common.go

// Copyright 2014 The rkt Authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

// Package common defines values shared by different parts
// of rkt (e.g. stage0 and stage1)
package common

import (
	"bufio"
	"errors"
	"fmt"
	"net"
	"os"
	"os/exec"
	"path/filepath"
	"runtime"
	"strconv"
	"strings"
	"syscall"
	"unsafe"

	"github.com/appc/spec/aci"
	"github.com/appc/spec/schema/types"
	"github.com/hashicorp/errwrap"

	"github.com/rkt/rkt/pkg/fileutil"
)

const (
	sharedVolumesDir = "/sharedVolumes"
	SharedVolumePerm = os.FileMode(0755)
	stage1Dir        = "/stage1"
	stage2Dir        = "/opt/stage2"
	AppsInfoDir      = "/appsinfo"

	EnvLockFd                    = "RKT_LOCK_FD"
	EnvSELinuxContext            = "RKT_SELINUX_CONTEXT"
	EnvSELinuxMountContext       = "RKT_SELINUX_MOUNT_CONTEXT"
	Stage1TreeStoreIDFilename    = "stage1TreeStoreID"
	AppTreeStoreIDFilename       = "treeStoreID"
	OverlayPreparedFilename      = "overlay-prepared"
	PrivateUsersPreparedFilename = "private-users-prepared"

	PrepareLock = "prepareLock"

	MetadataServicePort    = 18112
	MetadataServiceRegSock = "/run/rkt/metadata-svc.sock"

	APIServiceListenAddr = "localhost:15441"

	DefaultLocalConfigDir  = "/etc/rkt"
	DefaultSystemConfigDir = "/usr/lib/rkt"

	// Default perm bits for the regular files
	// within the stage1 directory. (e.g. image manifest,
	// pod manifest, stage1ID, etc).
	DefaultRegularFilePerm = os.FileMode(0640)

	// Default perm bits for the regular directories
	// within the stage1 directory.
	DefaultRegularDirPerm = os.FileMode(0750)

	// Enter command for crossing entrypoints.
	CrossingEnterCmd = "RKT_STAGE1_ENTERCMD"
	// Stage1 (PID) to enter, used by crossing entrypoints.
	CrossingEnterPID = "RKT_STAGE1_ENTERPID"
	// Stage2 (application name) to enter, optionally used by crossing entrypoints.
	CrossingEnterApp = "RKT_STAGE1_ENTERAPP"
)

const (
	FsMagicAUFS = 0x61756673 // https://goo.gl/CBwx43
	FsMagicZFS  = 0x2FC12FC1 // https://goo.gl/xTvzO5
)

// ErrOverlayUnsupported is the error determining whether OverlayFS is supported.
type ErrOverlayUnsupported string

func (e ErrOverlayUnsupported) Error() string {
	return string(e)
}

// Stage1ImagePath returns the path where the stage1 app image (unpacked ACI) is rooted,
// (i.e. where its contents are extracted during stage0).
func Stage1ImagePath(root string) string {
	return filepath.Join(root, stage1Dir)
}

// Stage1RootfsPath returns the path to the stage1 rootfs
func Stage1RootfsPath(root string) string {
	return filepath.Join(Stage1ImagePath(root), aci.RootfsDir)
}

// Stage1ManifestPath returns the path to the stage1's manifest file inside the expanded ACI.
func Stage1ManifestPath(root string) string {
	return filepath.Join(Stage1ImagePath(root), aci.ManifestFile)
}

// PodManifestPath returns the path in root to the Pod Manifest
func PodManifestPath(root string) string {
	return filepath.Join(root, "pod")
}

// PodCreatedPath returns the path in root to the Pod Created file used to
// denote the time of creation.
func PodCreatedPath(root string) string {
	return filepath.Join(root, "pod-created")
}

// PodManifestLockPath returns the path in root to the Pod Manifest lock file.
// This must be different from the PodManifestPath since mutations on the pod manifest file
// happen by overwriting the original file.
func PodManifestLockPath(root string) string {
	return filepath.Join(root, "pod.lck")
}

// AppsStatusesPathFromStage1Rootfs returns the path of the status dir for all apps.
// It receives the stage1 rootfs as parameter instead of the pod root.
func AppsStatusesPathFromStage1Rootfs(rootfs string) string {
	return filepath.Join(rootfs, "/rkt/status")
}

// AppsStatusesPath returns the path of the status dir for all apps.
func AppsStatusesPath(root string) string {
	return AppsStatusesPathFromStage1Rootfs(Stage1RootfsPath(root))
}

// AppStatusPath returns the path of the status file of an app.
func AppStatusPath(root, appName string) string {
	return filepath.Join(AppsStatusesPath(root), appName)
}

// AppStatusPathFromStage1Rootfs returns the path of the status file of an app.
// It receives the stage1 rootfs as parameter instead of the pod root.
func AppStatusPathFromStage1Rootfs(rootfs, appName string) string {
	return filepath.Join(AppsStatusesPathFromStage1Rootfs(rootfs), appName)
}

// AppCreatedPath returns the path of the ${appname}-created file, which is used to record
// the creation timestamp of the app.
func AppCreatedPath(root, appName string) string {
	return filepath.Join(AppsStatusesPath(root), fmt.Sprintf("%s-created", appName))
}

// AppCreatedPathFromStage1Rootfs returns the path of the ${appname}-created file,
// which is used to record the creation timestamp of the app.
// It receives the stage1 rootfs as parameter instead of the pod root.
func AppCreatedPathFromStage1Rootfs(rootfs, appName string) string {
	return filepath.Join(AppsStatusesPathFromStage1Rootfs(rootfs), fmt.Sprintf("%s-created", appName))
}

// AppStartedPath returns the path of the ${appname}-started file, which is used to record
// the start timestamp of the app.
func AppStartedPath(root, appName string) string {
	return filepath.Join(AppsStatusesPath(root), fmt.Sprintf("%s-started", appName))
}

// AppStartedPathFromStage1Rootfs returns the path of the ${appname}-started file, which is used to record
// the start timestamp of the app.
// It receives the stage1 rootfs as parameter instead of the pod root.
func AppStartedPathFromStage1Rootfs(rootfs, appName string) string {
	return filepath.Join(AppsStatusesPathFromStage1Rootfs(rootfs), fmt.Sprintf("%s-started", appName))
}

// AppsPath returns the path where the apps within a pod live.
func AppsPath(root string) string {
	return filepath.Join(Stage1RootfsPath(root), stage2Dir)
}

// AppPath returns the path to an app's rootfs.
func AppPath(root string, appName types.ACName) string {
	return filepath.Join(AppsPath(root), appName.String())
}

// AppRootfsPath returns the path to an app's rootfs.
func AppRootfsPath(root string, appName types.ACName) string {
	return filepath.Join(AppPath(root, appName), aci.RootfsDir)
}

// RelAppPath returns the path of an app relative to the stage1 chroot.
func RelAppPath(appName types.ACName) string {
	return filepath.Join(stage2Dir, appName.String())
}

// RelAppRootfsPath returns the path of an app's rootfs relative to the stage1 chroot.
func RelAppRootfsPath(appName types.ACName) string {
	return filepath.Join(RelAppPath(appName), aci.RootfsDir)
}

// ImageManifestPath returns the path to the app's manifest file of a pod.
func ImageManifestPath(root string, appName types.ACName) string {
	return filepath.Join(AppPath(root, appName), aci.ManifestFile)
}

// AppsInfoPath returns the path to the appsinfo directory of a pod.
func AppsInfoPath(root string) string {
	return filepath.Join(root, AppsInfoDir)
}

// AppInfoPath returns the path to the app's appsinfo directory of a pod.
func AppInfoPath(root string, appName types.ACName) string {
	return filepath.Join(AppsInfoPath(root), appName.String())
}

// AppTreeStoreIDPath returns the path to the app's treeStoreID file of a pod.
func AppTreeStoreIDPath(root string, appName types.ACName) string {
	return filepath.Join(AppInfoPath(root, appName), AppTreeStoreIDFilename)
}

// AppImageManifestPath returns the path to the app's ImageManifest file
func AppImageManifestPath(root string, appName types.ACName) string {
	return filepath.Join(AppInfoPath(root, appName), aci.ManifestFile)
}

// SharedVolumesPath returns the path to the shared (empty) volumes of a pod.
func SharedVolumesPath(root string) string {
	return filepath.Join(root, sharedVolumesDir)
}

// CreateSharedVolumesPath ensures the sharedVolumePath for the pod root passed
// in exists. It returns the shared volume path or an error.
func CreateSharedVolumesPath(root string) (string, error) {
	sharedVolPath := SharedVolumesPath(root)

	if err := os.MkdirAll(sharedVolPath, SharedVolumePerm); err != nil {
		return "", errwrap.Wrap(errors.New("could not create shared volumes directory"), err)
	}
	// In case it already existed and we didn't make it, ensure permissions are
	// what the caller expects them to be.
	if err := os.Chmod(sharedVolPath, SharedVolumePerm); err != nil {
		return "", errwrap.Wrap(fmt.Errorf("could not change permissions of %q", sharedVolPath), err)
	}

	return sharedVolPath, nil
}

// MetadataServicePublicURL returns the public URL used to host the metadata service
func MetadataServicePublicURL(ip net.IP, token string) string {
	return fmt.Sprintf("http://%v:%v/%v", ip, MetadataServicePort, token)
}

func GetRktLockFD() (int, error) {
	if v := os.Getenv(EnvLockFd); v != "" {
		fd, err := strconv.ParseUint(v, 10, 32)
		if err != nil {
			return -1, err
		}
		return int(fd), nil
	}
	return -1, fmt.Errorf("%v env var is not set", EnvLockFd)
}

// SupportsUserNS returns whether the kernel has CONFIG_USER_NS set
func SupportsUserNS() bool {
	if _, err := os.Stat("/proc/self/uid_map"); err == nil {
		return true
	}

	return false
}

// NetList implements the flag.Value interface to allow specification of --net with and without values
// Example: --net="all,net1:k1=v1;k2=v2,net2:l1=w1"
type NetList struct {
	mapping map[string]string
}

func (l *NetList) String() string {
	return strings.Join(l.Strings(), ",")
}

func (l *NetList) Set(value string) error {
	if l.mapping == nil {
		l.mapping = make(map[string]string)
	}
	for _, s := range strings.Split(value, ",") {
		netArgsPair := strings.Split(s, ":")
		netName := netArgsPair[0]

		if netName == "" {
			return fmt.Errorf("netname must not be empty")
		}

		if _, duplicate := l.mapping[netName]; duplicate {
			return fmt.Errorf("found duplicate netname %q", netName)
		}

		switch {
		case len(netArgsPair) == 1:
			l.mapping[netName] = ""
		case len(netArgsPair) == 2:
			if netName == "all" ||
				netName == "host" {
				return fmt.Errorf("arguments are not supported by special netname %q", netName)
			}
			l.mapping[netName] = netArgsPair[1]
		case len(netArgsPair) > 2:
			return fmt.Errorf("network %q provided with invalid arguments: %v", netName, netArgsPair[1:])
		default:
			return fmt.Errorf("unexpected case when processing network %q", s)
		}
	}
	return nil
}

func (l *NetList) Type() string {
	return "netList"
}

func (l *NetList) Strings() []string {
	if len(l.mapping) == 0 {
		return []string{"default"}
	}

	var list []string
	for k, v := range l.mapping {
		if v == "" {
			list = append(list, k)
		} else {
			list = append(list, fmt.Sprintf("%s:%s", k, v))
		}
	}
	return list
}

func (l *NetList) StringsOnlyNames() (list []string) {
	for k := range l.mapping {
		list = append(list, k)
	}

	return
}

// Check if host networking has been requested
func (l *NetList) Host() bool {
	return l.Specific("host")
}

// Check if 'none' (loopback only) networking has been requested
func (l *NetList) None() bool {
	return l.Specific("none")
}

// Check if the container needs to be put in a separate network namespace
func (l *NetList) Contained() bool {
	return !l.Host() && len(l.mapping) > 0
}

func (l *NetList) Specific(net string) bool {
	_, exists := l.mapping[net]
	return exists
}

func (l *NetList) SpecificArgs(net string) string {
	return l.mapping[net]
}

func (l *NetList) All() bool {
	return l.Specific("all")
}

// LookupPath search for bin in paths. If found, it returns its absolute path,
// if not, an error
func LookupPath(bin string, paths string) (string, error) {
	pathsArr := filepath.SplitList(paths)
	for _, path := range pathsArr {
		binPath := filepath.Join(path, bin)
		binAbsPath, err := filepath.Abs(binPath)
		if err != nil {
			return "", fmt.Errorf("unable to find absolute path for %s", binPath)
		}
		if fileutil.IsExecutable(binAbsPath) {
			return binAbsPath, nil
		}
	}
	return "", fmt.Errorf("unable to find %q in %q", bin, paths)
}

// SystemdVersion parses and returns the version of a given systemd binary
func SystemdVersion(systemdBinaryPath string) (int, error) {
	versionBytes, err := exec.Command(systemdBinaryPath, "--version").CombinedOutput()
	if err != nil {
		return -1, errwrap.Wrap(fmt.Errorf("unable to probe %s version", systemdBinaryPath), err)
	}
	versionStr := strings.SplitN(string(versionBytes), "\n", 2)[0]
	var version int
	n, err := fmt.Sscanf(versionStr, "systemd %d", &version)
	if err != nil || n != 1 {
		return -1, fmt.Errorf("cannot parse version: %q", versionStr)
	}

	return version, nil
}

// SupportsOverlay returns whether the operating system generally supports OverlayFS,
// returning an instance of ErrOverlayUnsupported which encodes the reason.
// It is sufficient to check for nil if the reason is not of interest.
func SupportsOverlay() error {
	// ignore exec.Command error, modprobe may not be present on the system,
	// or the kernel module will fail to load.
	// we'll find out by reading the side effect in /proc/filesystems
	_ = exec.Command("modprobe", "overlay").Run()

	f, err := os.Open("/proc/filesystems")
	if err != nil {
		// don't use errwrap so consumers can type-check on ErrOverlayUnsupported
		return ErrOverlayUnsupported(fmt.Sprintf("cannot open /proc/filesystems: %v", err))
	}
	defer f.Close()

	s := bufio.NewScanner(f)
	for s.Scan() {
		if s.Text() == "nodev\toverlay" {
			return nil
		}
	}

	return ErrOverlayUnsupported("overlay entry not present in /proc/filesystems")
}

// PathSupportsOverlay checks whether the given path is compatible with OverlayFS.
// This method also calls SupportsOverlay().
//
// It returns an instance of ErrOverlayUnsupported if OverlayFS is not supported
// or any other error if determining overlay support failed.
func PathSupportsOverlay(path string) error {
	if err := SupportsOverlay(); err != nil {
		// don't wrap since SupportsOverlay already returns ErrOverlayUnsupported
		return err
	}

	var data syscall.Statfs_t
	if err := syscall.Statfs(path, &data); err != nil {
		return errwrap.Wrap(fmt.Errorf("cannot statfs %q", path), err)
	}

	switch data.Type {
	case FsMagicAUFS:
		return ErrOverlayUnsupported("unsupported filesystem: aufs")
	case FsMagicZFS:
		return ErrOverlayUnsupported("unsupported filesystem: zfs")
	}

	dir, err := os.OpenFile(path, syscall.O_RDONLY|syscall.O_DIRECTORY, 0755)
	if err != nil {
		return errwrap.Wrap(fmt.Errorf("cannot open %q", path), err)
	}
	defer dir.Close()

	buf := make([]byte, 4096)
	// ReadDirent forwards to the raw syscall getdents(3),
	// passing the buffer size.
	n, err := syscall.ReadDirent(int(dir.Fd()), buf)
	if err != nil {
		return errwrap.Wrap(fmt.Errorf("cannot read directory %q", path), err)
	}

	offset := 0
	for offset < n {
		// offset overflow cannot happen, because Reclen
		// is being maintained by getdents(3), considering the buffer size.
		dirent := (*syscall.Dirent)(unsafe.Pointer(&buf[offset]))
		offset += int(dirent.Reclen)

		if dirent.Ino == 0 { // File absent in directory.
			continue
		}

		if dirent.Type == syscall.DT_UNKNOWN {
			return ErrOverlayUnsupported("unsupported filesystem: missing d_type support")
		}
	}

	return nil
}

// RemoveEmptyLines removes empty lines from the given string
// and breaks it up into a list of strings at newline characters
func RemoveEmptyLines(str string) []string {
	lines := make([]string, 0)

	for _, v := range strings.Split(str, "\n") {
		if len(v) > 0 {
			lines = append(lines, v)
		}
	}

	return lines
}

// GetExitStatus converts an error to an exit status. If it wasn't an exit
// status != 0 it returns the same error that it was called with
func GetExitStatus(err error) (int, error) {
	if err == nil {
		return 0, nil
	}
	if exiterr, ok := err.(*exec.ExitError); ok {
		// the program has exited with an exit code != 0
		if status, ok := exiterr.Sys().(syscall.WaitStatus); ok {
			return status.ExitStatus(), nil
		}
	}
	return -1, err
}

// GetOS returns the current ACI operating system (linux, windows etc...)
func GetOS() string {
	os, _ := GetOSArch()
	return os
}

// GetArch returns the current ACI architecture.
func GetArch() string {
	_, arch := GetOSArch()
	return arch
}

// reference to GOARM, needs to be globally, if in GetArch() function it resolves to zero
//go:linkname goarm runtime.goarm
var goarm uint8

func GetOSArch() (os string, arch string) {
	arch = runtime.GOARCH
	flavor := ""
	if arch == "arm" {
		flavor = strconv.Itoa(int(goarm))
	}
	os, arch, _ = types.ToAppcOSArch(runtime.GOOS, arch, flavor)
	return os, arch
}

// ImageNameToAppName converts the full name of image to an app name without special
// characters - we use it as a default app name when specyfing it is optional
func ImageNameToAppName(name types.ACIdentifier) (*types.ACName, error) {
	parts := strings.Split(name.String(), "/")
	last := parts[len(parts)-1]

	sn, err := types.SanitizeACName(last)
	if err != nil {
		return nil, err
	}

	return types.MustACName(sn), nil
}