Files
podman/utils/utils.go
Paul Holzinger 8d0fb0a4ed move rootless netns slirp4netns process to systemd user.slice
When running podman inside systemd user units, it is possible that
systemd kills the rootless netns slirp4netns process because it was
started in the default unit cgroup. When the unit is stopped all
processes in that cgroup are killed. Since the slirp4netns process is
run once for all containers it should not be killed. To make sure
systemd will not kill the process we move it to the user.slice.

Fixes #13153

Signed-off-by: Paul Holzinger <pholzing@redhat.com>
2022-02-07 17:38:53 +01:00

270 lines
7.2 KiB
Go

package utils
import (
"bytes"
"crypto/rand"
"fmt"
"io"
"io/ioutil"
"os"
"os/exec"
"strconv"
"strings"
"sync"
"github.com/containers/common/pkg/cgroups"
"github.com/containers/podman/v4/libpod/define"
"github.com/containers/storage/pkg/archive"
"github.com/godbus/dbus/v5"
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
)
// ExecCmd executes a command with args and returns its output as a string along
// with an error, if any
func ExecCmd(name string, args ...string) (string, error) {
cmd := exec.Command(name, args...)
var stdout bytes.Buffer
var stderr bytes.Buffer
cmd.Stdout = &stdout
cmd.Stderr = &stderr
err := cmd.Run()
if err != nil {
return "", fmt.Errorf("`%v %v` failed: %v %v (%v)", name, strings.Join(args, " "), stderr.String(), stdout.String(), err)
}
return stdout.String(), nil
}
// ExecCmdWithStdStreams execute a command with the specified standard streams.
func ExecCmdWithStdStreams(stdin io.Reader, stdout, stderr io.Writer, env []string, name string, args ...string) error {
cmd := exec.Command(name, args...)
cmd.Stdin = stdin
cmd.Stdout = stdout
cmd.Stderr = stderr
cmd.Env = env
err := cmd.Run()
if err != nil {
return fmt.Errorf("`%v %v` failed: %v", name, strings.Join(args, " "), err)
}
return nil
}
// ErrDetach is an error indicating that the user manually detached from the
// container.
var ErrDetach = define.ErrDetach
// CopyDetachable is similar to io.Copy but support a detach key sequence to break out.
func CopyDetachable(dst io.Writer, src io.Reader, keys []byte) (written int64, err error) {
buf := make([]byte, 32*1024)
for {
nr, er := src.Read(buf)
if nr > 0 {
preservBuf := []byte{}
for i, key := range keys {
preservBuf = append(preservBuf, buf[0:nr]...)
if nr != 1 || buf[0] != key {
break
}
if i == len(keys)-1 {
return 0, ErrDetach
}
nr, er = src.Read(buf)
}
var nw int
var ew error
if len(preservBuf) > 0 {
nw, ew = dst.Write(preservBuf)
nr = len(preservBuf)
} else {
nw, ew = dst.Write(buf[0:nr])
}
if nw > 0 {
written += int64(nw)
}
if ew != nil {
err = ew
break
}
if nr != nw {
err = io.ErrShortWrite
break
}
}
if er != nil {
if er != io.EOF {
err = er
}
break
}
}
return written, err
}
// UntarToFileSystem untars an os.file of a tarball to a destination in the filesystem
func UntarToFileSystem(dest string, tarball *os.File, options *archive.TarOptions) error {
logrus.Debugf("untarring %s", tarball.Name())
return archive.Untar(tarball, dest, options)
}
// Creates a new tar file and wrties bytes from io.ReadCloser
func CreateTarFromSrc(source string, dest string) error {
file, err := os.Create(dest)
if err != nil {
return errors.Wrapf(err, "Could not create tarball file '%s'", dest)
}
defer file.Close()
return TarToFilesystem(source, file)
}
// TarToFilesystem creates a tarball from source and writes to an os.file
// provided
func TarToFilesystem(source string, tarball *os.File) error {
tb, err := Tar(source)
if err != nil {
return err
}
_, err = io.Copy(tarball, tb)
if err != nil {
return err
}
logrus.Debugf("wrote tarball file %s", tarball.Name())
return nil
}
// Tar creates a tarball from source and returns a readcloser of it
func Tar(source string) (io.ReadCloser, error) {
logrus.Debugf("creating tarball of %s", source)
return archive.Tar(source, archive.Uncompressed)
}
// RemoveScientificNotationFromFloat returns a float without any
// scientific notation if the number has any.
// golang does not handle conversion of float64s that have scientific
// notation in them and otherwise stinks. please replace this if you have
// a better implementation.
func RemoveScientificNotationFromFloat(x float64) (float64, error) {
bigNum := strconv.FormatFloat(x, 'g', -1, 64)
breakPoint := strings.IndexAny(bigNum, "Ee")
if breakPoint > 0 {
bigNum = bigNum[:breakPoint]
}
result, err := strconv.ParseFloat(bigNum, 64)
if err != nil {
return x, errors.Wrapf(err, "unable to remove scientific number from calculations")
}
return result, nil
}
var (
runsOnSystemdOnce sync.Once
runsOnSystemd bool
)
// RunsOnSystemd returns whether the system is using systemd
func RunsOnSystemd() bool {
runsOnSystemdOnce.Do(func() {
initCommand, err := ioutil.ReadFile("/proc/1/comm")
// On errors, default to systemd
runsOnSystemd = err != nil || strings.TrimRight(string(initCommand), "\n") == "systemd"
})
return runsOnSystemd
}
func moveProcessPIDFileToScope(pidPath, slice, scope string) error {
data, err := ioutil.ReadFile(pidPath)
if err != nil {
// do not raise an error if the file doesn't exist
if os.IsNotExist(err) {
return nil
}
return errors.Wrapf(err, "cannot read pid file %s", pidPath)
}
pid, err := strconv.ParseUint(string(data), 10, 0)
if err != nil {
return errors.Wrapf(err, "cannot parse pid file %s", pidPath)
}
return moveProcessToScope(int(pid), slice, scope)
}
func moveProcessToScope(pid int, slice, scope string) error {
err := RunUnderSystemdScope(int(pid), slice, scope)
// If the PID is not valid anymore, do not return an error.
if dbusErr, ok := err.(dbus.Error); ok {
if dbusErr.Name == "org.freedesktop.DBus.Error.UnixProcessIdUnknown" {
return nil
}
}
return err
}
// MoveRootlessNetnsSlirpProcessToUserSlice moves the slirp4netns process for the rootless netns
// into a different scope so that systemd does not kill it with a container.
func MoveRootlessNetnsSlirpProcessToUserSlice(pid int) error {
randBytes := make([]byte, 4)
_, err := rand.Read(randBytes)
if err != nil {
return err
}
return moveProcessToScope(pid, "user.slice", fmt.Sprintf("rootless-netns-%x.scope", randBytes))
}
// MovePauseProcessToScope moves the pause process used for rootless mode to keep the namespaces alive to
// a separate scope.
func MovePauseProcessToScope(pausePidPath string) {
var err error
for i := 0; i < 10; i++ {
randBytes := make([]byte, 4)
_, err = rand.Read(randBytes)
if err != nil {
logrus.Errorf("failed to read random bytes: %v", err)
continue
}
err = moveProcessPIDFileToScope(pausePidPath, "user.slice", fmt.Sprintf("podman-pause-%x.scope", randBytes))
if err == nil {
return
}
}
if err != nil {
unified, err2 := cgroups.IsCgroup2UnifiedMode()
if err2 != nil {
logrus.Warnf("Failed to detect if running with cgroup unified: %v", err)
}
if RunsOnSystemd() && unified {
logrus.Warnf("Failed to add pause process to systemd sandbox cgroup: %v", err)
} else {
logrus.Debugf("Failed to add pause process to systemd sandbox cgroup: %v", err)
}
}
}
// CreateSCPCommand takes an existing command, appends the given arguments and returns a configured podman command for image scp
func CreateSCPCommand(cmd *exec.Cmd, command []string) *exec.Cmd {
cmd.Args = append(cmd.Args, command...)
cmd.Env = os.Environ()
cmd.Stderr = os.Stderr
cmd.Stdout = os.Stdout
return cmd
}
// LoginUser starts the user process on the host so that image scp can use systemd-run
func LoginUser(user string) (*exec.Cmd, error) {
sleep, err := exec.LookPath("sleep")
if err != nil {
return nil, err
}
machinectl, err := exec.LookPath("machinectl")
if err != nil {
return nil, err
}
cmd := exec.Command(machinectl, "shell", "-q", user+"@.host", sleep, "inf")
err = cmd.Start()
return cmd, err
}