Support time namespace

"time" namespace was introduced in Linux v5.6
support new time namespace to set boottime and monotonic time offset

Example runtime spec

"timeOffsets": {
    "monotonic": {
        "secs": 172800,
        "nanosecs": 0
    },
    "boottime": {
        "secs": 604800,
        "nanosecs": 0
    }
}

Signed-off-by: Chethan Suresh <[email protected]>

Author: chethanah

libcontainer/configs/config.go

@@ -212,6 +212,9 @@ type Config struct {
 	// RootlessCgroups is set when unlikely to have the full access to cgroups.
 	// When RootlessCgroups is set, cgroups errors are ignored.
 	RootlessCgroups bool `json:"rootless_cgroups,omitempty"`
+
+	// TimeOffsets specifies the offset for supporting time namespaces.
+	TimeOffsets map[string]specs.LinuxTimeOffset `json:"timeOffsets,omitempty"`
 }
 
 type (

libcontainer/configs/namespaces_linux.go

@@ -14,6 +14,7 @@ const (
 	NEWIPC    NamespaceType = "NEWIPC"
 	NEWUSER   NamespaceType = "NEWUSER"
 	NEWCGROUP NamespaceType = "NEWCGROUP"
+	NEWTIME   NamespaceType = "NEWTIME"
 )
 
 var (
@@ -38,6 +39,8 @@ func NsName(ns NamespaceType) string {
 		return "uts"
 	case NEWCGROUP:
 		return "cgroup"
+	case NEWTIME:
+		return "time"
 	}
 	return ""
 }
@@ -72,6 +75,7 @@ func NamespaceTypes() []NamespaceType {
 		NEWPID,
 		NEWNS,
 		NEWCGROUP,
+		NEWTIME,
 	}
 }
 

libcontainer/configs/namespaces_syscall.go

@@ -17,6 +17,7 @@ var namespaceInfo = map[NamespaceType]int{
 	NEWUTS:    unix.CLONE_NEWUTS,
 	NEWPID:    unix.CLONE_NEWPID,
 	NEWCGROUP: unix.CLONE_NEWCGROUP,
+	NEWTIME:   unix.CLONE_NEWTIME,
 }
 
 // CloneFlags parses the container's Namespaces options to set the correct

libcontainer/configs/validate/validator.go

@@ -115,6 +115,12 @@ func namespaces(config *configs.Config) error {
 		}
 	}
 
+	if config.Namespaces.Contains(configs.NEWTIME) {
+		if _, err := os.Stat("/proc/self/timens_offsets"); os.IsNotExist(err) {
+			return errors.New("time namespaces aren't enabled in the kernel")
+		}
+	}
+
 	return nil
 }
 

libcontainer/container_linux.go

@@ -2114,6 +2114,16 @@ func encodeIDMapping(idMap []configs.IDMap) ([]byte, error) {
 	return data.Bytes(), nil
 }
 
+func encodeTimeNs(timeOffsets specs.LinuxTimeOffset) ([]byte, error) {
+	data := bytes.NewBuffer(nil)
+	line := fmt.Sprintf("%d %d\n", timeOffsets.Secs, timeOffsets.Nanosecs)
+	if _, err := data.WriteString(line); err != nil {
+		return nil, err
+	}
+
+	return data.Bytes(), nil
+}
+
 // netlinkError is an error wrapper type for use by custom netlink message
 // types. Panics with errors are wrapped in netlinkError so that the recover
 // in bootstrapData can distinguish intentional panics.
@@ -2246,6 +2256,34 @@ func (c *Container) bootstrapData(cloneFlags uintptr, nsMaps map[configs.Namespa
 		})
 	}
 
+	// write boottime and monotonic offsets
+	// 0 is default value in /proc/PID/timens_offsets
+	// if 0, do not bootstrap data
+	if c.config.TimeOffsets != nil {
+
+		if c.config.TimeOffsets["boottime"].Nanosecs != 0 || c.config.TimeOffsets["boottime"].Secs != 0 {
+			b, err := encodeTimeNs(c.config.TimeOffsets["boottime"])
+			if err != nil {
+				return nil, err
+			}
+			r.AddData(&Bytemsg{
+				Type:  BootTimeNsAttr,
+				Value: append([]byte("boottime "), b...),
+			})
+		}
+
+		if c.config.TimeOffsets["monotonic"].Nanosecs != 0 || c.config.TimeOffsets["monotonic"].Secs != 0 {
+			b, err := encodeTimeNs(c.config.TimeOffsets["monotonic"])
+			if err != nil {
+				return nil, err
+			}
+			r.AddData(&Bytemsg{
+				Type:  MonotonicNsAttr,
+				Value: append([]byte("monotonic "), b...),
+			})
+		}
+	}
+
 	return bytes.NewReader(r.Serialize()), nil
 }
 

libcontainer/message_linux.go

@@ -22,6 +22,8 @@ const (
 	UidmapPathAttr   uint16 = 27288
 	GidmapPathAttr   uint16 = 27289
 	MountSourcesAttr uint16 = 27290
+	BootTimeNsAttr   uint16 = 27291
+	MonotonicNsAttr  uint16 = 27292
 )
 
 type Int32msg struct {

libcontainer/nsenter/namespace.h

@@ -28,5 +28,8 @@
 #ifndef CLONE_NEWNET
 #	define CLONE_NEWNET 0x40000000 /* New network namespace */
 #endif
+#ifndef CLONE_NEWTIME
+#	define CLONE_NEWTIME 0x00000080	/* New time namespace */
+#endif
 
 #endif /* NSENTER_NAMESPACE_H */

libcontainer/nsenter/nsexec.c

@@ -95,6 +95,13 @@ struct nlconfig_t {
 	/* Mount sources opened outside the container userns. */
 	char *mountsources;
 	size_t mountsources_len;
+
+	/* Time NS settings for boottime and monotonic */
+	char *monotonic;
+	size_t monotonic_len;
+
+	char *boottime;
+	size_t boottime_len;
 };
 
 /*
@@ -112,6 +119,8 @@ struct nlconfig_t {
 #define UIDMAPPATH_ATTR		27288
 #define GIDMAPPATH_ATTR		27289
 #define MOUNT_SOURCES_ATTR	27290
+#define BOOTTIME_NS_ATTR	27291
+#define MONOTONIC_NS_ATTR	27292
 
 /*
  * Use the raw syscall for versions of glibc which don't include a function for
@@ -341,6 +350,8 @@ static int nsflag(char *name)
 		return CLONE_NEWUSER;
 	else if (!strcmp(name, "uts"))
 		return CLONE_NEWUTS;
+	else if (!strcmp(name, "time"))
+		return CLONE_NEWTIME;
 
 	/* If we don't recognise a name, fallback to 0. */
 	return 0;
@@ -431,6 +442,14 @@ static void nl_parse(int fd, struct nlconfig_t *config)
 			config->mountsources = current;
 			config->mountsources_len = payload_len;
 			break;
+		case BOOTTIME_NS_ATTR:
+			config->boottime = current;
+			config->boottime_len = payload_len;
+			break;
+		case MONOTONIC_NS_ATTR:
+			config->monotonic = current;
+			config->monotonic_len = payload_len;
+			break;
 		default:
 			bail("unknown netlink message type %d", nlattr->nla_type);
 		}
@@ -641,6 +660,17 @@ void try_unshare(int flags, const char *msg)
 	bail("failed to unshare %s", msg);
 }
 
+static void update_timens(char *map, size_t map_len)
+{
+	if (map == NULL || map_len == 0)
+		return;
+	write_log(DEBUG, "update /proc/self/timens_offsets to '%s'", map);
+	if (write_file(map, map_len, "/proc/self/timens_offsets") < 0) {
+		if (errno != EPERM)
+			bail("failed to update /proc/self/timens_offsets");
+	}
+}
+
 void nsexec(void)
 {
 	int pipenum;
@@ -1053,6 +1083,14 @@ void nsexec(void)
 					bail("failed to sync with parent: SYNC_MOUNTSOURCES_ACK: got %u", s);
 			}
 
+			/*
+			 * Update timens offsets
+			 * set boottime and monotonic offsets
+			 */
+			write_log(DEBUG, "set timens offsets %s", config.boottime);
+			update_timens(config.boottime, config.boottime_len);
+			update_timens(config.monotonic, config.monotonic_len);
+
 			/*
 			 * TODO: What about non-namespace clone flags that we're dropping here?
 			 *

libcontainer/specconv/spec_linux.go

@@ -49,6 +49,7 @@ func initMaps() {
 			specs.IPCNamespace:     configs.NEWIPC,
 			specs.UTSNamespace:     configs.NEWUTS,
 			specs.CgroupNamespace:  configs.NEWCGROUP,
+			specs.TimeNamespace:    configs.NEWTIME,
 		}
 
 		mountPropagationMapping = map[string]int{
@@ -433,6 +434,11 @@ func CreateLibcontainerConfig(opts *CreateOpts) (*configs.Config, error) {
 				MemBwSchema:   spec.Linux.IntelRdt.MemBwSchema,
 			}
 		}
+
+		// update timens offsets
+		if spec.Linux.TimeOffsets != nil {
+			config.TimeOffsets = spec.Linux.TimeOffsets
+		}
 	}
 
 	// Set the host UID that should own the container's cgroup.
@@ -492,6 +498,7 @@ func CreateLibcontainerConfig(opts *CreateOpts) (*configs.Config, error) {
 			}
 		}
 	}
+
 	createHooks(spec, config)
 	config.Version = specs.Version
 	return config, nil