package createconfig import ( "net" "os" "strconv" "strings" "syscall" "github.com/containers/image/manifest" "github.com/containers/libpod/libpod" "github.com/containers/libpod/libpod/define" "github.com/containers/libpod/pkg/namespaces" "github.com/containers/storage" "github.com/cri-o/ocicni/pkg/ocicni" "github.com/docker/go-connections/nat" spec "github.com/opencontainers/runtime-spec/specs-go" "github.com/opencontainers/runtime-tools/generate" "github.com/pkg/errors" "github.com/sirupsen/logrus" ) // Type constants const ( bps = iota iops ) // CreateResourceConfig represents resource elements in CreateConfig // structures type CreateResourceConfig struct { BlkioWeight uint16 // blkio-weight BlkioWeightDevice []string // blkio-weight-device CPUPeriod uint64 // cpu-period CPUQuota int64 // cpu-quota CPURtPeriod uint64 // cpu-rt-period CPURtRuntime int64 // cpu-rt-runtime CPUShares uint64 // cpu-shares CPUs float64 // cpus CPUsetCPUs string CPUsetMems string // cpuset-mems DeviceReadBps []string // device-read-bps DeviceReadIOps []string // device-read-iops DeviceWriteBps []string // device-write-bps DeviceWriteIOps []string // device-write-iops DisableOomKiller bool // oom-kill-disable KernelMemory int64 // kernel-memory Memory int64 //memory MemoryReservation int64 // memory-reservation MemorySwap int64 //memory-swap MemorySwappiness int // memory-swappiness OomScoreAdj int //oom-score-adj PidsLimit int64 // pids-limit ShmSize int64 Ulimit []string //ulimit } // CreateConfig is a pre OCI spec structure. It represents user input from varlink or the CLI type CreateConfig struct { Annotations map[string]string Args []string CapAdd []string // cap-add CapDrop []string // cap-drop CidFile string ConmonPidFile string Cgroupns string CgroupParent string // cgroup-parent Command []string // Full command that will be used UserCommand []string // User-entered command (or image CMD) Detach bool // detach Devices []string // device DNSOpt []string //dns-opt DNSSearch []string //dns-search DNSServers []string //dns Entrypoint []string //entrypoint Env map[string]string //env ExposedPorts map[nat.Port]struct{} GroupAdd []string // group-add HealthCheck *manifest.Schema2HealthConfig NoHosts bool HostAdd []string //add-host Hostname string //hostname HTTPProxy bool Init bool // init InitPath string //init-path Image string ImageID string BuiltinImgVolumes map[string]struct{} // volumes defined in the image config IDMappings *storage.IDMappingOptions ImageVolumeType string // how to handle the image volume, either bind, tmpfs, or ignore Interactive bool //interactive IpcMode namespaces.IpcMode //ipc IP6Address string //ipv6 IPAddress string //ip Labels map[string]string //label LinkLocalIP []string // link-local-ip LogDriver string // log-driver LogDriverOpt []string // log-opt MacAddress string //mac-address Name string //name NetMode namespaces.NetworkMode //net Network string //network NetworkAlias []string //network-alias PidMode namespaces.PidMode //pid Pod string //pod CgroupMode namespaces.CgroupMode //cgroup PortBindings nat.PortMap Privileged bool //privileged Publish []string //publish PublishAll bool //publish-all Quiet bool //quiet ReadOnlyRootfs bool //read-only ReadOnlyTmpfs bool //read-only-tmpfs Resources CreateResourceConfig RestartPolicy string Rm bool //rm StopSignal syscall.Signal // stop-signal StopTimeout uint // stop-timeout Sysctl map[string]string //sysctl Systemd bool Tmpfs []string // tmpfs Tty bool //tty UsernsMode namespaces.UsernsMode //userns User string //user UtsMode namespaces.UTSMode //uts Mounts []spec.Mount MountsFlag []string // mounts NamedVolumes []*libpod.ContainerNamedVolume Volumes []string //volume VolumesFrom []string WorkDir string //workdir LabelOpts []string //SecurityOpts NoNewPrivs bool //SecurityOpts ApparmorProfile string //SecurityOpts SeccompProfilePath string //SecurityOpts SecurityOpts []string Rootfs string Syslog bool // Whether to enable syslog on exit commands } func u32Ptr(i int64) *uint32 { u := uint32(i); return &u } func fmPtr(i int64) *os.FileMode { fm := os.FileMode(i); return &fm } // CreateBlockIO returns a LinuxBlockIO struct from a CreateConfig func (c *CreateConfig) CreateBlockIO() (*spec.LinuxBlockIO, error) { return c.createBlockIO() } func (c *CreateConfig) createExitCommand(runtime *libpod.Runtime) ([]string, error) { config, err := runtime.GetConfig() if err != nil { return nil, err } cmd, _ := os.Executable() command := []string{cmd, "--root", config.StorageConfig.GraphRoot, "--runroot", config.StorageConfig.RunRoot, "--log-level", logrus.GetLevel().String(), "--cgroup-manager", config.CgroupManager, "--tmpdir", config.TmpDir, } if config.OCIRuntime != "" { command = append(command, []string{"--runtime", config.OCIRuntime}...) } if config.StorageConfig.GraphDriverName != "" { command = append(command, []string{"--storage-driver", config.StorageConfig.GraphDriverName}...) } for _, opt := range config.StorageConfig.GraphDriverOptions { command = append(command, []string{"--storage-opt", opt}...) } if config.EventsLogger != "" { command = append(command, []string{"--events-backend", config.EventsLogger}...) } if c.Syslog { command = append(command, "--syslog", "true") } command = append(command, []string{"container", "cleanup"}...) if c.Rm { command = append(command, "--rm") } return command, nil } // GetContainerCreateOptions takes a CreateConfig and returns a slice of CtrCreateOptions func (c *CreateConfig) getContainerCreateOptions(runtime *libpod.Runtime, pod *libpod.Pod, mounts []spec.Mount, namedVolumes []*libpod.ContainerNamedVolume) ([]libpod.CtrCreateOption, error) { var options []libpod.CtrCreateOption var portBindings []ocicni.PortMapping var err error if c.Interactive { options = append(options, libpod.WithStdin()) } if c.Systemd && (strings.HasSuffix(c.Command[0], "init") || strings.HasSuffix(c.Command[0], "systemd")) { options = append(options, libpod.WithSystemd()) } if c.Name != "" { logrus.Debugf("setting container name %s", c.Name) options = append(options, libpod.WithName(c.Name)) } if c.Pod != "" { logrus.Debugf("adding container to pod %s", c.Pod) options = append(options, runtime.WithPod(pod)) } if len(c.PortBindings) > 0 { portBindings, err = c.CreatePortBindings() if err != nil { return nil, errors.Wrapf(err, "unable to create port bindings") } } if len(mounts) != 0 || len(namedVolumes) != 0 { destinations := []string{} // Take all mount and named volume destinations. for _, mount := range mounts { destinations = append(destinations, mount.Destination) } for _, volume := range namedVolumes { destinations = append(destinations, volume.Dest) } options = append(options, libpod.WithUserVolumes(destinations)) } if len(namedVolumes) != 0 { options = append(options, libpod.WithNamedVolumes(namedVolumes)) } if len(c.UserCommand) != 0 { options = append(options, libpod.WithCommand(c.UserCommand)) } // Add entrypoint unconditionally // If it's empty it's because it was explicitly set to "" or the image // does not have one options = append(options, libpod.WithEntrypoint(c.Entrypoint)) networks := make([]string, 0) userNetworks := c.NetMode.UserDefined() if IsPod(userNetworks) { userNetworks = "" } if userNetworks != "" { for _, netName := range strings.Split(userNetworks, ",") { if netName == "" { return nil, errors.Wrapf(err, "container networks %q invalid", networks) } networks = append(networks, netName) } } if c.NetMode.IsNS() { ns := c.NetMode.NS() if ns == "" { return nil, errors.Errorf("invalid empty user-defined network namespace") } _, err := os.Stat(ns) if err != nil { return nil, err } } else if c.NetMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.NetMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.NetMode.Container()) } options = append(options, libpod.WithNetNSFrom(connectedCtr)) } else if !c.NetMode.IsHost() && !c.NetMode.IsNone() { hasUserns := c.UsernsMode.IsContainer() || c.UsernsMode.IsNS() || len(c.IDMappings.UIDMap) > 0 || len(c.IDMappings.GIDMap) > 0 postConfigureNetNS := c.NetMode.IsSlirp4netns() || (hasUserns && !c.UsernsMode.IsHost()) options = append(options, libpod.WithNetNS(portBindings, postConfigureNetNS, string(c.NetMode), networks)) } if c.CgroupMode.IsNS() { ns := c.CgroupMode.NS() if ns == "" { return nil, errors.Errorf("invalid empty user-defined network namespace") } _, err := os.Stat(ns) if err != nil { return nil, err } } else if c.CgroupMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.CgroupMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.CgroupMode.Container()) } options = append(options, libpod.WithCgroupNSFrom(connectedCtr)) } if c.UsernsMode.IsNS() { ns := c.UsernsMode.NS() if ns == "" { return nil, errors.Errorf("invalid empty user-defined user namespace") } _, err := os.Stat(ns) if err != nil { return nil, err } options = append(options, libpod.WithIDMappings(*c.IDMappings)) } else if c.UsernsMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.UsernsMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.UsernsMode.Container()) } options = append(options, libpod.WithUserNSFrom(connectedCtr)) } else { options = append(options, libpod.WithIDMappings(*c.IDMappings)) } if c.PidMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.PidMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.PidMode.Container()) } options = append(options, libpod.WithPIDNSFrom(connectedCtr)) } if c.IpcMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.IpcMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.IpcMode.Container()) } options = append(options, libpod.WithIPCNSFrom(connectedCtr)) } if IsPod(string(c.UtsMode)) { options = append(options, libpod.WithUTSNSFromPod(pod)) } if c.UtsMode.IsContainer() { connectedCtr, err := runtime.LookupContainer(c.UtsMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.UtsMode.Container()) } options = append(options, libpod.WithUTSNSFrom(connectedCtr)) } // TODO: MNT, USER, CGROUP options = append(options, libpod.WithStopSignal(c.StopSignal)) options = append(options, libpod.WithStopTimeout(c.StopTimeout)) if len(c.DNSSearch) > 0 { options = append(options, libpod.WithDNSSearch(c.DNSSearch)) } if len(c.DNSServers) > 0 { if len(c.DNSServers) == 1 && strings.ToLower(c.DNSServers[0]) == "none" { options = append(options, libpod.WithUseImageResolvConf()) } else { options = append(options, libpod.WithDNS(c.DNSServers)) } } if len(c.DNSOpt) > 0 { options = append(options, libpod.WithDNSOption(c.DNSOpt)) } if c.NoHosts { options = append(options, libpod.WithUseImageHosts()) } if len(c.HostAdd) > 0 && !c.NoHosts { options = append(options, libpod.WithHosts(c.HostAdd)) } logPath := getLoggingPath(c.LogDriverOpt) if logPath != "" { options = append(options, libpod.WithLogPath(logPath)) } if c.LogDriver != "" { options = append(options, libpod.WithLogDriver(c.LogDriver)) } if c.IPAddress != "" { ip := net.ParseIP(c.IPAddress) if ip == nil { return nil, errors.Wrapf(define.ErrInvalidArg, "cannot parse %s as IP address", c.IPAddress) } else if ip.To4() == nil { return nil, errors.Wrapf(define.ErrInvalidArg, "%s is not an IPv4 address", c.IPAddress) } options = append(options, libpod.WithStaticIP(ip)) } options = append(options, libpod.WithPrivileged(c.Privileged)) useImageVolumes := c.ImageVolumeType == TypeBind // Gather up the options for NewContainer which consist of With... funcs options = append(options, libpod.WithRootFSFromImage(c.ImageID, c.Image, useImageVolumes)) options = append(options, libpod.WithSecLabels(c.LabelOpts)) options = append(options, libpod.WithConmonPidFile(c.ConmonPidFile)) options = append(options, libpod.WithLabels(c.Labels)) options = append(options, libpod.WithUser(c.User)) if c.IpcMode.IsHost() { options = append(options, libpod.WithShmDir("/dev/shm")) } else if c.IpcMode.IsContainer() { ctr, err := runtime.LookupContainer(c.IpcMode.Container()) if err != nil { return nil, errors.Wrapf(err, "container %q not found", c.IpcMode.Container()) } options = append(options, libpod.WithShmDir(ctr.ShmDir())) } options = append(options, libpod.WithShmSize(c.Resources.ShmSize)) options = append(options, libpod.WithGroups(c.GroupAdd)) if c.Rootfs != "" { options = append(options, libpod.WithRootFS(c.Rootfs)) } // Default used if not overridden on command line if c.CgroupParent != "" { options = append(options, libpod.WithCgroupParent(c.CgroupParent)) } if c.RestartPolicy != "" { if c.RestartPolicy == "unless-stopped" { return nil, errors.Wrapf(define.ErrInvalidArg, "the unless-stopped restart policy is not supported") } split := strings.Split(c.RestartPolicy, ":") if len(split) > 1 { numTries, err := strconv.Atoi(split[1]) if err != nil { return nil, errors.Wrapf(err, "%s is not a valid number of retries for restart policy", split[1]) } if numTries < 0 { return nil, errors.Wrapf(define.ErrInvalidArg, "restart policy requires a positive number of retries") } options = append(options, libpod.WithRestartRetries(uint(numTries))) } options = append(options, libpod.WithRestartPolicy(split[0])) } // Always use a cleanup process to clean up Podman after termination exitCmd, err := c.createExitCommand(runtime) if err != nil { return nil, err } options = append(options, libpod.WithExitCommand(exitCmd)) if c.HealthCheck != nil { options = append(options, libpod.WithHealthCheck(c.HealthCheck)) logrus.Debugf("New container has a health check") } return options, nil } // CreatePortBindings iterates ports mappings and exposed ports into a format CNI understands func (c *CreateConfig) CreatePortBindings() ([]ocicni.PortMapping, error) { return NatToOCIPortBindings(c.PortBindings) } // NatToOCIPortBindings iterates a nat.portmap slice and creates []ocicni portmapping slice func NatToOCIPortBindings(ports nat.PortMap) ([]ocicni.PortMapping, error) { var portBindings []ocicni.PortMapping for containerPb, hostPb := range ports { var pm ocicni.PortMapping pm.ContainerPort = int32(containerPb.Int()) for _, i := range hostPb { var hostPort int var err error pm.HostIP = i.HostIP if i.HostPort == "" { hostPort = containerPb.Int() } else { hostPort, err = strconv.Atoi(i.HostPort) if err != nil { return nil, errors.Wrapf(err, "unable to convert host port to integer") } } pm.HostPort = int32(hostPort) pm.Protocol = containerPb.Proto() portBindings = append(portBindings, pm) } } return portBindings, nil } // AddPrivilegedDevices iterates through host devices and adds all // host devices to the spec func (c *CreateConfig) AddPrivilegedDevices(g *generate.Generator) error { return c.addPrivilegedDevices(g) }