cgroups: join cgroup v2 when using hybrid mode

Currently the parent process of the container is moved to the right
cgroup v2 tree when systemd is using a hybrid model (last line with 0::):

$ runc --systemd-cgroup run myid
/ # cat /proc/self/cgroup
12:cpuset:/system.slice/runc-myid.scope
11:blkio:/system.slice/runc-myid.scope
10:devices:/system.slice/runc-myid.scope
9:hugetlb:/system.slice/runc-myid.scope
8:memory:/system.slice/runc-myid.scope
7:rdma:/
6:perf_event:/system.slice/runc-myid.scope
5:net_cls,net_prio:/system.slice/runc-myid.scope
4:freezer:/system.slice/runc-myid.scope
3:pids:/system.slice/runc-myid.scope
2:cpu,cpuacct:/system.slice/runc-myid.scope
1:name=systemd:/system.slice/runc-myid.scope
0::/system.slice/runc-myid.scope

However, if a second process is executed in the same container, it is
not moved to the right cgroup v2 tree:

$ runc exec myid /bin/sh -c 'cat /proc/self/cgroup'
12:cpuset:/system.slice/runc-myid.scope
11:blkio:/system.slice/runc-myid.scope
10:devices:/system.slice/runc-myid.scope
9:hugetlb:/system.slice/runc-myid.scope
8:memory:/system.slice/runc-myid.scope
7:rdma:/
6:perf_event:/system.slice/runc-myid.scope
5:net_cls,net_prio:/system.slice/runc-myid.scope
4:freezer:/system.slice/runc-myid.scope
3:pids:/system.slice/runc-myid.scope
2:cpu,cpuacct:/system.slice/runc-myid.scope
1:name=systemd:/system.slice/runc-myid.scope
0::/user.slice/user-1000.slice/session-8.scope

This commit makes that processes executed with exec are placed into the
right cgroup v2 tree. The implementation checks if systemd is using a
hybrid mode (by checking if cgroups v2 is mounted in
/sys/fs/cgroup/unified), if yes, the path of the cgroup v2 slice for
this container is saved into the cgroup path list.

The fs group driver has a similar issue, in this case none of the runc
run or runc exec commands put the process in the right cgroups v2. This
commit also fixes that.

Having the processes of the container in its own cgroup v2 is useful
for any BPF programs that rely on bpf_get_current_cgroup_id(), like
https://github.com/kinvolk/inspektor-gadget/ for instance.

[@kolyshkin: rebased]

Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>

libcontainer/cgroups/fs/fs.go

@@ -35,6 +35,14 @@ var (
 
 var errSubsystemDoesNotExist = errors.New("cgroup: subsystem does not exist")
 
+func init() {
+	// If using cgroups-hybrid mode then add a "" controller indicating
+	// it should join the cgroups v2.
+	if cgroups.IsCgroup2HybridMode() {
+		subsystems = append(subsystems, &NameGroup{GroupName: "", Join: true})
+	}
+}
+
 type subsystem interface {
 	// Name returns the name of the subsystem.
 	Name() string

libcontainer/cgroups/systemd/v1.go

@@ -143,6 +143,19 @@ func initPaths(c *configs.Cgroup) (map[string]string, error) {
 		}
 		paths[s.Name()] = subsystemPath
 	}
+
+	// If systemd is using cgroups-hybrid mode then add the slice path of
+	// this container to the paths so the following process executed with
+	// "runc exec" joins that cgroup as well.
+	if cgroups.IsCgroup2HybridMode() {
+		// "" means cgroup-hybrid path
+		cgroupsHybridPath, err := getSubsystemPath(slice, unit, "")
+		if err != nil && cgroups.IsNotFound(err) {
+			return nil, err
+		}
+		paths[""] = cgroupsHybridPath
+	}
+
 	return paths, nil
 }
 

libcontainer/cgroups/utils.go

@@ -21,11 +21,14 @@ import (
 const (
 	CgroupProcesses   = "cgroup.procs"
 	unifiedMountpoint = "/sys/fs/cgroup"
+	hybridMountpoint  = "/sys/fs/cgroup/unified"
 )
 
 var (
 	isUnifiedOnce sync.Once
 	isUnified     bool
+	isHybridOnce  sync.Once
+	isHybrid      bool
 )
 
 // IsCgroup2UnifiedMode returns whether we are running in cgroup v2 unified mode.
@@ -47,6 +50,24 @@ func IsCgroup2UnifiedMode() bool {
 	return isUnified
 }
 
+// IsCgroup2HybridMode returns whether we are running in cgroup v2 hybrid mode.
+func IsCgroup2HybridMode() bool {
+	isHybridOnce.Do(func() {
+		var st unix.Statfs_t
+		err := unix.Statfs(hybridMountpoint, &st)
+		if err != nil {
+			if os.IsNotExist(err) {
+				// ignore the "not found" error
+				isHybrid = false
+				return
+			}
+			panic(fmt.Sprintf("cannot statfs cgroup root: %s", err))
+		}
+		isHybrid = st.Type == unix.CGROUP2_SUPER_MAGIC
+	})
+	return isHybrid
+}
+
 type Mount struct {
 	Mountpoint string
 	Root       string

libcontainer/cgroups/v1_utils.go

@@ -113,6 +113,11 @@ func FindCgroupMountpoint(cgroupPath, subsystem string) (string, error) {
 		return "", errUnified
 	}
 
+	// If subsystem is empty, we look for the cgroupv2 hybrid path.
+	if len(subsystem) == 0 {
+		return hybridMountpoint, nil
+	}
+
 	// Avoid parsing mountinfo by trying the default path first, if possible.
 	if path := tryDefaultPath(cgroupPath, subsystem); path != "" {
 		return path, nil
@@ -223,6 +228,11 @@ func GetOwnCgroupPath(subsystem string) (string, error) {
 		return "", err
 	}
 
+	// If subsystem is empty, we look for the cgroupv2 hybrid path.
+	if len(subsystem) == 0 {
+		return hybridMountpoint, nil
+	}
+
 	return getCgroupPathHelper(subsystem, cgroup)
 }