libct/cg/fs/freezer: fix freezing race

Before this commit, Set() used GetState() to check the freezer state
and retry the operation if the actual state still differs from requested.
This should help with the situation when a new process (such as one
added by runc exec) is added to the container's cgroup while it's being
freezed by the kernel, but it's not working as it should.

The problem is, GetState() never returns FREEZING state, looping until
the state is either FROZEN or THAWED, so Set() does not have a chance
to repeate the freeze attempt.

As a result, the container might end up stuck in a FREEZING state,
with GetState() never returning (which in turn blocks some other
operations).

One way to fix this would be to have GetState returning FREEZING state
instead of retrying ad infinitum. It would result in changing the public
API, and no callers of GetState expects it to return this.

To fix, let's not use GetState() from Set(). Instead, read the
freezer.state file directly and act accordingly -- return success
on FROZEN, retry on FREEZING, and error out on any other (unexpected)
value.

While at it, further improve the code:
 - limit the number of retries;
 - if retries are exceeded, thaw and return an error;
 - don't retry (or read the state back) on THAW.

I played a lot with various reproducers for this bug, including

 - parallel runc execs and runc pause/resumes
 - parallel runc execs and runc --systemd-cgroup update
   (the latter performs freeze/unfreeze);
 - continuously running /bin/printf inside container
   in parallel with runc pause/resume;
 - running pthread bomb (from criu test suite) in parallel
   with runc pause/resume;

and I was not able to make freeze work 100%, meaning sometimes
runc pause fails, or runc --systemd-cgroup update produces a warning.

With that said, it's still a big improvement over the previous
state of affairs where container is stuck in FREEZING state,
and GetState() (and all its users) are also stuck.

Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>

libcontainer/cgroups/fs/freezer.go

@@ -28,33 +28,51 @@ func (s *FreezerGroup) Apply(path string, d *cgroupData) error {
 
 func (s *FreezerGroup) Set(path string, cgroup *configs.Cgroup) error {
 	switch cgroup.Resources.Freezer {
-	case configs.Frozen, configs.Thawed:
-		for {
-			// In case this loop does not exit because it doesn't get the expected
-			// state, let's write again this state, hoping it's going to be properly
-			// set this time. Otherwise, this loop could run infinitely, waiting for
-			// a state change that would never happen.
-			if err := fscommon.WriteFile(path, "freezer.state", string(cgroup.Resources.Freezer)); err != nil {
+	case configs.Frozen:
+		// As per older kernel docs (freezer-subsystem.txt before
+		// kernel commit ef9fe980c6fcc1821), if FREEZING is seen,
+		// userspace should either retry or thaw. While current
+		// kernel cgroup v1 docs no longer mention a need to retry,
+		// the kernel (tested on v5.4, Ubuntu 20.04) can't reliably
+		// freeze a cgroup while new processes keep appearing in it
+		// (either via fork/clone or by writing new PIDs to
+		// cgroup.procs).
+		//
+		// The particulars in the code below (number of retries
+		// and no sleep in between) are chosen empirically to
+		// minimize the number of failured freezes.
+		for i := 0; i < 1000; i++ {
+			if err := fscommon.WriteFile(path, "freezer.state", string(configs.Frozen)); err != nil {
 				return err
 			}
 
-			state, err := s.GetState(path)
+			state, err := fscommon.ReadFile(path, "freezer.state")
 			if err != nil {
 				return err
 			}
-			if state == cgroup.Resources.Freezer {
-				break
+			state = strings.TrimSpace(state)
+			switch state {
+			case "FREEZING":
+				continue
+			case string(configs.Frozen):
+				return nil
+			default:
+				// should never happen
+				return fmt.Errorf("unexpected state %s while freezing", strings.TrimSpace(state))
 			}
-
-			time.Sleep(1 * time.Millisecond)
 		}
+		// Despite our best efforts, it got stuck in FREEZING.
+		// Leaving it in this state is bad and dangerous, so
+		// let's (try to) thaw it back and error out.
+		_ = fscommon.WriteFile(path, "freezer.state", string(configs.Thawed))
+		return errors.New("unable to freeze")
+	case configs.Thawed:
+		return fscommon.WriteFile(path, "freezer.state", string(configs.Thawed))
 	case configs.Undefined:
 		return nil
 	default:
 		return fmt.Errorf("Invalid argument '%s' to freezer.state", string(cgroup.Resources.Freezer))
 	}
-
-	return nil
 }
 
 func (s *FreezerGroup) GetStats(path string, stats *cgroups.Stats) error {