can invalidate the next mount's destination.
// this can happen when a user specifies mounts within other mounts to cause breakouts or other
// evil stuff to try to escape the container's rootfs.
if dest, err = symlink.FollowSymlinkInScope(filepath.Join(rootfs, m.Destination), rootfs); err != nil {
return err
}
if err := checkMountDestination(rootfs, dest); err != nil {
return err
}
// update the mount with the correct dest after symlinks are resolved.
m.Destination = dest
if err := createIfNotExists(dest, stat.IsDir()); err != nil {
return err
}
if err := mountPropagate(m, rootfs, mountLabel); err != nil {
return err
}
// bind mount won't change mount options, we need remount to make mount options effective.
// first check that we have non-default options required before attempting a remount
if m.Flags&^(syscall.MS_REC|syscall.MS_REMOUNT|syscall.MS_BIND) != 0 {
// only remount if unique mount options are set
if err := remount(m, rootfs); err != nil {
return err
}
}
if m.Relabel != "" {
if err := label.Validate(m.Relabel); err != nil {
return err
}
shared := label.IsShared(m.Relabel)
if err := label.Relabel(m.Source, mountLabel, shared); err != nil {
return err
}
}
case "cgroup":
binds, err := getCgroupMounts(m)
if err != nil {
return err
}
var merged []string
for _, b := range binds {
ss := filepath.Base(b.Destination)
if strings.Contains(ss, ",") {
merged = append(merged, ss)
}
}
tmpfs := &configs.Mount{
Source: "tmpfs",
Device: "tmpfs",
Destination: m.Destination,
Flags: defaultMountFlags,
Data: "mode=755",
PropagationFlags: m.PropagationFlags,
}
if err := mountToRootfs(tmpfs, rootfs, mountLabel); err != nil {
return err
}
for _, b := range binds {
if err := mountToRootfs(b, rootfs, mountLabel); err != nil {
return err
}
}
// create symlinks for merged cgroups
cwd, err := os.Getwd()
if err != nil {
return err
}
if err := os.Chdir(filepath.Join(rootfs, m.Destination)); err != nil {
return err
}
for _, mc := range merged {
for _, ss := range strings.Split(mc, ",") {
if err := os.Symlink(mc, ss); err != nil {
// if cgroup already exists, then okay(it could have been created before)
if os.IsExist(err) {
continue
}
os.Chdir(cwd)
return err
}
}
}
if err := os.Chdir(cwd); err != nil {
return err
}
if m.Flags&syscall.MS_RDONLY != 0 {
// remount cgroup root as readonly
mcgrouproot := &configs.Mount{
Destination: m.Destination,
Flags: defaultMountFlags | syscall.MS_RDONLY,
}
if err := remount(mcgrouproot, rootfs); err != nil {
return err
}
}
default:
return fmt.Errorf("unknown mount device %q to %q", m.Device, m.Destination)
}
return nil
}
func getCgroupMounts(m *configs.Mount) ([]*configs.Mount, error) {
mounts, err := cgroups.GetCgroupMounts()
if err != nil {
return nil, err
}
cgroupPaths, err := cgroups.ParseCgroupFile("/proc/self/cgroup")
if err != nil {
return nil, err
}
var binds []*configs.Mount
for _, mm := range mounts {
dir, err := mm.GetThisCgroupDir(cgroupPaths)
if err != nil {
return nil, err
}
relDir, err := filepath.Rel(mm.Root, dir)
if err != nil {
return nil, err
}
binds = append(binds, &configs.Mount{
Device: "bind",
Source: filepath.Join(mm.Mountpoint, relDir),
Destination: filepath.Join(m.Destination, strings.Join(mm.Subsystems, ",")),
Flags: syscall.MS_BIND | syscall.MS_REC | m.Flags,
PropagationFlags: m.PropagationFlags,
})
}
return binds, nil
}
// checkMountDestination checks to ensure that the mount destination is not over the top of /proc.
// dest is required to be an abs path and have any symlinks resolved before calling this function.
func checkMountDestination(rootfs, dest string) error {
if filepath.Clean(rootfs) == filepath.Clean(dest) {
return fmt.Errorf("mounting into / is prohibited")
}
invalidDestinations := []string{
"/proc",
}
for _, invalid := range invalidDestinations {
path, err := filepath.Rel(filepath.Join(rootfs, invalid), dest)
if err != nil {
return err
}
if path == "." || !strings.HasPrefix(path, "..") {
return fmt.Errorf("%q cannot be mounted because it is located inside %q", dest, invalid)
}
}
return nil
}
func setupDevSymlinks(rootfs string) error {
var links = [][2]string{
{"/proc/self/fd", "/dev/fd"},
{"/proc/self/fd/0", "/dev/stdin"},
{"/proc/self/fd/1", "/dev/stdout"},
{"/proc/self/fd/2", "/dev/stderr"},
}
// kcore support can be toggled with CONFIG_PROC_KCORE; only create a symlink
// in /dev if it exists in /proc.
if _, err := os.Stat("/proc/kcore"); err == nil {
links = append(links, [2]string{"/proc/kcore", "/dev/kcore"})
}
for _, link := range links {
var (
src = link[0]
dst = filepath.Join(rootfs, link[1])
)
if err := os.Symlink(src, dst); err != nil && !os.IsExist(err) {
return fmt.Errorf("symlink %s %s %s", src, dst, err)
}
}
return nil
}
// If stdin, stdout, and/or stderr are pointing to `/dev/null` in the parent's rootfs
// this method will make them point to `/dev/null` in this container's rootfs. This
// needs to be called after we chroot/pivot into the container's rootfs so that any
// symlinks are resolved locally.
func reOpenDevNull() error {
var stat, devNullStat syscall.Stat_t
file, err := os.OpenFile("/dev/null", os.O_RDWR, 0)
if err != nil {
return fmt.Errorf("Failed to open /dev/null - %s", err)
}
defer file.Close()
if err := syscall.Fstat(int(file.Fd()), &devNullStat); err != nil {
return err
}
for fd := 0; fd < 3; fd++ {
if err := syscall.Fstat(fd, &stat); err != nil {
return err
}
if stat.Rdev == devNullStat.Rdev {
// Close and re-open the fd.
if err := syscall.Dup3(int(file.Fd()), fd, 0); err != nil {
return err
}
}
}
return nil
}
// Create the device nodes in the container.
func createDevices(config *configs.Config) error {
oldMask := syscall.Umask(0000)
for _, node := range config.Devices {
// containers running in a user namespace are not allowed to mknod
// devices so we can just bind mount it from the host.
if err := createDeviceNode(config.Rootfs, node, config.Namespaces.Contains(configs.NEWUSER)); err != nil {
syscall.Umask(oldMask)
return err
}
}
syscall.Umask(oldMask)
return nil
}
func bindMountDeviceNode(dest string, node *configs.Device) error {
f, err := os.Create(dest)
if err != nil && !os.IsExist(err) {
return err
}
if f != nil {
f.Close()
}
return syscall.Mount(node.Path, dest, "bind", syscall.MS_BIND, "")
}
// Creates the device node in the rootfs of the container.
func createDeviceNode(rootfs string, node *configs.Device, bind bool) error {
dest := filepath.Join(rootfs, node.Path)
if err := os.MkdirAll(filepath.Dir(dest), 0755); err != nil {
return err
}
if bind {
return bindMountDeviceNode(dest, node)
}
if err := mknodDevice(dest, node); err != nil {
if os.IsExist(err) {
return nil
} else if os.IsPermission(err) {
return bindMountDeviceNode(dest, node)
}
return err
}
return nil
}
func mknodDevice(dest string, node *configs.Device) error {
fileMode := node.FileMode
switch node.Type {
case 'c':
fileMode |= syscall.S_IFCHR
case 'b':
fileMode |= syscall.S_IFBLK
default:
return fmt.Errorf("%c is not a valid device type for device %s", node.Type, node.Path)
}
if err := syscall.Mknod(dest, uint32(fileMode), node.Mkdev()); err != nil {
return err
}
return syscall.Chown(dest, int(node.Uid), int(node.Gid))
}
func getMountInfo(mountinfo []*mount.Info, dir string) *mount.Info {
for _, m := range mountinfo {
if m.Mountpoint == dir {
return m
}
}
return nil
}
// Get the parent mount point of directory passed in as argument. Also return
// optional fields.
func getParentMount(rootfs string) (string, string, error) {
var path string
mountinfos, err := mount.GetMounts()
if err != nil {
return "", "", err
}
mountinfo := getMountInfo(mountinfos, rootfs)
if mountinfo != nil {
return rootfs, mountinfo.Optional, nil
}
path = rootfs
for {
path = filepath.Dir(path)
mountinfo = getMountInfo(mountinfos, path)
if mountinfo != nil {
return path, mountinfo.Optional, nil
}
if path == "/" {
break
}
}
// If we are here, we did not find parent mount. Something is wrong.
return "", "", fmt.Errorf("Could not find parent mount of %s", rootfs)
}
// Make parent mount private if it was shared
func rootfsParentMountPrivate(config *configs.Config) error {
sharedMount := false
parentMount, optionalOpts, err := getParentMount(config.Rootfs)
if err != nil {
return err
}
optsSplit := strings.Split(optionalOpts, " ")
for _, opt := range optsSplit {
if strings.HasPrefix(opt, "shared:") {
sharedMount = true
break
}
}
// Make parent mount PRIVATE if it was shared. It is needed for two
// reasons. First of all pivot_root() will fail if parent mount is
// shared. Secondly when we bind mount rootfs it will propagate to
// parent namespace and we don't want that to happen.
if sharedMount {
return syscall.Mount("", parentMount, "", syscall.MS_PRIVATE, "")
}
return nil
}
func prepareRoot(config *configs.Config) error {
flag := syscall.MS_SLAVE | syscall.MS_REC
if config.RootPropagation != 0 {
flag = config.RootPropagation
}
if err := syscall.Mount("", "/", "", uintptr(flag), ""); err != nil {
return err
}
if err := rootfsParentMountPrivate(config); err != nil {
return err
}
return syscall.Mount(config.Rootfs, config.Rootfs, "bind", syscall.MS_BIND|syscall.MS_REC, "")
}
func setReadonly() error {
return syscall.Mount("/", "/", "bind", syscall.MS_BIND|syscall.MS_REMOUNT|syscall.MS_RDONLY|syscall.MS_REC, "")
}
func setupPtmx(config *configs.Config, console *linuxConsole) error {
ptmx := filepath.Join(config.Rootfs, "dev/ptmx")
if err := os.Remove(ptmx); err != nil && !os.IsNotExist(err) {
return err
}
if err := os.Symlink("pts/ptmx", ptmx); err != nil {
return fmt.Errorf("symlink dev ptmx %s", err)
}
if console != nil {
return console.mount(config.Rootfs, config.MountLabel)
}
return nil
}
func pivotRoot(rootfs, pivotBaseDir string) error {
if pivotBaseDir == "" {
pivotBaseDir = "/"
}
tmpDir := filepath.Join(rootfs, pivotBaseDir)
if err := os.MkdirAll(tmpDir, 0755); err != nil {
return fmt.Errorf("can't create tmp dir %s, error %v", tmpDir, err)
}
pivotDir, err := ioutil.TempDir(tmpDir, ".pivot_root")
if err != nil {
return fmt.Errorf("can't create pivot_root dir %s, error %v", pivotDir, err)
}
if err := syscall.PivotRoot(rootfs, pivotDir); err != nil {
return fmt.Errorf("pivot_root %s", err)
}
if err := syscall.Chdir("/"); err != nil {
return fmt.Errorf("chdir / %s", err)
}
// path to pivot dir now changed, update
pivotDir = filepath.Join(pivotBaseDir, filepath.Base(pivotDir))
// Make pivotDir rprivate to make sure any of the unmounts don't
// propagate to parent.
if err := syscall.Mount("", pivotDir, "", syscall.MS_PRIVATE|syscall.MS_REC, ""); err != nil {
return err
}
if err := syscall.Unmount(pivotDir, syscall.MNT_DETACH); err != nil {
return fmt.Errorf("unmount pivot_root dir %s", err)
}
return os.Remove(pivotDir)
}
func msMoveRoot(rootfs string) error {
if err := syscall.Mount(rootfs, "/", "", syscall.MS_MOVE, ""); err != nil {
return err
}
if err := syscall.Chroot("."); err != nil {
return err
}
return syscall.Chdir("/")
}
// createIfNotExists creates a file or a directory only if it does not already exist.
func createIfNotExists(path string, isDir bool) error {
if _, err := os.Stat(path); err != nil {
if os.IsNotExist(err) {
if isDir {
return os.MkdirAll(path, 0755)
}
if err := os.MkdirAll(filepath.Dir(path), 0755); err != nil {
return err
}
f, err := os.OpenFile(path, os.O_CREATE, 0755)
if err != nil {
return err
}
f.Close()
}
}
return nil
}
// remountReadonly will bind over the top of an existing path and ensure that it is read-only.
func remountReadonly(path string) error {
for i := 0; i < 5; i++ {
if err := syscall.Mount("", path, "", syscall.MS_REMOUNT|syscall.MS_RDONLY, ""); err != nil && !os.IsNotExist(err) {
switch err {
case syscall.EINVAL:
// Probably not a mountpoint, use bind-mount
if err := syscall.Mount(path, path, "", syscall.MS_BIND, ""); err != nil {
return err
}
return syscall.Mount(path, path, "", syscall.MS_BIND|syscall.MS_REMOUNT|syscall.MS_RDONLY|syscall.MS_REC|defaultMountFlags, "")
case syscall.EBUSY:
time.Sleep(100 * time.Millisecond)
continue
default:
return err
}
}
return nil
}
return fmt.Errorf("unable to mount %s as readonly max retries reached", path)
}
// maskFile bind mounts /dev/null over the top of the specified path inside a container
// to avoid security issues from processes reading information from non-namespace aware mounts ( proc/kcore ).
func maskFile(path string) error {
if err := syscall.Mount("/dev/null", path, "", syscall.MS_BIND, ""); err != nil && !os.IsNotExist(err) {
return err
}
return nil
}
// writeSystemProperty writes the value to a path under /proc/sys as determined from the key.
// For e.g. net.ipv4.ip_forward translated to /proc/sys/net/ipv4/ip_forward.
func writeSystemProperty(key, value string) error {
keyPath := strings.Replace(key, ".", "/", -1)
return ioutil.WriteFile(path.Join("/proc/sys", keyPath), []byte(value), 0644)
}
func remount(m *configs.Mount, rootfs string) error {
var (
dest = m.Destination
)
if !strings.HasPrefix(dest, rootfs) {
dest = filepath.Join(rootfs, dest)
}
if err := syscall.Mount(m.Source, dest, m.Device, uintptr(m.Flags|syscall.MS_REMOUNT), ""); err != nil {
return err
}
return nil
}
// Do the mount operation followed by additional mounts required to take care
// of propagation flags.
func mountPropagate(m *configs.Mount, rootfs string, mountLabel string) error {
var (
dest = m.Destination
data = label.FormatMountLabel(m.Data, mountLabel)
)
if !strings.HasPrefix(dest, rootfs) {
dest = filepath.Join(rootfs, dest)
}
if err := syscall.Mount(m.Source, dest, m.Device, uintptr(m.Flags), data); err != nil {
return err
}
for _, pflag := range m.PropagationFlags {
if err := syscall.Mount("", dest, "", uintptr(pflag), ""); err != nil {
return err
}
}
return nil
}

71
vendor/src/github.com/opencontainers/runc/libcontainer/seccomp/config.go vendored Normal file
View file

@ -0,0 +1,71 @@
package seccomp
import (
"fmt"
"github.com/opencontainers/runc/libcontainer/configs"
)
var operators = map[string]configs.Operator{
"SCMP_CMP_NE": configs.NotEqualTo,
"SCMP_CMP_LT": configs.LessThan,
"SCMP_CMP_LE": configs.LessThanOrEqualTo,
"SCMP_CMP_EQ": configs.EqualTo,
"SCMP_CMP_GE": configs.GreaterThanOrEqualTo,
"SCMP_CMP_GT": configs.GreaterThan,
"SCMP_CMP_MASKED_EQ": configs.MaskEqualTo,
}
var actions = map[string]configs.Action{
"SCMP_ACT_KILL": configs.Kill,
"SCMP_ACT_ERRNO": configs.Errno,
"SCMP_ACT_TRAP": configs.Trap,
"SCMP_ACT_ALLOW": configs.Allow,
"SCMP_ACT_TRACE": configs.Trace,
}
var archs = map[string]string{
"SCMP_ARCH_X86": "x86",
"SCMP_ARCH_X86_64": "amd64",
"SCMP_ARCH_X32": "x32",
"SCMP_ARCH_ARM": "arm",
"SCMP_ARCH_AARCH64": "arm64",
"SCMP_ARCH_MIPS": "mips",
"SCMP_ARCH_MIPS64": "mips64",
"SCMP_ARCH_MIPS64N32": "mips64n32",
"SCMP_ARCH_MIPSEL": "mipsel",
"SCMP_ARCH_MIPSEL64": "mipsel64",
"SCMP_ARCH_MIPSEL64N32": "mipsel64n32",
}
// ConvertStringToOperator converts a string into a Seccomp comparison operator.
// Comparison operators use the names they are assigned by Libseccomp's header.
// Attempting to convert a string that is not a valid operator results in an
// error.
func ConvertStringToOperator(in string) (configs.Operator, error) {
if op, ok := operators[in]; ok == true {
return op, nil
}
return 0, fmt.Errorf("string %s is not a valid operator for seccomp", in)
}
// ConvertStringToAction converts a string into a Seccomp rule match action.
// Actions use the names they are assigned in Libseccomp's header, though some
// (notable, SCMP_ACT_TRACE) are not available in this implementation and will
// return errors.
// Attempting to convert a string that is not a valid action results in an
// error.
func ConvertStringToAction(in string) (configs.Action, error) {
if act, ok := actions[in]; ok == true {
return act, nil
}
return 0, fmt.Errorf("string %s is not a valid action for seccomp", in)
}
// ConvertStringToArch converts a string into a Seccomp comparison arch.
func ConvertStringToArch(in string) (string, error) {
if arch, ok := archs[in]; ok == true {
return arch, nil
}
return "", fmt.Errorf("string %s is not a valid arch for seccomp", in)
}

180
vendor/src/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_linux.go vendored Normal file
View file

@ -0,0 +1,180 @@
// +build linux,cgo,seccomp
package seccomp
import (
"fmt"
"log"
"syscall"
"github.com/opencontainers/runc/libcontainer/configs"
libseccomp "github.com/seccomp/libseccomp-golang"
)
var (
actAllow = libseccomp.ActAllow
actTrap = libseccomp.ActTrap
actKill = libseccomp.ActKill
actTrace = libseccomp.ActTrace.SetReturnCode(int16(syscall.EPERM))
actErrno = libseccomp.ActErrno.SetReturnCode(int16(syscall.EPERM))
)
// Filters given syscalls in a container, preventing them from being used
// Started in the container init process, and carried over to all child processes
// Setns calls, however, require a separate invocation, as they are not children
// of the init until they join the namespace
func InitSeccomp(config *configs.Seccomp) error {
if config == nil {
return fmt.Errorf("cannot initialize Seccomp - nil config passed")
}
defaultAction, err := getAction(config.DefaultAction)
if err != nil {
return fmt.Errorf("error initializing seccomp - invalid default action")
}
filter, err := libseccomp.NewFilter(defaultAction)
if err != nil {
return fmt.Errorf("error creating filter: %s", err)
}
// Add extra architectures
for _, arch := range config.Architectures {
scmpArch, err := libseccomp.GetArchFromString(arch)
if err != nil {
return err
}
if err := filter.AddArch(scmpArch); err != nil {
return err
}
}
// Unset no new privs bit
if err := filter.SetNoNewPrivsBit(false); err != nil {
return fmt.Errorf("error setting no new privileges: %s", err)
}
// Add a rule for each syscall
for _, call := range config.Syscalls {
if call == nil {
return fmt.Errorf("encountered nil syscall while initializing Seccomp")
}
if err = matchCall(filter, call); err != nil {
return err
}
}
if err = filter.Load(); err != nil {
return fmt.Errorf("error loading seccomp filter into kernel: %s", err)
}
return nil
}
// Convert Libcontainer Action to Libseccomp ScmpAction
func getAction(act configs.Action) (libseccomp.ScmpAction, error) {
switch act {
case configs.Kill:
return actKill, nil
case configs.Errno:
return actErrno, nil
case configs.Trap:
return actTrap, nil
case configs.Allow:
return actAllow, nil
case configs.Trace:
return actTrace, nil
default:
return libseccomp.ActInvalid, fmt.Errorf("invalid action, cannot use in rule")
}
}
// Convert Libcontainer Operator to Libseccomp ScmpCompareOp
func getOperator(op configs.Operator) (libseccomp.ScmpCompareOp, error) {
switch op {
case configs.EqualTo:
return libseccomp.CompareEqual, nil
case configs.NotEqualTo:
return libseccomp.CompareNotEqual, nil
case configs.GreaterThan:
return libseccomp.CompareGreater, nil
case configs.GreaterThanOrEqualTo:
return libseccomp.CompareGreaterEqual, nil
case configs.LessThan:
return libseccomp.CompareLess, nil
case configs.LessThanOrEqualTo:
return libseccomp.CompareLessOrEqual, nil
case configs.MaskEqualTo:
return libseccomp.CompareMaskedEqual, nil
default:
return libseccomp.CompareInvalid, fmt.Errorf("invalid operator, cannot use in rule")
}
}
// Convert Libcontainer Arg to Libseccomp ScmpCondition
func getCondition(arg *configs.Arg) (libseccomp.ScmpCondition, error) {
cond := libseccomp.ScmpCondition{}
if arg == nil {
return cond, fmt.Errorf("cannot convert nil to syscall condition")
}
op, err := getOperator(arg.Op)
if err != nil {
return cond, err
}
return libseccomp.MakeCondition(arg.Index, op, arg.Value, arg.ValueTwo)
}
// Add a rule to match a single syscall
func matchCall(filter *libseccomp.ScmpFilter, call *configs.Syscall) error {
if call == nil || filter == nil {
return fmt.Errorf("cannot use nil as syscall to block")
}
if len(call.Name) == 0 {
return fmt.Errorf("empty string is not a valid syscall")
}
// If we can't resolve the syscall, assume it's not supported on this kernel
// Ignore it, don't error out
callNum, err := libseccomp.GetSyscallFromName(call.Name)
if err != nil {
log.Printf("Error resolving syscall name %s: %s - ignoring syscall.", call.Name, err)
return nil
}
// Convert the call's action to the libseccomp equivalent
callAct, err := getAction(call.Action)
if err != nil {
return err
}
// Unconditional match - just add the rule
if len(call.Args) == 0 {
if err = filter.AddRule(callNum, callAct); err != nil {
return err
}
} else {
// Conditional match - convert the per-arg rules into library format
conditions := []libseccomp.ScmpCondition{}
for _, cond := range call.Args {
newCond, err := getCondition(cond)
if err != nil {
return err
}
conditions = append(conditions, newCond)
}
if err = filter.AddRuleConditional(callNum, callAct, conditions); err != nil {
return err
}
}
return nil
}

19
vendor/src/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_unsupported.go vendored Normal file
View file

@ -0,0 +1,19 @@
// +build !linux !cgo !seccomp
package seccomp
import (
"errors"
"github.com/opencontainers/runc/libcontainer/configs"
)
var ErrSeccompNotEnabled = errors.New("seccomp: config provided but seccomp not supported")
// Seccomp not supported, do nothing
func InitSeccomp(config *configs.Seccomp) error {
if config != nil {
return ErrSeccompNotEnabled
}
return nil
}

11
vendor/src/github.com/opencontainers/runc/libcontainer/setgroups_linux.go vendored Normal file
View file

@ -0,0 +1,11 @@
// +build linux,go1.5
package libcontainer
import "syscall"
// Set the GidMappingsEnableSetgroups member to true, so the process's
// setgroups proc entry wont be set to 'deny' if GidMappings are set
func enableSetgroups(sys *syscall.SysProcAttr) {
sys.GidMappingsEnableSetgroups = true
}

44
vendor/src/github.com/opencontainers/runc/libcontainer/setns_init_linux.go vendored Normal file
View file

@ -0,0 +1,44 @@
// +build linux
package libcontainer
import (
"os"
"github.com/opencontainers/runc/libcontainer/apparmor"
"github.com/opencontainers/runc/libcontainer/label"
"github.com/opencontainers/runc/libcontainer/seccomp"
"github.com/opencontainers/runc/libcontainer/system"
)
// linuxSetnsInit performs the container's initialization for running a new process
// inside an existing container.
type linuxSetnsInit struct {
config *initConfig
}
func (l *linuxSetnsInit) Init() error {
if err := setupRlimits(l.config.Config); err != nil {
return err
}
if err := setOomScoreAdj(l.config.Config.OomScoreAdj); err != nil {
return err
}
if l.config.Config.Seccomp != nil {
if err := seccomp.InitSeccomp(l.config.Config.Seccomp); err != nil {
return err
}
}
if err := finalizeNamespace(l.config); err != nil {
return err
}
if err := apparmor.ApplyProfile(l.config.Config.AppArmorProfile); err != nil {
return err
}
if l.config.Config.ProcessLabel != "" {
if err := label.SetProcessLabel(l.config.Config.ProcessLabel); err != nil {
return err
}
}
return system.Execv(l.config.Args[0], l.config.Args[0:], os.Environ())
}

27
vendor/src/github.com/opencontainers/runc/libcontainer/stacktrace/capture.go vendored Normal file
View file

@ -0,0 +1,27 @@
package stacktrace
import "runtime"
// Caputure captures a stacktrace for the current calling go program
//
// skip is the number of frames to skip
func Capture(userSkip int) Stacktrace {
var (
skip = userSkip + 1 // add one for our own function
frames []Frame
prevPc uintptr = 0
)
for i := skip; ; i++ {
pc, file, line, ok := runtime.Caller(i)
//detect if caller is repeated to avoid loop, gccgo
//currently runs into a loop without this check
if !ok || pc == prevPc {
break
}
frames = append(frames, NewFrame(pc, file, line))
prevPc = pc
}
return Stacktrace{
Frames: frames,
}
}

38
vendor/src/github.com/opencontainers/runc/libcontainer/stacktrace/frame.go vendored Normal file
View file

@ -0,0 +1,38 @@
package stacktrace
import (
"path/filepath"
"runtime"
"strings"
)
// NewFrame returns a new stack frame for the provided information
func NewFrame(pc uintptr, file string, line int) Frame {
fn := runtime.FuncForPC(pc)
if fn == nil {
return Frame{}
}
pack, name := parseFunctionName(fn.Name())
return Frame{
Line: line,
File: filepath.Base(file),
Package: pack,
Function: name,
}
}
func parseFunctionName(name string) (string, string) {
i := strings.LastIndex(name, ".")
if i == -1 {
return "", name
}
return name[:i], name[i+1:]
}
// Frame contains all the information for a stack frame within a go program
type Frame struct {
File string
Function string
Package string
Line int
}

5
vendor/src/github.com/opencontainers/runc/libcontainer/stacktrace/stacktrace.go vendored Normal file
View file

@ -0,0 +1,5 @@
package stacktrace
type Stacktrace struct {
Frames []Frame
}

113
vendor/src/github.com/opencontainers/runc/libcontainer/standard_init_linux.go vendored Normal file
View file

@ -0,0 +1,113 @@
// +build linux
package libcontainer
import (
"os"
"syscall"
"github.com/opencontainers/runc/libcontainer/apparmor"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/label"
"github.com/opencontainers/runc/libcontainer/seccomp"
"github.com/opencontainers/runc/libcontainer/system"
)
type linuxStandardInit struct {
parentPid int
config *initConfig
}
func (l *linuxStandardInit) Init() error {
// join any namespaces via a path to the namespace fd if provided
if err := joinExistingNamespaces(l.config.Config.Namespaces); err != nil {
return err
}
var console *linuxConsole
if l.config.Console != "" {
console = newConsoleFromPath(l.config.Console)
if err := console.dupStdio(); err != nil {
return err
}
}
if _, err := syscall.Setsid(); err != nil {
return err
}
if console != nil {
if err := system.Setctty(); err != nil {
return err
}
}
if err := setupNetwork(l.config); err != nil {
return err
}
if err := setupRoute(l.config.Config); err != nil {
return err
}
if err := setupRlimits(l.config.Config); err != nil {
return err
}
if err := setOomScoreAdj(l.config.Config.OomScoreAdj); err != nil {
return err
}
label.Init()
// InitializeMountNamespace() can be executed only for a new mount namespace
if l.config.Config.Namespaces.Contains(configs.NEWNS) {
if err := setupRootfs(l.config.Config, console); err != nil {
return err
}
}
if hostname := l.config.Config.Hostname; hostname != "" {
if err := syscall.Sethostname([]byte(hostname)); err != nil {
return err
}
}
if err := apparmor.ApplyProfile(l.config.Config.AppArmorProfile); err != nil {
return err
}
if err := label.SetProcessLabel(l.config.Config.ProcessLabel); err != nil {
return err
}
for key, value := range l.config.Config.Sysctl {
if err := writeSystemProperty(key, value); err != nil {
return err
}
}
for _, path := range l.config.Config.ReadonlyPaths {
if err := remountReadonly(path); err != nil {
return err
}
}
for _, path := range l.config.Config.MaskPaths {
if err := maskFile(path); err != nil {
return err
}
}
pdeath, err := system.GetParentDeathSignal()
if err != nil {
return err
}
if l.config.Config.Seccomp != nil {
if err := seccomp.InitSeccomp(l.config.Config.Seccomp); err != nil {
return err
}
}
if err := finalizeNamespace(l.config); err != nil {
return err
}
// finalizeNamespace can change user/group which clears the parent death
// signal, so we restore it here.
if err := pdeath.Restore(); err != nil {
return err
}
// compare the parent from the inital start of the init process and make sure that it did not change.
// if the parent changes that means it died and we were reparened to something else so we should
// just kill ourself and not cause problems for someone else.
if syscall.Getppid() != l.parentPid {
return syscall.Kill(syscall.Getpid(), syscall.SIGKILL)
}
return system.Execv(l.config.Args[0], l.config.Args[0:], os.Environ())
}

217
vendor/src/github.com/opencontainers/runc/libcontainer/state_linux.go vendored Normal file
View file

@ -0,0 +1,217 @@
// +build linux
package libcontainer
import (
"fmt"
"os"
"path/filepath"
"github.com/Sirupsen/logrus"
"github.com/opencontainers/runc/libcontainer/configs"
)
func newStateTransitionError(from, to containerState) error {
return &stateTransitionError{
From: from.status().String(),
To: to.status().String(),
}
}
// stateTransitionError is returned when an invalid state transition happens from one
// state to another.
type stateTransitionError struct {
From string
To string
}
func (s *stateTransitionError) Error() string {
return fmt.Sprintf("invalid state transition from %s to %s", s.From, s.To)
}
type containerState interface {
transition(containerState) error
destroy() error
status() Status
}
func destroy(c *linuxContainer) error {
if !c.config.Namespaces.Contains(configs.NEWPID) {
if err := killCgroupProcesses(c.cgroupManager); err != nil {
logrus.Warn(err)
}
}
err := c.cgroupManager.Destroy()
if rerr := os.RemoveAll(c.root); err == nil {
err = rerr
}
c.initProcess = nil
if herr := runPoststopHooks(c); err == nil {
err = herr
}
return err
}
func runPoststopHooks(c *linuxContainer) error {
if c.config.Hooks != nil {
s := configs.HookState{
Version: c.config.Version,
ID: c.id,
Root: c.config.Rootfs,
}
for _, hook := range c.config.Hooks.Poststop {
if err := hook.Run(s); err != nil {
return err
}
}
}
return nil
}
// stoppedState represents a container is a stopped/destroyed state.
type stoppedState struct {
c *linuxContainer
}
func (b *stoppedState) status() Status {
return Destroyed
}
func (b *stoppedState) transition(s containerState) error {
switch s.(type) {
case *runningState:
b.c.state = s
return nil
case *restoredState:
b.c.state = s
return nil
case *stoppedState:
return nil
}
return newStateTransitionError(b, s)
}
func (b *stoppedState) destroy() error {
return destroy(b.c)
}
// runningState represents a container that is currently running.
type runningState struct {
c *linuxContainer
}
func (r *runningState) status() Status {
return Running
}
func (r *runningState) transition(s containerState) error {
switch s.(type) {
case *stoppedState:
running, err := r.c.isRunning()
if err != nil {
return err
}
if running {
return newGenericError(fmt.Errorf("container still running"), ContainerNotStopped)
}
r.c.state = s
return nil
case *pausedState:
r.c.state = s
return nil
case *runningState, *nullState:
return nil
}
return newStateTransitionError(r, s)
}
func (r *runningState) destroy() error {
running, err := r.c.isRunning()
if err != nil {
return err
}
if running {
return newGenericError(fmt.Errorf("container is not destroyed"), ContainerNotStopped)
}
return destroy(r.c)
}
// pausedState represents a container that is currently pause. It cannot be destroyed in a
// paused state and must transition back to running first.
type pausedState struct {
c *linuxContainer
}
func (p *pausedState) status() Status {
return Paused
}
func (p *pausedState) transition(s containerState) error {
switch s.(type) {
case *runningState:
p.c.state = s
return nil
case *pausedState:
return nil
}
return newStateTransitionError(p, s)
}
func (p *pausedState) destroy() error {
return newGenericError(fmt.Errorf("container is paused"), ContainerPaused)
}
// restoredState is the same as the running state but also has accociated checkpoint
// information that maybe need destroyed when the container is stopped and destory is called.
type restoredState struct {
imageDir string
c *linuxContainer
}
func (r *restoredState) status() Status {
return Running
}
func (r *restoredState) transition(s containerState) error {
switch s.(type) {
case *stoppedState:
return nil
case *runningState:
return nil
}
return newStateTransitionError(r, s)
}
func (r *restoredState) destroy() error {
if _, err := os.Stat(filepath.Join(r.c.root, "checkpoint")); err != nil {
if !os.IsNotExist(err) {
return err
}
}
return destroy(r.c)
}
// nullState is used whenever a container is restored, loaded, or setting additional
// processes inside and it should not be destroyed when it is exiting.
type nullState struct {
c *linuxContainer
s Status
}
func (n *nullState) status() Status {
return n.s
}
func (n *nullState) transition(s containerState) error {
switch s.(type) {
case *restoredState:
n.c.state = s
default:
// do nothing for null states
}
return nil
}
func (n *nullState) destroy() error {
return nil
}

15
vendor/src/github.com/opencontainers/runc/libcontainer/stats.go vendored Normal file
View file

@ -0,0 +1,15 @@
package libcontainer
type NetworkInterface struct {
// Name is the name of the network interface.
Name string
RxBytes uint64
RxPackets uint64
RxErrors uint64
RxDropped uint64
TxBytes uint64
TxPackets uint64
TxErrors uint64
TxDropped uint64
}

5
vendor/src/github.com/opencontainers/runc/libcontainer/stats_freebsd.go vendored Normal file
View file

@ -0,0 +1,5 @@
package libcontainer
type Stats struct {
Interfaces []*NetworkInterface
}

8
vendor/src/github.com/opencontainers/runc/libcontainer/stats_linux.go vendored Normal file
View file

@ -0,0 +1,8 @@
package libcontainer
import "github.com/opencontainers/runc/libcontainer/cgroups"
type Stats struct {
Interfaces []*NetworkInterface
CgroupStats *cgroups.Stats
}

5
vendor/src/github.com/opencontainers/runc/libcontainer/stats_windows.go vendored Normal file
View file

@ -0,0 +1,5 @@
package libcontainer
type Stats struct {
Interfaces []*NetworkInterface
}

77
vendor/src/github.com/opencontainers/runc/libcontainer/system/linux.go vendored Normal file
View file

@ -0,0 +1,77 @@
// +build linux
package system
import (
"os/exec"
"syscall"
"unsafe"
)
type ParentDeathSignal int
func (p ParentDeathSignal) Restore() error {
if p == 0 {
return nil
}
current, err := GetParentDeathSignal()
if err != nil {
return err
}
if p == current {
return nil
}
return p.Set()
}
func (p ParentDeathSignal) Set() error {
return SetParentDeathSignal(uintptr(p))
}
func Execv(cmd string, args []string, env []string) error {
name, err := exec.LookPath(cmd)
if err != nil {
return err
}
return syscall.Exec(name, args, env)
}
func SetParentDeathSignal(sig uintptr) error {
if _, _, err := syscall.RawSyscall(syscall.SYS_PRCTL, syscall.PR_SET_PDEATHSIG, sig, 0); err != 0 {
return err
}
return nil
}
func GetParentDeathSignal() (ParentDeathSignal, error) {
var sig int
_, _, err := syscall.RawSyscall(syscall.SYS_PRCTL, syscall.PR_GET_PDEATHSIG, uintptr(unsafe.Pointer(&sig)), 0)
if err != 0 {
return -1, err
}
return ParentDeathSignal(sig), nil
}
func SetKeepCaps() error {
if _, _, err := syscall.RawSyscall(syscall.SYS_PRCTL, syscall.PR_SET_KEEPCAPS, 1, 0); err != 0 {
return err
}
return nil
}
func ClearKeepCaps() error {
if _, _, err := syscall.RawSyscall(syscall.SYS_PRCTL, syscall.PR_SET_KEEPCAPS, 0, 0); err != 0 {
return err
}
return nil
}
func Setctty() error {
if _, _, err := syscall.RawSyscall(syscall.SYS_IOCTL, 0, uintptr(syscall.TIOCSCTTY), 0); err != 0 {
return err
}
return nil
}

27
vendor/src/github.com/opencontainers/runc/libcontainer/system/proc.go vendored Normal file
View file

@ -0,0 +1,27 @@
package system
import (
"io/ioutil"
"path/filepath"
"strconv"
"strings"
)
// look in /proc to find the process start time so that we can verify
// that this pid has started after ourself
func GetProcessStartTime(pid int) (string, error) {
data, err := ioutil.ReadFile(filepath.Join("/proc", strconv.Itoa(pid), "stat"))
if err != nil {
return "", err
}
parts := strings.Split(string(data), " ")
// the starttime is located at pos 22
// from the man page
//
// starttime %llu (was %lu before Linux 2.6)
// (22) The time the process started after system boot. In kernels before Linux 2.6, this
// value was expressed in jiffies. Since Linux 2.6, the value is expressed in clock ticks
// (divide by sysconf(_SC_CLK_TCK)).
return parts[22-1], nil // starts at 1
}

40
vendor/src/github.com/opencontainers/runc/libcontainer/system/setns_linux.go vendored Normal file
View file

@ -0,0 +1,40 @@
package system
import (
"fmt"
"runtime"
"syscall"
)
// Via http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=7b21fddd087678a70ad64afc0f632e0f1071b092
//
// We need different setns values for the different platforms and arch
// We are declaring the macro here because the SETNS syscall does not exist in th stdlib
var setNsMap = map[string]uintptr{
"linux/386": 346,
"linux/arm64": 268,
"linux/amd64": 308,
"linux/arm": 375,
"linux/ppc": 350,
"linux/ppc64": 350,
"linux/ppc64le": 350,
"linux/s390x": 339,
}
var sysSetns = setNsMap[fmt.Sprintf("%s/%s", runtime.GOOS, runtime.GOARCH)]
func SysSetns() uint32 {
return uint32(sysSetns)
}
func Setns(fd uintptr, flags uintptr) error {
ns, exists := setNsMap[fmt.Sprintf("%s/%s", runtime.GOOS, runtime.GOARCH)]
if !exists {
return fmt.Errorf("unsupported platform %s/%s", runtime.GOOS, runtime.GOARCH)
}
_, _, err := syscall.RawSyscall(ns, fd, flags, 0)
if err != 0 {
return err
}
return nil
}

25
vendor/src/github.com/opencontainers/runc/libcontainer/system/syscall_linux_386.go vendored Normal file
View file

@ -0,0 +1,25 @@
// +build linux,386
package system
import (
"syscall"
)
// Setuid sets the uid of the calling thread to the specified uid.
func Setuid(uid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETUID, uintptr(uid), 0, 0)
if e1 != 0 {
err = e1
}
return
}
// Setgid sets the gid of the calling thread to the specified gid.
func Setgid(gid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETGID32, uintptr(gid), 0, 0)
if e1 != 0 {
err = e1
}
return
}

25
vendor/src/github.com/opencontainers/runc/libcontainer/system/syscall_linux_64.go vendored Normal file
View file

@ -0,0 +1,25 @@
// +build linux,arm64 linux,amd64 linux,ppc linux,ppc64 linux,ppc64le linux,s390x
package system
import (
"syscall"
)
// Setuid sets the uid of the calling thread to the specified uid.
func Setuid(uid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETUID, uintptr(uid), 0, 0)
if e1 != 0 {
err = e1
}
return
}
// Setgid sets the gid of the calling thread to the specified gid.
func Setgid(gid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETGID, uintptr(gid), 0, 0)
if e1 != 0 {
err = e1
}
return
}

25
vendor/src/github.com/opencontainers/runc/libcontainer/system/syscall_linux_arm.go vendored Normal file
View file

@ -0,0 +1,25 @@
// +build linux,arm
package system
import (
"syscall"
)
// Setuid sets the uid of the calling thread to the specified uid.
func Setuid(uid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETUID32, uintptr(uid), 0, 0)
if e1 != 0 {
err = e1
}
return
}
// Setgid sets the gid of the calling thread to the specified gid.
func Setgid(gid int) (err error) {
_, _, e1 := syscall.RawSyscall(syscall.SYS_SETGID32, uintptr(gid), 0, 0)
if e1 != 0 {
err = e1
}
return
}

12
vendor/src/github.com/opencontainers/runc/libcontainer/system/sysconfig.go vendored Normal file
View file

@ -0,0 +1,12 @@
// +build cgo,linux cgo,freebsd
package system
/*
#include <unistd.h>
*/
import "C"
func GetClockTicks() int {
return int(C.sysconf(C._SC_CLK_TCK))
}

15
vendor/src/github.com/opencontainers/runc/libcontainer/system/sysconfig_notcgo.go vendored Normal file
View file

@ -0,0 +1,15 @@
// +build !cgo windows
package system
func GetClockTicks() int {
// TODO figure out a better alternative for platforms where we're missing cgo
//
// TODO Windows. This could be implemented using Win32 QueryPerformanceFrequency().
// https://msdn.microsoft.com/en-us/library/windows/desktop/ms644905(v=vs.85).aspx
//
// An example of its usage can be found here.
// https://msdn.microsoft.com/en-us/library/windows/desktop/dn553408(v=vs.85).aspx
return 100
}

Some files were not shown because too many files have changed in this diff Show more