github/skopeo
1
0
Fork 0
mirror of https://github.com/containers/skopeo.git synced 2025-09-22 10:27:08 +00:00
Code Issues Projects Releases Wiki Activity

migrate to go modules

Browse Source 
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
This commit is contained in:
Valentin Rothberg
2019-06-18 11:47:51 +02:00
parent ea49bfc2b4
commit 033b290217
447 changed files with 40519 additions and 15133 deletions
Download Patch File Download Diff File Expand all files Collapse all files

270
vendor/github.com/opencontainers/runc/README.md generated vendored
View File

@@ -1,270 +0,0 @@
# runc
[![Build Status](https://travis-ci.org/opencontainers/runc.svg?branch=master)](https://travis-ci.org/opencontainers/runc)
[![Go Report Card](https://goreportcard.com/badge/github.com/opencontainers/runc)](https://goreportcard.com/report/github.com/opencontainers/runc)
[![GoDoc](https://godoc.org/github.com/opencontainers/runc?status.svg)](https://godoc.org/github.com/opencontainers/runc)
## Introduction
`runc` is a CLI tool for spawning and running containers according to the OCI specification.
## Releases
`runc` depends on and tracks the [runtime-spec](https://github.com/opencontainers/runtime-spec) repository.
We will try to make sure that `runc` and the OCI specification major versions stay in lockstep.
This means that `runc` 1.0.0 should implement the 1.0 version of the specification.
You can find official releases of `runc` on the [release](https://github.com/opencontainers/runc/releases) page.
### Security
If you wish to report a security issue, please disclose the issue responsibly
to security@opencontainers.org.
## Building
`runc` currently supports the Linux platform with various architecture support.
It must be built with Go version 1.6 or higher in order for some features to function properly.
In order to enable seccomp support you will need to install `libseccomp` on your platform.
> e.g. `libseccomp-devel` for CentOS, or `libseccomp-dev` for Ubuntu
Otherwise, if you do not want to build `runc` with seccomp support you can add `BUILDTAGS=""` when running make.
```bash
# create a 'github.com/opencontainers' in your GOPATH/src
cd github.com/opencontainers
git clone https://github.com/opencontainers/runc
cd runc
make
sudo make install
```
You can also use `go get` to install to your `GOPATH`, assuming that you have a `github.com` parent folder already created under `src`:
```bash
go get github.com/opencontainers/runc
cd $GOPATH/src/github.com/opencontainers/runc
make
sudo make install
```
`runc` will be installed to `/usr/local/sbin/runc` on your system.
#### Build Tags
`runc` supports optional build tags for compiling support of various features.
To add build tags to the make option the `BUILDTAGS` variable must be set.
```bash
make BUILDTAGS='seccomp apparmor'
```
| Build Tag | Feature | Dependency |
|-----------|------------------------------------|-------------|
| seccomp | Syscall filtering | libseccomp |
| selinux | selinux process and mount labeling | <none> |
| apparmor | apparmor profile support | <none> |
| ambient | ambient capability support | kernel 4.3 |
| nokmem | disable kernel memory account | <none> |
### Running the test suite
`runc` currently supports running its test suite via Docker.
To run the suite just type `make test`.
```bash
make test
```
There are additional make targets for running the tests outside of a container but this is not recommended as the tests are written with the expectation that they can write and remove anywhere.
You can run a specific test case by setting the `TESTFLAGS` variable.
```bash
# make test TESTFLAGS="-run=SomeTestFunction"
```
You can run a specific integration test by setting the `TESTPATH` variable.
```bash
# make test TESTPATH="/checkpoint.bats"
```
You can run a test in your proxy environment by setting `DOCKER_BUILD_PROXY` and `DOCKER_RUN_PROXY` variables.
```bash
# make test DOCKER_BUILD_PROXY="--build-arg HTTP_PROXY=http://yourproxy/" DOCKER_RUN_PROXY="-e HTTP_PROXY=http://yourproxy/"
```
### Dependencies Management
`runc` uses [vndr](https://github.com/LK4D4/vndr) for dependencies management.
Please refer to [vndr](https://github.com/LK4D4/vndr) for how to add or update
new dependencies.
## Using runc
### Creating an OCI Bundle
In order to use runc you must have your container in the format of an OCI bundle.
If you have Docker installed you can use its `export` method to acquire a root filesystem from an existing Docker container.
```bash
# create the top most bundle directory
mkdir /mycontainer
cd /mycontainer
# create the rootfs directory
mkdir rootfs
# export busybox via Docker into the rootfs directory
docker export $(docker create busybox) | tar -C rootfs -xvf -
```
After a root filesystem is populated you just generate a spec in the format of a `config.json` file inside your bundle.
`runc` provides a `spec` command to generate a base template spec that you are then able to edit.
To find features and documentation for fields in the spec please refer to the [specs](https://github.com/opencontainers/runtime-spec) repository.
```bash
runc spec
```
### Running Containers
Assuming you have an OCI bundle from the previous step you can execute the container in two different ways.
The first way is to use the convenience command `run` that will handle creating, starting, and deleting the container after it exits.
```bash
# run as root
cd /mycontainer
runc run mycontainerid
```
If you used the unmodified `runc spec` template this should give you a `sh` session inside the container.
The second way to start a container is using the specs lifecycle operations.
This gives you more power over how the container is created and managed while it is running.
This will also launch the container in the background so you will have to edit the `config.json` to remove the `terminal` setting for the simple examples here.
Your process field in the `config.json` should look like this below with `"terminal": false` and `"args": ["sleep", "5"]`.
```json
"process": {
"terminal": false,
"user": {
"uid": 0,
"gid": 0
},
"args": [
"sleep", "5"
],
"env": [
"PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
"TERM=xterm"
],
"cwd": "/",
"capabilities": {
"bounding": [
"CAP_AUDIT_WRITE",
"CAP_KILL",
"CAP_NET_BIND_SERVICE"
],
"effective": [
"CAP_AUDIT_WRITE",
"CAP_KILL",
"CAP_NET_BIND_SERVICE"
],
"inheritable": [
"CAP_AUDIT_WRITE",
"CAP_KILL",
"CAP_NET_BIND_SERVICE"
],
"permitted": [
"CAP_AUDIT_WRITE",
"CAP_KILL",
"CAP_NET_BIND_SERVICE"
],
"ambient": [
"CAP_AUDIT_WRITE",
"CAP_KILL",
"CAP_NET_BIND_SERVICE"
]
},
"rlimits": [
{
"type": "RLIMIT_NOFILE",
"hard": 1024,
"soft": 1024
}
],
"noNewPrivileges": true
},
```
Now we can go through the lifecycle operations in your shell.
```bash
# run as root
cd /mycontainer
runc create mycontainerid
# view the container is created and in the "created" state
runc list
# start the process inside the container
runc start mycontainerid
# after 5 seconds view that the container has exited and is now in the stopped state
runc list
# now delete the container
runc delete mycontainerid
```
This allows higher level systems to augment the containers creation logic with setup of various settings after the container is created and/or before it is deleted. For example, the container's network stack is commonly set up after `create` but before `start`.
#### Rootless containers
`runc` has the ability to run containers without root privileges. This is called `rootless`. You need to pass some parameters to `runc` in order to run rootless containers. See below and compare with the previous version. Run the following commands as an ordinary user:
```bash
# Same as the first example
mkdir ~/mycontainer
cd ~/mycontainer
mkdir rootfs
docker export $(docker create busybox) | tar -C rootfs -xvf -
# The --rootless parameter instructs runc spec to generate a configuration for a rootless container, which will allow you to run the container as a non-root user.
runc spec --rootless
# The --root parameter tells runc where to store the container state. It must be writable by the user.
runc --root /tmp/runc run mycontainerid
```
#### Supervisors
`runc` can be used with process supervisors and init systems to ensure that containers are restarted when they exit.
An example systemd unit file looks something like this.
```systemd
[Unit]
Description=Start My Container
[Service]
Type=forking
ExecStart=/usr/local/sbin/runc run -d --pid-file /run/mycontainerid.pid mycontainerid
ExecStopPost=/usr/local/sbin/runc delete mycontainerid
WorkingDirectory=/mycontainer
PIDFile=/run/mycontainerid.pid
[Install]
WantedBy=multi-user.target
```
## License
The code and docs are released under the [Apache 2.0 license](LICENSE).

330
vendor/github.com/opencontainers/runc/libcontainer/README.md generated vendored
View File

@@ -1,330 +0,0 @@
# libcontainer
[![GoDoc](https://godoc.org/github.com/opencontainers/runc/libcontainer?status.svg)](https://godoc.org/github.com/opencontainers/runc/libcontainer)
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".
In addition to the go init function the early stage bootstrap is handled by importing
[nsenter](https://github.com/opencontainers/runc/blob/master/libcontainer/nsenter/README.md).
```go
import (
_ "github.com/opencontainers/runc/libcontainer/nsenter"
)
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 := unix.MS_NOEXEC | unix.MS_NOSUID | unix.MS_NODEV
config := &configs.Config{
Rootfs: "/your/path/to/rootfs",
Capabilities: &configs.Capabilities{
Bounding: []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",
},
Effective: []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",
},
Inheritable: []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",
},
Permitted: []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",
},
Ambient: []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},
{Type: configs.NEWCGROUP},
}),
Cgroups: &configs.Cgroup{
Name: "test-container",
Parent: "system",
Resources: &configs.Resources{
MemorySwappiness: nil,
AllowAllDevices: nil,
AllowedDevices: configs.DefaultAllowedDevices,
},
},
MaskPaths: []string{
"/proc/kcore",
"/sys/firmware",
},
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: unix.MS_NOSUID | unix.MS_STRICTATIME,
Data: "mode=755",
},
{
Source: "devpts",
Destination: "/dev/pts",
Device: "devpts",
Flags: unix.MS_NOSUID | unix.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 | unix.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: unix.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.Run(process)
if err != nil {
container.Destroy()
logrus.Fatal(err)
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)
// update container resource constraints.
container.Set(config)
// get current status of the container.
status, err := container.Status()
// get current container's state information.
state, err := container.State()
```
#### 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.
The code and documentation are released under the [Apache 2.0 license](../LICENSE).
The documentation is also released under Creative Commons Attribution 4.0 International License.
You may obtain a copy of the license, titled CC-BY-4.0, at http://creativecommons.org/licenses/by/4.0/.

44
vendor/github.com/opencontainers/runc/libcontainer/nsenter/README.md generated vendored
View File

@@ -1,44 +0,0 @@
## nsenter
The `nsenter` package registers a special init constructor that is called before
the Go runtime has a chance to boot. This provides us the ability to `setns` on
existing namespaces and avoid the issues that the Go runtime has with multiple
threads. This constructor will be called if this package is registered,
imported, in your go application.
The `nsenter` package will `import "C"` and it uses [cgo](https://golang.org/cmd/cgo/)
package. In cgo, if the import of "C" is immediately preceded by a comment, that comment,
called the preamble, is used as a header when compiling the C parts of the package.
So every time we import package `nsenter`, the C code function `nsexec()` would be
called. And package `nsenter` is only imported in `init.go`, so every time the runc
`init` command is invoked, that C code is run.
Because `nsexec()` must be run before the Go runtime in order to use the
Linux kernel namespace, you must `import` this library into a package if
you plan to use `libcontainer` directly. Otherwise Go will not execute
the `nsexec()` constructor, which means that the re-exec will not cause
the namespaces to be joined. You can import it like this:
```go
import _ "github.com/opencontainers/runc/libcontainer/nsenter"
```
`nsexec()` will first get the file descriptor number for the init pipe
from the environment variable `_LIBCONTAINER_INITPIPE` (which was opened
by the parent and kept open across the fork-exec of the `nsexec()` init
process). The init pipe is used to read bootstrap data (namespace paths,
clone flags, uid and gid mappings, and the console path) from the parent
process. `nsexec()` will then call `setns(2)` to join the namespaces
provided in the bootstrap data (if available), `clone(2)` a child process
with the provided clone flags, update the user and group ID mappings, do
some further miscellaneous setup steps, and then send the PID of the
child process to the parent of the `nsexec()` "caller". Finally,
the parent `nsexec()` will exit and the child `nsexec()` process will
return to allow the Go runtime take over.
NOTE: We do both `setns(2)` and `clone(2)` even if we don't have any
`CLONE_NEW*` clone flags because we must fork a new process in order to
enter the PID namespace.

32
vendor/github.com/opencontainers/runc/libcontainer/nsenter/namespace.h generated vendored
View File

@@ -1,32 +0,0 @@
#ifndef NSENTER_NAMESPACE_H
#define NSENTER_NAMESPACE_H
#ifndef _GNU_SOURCE
# define _GNU_SOURCE
#endif
#include <sched.h>
/* All of these are taken from include/uapi/linux/sched.h */
#ifndef CLONE_NEWNS
# define CLONE_NEWNS 0x00020000 /* New mount namespace group */
#endif
#ifndef CLONE_NEWCGROUP
# define CLONE_NEWCGROUP 0x02000000 /* New cgroup namespace */
#endif
#ifndef CLONE_NEWUTS
# define CLONE_NEWUTS 0x04000000 /* New utsname namespace */
#endif
#ifndef CLONE_NEWIPC
# define CLONE_NEWIPC 0x08000000 /* New ipc namespace */
#endif
#ifndef CLONE_NEWUSER
# define CLONE_NEWUSER 0x10000000 /* New user namespace */
#endif
#ifndef CLONE_NEWPID
# define CLONE_NEWPID 0x20000000 /* New pid namespace */
#endif
#ifndef CLONE_NEWNET
# define CLONE_NEWNET 0x40000000 /* New network namespace */
#endif
#endif /* NSENTER_NAMESPACE_H */

12
vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go generated vendored
View File

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

25
vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go generated vendored
View File

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

5
vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go generated vendored
View File

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

995
vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c generated vendored
View File

@@ -1,995 +0,0 @@
#define _GNU_SOURCE
#include <endian.h>
#include <errno.h>
#include <fcntl.h>
#include <grp.h>
#include <sched.h>
#include <setjmp.h>
#include <signal.h>
#include <stdarg.h>
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <stdbool.h>
#include <string.h>
#include <unistd.h>
#include <sys/ioctl.h>
#include <sys/prctl.h>
#include <sys/socket.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <linux/limits.h>
#include <linux/netlink.h>
#include <linux/types.h>
/* Get all of the CLONE_NEW* flags. */
#include "namespace.h"
/* Synchronisation values. */
enum sync_t {
SYNC_USERMAP_PLS = 0x40, /* Request parent to map our users. */
SYNC_USERMAP_ACK = 0x41, /* Mapping finished by the parent. */
SYNC_RECVPID_PLS = 0x42, /* Tell parent we're sending the PID. */
SYNC_RECVPID_ACK = 0x43, /* PID was correctly received by parent. */
SYNC_GRANDCHILD = 0x44, /* The grandchild is ready to run. */
SYNC_CHILD_READY = 0x45, /* The child or grandchild is ready to return. */
/* XXX: This doesn't help with segfaults and other such issues. */
SYNC_ERR = 0xFF, /* Fatal error, no turning back. The error code follows. */
};
/*
* Synchronisation value for cgroup namespace setup.
* The same constant is defined in process_linux.go as "createCgroupns".
*/
#define CREATECGROUPNS 0x80
/* longjmp() arguments. */
#define JUMP_PARENT 0x00
#define JUMP_CHILD 0xA0
#define JUMP_INIT 0xA1
/* JSON buffer. */
#define JSON_MAX 4096
/* Assume the stack grows down, so arguments should be above it. */
struct clone_t {
/*
* Reserve some space for clone() to locate arguments
* and retcode in this place
*/
char stack[4096] __attribute__ ((aligned(16)));
char stack_ptr[0];
/* There's two children. This is used to execute the different code. */
jmp_buf *env;
int jmpval;
};
struct nlconfig_t {
char *data;
/* Process settings. */
uint32_t cloneflags;
char *oom_score_adj;
size_t oom_score_adj_len;
/* User namespace settings. */
char *uidmap;
size_t uidmap_len;
char *gidmap;
size_t gidmap_len;
char *namespaces;
size_t namespaces_len;
uint8_t is_setgroup;
/* Rootless container settings. */
uint8_t is_rootless_euid; /* boolean */
char *uidmappath;
size_t uidmappath_len;
char *gidmappath;
size_t gidmappath_len;
};
/*
* List of netlink message types sent to us as part of bootstrapping the init.
* These constants are defined in libcontainer/message_linux.go.
*/
#define INIT_MSG 62000
#define CLONE_FLAGS_ATTR 27281
#define NS_PATHS_ATTR 27282
#define UIDMAP_ATTR 27283
#define GIDMAP_ATTR 27284
#define SETGROUP_ATTR 27285
#define OOM_SCORE_ADJ_ATTR 27286
#define ROOTLESS_EUID_ATTR 27287
#define UIDMAPPATH_ATTR 27288
#define GIDMAPPATH_ATTR 27289
/*
* Use the raw syscall for versions of glibc which don't include a function for
* it, namely (glibc 2.12).
*/
#if __GLIBC__ == 2 && __GLIBC_MINOR__ < 14
# define _GNU_SOURCE
# include "syscall.h"
# if !defined(SYS_setns) && defined(__NR_setns)
# define SYS_setns __NR_setns
# endif
#ifndef SYS_setns
# error "setns(2) syscall not supported by glibc version"
#endif
int setns(int fd, int nstype)
{
return syscall(SYS_setns, fd, nstype);
}
#endif
/* XXX: This is ugly. */
static int syncfd = -1;
/* TODO(cyphar): Fix this so it correctly deals with syncT. */
#define bail(fmt, ...) \
do { \
int ret = __COUNTER__ + 1; \
fprintf(stderr, "nsenter: " fmt ": %m\n", ##__VA_ARGS__); \
if (syncfd >= 0) { \
enum sync_t s = SYNC_ERR; \
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) \
fprintf(stderr, "nsenter: failed: write(s)"); \
if (write(syncfd, &ret, sizeof(ret)) != sizeof(ret)) \
fprintf(stderr, "nsenter: failed: write(ret)"); \
} \
exit(ret); \
} while(0)
static int write_file(char *data, size_t data_len, char *pathfmt, ...)
{
int fd, len, ret = 0;
char path[PATH_MAX];
va_list ap;
va_start(ap, pathfmt);
len = vsnprintf(path, PATH_MAX, pathfmt, ap);
va_end(ap);
if (len < 0)
return -1;
fd = open(path, O_RDWR);
if (fd < 0) {
return -1;
}
len = write(fd, data, data_len);
if (len != data_len) {
ret = -1;
goto out;
}
out:
close(fd);
return ret;
}
enum policy_t {
SETGROUPS_DEFAULT = 0,
SETGROUPS_ALLOW,
SETGROUPS_DENY,
};
/* This *must* be called before we touch gid_map. */
static void update_setgroups(int pid, enum policy_t setgroup)
{
char *policy;
switch (setgroup) {
case SETGROUPS_ALLOW:
policy = "allow";
break;
case SETGROUPS_DENY:
policy = "deny";
break;
case SETGROUPS_DEFAULT:
default:
/* Nothing to do. */
return;
}
if (write_file(policy, strlen(policy), "/proc/%d/setgroups", pid) < 0) {
/*
* If the kernel is too old to support /proc/pid/setgroups,
* open(2) or write(2) will return ENOENT. This is fine.
*/
if (errno != ENOENT)
bail("failed to write '%s' to /proc/%d/setgroups", policy, pid);
}
}
static int try_mapping_tool(const char *app, int pid, char *map, size_t map_len)
{
int child;
/*
* If @app is NULL, execve will segfault. Just check it here and bail (if
* we're in this path, the caller is already getting desperate and there
* isn't a backup to this failing). This usually would be a configuration
* or programming issue.
*/
if (!app)
bail("mapping tool not present");
child = fork();
if (child < 0)
bail("failed to fork");
if (!child) {
#define MAX_ARGV 20
char *argv[MAX_ARGV];
char *envp[] = { NULL };
char pid_fmt[16];
int argc = 0;
char *next;
snprintf(pid_fmt, 16, "%d", pid);
argv[argc++] = (char *)app;
argv[argc++] = pid_fmt;
/*
* Convert the map string into a list of argument that
* newuidmap/newgidmap can understand.
*/
while (argc < MAX_ARGV) {
if (*map == '\0') {
argv[argc++] = NULL;
break;
}
argv[argc++] = map;
next = strpbrk(map, "\n ");
if (next == NULL)
break;
*next++ = '\0';
map = next + strspn(next, "\n ");
}
execve(app, argv, envp);
bail("failed to execv");
} else {
int status;
while (true) {
if (waitpid(child, &status, 0) < 0) {
if (errno == EINTR)
continue;
bail("failed to waitpid");
}
if (WIFEXITED(status) || WIFSIGNALED(status))
return WEXITSTATUS(status);
}
}
return -1;
}
static void update_uidmap(const char *path, int pid, char *map, size_t map_len)
{
if (map == NULL || map_len <= 0)
return;
if (write_file(map, map_len, "/proc/%d/uid_map", pid) < 0) {
if (errno != EPERM)
bail("failed to update /proc/%d/uid_map", pid);
if (try_mapping_tool(path, pid, map, map_len))
bail("failed to use newuid map on %d", pid);
}
}
static void update_gidmap(const char *path, int pid, char *map, size_t map_len)
{
if (map == NULL || map_len <= 0)
return;
if (write_file(map, map_len, "/proc/%d/gid_map", pid) < 0) {
if (errno != EPERM)
bail("failed to update /proc/%d/gid_map", pid);
if (try_mapping_tool(path, pid, map, map_len))
bail("failed to use newgid map on %d", pid);
}
}
static void update_oom_score_adj(char *data, size_t len)
{
if (data == NULL || len <= 0)
return;
if (write_file(data, len, "/proc/self/oom_score_adj") < 0)
bail("failed to update /proc/self/oom_score_adj");
}
/* A dummy function that just jumps to the given jumpval. */
static int child_func(void *arg) __attribute__ ((noinline));
static int child_func(void *arg)
{
struct clone_t *ca = (struct clone_t *)arg;
longjmp(*ca->env, ca->jmpval);
}
static int clone_parent(jmp_buf *env, int jmpval) __attribute__ ((noinline));
static int clone_parent(jmp_buf *env, int jmpval)
{
struct clone_t ca = {
.env = env,
.jmpval = jmpval,
};
return clone(child_func, ca.stack_ptr, CLONE_PARENT | SIGCHLD, &ca);
}
/*
* Gets the init pipe fd from the environment, which is used to read the
* bootstrap data and tell the parent what the new pid is after we finish
* setting up the environment.
*/
static int initpipe(void)
{
int pipenum;
char *initpipe, *endptr;
initpipe = getenv("_LIBCONTAINER_INITPIPE");
if (initpipe == NULL || *initpipe == '\0')
return -1;
pipenum = strtol(initpipe, &endptr, 10);
if (*endptr != '\0')
bail("unable to parse _LIBCONTAINER_INITPIPE");
return pipenum;
}
/* Returns the clone(2) flag for a namespace, given the name of a namespace. */
static int nsflag(char *name)
{
if (!strcmp(name, "cgroup"))
return CLONE_NEWCGROUP;
else if (!strcmp(name, "ipc"))
return CLONE_NEWIPC;
else if (!strcmp(name, "mnt"))
return CLONE_NEWNS;
else if (!strcmp(name, "net"))
return CLONE_NEWNET;
else if (!strcmp(name, "pid"))
return CLONE_NEWPID;
else if (!strcmp(name, "user"))
return CLONE_NEWUSER;
else if (!strcmp(name, "uts"))
return CLONE_NEWUTS;
/* If we don't recognise a name, fallback to 0. */
return 0;
}
static uint32_t readint32(char *buf)
{
return *(uint32_t *) buf;
}
static uint8_t readint8(char *buf)
{
return *(uint8_t *) buf;
}
static void nl_parse(int fd, struct nlconfig_t *config)
{
size_t len, size;
struct nlmsghdr hdr;
char *data, *current;
/* Retrieve the netlink header. */
len = read(fd, &hdr, NLMSG_HDRLEN);
if (len != NLMSG_HDRLEN)
bail("invalid netlink header length %zu", len);
if (hdr.nlmsg_type == NLMSG_ERROR)
bail("failed to read netlink message");
if (hdr.nlmsg_type != INIT_MSG)
bail("unexpected msg type %d", hdr.nlmsg_type);
/* Retrieve data. */
size = NLMSG_PAYLOAD(&hdr, 0);
current = data = malloc(size);
if (!data)
bail("failed to allocate %zu bytes of memory for nl_payload", size);
len = read(fd, data, size);
if (len != size)
bail("failed to read netlink payload, %zu != %zu", len, size);
/* Parse the netlink payload. */
config->data = data;
while (current < data + size) {
struct nlattr *nlattr = (struct nlattr *)current;
size_t payload_len = nlattr->nla_len - NLA_HDRLEN;
/* Advance to payload. */
current += NLA_HDRLEN;
/* Handle payload. */
switch (nlattr->nla_type) {
case CLONE_FLAGS_ATTR:
config->cloneflags = readint32(current);
break;
case ROOTLESS_EUID_ATTR:
config->is_rootless_euid = readint8(current); /* boolean */
break;
case OOM_SCORE_ADJ_ATTR:
config->oom_score_adj = current;
config->oom_score_adj_len = payload_len;
break;
case NS_PATHS_ATTR:
config->namespaces = current;
config->namespaces_len = payload_len;
break;
case UIDMAP_ATTR:
config->uidmap = current;
config->uidmap_len = payload_len;
break;
case GIDMAP_ATTR:
config->gidmap = current;
config->gidmap_len = payload_len;
break;
case UIDMAPPATH_ATTR:
config->uidmappath = current;
config->uidmappath_len = payload_len;
break;
case GIDMAPPATH_ATTR:
config->gidmappath = current;
config->gidmappath_len = payload_len;
break;
case SETGROUP_ATTR:
config->is_setgroup = readint8(current);
break;
default:
bail("unknown netlink message type %d", nlattr->nla_type);
}
current += NLA_ALIGN(payload_len);
}
}
void nl_free(struct nlconfig_t *config)
{
free(config->data);
}
void join_namespaces(char *nslist)
{
int num = 0, i;
char *saveptr = NULL;
char *namespace = strtok_r(nslist, ",", &saveptr);
struct namespace_t {
int fd;
int ns;
char type[PATH_MAX];
char path[PATH_MAX];
} *namespaces = NULL;
if (!namespace || !strlen(namespace) || !strlen(nslist))
bail("ns paths are empty");
/*
* We have to open the file descriptors first, since after
* we join the mnt namespace we might no longer be able to
* access the paths.
*/
do {
int fd;
char *path;
struct namespace_t *ns;
/* Resize the namespace array. */
namespaces = realloc(namespaces, ++num * sizeof(struct namespace_t));
if (!namespaces)
bail("failed to reallocate namespace array");
ns = &namespaces[num - 1];
/* Split 'ns:path'. */
path = strstr(namespace, ":");
if (!path)
bail("failed to parse %s", namespace);
*path++ = '\0';
fd = open(path, O_RDONLY);
if (fd < 0)
bail("failed to open %s", path);
ns->fd = fd;
ns->ns = nsflag(namespace);
strncpy(ns->path, path, PATH_MAX - 1);
ns->path[PATH_MAX - 1] = '\0';
} while ((namespace = strtok_r(NULL, ",", &saveptr)) != NULL);
/*
* The ordering in which we join namespaces is important. We should
* always join the user namespace *first*. This is all guaranteed
* from the container_linux.go side of this, so we're just going to
* follow the order given to us.
*/
for (i = 0; i < num; i++) {
struct namespace_t ns = namespaces[i];
if (setns(ns.fd, ns.ns) < 0)
bail("failed to setns to %s", ns.path);
close(ns.fd);
}
free(namespaces);
}
void nsexec(void)
{
int pipenum;
jmp_buf env;
int sync_child_pipe[2], sync_grandchild_pipe[2];
struct nlconfig_t config = { 0 };
/*
* If we don't have an init pipe, just return to the go routine.
* We'll only get an init pipe for start or exec.
*/
pipenum = initpipe();
if (pipenum == -1)
return;
/* Parse all of the netlink configuration. */
nl_parse(pipenum, &config);
/* Set oom_score_adj. This has to be done before !dumpable because
* /proc/self/oom_score_adj is not writeable unless you're an privileged
* user (if !dumpable is set). All children inherit their parent's
* oom_score_adj value on fork(2) so this will always be propagated
* properly.
*/
update_oom_score_adj(config.oom_score_adj, config.oom_score_adj_len);
/*
* Make the process non-dumpable, to avoid various race conditions that
* could cause processes in namespaces we're joining to access host
* resources (or potentially execute code).
*
* However, if the number of namespaces we are joining is 0, we are not
* going to be switching to a different security context. Thus setting
* ourselves to be non-dumpable only breaks things (like rootless
* containers), which is the recommendation from the kernel folks.
*/
if (config.namespaces) {
if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
bail("failed to set process as non-dumpable");
}
/* Pipe so we can tell the child when we've finished setting up. */
if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_child_pipe) < 0)
bail("failed to setup sync pipe between parent and child");
/*
* We need a new socketpair to sync with grandchild so we don't have
* race condition with child.
*/
if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_grandchild_pipe) < 0)
bail("failed to setup sync pipe between parent and grandchild");
/* TODO: Currently we aren't dealing with child deaths properly. */
/*
* Okay, so this is quite annoying.
*
* In order for this unsharing code to be more extensible we need to split
* up unshare(CLONE_NEWUSER) and clone() in various ways. The ideal case
* would be if we did clone(CLONE_NEWUSER) and the other namespaces
* separately, but because of SELinux issues we cannot really do that. But
* we cannot just dump the namespace flags into clone(...) because several
* usecases (such as rootless containers) require more granularity around
* the namespace setup. In addition, some older kernels had issues where
* CLONE_NEWUSER wasn't handled before other namespaces (but we cannot
* handle this while also dealing with SELinux so we choose SELinux support
* over broken kernel support).
*
* However, if we unshare(2) the user namespace *before* we clone(2), then
* all hell breaks loose.
*
* The parent no longer has permissions to do many things (unshare(2) drops
* all capabilities in your old namespace), and the container cannot be set
* up to have more than one {uid,gid} mapping. This is obviously less than
* ideal. In order to fix this, we have to first clone(2) and then unshare.
*
* Unfortunately, it's not as simple as that. We have to fork to enter the
* PID namespace (the PID namespace only applies to children). Since we'll
* have to double-fork, this clone_parent() call won't be able to get the
* PID of the _actual_ init process (without doing more synchronisation than
* I can deal with at the moment). So we'll just get the parent to send it
* for us, the only job of this process is to update
* /proc/pid/{setgroups,uid_map,gid_map}.
*
* And as a result of the above, we also need to setns(2) in the first child
* because if we join a PID namespace in the topmost parent then our child
* will be in that namespace (and it will not be able to give us a PID value
* that makes sense without resorting to sending things with cmsg).
*
* This also deals with an older issue caused by dumping cloneflags into
* clone(2): On old kernels, CLONE_PARENT didn't work with CLONE_NEWPID, so
* we have to unshare(2) before clone(2) in order to do this. This was fixed
* in upstream commit 1f7f4dde5c945f41a7abc2285be43d918029ecc5, and was
* introduced by 40a0d32d1eaffe6aac7324ca92604b6b3977eb0e. As far as we're
* aware, the last mainline kernel which had this bug was Linux 3.12.
* However, we cannot comment on which kernels the broken patch was
* backported to.
*
* -- Aleksa "what has my life come to?" Sarai
*/
switch (setjmp(env)) {
/*
* Stage 0: We're in the parent. Our job is just to create a new child
* (stage 1: JUMP_CHILD) process and write its uid_map and
* gid_map. That process will go on to create a new process, then
* it will send us its PID which we will send to the bootstrap
* process.
*/
case JUMP_PARENT:{
int len;
pid_t child, first_child = -1;
bool ready = false;
/* For debugging. */
prctl(PR_SET_NAME, (unsigned long)"runc:[0:PARENT]", 0, 0, 0);
/* Start the process of getting a container. */
child = clone_parent(&env, JUMP_CHILD);
if (child < 0)
bail("unable to fork: child_func");
/*
* State machine for synchronisation with the children.
*
* Father only return when both child and grandchild are
* ready, so we can receive all possible error codes
* generated by children.
*/
while (!ready) {
enum sync_t s;
int ret;
syncfd = sync_child_pipe[1];
close(sync_child_pipe[0]);
if (read(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with child: next state");
switch (s) {
case SYNC_ERR:
/* We have to mirror the error code of the child. */
if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
bail("failed to sync with child: read(error code)");
exit(ret);
case SYNC_USERMAP_PLS:
/*
* Enable setgroups(2) if we've been asked to. But we also
* have to explicitly disable setgroups(2) if we're
* creating a rootless container for single-entry mapping.
* i.e. config.is_setgroup == false.
* (this is required since Linux 3.19).
*
* For rootless multi-entry mapping, config.is_setgroup shall be true and
* newuidmap/newgidmap shall be used.
*/
if (config.is_rootless_euid && !config.is_setgroup)
update_setgroups(child, SETGROUPS_DENY);
/* Set up mappings. */
update_uidmap(config.uidmappath, child, config.uidmap, config.uidmap_len);
update_gidmap(config.gidmappath, child, config.gidmap, config.gidmap_len);
s = SYNC_USERMAP_ACK;
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(child, SIGKILL);
bail("failed to sync with child: write(SYNC_USERMAP_ACK)");
}
break;
case SYNC_RECVPID_PLS:{
first_child = child;
/* Get the init_func pid. */
if (read(syncfd, &child, sizeof(child)) != sizeof(child)) {
kill(first_child, SIGKILL);
bail("failed to sync with child: read(childpid)");
}
/* Send ACK. */
s = SYNC_RECVPID_ACK;
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(first_child, SIGKILL);
kill(child, SIGKILL);
bail("failed to sync with child: write(SYNC_RECVPID_ACK)");
}
/* Send the init_func pid back to our parent.
*
* Send the init_func pid and the pid of the first child back to our parent.
* We need to send both back because we can't reap the first child we created (CLONE_PARENT).
* It becomes the responsibility of our parent to reap the first child.
*/
len = dprintf(pipenum, "{\"pid\": %d, \"pid_first\": %d}\n", child, first_child);
if (len < 0) {
kill(child, SIGKILL);
bail("unable to generate JSON for child pid");
}
}
break;
case SYNC_CHILD_READY:
ready = true;
break;
default:
bail("unexpected sync value: %u", s);
}
}
/* Now sync with grandchild. */
ready = false;
while (!ready) {
enum sync_t s;
int ret;
syncfd = sync_grandchild_pipe[1];
close(sync_grandchild_pipe[0]);
s = SYNC_GRANDCHILD;
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(child, SIGKILL);
bail("failed to sync with child: write(SYNC_GRANDCHILD)");
}
if (read(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with child: next state");
switch (s) {
case SYNC_ERR:
/* We have to mirror the error code of the child. */
if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
bail("failed to sync with child: read(error code)");
exit(ret);
case SYNC_CHILD_READY:
ready = true;
break;
default:
bail("unexpected sync value: %u", s);
}
}
exit(0);
}
/*
* Stage 1: We're in the first child process. Our job is to join any
* provided namespaces in the netlink payload and unshare all
* of the requested namespaces. If we've been asked to
* CLONE_NEWUSER, we will ask our parent (stage 0) to set up
* our user mappings for us. Then, we create a new child
* (stage 2: JUMP_INIT) for PID namespace. We then send the
* child's PID to our parent (stage 0).
*/
case JUMP_CHILD:{
pid_t child;
enum sync_t s;
/* We're in a child and thus need to tell the parent if we die. */
syncfd = sync_child_pipe[0];
close(sync_child_pipe[1]);
/* For debugging. */
prctl(PR_SET_NAME, (unsigned long)"runc:[1:CHILD]", 0, 0, 0);
/*
* We need to setns first. We cannot do this earlier (in stage 0)
* because of the fact that we forked to get here (the PID of
* [stage 2: JUMP_INIT]) would be meaningless). We could send it
* using cmsg(3) but that's just annoying.
*/
if (config.namespaces)
join_namespaces(config.namespaces);
/*
* Deal with user namespaces first. They are quite special, as they
* affect our ability to unshare other namespaces and are used as
* context for privilege checks.
*
* We don't unshare all namespaces in one go. The reason for this
* is that, while the kernel documentation may claim otherwise,
* there are certain cases where unsharing all namespaces at once
* will result in namespace objects being owned incorrectly.
* Ideally we should just fix these kernel bugs, but it's better to
* be safe than sorry, and fix them separately.
*
* A specific case of this is that the SELinux label of the
* internal kern-mount that mqueue uses will be incorrect if the
* UTS namespace is cloned before the USER namespace is mapped.
* I've also heard of similar problems with the network namespace
* in some scenarios. This also mirrors how LXC deals with this
* problem.
*/
if (config.cloneflags & CLONE_NEWUSER) {
if (unshare(CLONE_NEWUSER) < 0)
bail("failed to unshare user namespace");
config.cloneflags &= ~CLONE_NEWUSER;
/*
* We don't have the privileges to do any mapping here (see the
* clone_parent rant). So signal our parent to hook us up.
*/
/* Switching is only necessary if we joined namespaces. */
if (config.namespaces) {
if (prctl(PR_SET_DUMPABLE, 1, 0, 0, 0) < 0)
bail("failed to set process as dumpable");
}
s = SYNC_USERMAP_PLS;
if (write(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with parent: write(SYNC_USERMAP_PLS)");
/* ... wait for mapping ... */
if (read(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with parent: read(SYNC_USERMAP_ACK)");
if (s != SYNC_USERMAP_ACK)
bail("failed to sync with parent: SYNC_USERMAP_ACK: got %u", s);
/* Switching is only necessary if we joined namespaces. */
if (config.namespaces) {
if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
bail("failed to set process as dumpable");
}
/* Become root in the namespace proper. */
if (setresuid(0, 0, 0) < 0)
bail("failed to become root in user namespace");
}
/*
* Unshare all of the namespaces. Now, it should be noted that this
* ordering might break in the future (especially with rootless
* containers). But for now, it's not possible to split this into
* CLONE_NEWUSER + [the rest] because of some RHEL SELinux issues.
*
* Note that we don't merge this with clone() because there were
* some old kernel versions where clone(CLONE_PARENT | CLONE_NEWPID)
* was broken, so we'll just do it the long way anyway.
*/
if (unshare(config.cloneflags & ~CLONE_NEWCGROUP) < 0)
bail("failed to unshare namespaces");
/*
* TODO: What about non-namespace clone flags that we're dropping here?
*
* We fork again because of PID namespace, setns(2) or unshare(2) don't
* change the PID namespace of the calling process, because doing so
* would change the caller's idea of its own PID (as reported by getpid()),
* which would break many applications and libraries, so we must fork
* to actually enter the new PID namespace.
*/
child = clone_parent(&env, JUMP_INIT);
if (child < 0)
bail("unable to fork: init_func");
/* Send the child to our parent, which knows what it's doing. */
s = SYNC_RECVPID_PLS;
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(child, SIGKILL);
bail("failed to sync with parent: write(SYNC_RECVPID_PLS)");
}
if (write(syncfd, &child, sizeof(child)) != sizeof(child)) {
kill(child, SIGKILL);
bail("failed to sync with parent: write(childpid)");
}
/* ... wait for parent to get the pid ... */
if (read(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(child, SIGKILL);
bail("failed to sync with parent: read(SYNC_RECVPID_ACK)");
}
if (s != SYNC_RECVPID_ACK) {
kill(child, SIGKILL);
bail("failed to sync with parent: SYNC_RECVPID_ACK: got %u", s);
}
s = SYNC_CHILD_READY;
if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
kill(child, SIGKILL);
bail("failed to sync with parent: write(SYNC_CHILD_READY)");
}
/* Our work is done. [Stage 2: JUMP_INIT] is doing the rest of the work. */
exit(0);
}
/*
* Stage 2: We're the final child process, and the only process that will
* actually return to the Go runtime. Our job is to just do the
* final cleanup steps and then return to the Go runtime to allow
* init_linux.go to run.
*/
case JUMP_INIT:{
/*
* We're inside the child now, having jumped from the
* start_child() code after forking in the parent.
*/
enum sync_t s;
/* We're in a child and thus need to tell the parent if we die. */
syncfd = sync_grandchild_pipe[0];
close(sync_grandchild_pipe[1]);
close(sync_child_pipe[0]);
close(sync_child_pipe[1]);
/* For debugging. */
prctl(PR_SET_NAME, (unsigned long)"runc:[2:INIT]", 0, 0, 0);
if (read(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with parent: read(SYNC_GRANDCHILD)");
if (s != SYNC_GRANDCHILD)
bail("failed to sync with parent: SYNC_GRANDCHILD: got %u", s);
if (setsid() < 0)
bail("setsid failed");
if (setuid(0) < 0)
bail("setuid failed");
if (setgid(0) < 0)
bail("setgid failed");
if (!config.is_rootless_euid && config.is_setgroup) {
if (setgroups(0, NULL) < 0)
bail("setgroups failed");
}
/* ... wait until our topmost parent has finished cgroup setup in p.manager.Apply() ... */
if (config.cloneflags & CLONE_NEWCGROUP) {
uint8_t value;
if (read(pipenum, &value, sizeof(value)) != sizeof(value))
bail("read synchronisation value failed");
if (value == CREATECGROUPNS) {
if (unshare(CLONE_NEWCGROUP) < 0)
bail("failed to unshare cgroup namespace");
} else
bail("received unknown synchronisation value");
}
s = SYNC_CHILD_READY;
if (write(syncfd, &s, sizeof(s)) != sizeof(s))
bail("failed to sync with patent: write(SYNC_CHILD_READY)");
/* Close sync pipes. */
close(sync_grandchild_pipe[0]);
/* Free netlink data. */
nl_free(&config);
/* Finish executing, let the Go runtime take over. */
return;
}
default:
bail("unexpected jump value");
}
/* Should never be reached. */
bail("should never be reached");
}

2
vendor/github.com/opencontainers/runc/libcontainer/user/MAINTAINERS generated vendored Normal file
View File

@@ -0,0 +1,2 @@
Tianon Gravi <admwiggin@gmail.com> (@tianon)
Aleksa Sarai <cyphar@cyphar.com> (@cyphar)

25
vendor/github.com/opencontainers/runc/vendor.conf generated vendored
View File

@@ -1,25 +0,0 @@
# OCI runtime-spec. When updating this, make sure you use a version tag rather
# than a commit ID so it's much more obvious what version of the spec we are
# using.
github.com/opencontainers/runtime-spec 5684b8af48c1ac3b1451fa499724e30e3c20a294
# Core libcontainer functionality.
github.com/mrunalp/fileutils ed869b029674c0e9ce4c0dfa781405c2d9946d08
github.com/opencontainers/selinux v1.0.0-rc1
github.com/seccomp/libseccomp-golang 84e90a91acea0f4e51e62bc1a75de18b1fc0790f
github.com/sirupsen/logrus a3f95b5c423586578a4e099b11a46c2479628cac
github.com/syndtr/gocapability db04d3cc01c8b54962a58ec7e491717d06cfcc16
github.com/vishvananda/netlink 1e2e08e8a2dcdacaae3f14ac44c5cfa31361f270
# systemd integration.
github.com/coreos/go-systemd v14
github.com/coreos/pkg v3
github.com/godbus/dbus v3
github.com/golang/protobuf 18c9bb3261723cd5401db4d0c9fbc5c3b6c70fe8
# Command-line interface.
github.com/cyphar/filepath-securejoin v0.2.1
github.com/docker/go-units v0.2.0
github.com/urfave/cli d53eb991652b1d438abdd34ce4bfa3ef1539108e
golang.org/x/sys 7ddbeae9ae08c6a06a59597f0c9edbc5ff2444ce https://github.com/golang/sys
# console dependencies
github.com/containerd/console 2748ece16665b45a47f884001d5831ec79703880
github.com/pkg/errors v0.8.0