package container import ( "bytes" "encoding/json" "fmt" "os" "path" "path/filepath" "reflect" "regexp" "strconv" "strings" "time" "github.com/docker/cli/cli/command" "github.com/docker/cli/cli/compose/loader" "github.com/docker/cli/opts" "github.com/docker/docker/api/types/container" mounttypes "github.com/docker/docker/api/types/mount" networktypes "github.com/docker/docker/api/types/network" "github.com/docker/docker/api/types/strslice" "github.com/docker/docker/errdefs" "github.com/docker/go-connections/nat" "github.com/pkg/errors" "github.com/sirupsen/logrus" "github.com/spf13/pflag" cdi "tags.cncf.io/container-device-interface/pkg/parser" ) const ( // TODO(thaJeztah): define these in the API-types, or query available defaults // from the daemon, or require "local" profiles to be an absolute path or // relative paths starting with "./". The daemon-config has consts for this // but we don't want to import that package: // https://github.com/moby/moby/blob/v23.0.0/daemon/config/config.go#L63-L67 // seccompProfileDefault is the built-in default seccomp profile. seccompProfileDefault = "builtin" // seccompProfileUnconfined is a special profile name for seccomp to use an // "unconfined" seccomp profile. seccompProfileUnconfined = "unconfined" ) var deviceCgroupRuleRegexp = regexp.MustCompile(`^[acb] ([0-9]+|\*):([0-9]+|\*) [rwm]{1,3}$`) // containerOptions is a data object with all the options for creating a container type containerOptions struct { attach opts.ListOpts volumes opts.ListOpts tmpfs opts.ListOpts mounts opts.MountOpt blkioWeightDevice opts.WeightdeviceOpt deviceReadBps opts.ThrottledeviceOpt deviceWriteBps opts.ThrottledeviceOpt links opts.ListOpts aliases opts.ListOpts linkLocalIPs opts.ListOpts deviceReadIOps opts.ThrottledeviceOpt deviceWriteIOps opts.ThrottledeviceOpt env opts.ListOpts labels opts.ListOpts deviceCgroupRules opts.ListOpts devices opts.ListOpts gpus opts.GpuOpts ulimits *opts.UlimitOpt sysctls *opts.MapOpts publish opts.ListOpts expose opts.ListOpts dns opts.ListOpts dnsSearch opts.ListOpts dnsOptions opts.ListOpts extraHosts opts.ListOpts volumesFrom opts.ListOpts envFile opts.ListOpts capAdd opts.ListOpts capDrop opts.ListOpts groupAdd opts.ListOpts securityOpt opts.ListOpts storageOpt opts.ListOpts labelsFile opts.ListOpts loggingOpts opts.ListOpts privileged bool pidMode string utsMode string usernsMode string cgroupnsMode string publishAll bool stdin bool tty bool oomKillDisable bool oomScoreAdj int containerIDFile string entrypoint string hostname string domainname string memory opts.MemBytes memoryReservation opts.MemBytes memorySwap opts.MemSwapBytes kernelMemory opts.MemBytes user string workingDir string cpuCount int64 cpuShares int64 cpuPercent int64 cpuPeriod int64 cpuRealtimePeriod int64 cpuRealtimeRuntime int64 cpuQuota int64 cpus opts.NanoCPUs cpusetCpus string cpusetMems string blkioWeight uint16 ioMaxBandwidth opts.MemBytes ioMaxIOps uint64 swappiness int64 netMode opts.NetworkOpt macAddress string ipv4Address string ipv6Address string ipcMode string pidsLimit int64 restartPolicy string readonlyRootfs bool loggingDriver string cgroupParent string volumeDriver string stopSignal string stopTimeout int isolation string shmSize opts.MemBytes noHealthcheck bool healthCmd string healthInterval time.Duration healthTimeout time.Duration healthStartPeriod time.Duration healthStartInterval time.Duration healthRetries int runtime string autoRemove bool init bool annotations *opts.MapOpts Image string Args []string } // addFlags adds all command line flags that will be used by parse to the FlagSet func addFlags(flags *pflag.FlagSet) *containerOptions { copts := &containerOptions{ aliases: opts.NewListOpts(nil), attach: opts.NewListOpts(validateAttach), blkioWeightDevice: opts.NewWeightdeviceOpt(opts.ValidateWeightDevice), capAdd: opts.NewListOpts(nil), capDrop: opts.NewListOpts(nil), dns: opts.NewListOpts(opts.ValidateIPAddress), dnsOptions: opts.NewListOpts(nil), dnsSearch: opts.NewListOpts(opts.ValidateDNSSearch), deviceCgroupRules: opts.NewListOpts(validateDeviceCgroupRule), deviceReadBps: opts.NewThrottledeviceOpt(opts.ValidateThrottleBpsDevice), deviceReadIOps: opts.NewThrottledeviceOpt(opts.ValidateThrottleIOpsDevice), deviceWriteBps: opts.NewThrottledeviceOpt(opts.ValidateThrottleBpsDevice), deviceWriteIOps: opts.NewThrottledeviceOpt(opts.ValidateThrottleIOpsDevice), devices: opts.NewListOpts(nil), // devices can only be validated after we know the server OS env: opts.NewListOpts(opts.ValidateEnv), envFile: opts.NewListOpts(nil), expose: opts.NewListOpts(nil), extraHosts: opts.NewListOpts(opts.ValidateExtraHost), groupAdd: opts.NewListOpts(nil), labels: opts.NewListOpts(opts.ValidateLabel), labelsFile: opts.NewListOpts(nil), linkLocalIPs: opts.NewListOpts(nil), links: opts.NewListOpts(opts.ValidateLink), loggingOpts: opts.NewListOpts(nil), publish: opts.NewListOpts(nil), securityOpt: opts.NewListOpts(nil), storageOpt: opts.NewListOpts(nil), sysctls: opts.NewMapOpts(nil, opts.ValidateSysctl), tmpfs: opts.NewListOpts(nil), ulimits: opts.NewUlimitOpt(nil), volumes: opts.NewListOpts(nil), volumesFrom: opts.NewListOpts(nil), annotations: opts.NewMapOpts(nil, nil), } // General purpose flags flags.VarP(&copts.attach, "attach", "a", "Attach to STDIN, STDOUT or STDERR") flags.Var(&copts.deviceCgroupRules, "device-cgroup-rule", "Add a rule to the cgroup allowed devices list") flags.Var(&copts.devices, "device", "Add a host device to the container") flags.Var(&copts.gpus, "gpus", "GPU devices to add to the container ('all' to pass all GPUs)") flags.SetAnnotation("gpus", "version", []string{"1.40"}) flags.VarP(&copts.env, "env", "e", "Set environment variables") flags.Var(&copts.envFile, "env-file", "Read in a file of environment variables") flags.StringVar(&copts.entrypoint, "entrypoint", "", "Overwrite the default ENTRYPOINT of the image") flags.Var(&copts.groupAdd, "group-add", "Add additional groups to join") flags.StringVarP(&copts.hostname, "hostname", "h", "", "Container host name") flags.StringVar(&copts.domainname, "domainname", "", "Container NIS domain name") flags.BoolVarP(&copts.stdin, "interactive", "i", false, "Keep STDIN open even if not attached") flags.VarP(&copts.labels, "label", "l", "Set meta data on a container") flags.Var(&copts.labelsFile, "label-file", "Read in a line delimited file of labels") flags.BoolVar(&copts.readonlyRootfs, "read-only", false, "Mount the container's root filesystem as read only") flags.StringVar(&copts.restartPolicy, "restart", string(container.RestartPolicyDisabled), "Restart policy to apply when a container exits") flags.StringVar(&copts.stopSignal, "stop-signal", "", "Signal to stop the container") flags.IntVar(&copts.stopTimeout, "stop-timeout", 0, "Timeout (in seconds) to stop a container") flags.SetAnnotation("stop-timeout", "version", []string{"1.25"}) flags.Var(copts.sysctls, "sysctl", "Sysctl options") flags.BoolVarP(&copts.tty, "tty", "t", false, "Allocate a pseudo-TTY") flags.Var(copts.ulimits, "ulimit", "Ulimit options") flags.StringVarP(&copts.user, "user", "u", "", "Username or UID (format: [:])") flags.StringVarP(&copts.workingDir, "workdir", "w", "", "Working directory inside the container") flags.BoolVar(&copts.autoRemove, "rm", false, "Automatically remove the container when it exits") // Security flags.Var(&copts.capAdd, "cap-add", "Add Linux capabilities") flags.Var(&copts.capDrop, "cap-drop", "Drop Linux capabilities") flags.BoolVar(&copts.privileged, "privileged", false, "Give extended privileges to this container") flags.Var(&copts.securityOpt, "security-opt", "Security Options") flags.StringVar(&copts.usernsMode, "userns", "", "User namespace to use") flags.StringVar(&copts.cgroupnsMode, "cgroupns", "", `Cgroup namespace to use (host|private) 'host': Run the container in the Docker host's cgroup namespace 'private': Run the container in its own private cgroup namespace '': Use the cgroup namespace as configured by the default-cgroupns-mode option on the daemon (default)`) flags.SetAnnotation("cgroupns", "version", []string{"1.41"}) // Network and port publishing flag flags.Var(&copts.extraHosts, "add-host", "Add a custom host-to-IP mapping (host:ip)") flags.Var(&copts.dns, "dns", "Set custom DNS servers") // We allow for both "--dns-opt" and "--dns-option", although the latter is the recommended way. // This is to be consistent with service create/update flags.Var(&copts.dnsOptions, "dns-opt", "Set DNS options") flags.Var(&copts.dnsOptions, "dns-option", "Set DNS options") flags.MarkHidden("dns-opt") flags.Var(&copts.dnsSearch, "dns-search", "Set custom DNS search domains") flags.Var(&copts.expose, "expose", "Expose a port or a range of ports") flags.StringVar(&copts.ipv4Address, "ip", "", "IPv4 address (e.g., 172.30.100.104)") flags.StringVar(&copts.ipv6Address, "ip6", "", "IPv6 address (e.g., 2001:db8::33)") flags.Var(&copts.links, "link", "Add link to another container") flags.Var(&copts.linkLocalIPs, "link-local-ip", "Container IPv4/IPv6 link-local addresses") flags.StringVar(&copts.macAddress, "mac-address", "", "Container MAC address (e.g., 92:d0:c6:0a:29:33)") flags.VarP(&copts.publish, "publish", "p", "Publish a container's port(s) to the host") flags.BoolVarP(&copts.publishAll, "publish-all", "P", false, "Publish all exposed ports to random ports") // We allow for both "--net" and "--network", although the latter is the recommended way. flags.Var(&copts.netMode, "net", "Connect a container to a network") flags.Var(&copts.netMode, "network", "Connect a container to a network") flags.MarkHidden("net") // We allow for both "--net-alias" and "--network-alias", although the latter is the recommended way. flags.Var(&copts.aliases, "net-alias", "Add network-scoped alias for the container") flags.Var(&copts.aliases, "network-alias", "Add network-scoped alias for the container") flags.MarkHidden("net-alias") // Logging and storage flags.StringVar(&copts.loggingDriver, "log-driver", "", "Logging driver for the container") flags.StringVar(&copts.volumeDriver, "volume-driver", "", "Optional volume driver for the container") flags.Var(&copts.loggingOpts, "log-opt", "Log driver options") flags.Var(&copts.storageOpt, "storage-opt", "Storage driver options for the container") flags.Var(&copts.tmpfs, "tmpfs", "Mount a tmpfs directory") flags.Var(&copts.volumesFrom, "volumes-from", "Mount volumes from the specified container(s)") flags.VarP(&copts.volumes, "volume", "v", "Bind mount a volume") flags.Var(&copts.mounts, "mount", "Attach a filesystem mount to the container") // Health-checking flags.StringVar(&copts.healthCmd, "health-cmd", "", "Command to run to check health") flags.DurationVar(&copts.healthInterval, "health-interval", 0, "Time between running the check (ms|s|m|h) (default 0s)") flags.IntVar(&copts.healthRetries, "health-retries", 0, "Consecutive failures needed to report unhealthy") flags.DurationVar(&copts.healthTimeout, "health-timeout", 0, "Maximum time to allow one check to run (ms|s|m|h) (default 0s)") flags.DurationVar(&copts.healthStartPeriod, "health-start-period", 0, "Start period for the container to initialize before starting health-retries countdown (ms|s|m|h) (default 0s)") flags.SetAnnotation("health-start-period", "version", []string{"1.29"}) flags.DurationVar(&copts.healthStartInterval, "health-start-interval", 0, "Time between running the check during the start period (ms|s|m|h) (default 0s)") flags.SetAnnotation("health-start-interval", "version", []string{"1.44"}) flags.BoolVar(&copts.noHealthcheck, "no-healthcheck", false, "Disable any container-specified HEALTHCHECK") // Resource management flags.Uint16Var(&copts.blkioWeight, "blkio-weight", 0, "Block IO (relative weight), between 10 and 1000, or 0 to disable (default 0)") flags.Var(&copts.blkioWeightDevice, "blkio-weight-device", "Block IO weight (relative device weight)") flags.StringVar(&copts.containerIDFile, "cidfile", "", "Write the container ID to the file") flags.StringVar(&copts.cpusetCpus, "cpuset-cpus", "", "CPUs in which to allow execution (0-3, 0,1)") flags.StringVar(&copts.cpusetMems, "cpuset-mems", "", "MEMs in which to allow execution (0-3, 0,1)") flags.Int64Var(&copts.cpuCount, "cpu-count", 0, "CPU count (Windows only)") flags.SetAnnotation("cpu-count", "ostype", []string{"windows"}) flags.Int64Var(&copts.cpuPercent, "cpu-percent", 0, "CPU percent (Windows only)") flags.SetAnnotation("cpu-percent", "ostype", []string{"windows"}) flags.Int64Var(&copts.cpuPeriod, "cpu-period", 0, "Limit CPU CFS (Completely Fair Scheduler) period") flags.Int64Var(&copts.cpuQuota, "cpu-quota", 0, "Limit CPU CFS (Completely Fair Scheduler) quota") flags.Int64Var(&copts.cpuRealtimePeriod, "cpu-rt-period", 0, "Limit CPU real-time period in microseconds") flags.SetAnnotation("cpu-rt-period", "version", []string{"1.25"}) flags.Int64Var(&copts.cpuRealtimeRuntime, "cpu-rt-runtime", 0, "Limit CPU real-time runtime in microseconds") flags.SetAnnotation("cpu-rt-runtime", "version", []string{"1.25"}) flags.Int64VarP(&copts.cpuShares, "cpu-shares", "c", 0, "CPU shares (relative weight)") flags.Var(&copts.cpus, "cpus", "Number of CPUs") flags.SetAnnotation("cpus", "version", []string{"1.25"}) flags.Var(&copts.deviceReadBps, "device-read-bps", "Limit read rate (bytes per second) from a device") flags.Var(&copts.deviceReadIOps, "device-read-iops", "Limit read rate (IO per second) from a device") flags.Var(&copts.deviceWriteBps, "device-write-bps", "Limit write rate (bytes per second) to a device") flags.Var(&copts.deviceWriteIOps, "device-write-iops", "Limit write rate (IO per second) to a device") flags.Var(&copts.ioMaxBandwidth, "io-maxbandwidth", "Maximum IO bandwidth limit for the system drive (Windows only)") flags.SetAnnotation("io-maxbandwidth", "ostype", []string{"windows"}) flags.Uint64Var(&copts.ioMaxIOps, "io-maxiops", 0, "Maximum IOps limit for the system drive (Windows only)") flags.SetAnnotation("io-maxiops", "ostype", []string{"windows"}) flags.Var(&copts.kernelMemory, "kernel-memory", "Kernel memory limit") flags.VarP(&copts.memory, "memory", "m", "Memory limit") flags.Var(&copts.memoryReservation, "memory-reservation", "Memory soft limit") flags.Var(&copts.memorySwap, "memory-swap", "Swap limit equal to memory plus swap: '-1' to enable unlimited swap") flags.Int64Var(&copts.swappiness, "memory-swappiness", -1, "Tune container memory swappiness (0 to 100)") flags.BoolVar(&copts.oomKillDisable, "oom-kill-disable", false, "Disable OOM Killer") flags.IntVar(&copts.oomScoreAdj, "oom-score-adj", 0, "Tune host's OOM preferences (-1000 to 1000)") flags.Int64Var(&copts.pidsLimit, "pids-limit", 0, "Tune container pids limit (set -1 for unlimited)") // Low-level execution (cgroups, namespaces, ...) flags.StringVar(&copts.cgroupParent, "cgroup-parent", "", "Optional parent cgroup for the container") flags.StringVar(&copts.ipcMode, "ipc", "", "IPC mode to use") flags.StringVar(&copts.isolation, "isolation", "", "Container isolation technology") flags.StringVar(&copts.pidMode, "pid", "", "PID namespace to use") flags.Var(&copts.shmSize, "shm-size", "Size of /dev/shm") flags.StringVar(&copts.utsMode, "uts", "", "UTS namespace to use") flags.StringVar(&copts.runtime, "runtime", "", "Runtime to use for this container") flags.BoolVar(&copts.init, "init", false, "Run an init inside the container that forwards signals and reaps processes") flags.SetAnnotation("init", "version", []string{"1.25"}) flags.Var(copts.annotations, "annotation", "Add an annotation to the container (passed through to the OCI runtime)") flags.SetAnnotation("annotation", "version", []string{"1.43"}) return copts } type containerConfig struct { Config *container.Config HostConfig *container.HostConfig NetworkingConfig *networktypes.NetworkingConfig } // parse parses the args for the specified command and generates a Config, // a HostConfig and returns them with the specified command. // If the specified args are not valid, it will return an error. // //nolint:gocyclo func parse(flags *pflag.FlagSet, copts *containerOptions, serverOS string) (*containerConfig, error) { var ( attachStdin = copts.attach.Get("stdin") attachStdout = copts.attach.Get("stdout") attachStderr = copts.attach.Get("stderr") ) // Validate the input mac address if copts.macAddress != "" { if _, err := opts.ValidateMACAddress(copts.macAddress); err != nil { return nil, errors.Errorf("%s is not a valid mac address", copts.macAddress) } } if copts.stdin { attachStdin = true } // If -a is not set, attach to stdout and stderr if copts.attach.Len() == 0 { attachStdout = true attachStderr = true } var err error swappiness := copts.swappiness if swappiness != -1 && (swappiness < 0 || swappiness > 100) { return nil, errors.Errorf("invalid value: %d. Valid memory swappiness range is 0-100", swappiness) } mounts := copts.mounts.Value() if len(mounts) > 0 && copts.volumeDriver != "" { logrus.Warn("`--volume-driver` is ignored for volumes specified via `--mount`. Use `--mount type=volume,volume-driver=...` instead.") } var binds []string volumes := copts.volumes.GetMap() // add any bind targets to the list of container volumes for bind := range copts.volumes.GetMap() { parsed, err := loader.ParseVolume(bind) if err != nil { return nil, err } if parsed.Source != "" { toBind := bind if parsed.Type == string(mounttypes.TypeBind) { if hostPart, targetPath, ok := strings.Cut(bind, ":"); ok { if strings.HasPrefix(hostPart, "."+string(filepath.Separator)) || hostPart == "." { if absHostPart, err := filepath.Abs(hostPart); err == nil { hostPart = absHostPart } } toBind = hostPart + ":" + targetPath } } // after creating the bind mount we want to delete it from the copts.volumes values because // we do not want bind mounts being committed to image configs binds = append(binds, toBind) // We should delete from the map (`volumes`) here, as deleting from copts.volumes will not work if // there are duplicates entries. delete(volumes, bind) } } // Can't evaluate options passed into --tmpfs until we actually mount tmpfs := make(map[string]string) for _, t := range copts.tmpfs.GetAll() { k, v, _ := strings.Cut(t, ":") tmpfs[k] = v } var ( runCmd strslice.StrSlice entrypoint strslice.StrSlice ) if len(copts.Args) > 0 { runCmd = copts.Args } if copts.entrypoint != "" { entrypoint = strslice.StrSlice{copts.entrypoint} } else if flags.Changed("entrypoint") { // if `--entrypoint=` is parsed then Entrypoint is reset entrypoint = []string{""} } publishOpts := copts.publish.GetAll() var ( ports map[nat.Port]struct{} portBindings map[nat.Port][]nat.PortBinding convertedOpts []string ) convertedOpts, err = convertToStandardNotation(publishOpts) if err != nil { return nil, err } ports, portBindings, err = nat.ParsePortSpecs(convertedOpts) if err != nil { return nil, err } // Merge in exposed ports to the map of published ports for _, e := range copts.expose.GetAll() { if strings.Contains(e, ":") { return nil, errors.Errorf("invalid port format for --expose: %s", e) } // support two formats for expose, original format /[] // or /[] proto, port := nat.SplitProtoPort(e) // parse the start and end port and create a sequence of ports to expose // if expose a port, the start and end port are the same start, end, err := nat.ParsePortRange(port) if err != nil { return nil, errors.Errorf("invalid range format for --expose: %s, error: %s", e, err) } for i := start; i <= end; i++ { p, err := nat.NewPort(proto, strconv.FormatUint(i, 10)) if err != nil { return nil, err } if _, exists := ports[p]; !exists { ports[p] = struct{}{} } } } // validate and parse device mappings. Note we do late validation of the // device path (as opposed to during flag parsing), as at the time we are // parsing flags, we haven't yet sent a _ping to the daemon to determine // what operating system it is. deviceMappings := []container.DeviceMapping{} var cdiDeviceNames []string for _, device := range copts.devices.GetAll() { var ( validated string deviceMapping container.DeviceMapping err error ) if cdi.IsQualifiedName(device) { cdiDeviceNames = append(cdiDeviceNames, device) continue } validated, err = validateDevice(device, serverOS) if err != nil { return nil, err } deviceMapping, err = parseDevice(validated, serverOS) if err != nil { return nil, err } deviceMappings = append(deviceMappings, deviceMapping) } // collect all the environment variables for the container envVariables, err := opts.ReadKVEnvStrings(copts.envFile.GetAll(), copts.env.GetAll()) if err != nil { return nil, err } // collect all the labels for the container labels, err := opts.ReadKVStrings(copts.labelsFile.GetAll(), copts.labels.GetAll()) if err != nil { return nil, err } pidMode := container.PidMode(copts.pidMode) if !pidMode.Valid() { return nil, errors.Errorf("--pid: invalid PID mode") } utsMode := container.UTSMode(copts.utsMode) if !utsMode.Valid() { return nil, errors.Errorf("--uts: invalid UTS mode") } usernsMode := container.UsernsMode(copts.usernsMode) if !usernsMode.Valid() { return nil, errors.Errorf("--userns: invalid USER mode") } cgroupnsMode := container.CgroupnsMode(copts.cgroupnsMode) if !cgroupnsMode.Valid() { return nil, errors.Errorf("--cgroupns: invalid CGROUP mode") } restartPolicy, err := opts.ParseRestartPolicy(copts.restartPolicy) if err != nil { return nil, err } loggingOpts, err := parseLoggingOpts(copts.loggingDriver, copts.loggingOpts.GetAll()) if err != nil { return nil, err } securityOpts, err := parseSecurityOpts(copts.securityOpt.GetAll()) if err != nil { return nil, err } securityOpts, maskedPaths, readonlyPaths := parseSystemPaths(securityOpts) storageOpts, err := parseStorageOpts(copts.storageOpt.GetAll()) if err != nil { return nil, err } // Healthcheck var healthConfig *container.HealthConfig haveHealthSettings := copts.healthCmd != "" || copts.healthInterval != 0 || copts.healthTimeout != 0 || copts.healthStartPeriod != 0 || copts.healthRetries != 0 || copts.healthStartInterval != 0 if copts.noHealthcheck { if haveHealthSettings { return nil, errors.Errorf("--no-healthcheck conflicts with --health-* options") } healthConfig = &container.HealthConfig{Test: strslice.StrSlice{"NONE"}} } else if haveHealthSettings { var probe strslice.StrSlice if copts.healthCmd != "" { probe = []string{"CMD-SHELL", copts.healthCmd} } if copts.healthInterval < 0 { return nil, errors.Errorf("--health-interval cannot be negative") } if copts.healthTimeout < 0 { return nil, errors.Errorf("--health-timeout cannot be negative") } if copts.healthRetries < 0 { return nil, errors.Errorf("--health-retries cannot be negative") } if copts.healthStartPeriod < 0 { return nil, fmt.Errorf("--health-start-period cannot be negative") } if copts.healthStartInterval < 0 { return nil, fmt.Errorf("--health-start-interval cannot be negative") } healthConfig = &container.HealthConfig{ Test: probe, Interval: copts.healthInterval, Timeout: copts.healthTimeout, StartPeriod: copts.healthStartPeriod, StartInterval: copts.healthStartInterval, Retries: copts.healthRetries, } } deviceRequests := copts.gpus.Value() if len(cdiDeviceNames) > 0 { cdiDeviceRequest := container.DeviceRequest{ Driver: "cdi", DeviceIDs: cdiDeviceNames, } deviceRequests = append(deviceRequests, cdiDeviceRequest) } resources := container.Resources{ CgroupParent: copts.cgroupParent, Memory: copts.memory.Value(), MemoryReservation: copts.memoryReservation.Value(), MemorySwap: copts.memorySwap.Value(), MemorySwappiness: &copts.swappiness, KernelMemory: copts.kernelMemory.Value(), OomKillDisable: &copts.oomKillDisable, NanoCPUs: copts.cpus.Value(), CPUCount: copts.cpuCount, CPUPercent: copts.cpuPercent, CPUShares: copts.cpuShares, CPUPeriod: copts.cpuPeriod, CpusetCpus: copts.cpusetCpus, CpusetMems: copts.cpusetMems, CPUQuota: copts.cpuQuota, CPURealtimePeriod: copts.cpuRealtimePeriod, CPURealtimeRuntime: copts.cpuRealtimeRuntime, PidsLimit: &copts.pidsLimit, BlkioWeight: copts.blkioWeight, BlkioWeightDevice: copts.blkioWeightDevice.GetList(), BlkioDeviceReadBps: copts.deviceReadBps.GetList(), BlkioDeviceWriteBps: copts.deviceWriteBps.GetList(), BlkioDeviceReadIOps: copts.deviceReadIOps.GetList(), BlkioDeviceWriteIOps: copts.deviceWriteIOps.GetList(), IOMaximumIOps: copts.ioMaxIOps, IOMaximumBandwidth: uint64(copts.ioMaxBandwidth), Ulimits: copts.ulimits.GetList(), DeviceCgroupRules: copts.deviceCgroupRules.GetAll(), Devices: deviceMappings, DeviceRequests: deviceRequests, } config := &container.Config{ Hostname: copts.hostname, Domainname: copts.domainname, ExposedPorts: ports, User: copts.user, Tty: copts.tty, OpenStdin: copts.stdin, AttachStdin: attachStdin, AttachStdout: attachStdout, AttachStderr: attachStderr, Env: envVariables, Cmd: runCmd, Image: copts.Image, Volumes: volumes, MacAddress: copts.macAddress, Entrypoint: entrypoint, WorkingDir: copts.workingDir, Labels: opts.ConvertKVStringsToMap(labels), StopSignal: copts.stopSignal, Healthcheck: healthConfig, } if flags.Changed("stop-timeout") { config.StopTimeout = &copts.stopTimeout } hostConfig := &container.HostConfig{ Binds: binds, ContainerIDFile: copts.containerIDFile, OomScoreAdj: copts.oomScoreAdj, AutoRemove: copts.autoRemove, Privileged: copts.privileged, PortBindings: portBindings, Links: copts.links.GetAll(), PublishAllPorts: copts.publishAll, // Make sure the dns fields are never nil. // New containers don't ever have those fields nil, // but pre created containers can still have those nil values. // See https://github.com/docker/docker/pull/17779 // for a more detailed explanation on why we don't want that. DNS: copts.dns.GetAllOrEmpty(), DNSSearch: copts.dnsSearch.GetAllOrEmpty(), DNSOptions: copts.dnsOptions.GetAllOrEmpty(), ExtraHosts: copts.extraHosts.GetAll(), VolumesFrom: copts.volumesFrom.GetAll(), IpcMode: container.IpcMode(copts.ipcMode), NetworkMode: container.NetworkMode(copts.netMode.NetworkMode()), PidMode: pidMode, UTSMode: utsMode, UsernsMode: usernsMode, CgroupnsMode: cgroupnsMode, CapAdd: strslice.StrSlice(copts.capAdd.GetAll()), CapDrop: strslice.StrSlice(copts.capDrop.GetAll()), GroupAdd: copts.groupAdd.GetAll(), RestartPolicy: restartPolicy, SecurityOpt: securityOpts, StorageOpt: storageOpts, ReadonlyRootfs: copts.readonlyRootfs, LogConfig: container.LogConfig{Type: copts.loggingDriver, Config: loggingOpts}, VolumeDriver: copts.volumeDriver, Isolation: container.Isolation(copts.isolation), ShmSize: copts.shmSize.Value(), Resources: resources, Tmpfs: tmpfs, Sysctls: copts.sysctls.GetAll(), Runtime: copts.runtime, Mounts: mounts, MaskedPaths: maskedPaths, ReadonlyPaths: readonlyPaths, Annotations: copts.annotations.GetAll(), } if copts.autoRemove && !hostConfig.RestartPolicy.IsNone() { return nil, errors.Errorf("Conflicting options: --restart and --rm") } // only set this value if the user provided the flag, else it should default to nil if flags.Changed("init") { hostConfig.Init = &copts.init } // When allocating stdin in attached mode, close stdin at client disconnect if config.OpenStdin && config.AttachStdin { config.StdinOnce = true } networkingConfig := &networktypes.NetworkingConfig{ EndpointsConfig: make(map[string]*networktypes.EndpointSettings), } networkingConfig.EndpointsConfig, err = parseNetworkOpts(copts) if err != nil { return nil, err } // Put the endpoint-specific MacAddress of the "main" network attachment into the container Config for backward // compatibility with older daemons. if nw, ok := networkingConfig.EndpointsConfig[hostConfig.NetworkMode.NetworkName()]; ok { config.MacAddress = nw.MacAddress //nolint:staticcheck // ignore SA1019: field is deprecated, but still used on API < v1.44. } return &containerConfig{ Config: config, HostConfig: hostConfig, NetworkingConfig: networkingConfig, }, nil } // parseNetworkOpts converts --network advanced options to endpoint-specs, and combines // them with the old --network-alias and --links. If returns an error if conflicting options // are found. // // this function may return _multiple_ endpoints, which is not currently supported // by the daemon, but may be in future; it's up to the daemon to produce an error // in case that is not supported. func parseNetworkOpts(copts *containerOptions) (map[string]*networktypes.EndpointSettings, error) { var ( endpoints = make(map[string]*networktypes.EndpointSettings, len(copts.netMode.Value())) hasUserDefined, hasNonUserDefined bool ) if len(copts.netMode.Value()) == 0 { n := opts.NetworkAttachmentOpts{ Target: "default", } if err := applyContainerOptions(&n, copts); err != nil { return nil, err } ep, err := parseNetworkAttachmentOpt(n) if err != nil { return nil, err } endpoints["default"] = ep } for i, n := range copts.netMode.Value() { n := n if container.NetworkMode(n.Target).IsUserDefined() { hasUserDefined = true } else { hasNonUserDefined = true } if i == 0 { // The first network corresponds with what was previously the "only" // network, and what would be used when using the non-advanced syntax // `--network-alias`, `--link`, `--ip`, `--ip6`, and `--link-local-ip` // are set on this network, to preserve backward compatibility with // the non-advanced notation if err := applyContainerOptions(&n, copts); err != nil { return nil, err } } ep, err := parseNetworkAttachmentOpt(n) if err != nil { return nil, err } if _, ok := endpoints[n.Target]; ok { return nil, errdefs.InvalidParameter(errors.Errorf("network %q is specified multiple times", n.Target)) } // For backward compatibility: if no custom options are provided for the network, // and only a single network is specified, omit the endpoint-configuration // on the client (the daemon will still create it when creating the container) if i == 0 && len(copts.netMode.Value()) == 1 { if ep == nil || reflect.DeepEqual(*ep, networktypes.EndpointSettings{}) { continue } } endpoints[n.Target] = ep } if hasUserDefined && hasNonUserDefined { return nil, errdefs.InvalidParameter(errors.New("conflicting options: cannot attach both user-defined and non-user-defined network-modes")) } return endpoints, nil } func applyContainerOptions(n *opts.NetworkAttachmentOpts, copts *containerOptions) error { //nolint:gocyclo // TODO should we error if _any_ advanced option is used? (i.e. forbid to combine advanced notation with the "old" flags (`--network-alias`, `--link`, `--ip`, `--ip6`)? if len(n.Aliases) > 0 && copts.aliases.Len() > 0 { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --network-alias and per-network alias")) } if len(n.Links) > 0 && copts.links.Len() > 0 { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --link and per-network links")) } if n.IPv4Address != "" && copts.ipv4Address != "" { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --ip and per-network IPv4 address")) } if n.IPv6Address != "" && copts.ipv6Address != "" { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --ip6 and per-network IPv6 address")) } if n.MacAddress != "" && copts.macAddress != "" { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --mac-address and per-network MAC address")) } if len(n.LinkLocalIPs) > 0 && copts.linkLocalIPs.Len() > 0 { return errdefs.InvalidParameter(errors.New("conflicting options: cannot specify both --link-local-ip and per-network link-local IP addresses")) } if copts.aliases.Len() > 0 { n.Aliases = make([]string, copts.aliases.Len()) copy(n.Aliases, copts.aliases.GetAll()) } if n.Target != "default" && copts.links.Len() > 0 { n.Links = make([]string, copts.links.Len()) copy(n.Links, copts.links.GetAll()) } if copts.ipv4Address != "" { n.IPv4Address = copts.ipv4Address } if copts.ipv6Address != "" { n.IPv6Address = copts.ipv6Address } if copts.macAddress != "" { n.MacAddress = copts.macAddress } if copts.linkLocalIPs.Len() > 0 { n.LinkLocalIPs = make([]string, copts.linkLocalIPs.Len()) copy(n.LinkLocalIPs, copts.linkLocalIPs.GetAll()) } return nil } func parseNetworkAttachmentOpt(ep opts.NetworkAttachmentOpts) (*networktypes.EndpointSettings, error) { if strings.TrimSpace(ep.Target) == "" { return nil, errors.New("no name set for network") } if !container.NetworkMode(ep.Target).IsUserDefined() { if len(ep.Aliases) > 0 { return nil, errors.New("network-scoped aliases are only supported for user-defined networks") } if len(ep.Links) > 0 { return nil, errors.New("links are only supported for user-defined networks") } } epConfig := &networktypes.EndpointSettings{} epConfig.Aliases = append(epConfig.Aliases, ep.Aliases...) if len(ep.DriverOpts) > 0 { epConfig.DriverOpts = make(map[string]string) epConfig.DriverOpts = ep.DriverOpts } if len(ep.Links) > 0 { epConfig.Links = ep.Links } if ep.IPv4Address != "" || ep.IPv6Address != "" || len(ep.LinkLocalIPs) > 0 { epConfig.IPAMConfig = &networktypes.EndpointIPAMConfig{ IPv4Address: ep.IPv4Address, IPv6Address: ep.IPv6Address, LinkLocalIPs: ep.LinkLocalIPs, } } if ep.MacAddress != "" { if _, err := opts.ValidateMACAddress(ep.MacAddress); err != nil { return nil, errors.Errorf("%s is not a valid mac address", ep.MacAddress) } epConfig.MacAddress = ep.MacAddress } return epConfig, nil } func convertToStandardNotation(ports []string) ([]string, error) { optsList := []string{} for _, publish := range ports { if strings.Contains(publish, "=") { params := map[string]string{"protocol": "tcp"} for _, param := range strings.Split(publish, ",") { k, v, ok := strings.Cut(param, "=") if !ok || k == "" { return optsList, errors.Errorf("invalid publish opts format (should be name=value but got '%s')", param) } params[k] = v } optsList = append(optsList, fmt.Sprintf("%s:%s/%s", params["published"], params["target"], params["protocol"])) } else { optsList = append(optsList, publish) } } return optsList, nil } func parseLoggingOpts(loggingDriver string, loggingOpts []string) (map[string]string, error) { loggingOptsMap := opts.ConvertKVStringsToMap(loggingOpts) if loggingDriver == "none" && len(loggingOpts) > 0 { return map[string]string{}, errors.Errorf("invalid logging opts for driver %s", loggingDriver) } return loggingOptsMap, nil } // takes a local seccomp daemon, reads the file contents for sending to the daemon func parseSecurityOpts(securityOpts []string) ([]string, error) { for key, opt := range securityOpts { k, v, ok := strings.Cut(opt, "=") if !ok && k != "no-new-privileges" { k, v, ok = strings.Cut(opt, ":") } if (!ok || v == "") && k != "no-new-privileges" { // "no-new-privileges" is the only option that does not require a value. return securityOpts, errors.Errorf("Invalid --security-opt: %q", opt) } if k == "seccomp" { switch v { case seccompProfileDefault, seccompProfileUnconfined: // known special names for built-in profiles, nothing to do. default: // value may be a filename, in which case we send the profile's // content if it's valid JSON. f, err := os.ReadFile(v) if err != nil { return securityOpts, errors.Errorf("opening seccomp profile (%s) failed: %v", v, err) } b := bytes.NewBuffer(nil) if err := json.Compact(b, f); err != nil { return securityOpts, errors.Errorf("compacting json for seccomp profile (%s) failed: %v", v, err) } securityOpts[key] = fmt.Sprintf("seccomp=%s", b.Bytes()) } } } return securityOpts, nil } // parseSystemPaths checks if `systempaths=unconfined` security option is set, // and returns the `MaskedPaths` and `ReadonlyPaths` accordingly. An updated // list of security options is returned with this option removed, because the // `unconfined` option is handled client-side, and should not be sent to the // daemon. func parseSystemPaths(securityOpts []string) (filtered, maskedPaths, readonlyPaths []string) { filtered = securityOpts[:0] for _, opt := range securityOpts { if opt == "systempaths=unconfined" { maskedPaths = []string{} readonlyPaths = []string{} } else { filtered = append(filtered, opt) } } return filtered, maskedPaths, readonlyPaths } // parses storage options per container into a map func parseStorageOpts(storageOpts []string) (map[string]string, error) { m := make(map[string]string) for _, option := range storageOpts { k, v, ok := strings.Cut(option, "=") if !ok { return nil, errors.Errorf("invalid storage option") } m[k] = v } return m, nil } // parseDevice parses a device mapping string to a container.DeviceMapping struct func parseDevice(device, serverOS string) (container.DeviceMapping, error) { switch serverOS { case "linux": return parseLinuxDevice(device) case "windows": return parseWindowsDevice(device) } return container.DeviceMapping{}, errors.Errorf("unknown server OS: %s", serverOS) } // parseLinuxDevice parses a device mapping string to a container.DeviceMapping struct // knowing that the target is a Linux daemon func parseLinuxDevice(device string) (container.DeviceMapping, error) { var src, dst string permissions := "rwm" // We expect 3 parts at maximum; limit to 4 parts to detect invalid options. arr := strings.SplitN(device, ":", 4) switch len(arr) { case 3: permissions = arr[2] fallthrough case 2: if validDeviceMode(arr[1]) { permissions = arr[1] } else { dst = arr[1] } fallthrough case 1: src = arr[0] default: return container.DeviceMapping{}, errors.Errorf("invalid device specification: %s", device) } if dst == "" { dst = src } deviceMapping := container.DeviceMapping{ PathOnHost: src, PathInContainer: dst, CgroupPermissions: permissions, } return deviceMapping, nil } // parseWindowsDevice parses a device mapping string to a container.DeviceMapping struct // knowing that the target is a Windows daemon func parseWindowsDevice(device string) (container.DeviceMapping, error) { return container.DeviceMapping{PathOnHost: device}, nil } // validateDeviceCgroupRule validates a device cgroup rule string format // It will make sure 'val' is in the form: // // 'type major:minor mode' func validateDeviceCgroupRule(val string) (string, error) { if deviceCgroupRuleRegexp.MatchString(val) { return val, nil } return val, errors.Errorf("invalid device cgroup format '%s'", val) } // validDeviceMode checks if the mode for device is valid or not. // Valid mode is a composition of r (read), w (write), and m (mknod). func validDeviceMode(mode string) bool { legalDeviceMode := map[rune]bool{ 'r': true, 'w': true, 'm': true, } if mode == "" { return false } for _, c := range mode { if !legalDeviceMode[c] { return false } legalDeviceMode[c] = false } return true } // validateDevice validates a path for devices func validateDevice(val string, serverOS string) (string, error) { switch serverOS { case "linux": return validateLinuxPath(val, validDeviceMode) case "windows": // Windows does validation entirely server-side return val, nil } return "", errors.Errorf("unknown server OS: %s", serverOS) } // validateLinuxPath is the implementation of validateDevice knowing that the // target server operating system is a Linux daemon. // It will make sure 'val' is in the form: // // [host-dir:]container-path[:mode] // // It also validates the device mode. func validateLinuxPath(val string, validator func(string) bool) (string, error) { var containerPath string var mode string if strings.Count(val, ":") > 2 { return val, errors.Errorf("bad format for path: %s", val) } split := strings.SplitN(val, ":", 3) if split[0] == "" { return val, errors.Errorf("bad format for path: %s", val) } switch len(split) { case 1: containerPath = split[0] val = path.Clean(containerPath) case 2: if isValid := validator(split[1]); isValid { containerPath = split[0] mode = split[1] val = fmt.Sprintf("%s:%s", path.Clean(containerPath), mode) } else { containerPath = split[1] val = fmt.Sprintf("%s:%s", split[0], path.Clean(containerPath)) } case 3: containerPath = split[1] mode = split[2] if isValid := validator(split[2]); !isValid { return val, errors.Errorf("bad mode specified: %s", mode) } val = fmt.Sprintf("%s:%s:%s", split[0], containerPath, mode) } if !path.IsAbs(containerPath) { return val, errors.Errorf("%s is not an absolute path", containerPath) } return val, nil } // validateAttach validates that the specified string is a valid attach option. func validateAttach(val string) (string, error) { s := strings.ToLower(val) for _, str := range []string{"stdin", "stdout", "stderr"} { if s == str { return s, nil } } return val, errors.Errorf("valid streams are STDIN, STDOUT and STDERR") } func validateAPIVersion(c *containerConfig, serverAPIVersion string) error { for _, m := range c.HostConfig.Mounts { if err := command.ValidateMountWithAPIVersion(m, serverAPIVersion); err != nil { return err } } return nil }