DevFW-CICD/ingress-nginx-helm
6
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Switch to go modules

Browse source
This commit is contained in:
Manuel Alejandro de Brito Fontes 2019-04-15 08:34:23 -04:00
parent 461954facb
commit 1720059244
No known key found for this signature in database
GPG key ID: 786136016A8BA02A
763 changed files with 24896 additions and 177398 deletions
Download patch file Download diff file Expand all files Collapse all files

241
vendor/github.com/opencontainers/runc/libcontainer/README.md generated vendored
View file

@ -1,241 +0,0 @@
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
Because containers are spawned in a two step process you will need a binary that
will be executed as the init process for the container. In libcontainer, we use
the current binary (/proc/self/exe) to be executed as the init process, and use
arg "init", we call the first step process "bootstrap", so you always need a "init"
function as the entry of "bootstrap".
```go
func init() {
if len(os.Args) > 1 && os.Args[1] == "init" {
runtime.GOMAXPROCS(1)
runtime.LockOSThread()
factory, _ := libcontainer.New("")
if err := factory.StartInitialization(); err != nil {
logrus.Fatal(err)
}
panic("--this line should have never been executed, congratulations--")
}
}
```
Then to create a container you first have to initialize an instance of a factory
that will handle the creation and initialization for a container.
```go
factory, err := libcontainer.New("/var/lib/container", libcontainer.Cgroupfs, libcontainer.InitArgs(os.Args[0], "init"))
if err != nil {
logrus.Fatal(err)
return
}
```
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
defaultMountFlags := syscall.MS_NOEXEC | syscall.MS_NOSUID | syscall.MS_NODEV
config := &configs.Config{
Rootfs: "/your/path/to/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.NEWUSER},
{Type: configs.NEWNET},
}),
Cgroups: &configs.Cgroup{
Name: "test-container",
Parent: "system",
Resources: &configs.Resources{
MemorySwappiness: nil,
AllowAllDevices: false,
AllowedDevices: configs.DefaultAllowedDevices,
},
},
MaskPaths: []string{
"/proc/kcore",
},
ReadonlyPaths: []string{
"/proc/sys", "/proc/sysrq-trigger", "/proc/irq", "/proc/bus",
},
Devices: configs.DefaultAutoCreatedDevices,
Hostname: "testing",
Mounts: []*configs.Mount{
{
Source: "proc",
Destination: "/proc",
Device: "proc",
Flags: defaultMountFlags,
},
{
Source: "tmpfs",
Destination: "/dev",
Device: "tmpfs",
Flags: syscall.MS_NOSUID | syscall.MS_STRICTATIME,
Data: "mode=755",
},
{
Source: "devpts",
Destination: "/dev/pts",
Device: "devpts",
Flags: syscall.MS_NOSUID | syscall.MS_NOEXEC,
Data: "newinstance,ptmxmode=0666,mode=0620,gid=5",
},
{
Device: "tmpfs",
Source: "shm",
Destination: "/dev/shm",
Data: "mode=1777,size=65536k",
Flags: defaultMountFlags,
},
{
Source: "mqueue",
Destination: "/dev/mqueue",
Device: "mqueue",
Flags: defaultMountFlags,
},
{
Source: "sysfs",
Destination: "/sys",
Device: "sysfs",
Flags: defaultMountFlags | syscall.MS_RDONLY,
},
},
UidMappings: []configs.IDMap{
{
ContainerID: 0,
HostID: 1000,
Size: 65536,
},
},
GidMappings: []configs.IDMap{
{
ContainerID: 0,
HostID: 1000,
Size: 65536,
},
},
Networks: []*configs.Network{
{
Type: "loopback",
Address: "127.0.0.1/0",
Gateway: "localhost",
},
},
Rlimits: []configs.Rlimit{
{
Type: syscall.RLIMIT_NOFILE,
Hard: uint64(1025),
Soft: uint64(1025),
},
},
}
```
Once you have the configuration populated you can create a container:
```go
container, err := factory.Create("container-id", config)
if err != nil {
logrus.Fatal(err)
return
}
```
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 {
logrus.Fatal(err)
container.Destroy()
return
}
// wait for the process to finish.
_, err := process.Wait()
if err != nil {
logrus.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()
// send signal to container's init process.
container.Signal(signal)
```
#### 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.

335
vendor/github.com/opencontainers/runc/libcontainer/SPEC.md generated vendored
View file

@ -1,335 +0,0 @@
## 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 |
| pids | 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_SUSPEND | 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.

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

@ -1,69 +0,0 @@
// +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)
}

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

@ -1,10 +0,0 @@
// +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) {
}

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

@ -1,15 +0,0 @@
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/github.com/opencontainers/runc/libcontainer/console_freebsd.go generated vendored
View file

@ -1,13 +0,0 @@
// +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/github.com/opencontainers/runc/libcontainer/console_linux.go generated vendored
View file

@ -1,145 +0,0 @@
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/github.com/opencontainers/runc/libcontainer/console_windows.go generated vendored
View file

@ -1,30 +0,0 @@
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
}

144
vendor/github.com/opencontainers/runc/libcontainer/container.go generated vendored
View file

@ -1,144 +0,0 @@
// 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"
"time"
"github.com/opencontainers/runc/libcontainer/configs"
)
// The status of a container.
type Status int
const (
// The container exists but has not been run yet
Created Status = iota
// The container exists and is running.
Running
// The container exists, it is in the process of being paused.
Pausing
// The container exists, but all its processes are paused.
Paused
// The container does not exist.
Destroyed
)
func (s Status) String() string {
switch s {
case Created:
return "created"
case Running:
return "running"
case Pausing:
return "pausing"
case Paused:
return "paused"
case Destroyed:
return "destroyed"
default:
return "unknown"
}
}
// 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 in clock cycles since boot time.
InitProcessStartTime string `json:"init_process_start"`
// Created is the unix timestamp for the creation time of the container in UTC
Created time.Time `json:"created"`
// 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
}

1228
vendor/github.com/opencontainers/runc/libcontainer/container_linux.go generated vendored
View file

File diff suppressed because it is too large Load diff

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

@ -1,20 +0,0 @@
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
}

37
vendor/github.com/opencontainers/runc/libcontainer/criu_opts_unix.go generated vendored
View file

@ -1,37 +0,0 @@
// +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
EmptyNs uint32 // don't c/r properties for namespace from this mask
}

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

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

70
vendor/github.com/opencontainers/runc/libcontainer/error.go generated vendored
View file

@ -1,70 +0,0 @@
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
ContainerNotPaused
// Process errors
NoProcessOps
// 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"
case ContainerNotPaused:
return "Container is not paused"
case NoProcessOps:
return "No process operations"
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/github.com/opencontainers/runc/libcontainer/factory.go generated vendored
View file

@ -1,45 +0,0 @@
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
}

290
vendor/github.com/opencontainers/runc/libcontainer/factory_linux.go generated vendored
View file

@ -1,290 +0,0 @@
// +build linux
package libcontainer
import (
"encoding/json"
"fmt"
"os"
"os/exec"
"path/filepath"
"regexp"
"runtime/debug"
"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"
"github.com/opencontainers/runc/libcontainer/utils"
)
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,
created: state.Created,
}
c.state = &createdState{c: c, s: Created}
if err := c.refreshState(); err != nil {
return nil, err
}
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()
var i initer
defer func() {
// We have an error during the initialization of the container's init,
// send it back to the parent process in the form of an initError.
// If container's init successed, syscall.Exec will not return, hence
// this defer function will never be called.
if _, ok := i.(*linuxStandardInit); ok {
// Synchronisation only necessary for standard init.
if err := utils.WriteJSON(pipe, syncT{procError}); err != nil {
panic(err)
}
}
if err := utils.WriteJSON(pipe, newSystemError(err)); err != nil {
panic(err)
}
// ensure that this pipe is always closed
pipe.Close()
}()
defer func() {
if e := recover(); e != nil {
err = fmt.Errorf("panic from initialization: %v, %v", e, string(debug.Stack()))
}
}()
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
}

87
vendor/github.com/opencontainers/runc/libcontainer/generic_error.go generated vendored
View file

@ -1,87 +0,0 @@
package libcontainer
import (
"io"
"text/template"
"time"
"github.com/opencontainers/runc/libcontainer/stacktrace"
)
type syncType uint8
const (
procReady syncType = iota
procError
procRun
procHooks
procResume
)
type syncT struct {
Type syncType `json:"type"`
}
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 e.Message
}
func (e *genericError) Code() ErrorCode {
return e.ECode
}
func (e *genericError) Detail(w io.Writer) error {
return errorTemplate.Execute(w, e)
}

367
vendor/github.com/opencontainers/runc/libcontainer/init_linux.go generated vendored
View file

@ -1,367 +0,0 @@
// +build linux
package libcontainer
import (
"encoding/json"
"fmt"
"io"
"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"`
ProcessLabel string `json:"process_label"`
AppArmorProfile string `json:"apparmor_profile"`
NoNewPrivileges bool `json:"no_new_privileges"`
User string `json:"user"`
Config *configs.Config `json:"config"`
Console string `json:"console"`
Networks []*network `json:"network"`
PassedFilesCount int `json:"passed_files_count"`
ContainerId string `json:"containerid"`
Rlimits []configs.Rlimit `json:"rlimits"`
}
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{
pipe: pipe,
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
}
// syncParentReady sends to the given pipe a JSON payload which indicates that
// the init is ready to Exec the child process. It then waits for the parent to
// indicate that it is cleared to Exec.
func syncParentReady(pipe io.ReadWriter) error {
// Tell parent.
if err := utils.WriteJSON(pipe, syncT{procReady}); err != nil {
return err
}
// Wait for parent to give the all-clear.
var procSync syncT
if err := json.NewDecoder(pipe).Decode(&procSync); err != nil {
if err == io.EOF {
return fmt.Errorf("parent closed synchronisation channel")
}
if procSync.Type != procRun {
return fmt.Errorf("invalid synchronisation flag from parent")
}
}
return nil
}
// syncParentHooks sends to the given pipe a JSON payload which indicates that
// the parent should execute pre-start hooks. It then waits for the parent to
// indicate that it is cleared to resume.
func syncParentHooks(pipe io.ReadWriter) error {
// Tell parent.
if err := utils.WriteJSON(pipe, syncT{procHooks}); err != nil {
return err
}
// Wait for parent to give the all-clear.
var procSync syncT
if err := json.NewDecoder(pipe).Decode(&procSync); err != nil {
if err == io.EOF {
return fmt.Errorf("parent closed synchronisation channel")
}
if procSync.Type != procResume {
return fmt.Errorf("invalid synchronisation flag from parent")
}
}
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(limits []configs.Rlimit, pid int) error {
for _, rlimit := range limits {
if err := system.Prlimit(pid, rlimit.Type, syscall.Rlimit{Max: rlimit.Hard, Cur: rlimit.Soft}); err != nil {
return fmt.Errorf("error setting rlimit type %v: %v", rlimit.Type, err)
}
}
return nil
}
func setOomScoreAdj(oomScoreAdj int, pid int) error {
path := fmt.Sprintf("/proc/%d/oom_score_adj", pid)
return ioutil.WriteFile(path, []byte(strconv.Itoa(oomScoreAdj)), 0600)
}
// 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.GetAllPids()
if err != nil {
m.Freeze(configs.Thawed)
return err
}
for _, pid := range pids {
p, err := os.FindProcess(pid)
if err != nil {
logrus.Warn(err)
continue
}
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
}

88
vendor/github.com/opencontainers/runc/libcontainer/message_linux.go generated vendored
View file

@ -1,88 +0,0 @@
// +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
CloneFlagsAttr uint16 = 27281
ConsolePathAttr uint16 = 27282
NsPathsAttr uint16 = 27283
UidmapAttr uint16 = 27284
GidmapAttr uint16 = 27285
SetgroupAttr uint16 = 27286
// 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
}
type Boolmsg struct {
Type uint16
Value bool
}
func (msg *Boolmsg) 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)
if msg.Value {
buf[4] = 1
} else {
buf[4] = 0
}
return buf
}
func (msg *Boolmsg) Len() int {
return syscall_NLA_HDRLEN + 1
}

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

@ -1,259 +0,0 @@
// +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
}

89
vendor/github.com/opencontainers/runc/libcontainer/notify_linux.go generated vendored
View file

@ -1,89 +0,0 @@
// +build linux
package libcontainer
import (
"fmt"
"io/ioutil"
"os"
"path/filepath"
"syscall"
)
const oomCgroupName = "memory"
type PressureLevel uint
const (
LowPressure PressureLevel = iota
MediumPressure
CriticalPressure
)
func registerMemoryEvent(cgDir string, evName string, arg string) (<-chan struct{}, error) {
evFile, err := os.Open(filepath.Join(cgDir, evName))
if err != nil {
return nil, err
}
fd, _, syserr := syscall.RawSyscall(syscall.SYS_EVENTFD2, 0, syscall.FD_CLOEXEC, 0)
if syserr != 0 {
evFile.Close()
return nil, syserr
}
eventfd := os.NewFile(fd, "eventfd")
eventControlPath := filepath.Join(cgDir, "cgroup.event_control")
data := fmt.Sprintf("%d %d %s", eventfd.Fd(), evFile.Fd(), arg)
if err := ioutil.WriteFile(eventControlPath, []byte(data), 0700); err != nil {
eventfd.Close()
evFile.Close()
return nil, err
}
ch := make(chan struct{})
go func() {
defer func() {
close(ch)
eventfd.Close()
evFile.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
}
// notifyOnOOM returns channel on which you can expect event about OOM,
// if process died without OOM this channel will be closed.
func notifyOnOOM(paths map[string]string) (<-chan struct{}, error) {
dir := paths[oomCgroupName]
if dir == "" {
return nil, fmt.Errorf("path %q missing", oomCgroupName)
}
return registerMemoryEvent(dir, "memory.oom_control", "")
}
func notifyMemoryPressure(paths map[string]string, level PressureLevel) (<-chan struct{}, error) {
dir := paths[oomCgroupName]
if dir == "" {
return nil, fmt.Errorf("path %q missing", oomCgroupName)
}
if level > CriticalPressure {
return nil, fmt.Errorf("invalid pressure level %d", level)
}
levelStr := []string{"low", "medium", "critical"}[level]
return registerMemoryEvent(dir, "memory.pressure_level", levelStr)
}

121
vendor/github.com/opencontainers/runc/libcontainer/process.go generated vendored
View file

@ -1,121 +0,0 @@
package libcontainer
import (
"fmt"
"io"
"math"
"os"
"github.com/opencontainers/runc/libcontainer/configs"
)
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
// AppArmorProfile specifies the profile to apply to the process and is
// changed at the time the process is execed
AppArmorProfile string
// Label specifies the label to apply to the process. It is commonly used by selinux
Label string
// NoNewPrivileges controls whether processes can gain additional privileges.
NoNewPrivileges *bool
// 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 []configs.Rlimit
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"), NoProcessOps)
}
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"), NoProcessOps)
}
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"), NoProcessOps)
}
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
}

487
vendor/github.com/opencontainers/runc/libcontainer/process_linux.go generated vendored
View file

@ -1,487 +0,0 @@
// +build linux
package libcontainer
import (
"encoding/json"
"errors"
"fmt"
"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"
"github.com/opencontainers/runc/libcontainer/utils"
)
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)
}
}
// set oom_score_adj
if err := setOomScoreAdj(p.config.Config.OomScoreAdj, p.pid()); err != nil {
return newSystemError(err)
}
// set rlimits, this has to be done here because we lose permissions
// to raise the limits once we enter a user-namespace
if err := setupRlimits(p.config.Rlimits, p.pid()); err != nil {
return newSystemError(err)
}
if err := utils.WriteJSON(p.parentPipe, 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)
}
// Must be done after Shutdown so the child will exit and we can wait for it.
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
bootstrapData io.Reader
sharePidns bool
}
func (p *initProcess) pid() int {
return p.cmd.Process.Pid
}
func (p *initProcess) externalDescriptors() []string {
return p.fds
}
// 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.
// This is called by initProcess.start function
func (p *initProcess) execSetns() error {
status, err := p.cmd.Process.Wait()
if err != nil {
p.cmd.Wait()
return err
}
if !status.Success() {
p.cmd.Wait()
return &exec.ExitError{ProcessState: status}
}
var pid *pid
if err := json.NewDecoder(p.parentPipe).Decode(&pid); err != nil {
p.cmd.Wait()
return err
}
process, err := os.FindProcess(pid.Pid)
if err != nil {
return err
}
p.cmd.Process = process
return nil
}
func (p *initProcess) start() 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)
}
if _, err := io.Copy(p.parentPipe, p.bootstrapData); err != nil {
return err
}
if err := p.execSetns(); err != 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 err := p.createNetworkInterfaces(); err != nil {
return newSystemError(err)
}
if err := p.sendConfig(); err != nil {
return newSystemError(err)
}
var (
procSync syncT
sentRun bool
sentResume bool
ierr *genericError
)
dec := json.NewDecoder(p.parentPipe)
loop:
for {
if err := dec.Decode(&procSync); err != nil {
if err == io.EOF {
break loop
}
return newSystemError(err)
}
switch procSync.Type {
case procReady:
if err := p.manager.Set(p.config.Config); err != nil {
return newSystemError(err)
}
// set oom_score_adj
if err := setOomScoreAdj(p.config.Config.OomScoreAdj, p.pid()); err != nil {
return newSystemError(err)
}
// set rlimits, this has to be done here because we lose permissions
// to raise the limits once we enter a user-namespace
if err := setupRlimits(p.config.Rlimits, p.pid()); err != nil {
return newSystemError(err)
}
// call prestart hooks
if !p.config.Config.Namespaces.Contains(configs.NEWNS) {
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)
}
}
}
}
// Sync with child.
if err := utils.WriteJSON(p.parentPipe, syncT{procRun}); err != nil {
return newSystemError(err)
}
sentRun = true
case procHooks:
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,
BundlePath: utils.SearchLabels(p.config.Config.Labels, "bundle"),
}
for _, hook := range p.config.Config.Hooks.Prestart {
if err := hook.Run(s); err != nil {
return newSystemError(err)
}
}
}
// Sync with child.
if err := utils.WriteJSON(p.parentPipe, syncT{procResume}); err != nil {
return newSystemError(err)
}
sentResume = true
case procError:
// wait for the child process to fully complete and receive an error message
// if one was encoutered
if err := dec.Decode(&ierr); err != nil && err != io.EOF {
return newSystemError(err)
}
if ierr != nil {
break loop
}
// Programmer error.
panic("No error following JSON procError payload.")
default:
return newSystemError(fmt.Errorf("invalid JSON synchronisation payload from child"))
}
}
if !sentRun {
return newSystemError(fmt.Errorf("could not synchronise with container process: %v", ierr))
}
if p.config.Config.Namespaces.Contains(configs.NEWNS) && !sentResume {
return newSystemError(fmt.Errorf("could not synchronise after executing prestart hooks with container process"))
}
if err := syscall.Shutdown(int(p.parentPipe.Fd()), syscall.SHUT_WR); err != nil {
return newSystemError(err)
}
// Must be done after Shutdown so the child will exit and we can wait for it.
if ierr != nil {
p.wait()
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.sharePidns {
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 config to the container's init process, we don't use JSON Encode
// here because there might be a problem in JSON decoder in some cases, see:
// https://github.com/docker/docker/issues/14203#issuecomment-174177790
return utils.WriteJSON(p.parentPipe, p.config)
}
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/github.com/opencontainers/runc/libcontainer/restored_process.go generated vendored
View file

@ -1,122 +0,0 @@
// +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
}

725
vendor/github.com/opencontainers/runc/libcontainer/rootfs_linux.go generated vendored
View file

@ -1,725 +0,0 @@
// +build linux
package libcontainer
import (
"fmt"
"io"
"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"
"github.com/opencontainers/runc/libcontainer/system"
libcontainerUtils "github.com/opencontainers/runc/libcontainer/utils"
)
const defaultMountFlags = syscall.MS_NOEXEC | syscall.MS_NOSUID | syscall.MS_NODEV
// setupDev returns true if /dev needs to be set up.
func needsSetupDev(config *configs.Config) bool {
for _, m := range config.Mounts {
if m.Device == "bind" && (m.Destination == "/dev" || m.Destination == "/dev/") {
return false
}
}
return true
}
// setupRootfs sets up the devices, mount points, and filesystems for use inside a
// new mount namespace.
func setupRootfs(config *configs.Config, console *linuxConsole, pipe io.ReadWriter) (err error) {
if err := prepareRoot(config); err != nil {
return newSystemError(err)
}
setupDev := needsSetupDev(config)
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)
}
}
// Signal the parent to run the pre-start hooks.
// The hooks are run after the mounts are setup, but before we switch to the new
// root, so that the old root is still available in the hooks for any mount
// manipulations.
if err := syncParentHooks(pipe); err != nil {
return 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)
}
}
// remount dev as ro if specifed
for _, m := range config.Mounts {
if m.Destination == "/dev" {
if m.Flags&syscall.MS_RDONLY != 0 {
if err := remountReadonly(m.Destination); err != nil {
return newSystemError(err)
}
}
break
}
}
// set rootfs ( / ) as readonly
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)
}
return nil
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 "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:
if err := os.MkdirAll(dest, 0755); err != nil {
return err
}
return mountPropagate(m, rootfs, mountLabel)
}
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 libcontainerUtils.CleanPath(rootfs) == libcontainerUtils.CleanPath(dest) {
return fmt.Errorf("mounting into / is prohibited")
}
invalidDestinations := []string{
"/proc",
}
// White list, it should be sub directories of invalid destinations
validDestinations := []string{
// These entries can be bind mounted by files emulated by fuse,
// so commands like top, free displays stats in container.
"/proc/cpuinfo",
"/proc/diskstats",
"/proc/meminfo",
"/proc/stat",
"/proc/net/dev",
}
for _, valid := range validDestinations {
path, err := filepath.Rel(filepath.Join(rootfs, valid), dest)
if err != nil {
return err
}
if path == "." {
return nil
}
}
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/core"})
}
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 {
useBindMount := system.RunningInUserNS() || config.Namespaces.Contains(configs.NEWUSER)
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, useBindMount); 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) (err 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)
}
defer func() {
errVal := os.Remove(pivotDir)
if err == nil {
err = errVal
}
}()
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 nil
}
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)
flags = m.Flags
)
if dest == "/dev" {
flags &= ^syscall.MS_RDONLY
}
if !strings.HasPrefix(dest, rootfs) {
dest = filepath.Join(rootfs, dest)
}
if err := syscall.Mount(m.Source, dest, m.Device, uintptr(flags), data); err != nil {
return err
}
for _, pflag := range m.PropagationFlags {
if err := syscall.Mount("", dest, "", uintptr(pflag), ""); err != nil {
return err
}
}
return nil
}

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

@ -1,11 +0,0 @@
// +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
}

53
vendor/github.com/opencontainers/runc/libcontainer/setns_init_linux.go generated vendored
View file

@ -1,53 +0,0 @@
// +build linux
package libcontainer
import (
"fmt"
"os"
"github.com/opencontainers/runc/libcontainer/apparmor"
"github.com/opencontainers/runc/libcontainer/keys"
"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) getSessionRingName() string {
return fmt.Sprintf("_ses.%s", l.config.ContainerId)
}
func (l *linuxSetnsInit) Init() error {
// do not inherit the parent's session keyring
if _, err := keyctl.JoinSessionKeyring(l.getSessionRingName()); err != nil {
return err
}
if l.config.NoNewPrivileges {
if err := system.Prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); 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.AppArmorProfile); err != nil {
return err
}
if l.config.ProcessLabel != "" {
if err := label.SetProcessLabel(l.config.ProcessLabel); err != nil {
return err
}
}
return system.Execv(l.config.Args[0], l.config.Args[0:], os.Environ())
}

147
vendor/github.com/opencontainers/runc/libcontainer/standard_init_linux.go generated vendored
View file

@ -1,147 +0,0 @@
// +build linux
package libcontainer
import (
"fmt"
"io"
"os"
"syscall"
"github.com/opencontainers/runc/libcontainer/apparmor"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/keys"
"github.com/opencontainers/runc/libcontainer/label"
"github.com/opencontainers/runc/libcontainer/seccomp"
"github.com/opencontainers/runc/libcontainer/system"
)
type linuxStandardInit struct {
pipe io.ReadWriter
parentPid int
config *initConfig
}
func (l *linuxStandardInit) getSessionRingParams() (string, uint32, uint32) {
var newperms uint32
if l.config.Config.Namespaces.Contains(configs.NEWUSER) {
// with user ns we need 'other' search permissions
newperms = 0x8
} else {
// without user ns we need 'UID' search permissions
newperms = 0x80000
}
// create a unique per session container name that we can
// join in setns; however, other containers can also join it
return fmt.Sprintf("_ses.%s", l.config.ContainerId), 0xffffffff, newperms
}
// PR_SET_NO_NEW_PRIVS isn't exposed in Golang so we define it ourselves copying the value
// the kernel
const PR_SET_NO_NEW_PRIVS = 0x26
func (l *linuxStandardInit) Init() error {
ringname, keepperms, newperms := l.getSessionRingParams()
// do not inherit the parent's session keyring
sessKeyId, err := keyctl.JoinSessionKeyring(ringname)
if err != nil {
return err
}
// make session keyring searcheable
if err := keyctl.ModKeyringPerm(sessKeyId, keepperms, newperms); 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 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
}
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, l.pipe); 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.AppArmorProfile); err != nil {
return err
}
if err := label.SetProcessLabel(l.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.NoNewPrivileges {
if err := system.Prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); err != nil {
return err
}
}
// Tell our parent that we're ready to Execv. This must be done before the
// Seccomp rules have been applied, because we need to be able to read and
// write to a socket.
if err := syncParentReady(l.pipe); 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())
}

228
vendor/github.com/opencontainers/runc/libcontainer/state_linux.go generated vendored
View file

@ -1,228 +0,0 @@
// +build linux
package libcontainer
import (
"fmt"
"os"
"path/filepath"
"github.com/Sirupsen/logrus"
"github.com/opencontainers/runc/libcontainer/configs"
"github.com/opencontainers/runc/libcontainer/utils"
)
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
}
c.state = &stoppedState{c: c}
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,
BundlePath: utils.SearchLabels(c.config.Labels, "bundle"),
}
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:
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, *stoppedState:
p.c.state = s
return nil
case *pausedState:
return nil
}
return newStateTransitionError(p, s)
}
func (p *pausedState) destroy() error {
isRunning, err := p.c.isRunning()
if err != nil {
return err
}
if !isRunning {
if err := p.c.cgroupManager.Freeze(configs.Thawed); err != nil {
return err
}
return destroy(p.c)
}
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)
}
// createdState is used whenever a container is restored, loaded, or setting additional
// processes inside and it should not be destroyed when it is exiting.
type createdState struct {
c *linuxContainer
s Status
}
func (n *createdState) status() Status {
return n.s
}
func (n *createdState) transition(s containerState) error {
n.c.state = s
return nil
}
func (n *createdState) destroy() error {
if err := n.c.refreshState(); err != nil {
return err
}
return n.c.state.destroy()
}

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

@ -1,15 +0,0 @@
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/github.com/opencontainers/runc/libcontainer/stats_freebsd.go generated vendored
View file

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

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

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

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

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