diff --git a/virtcontainers/container.go b/virtcontainers/container.go
index a5ff44a557..780a4370e2 100644
--- a/virtcontainers/container.go
+++ b/virtcontainers/container.go
@@ -481,7 +481,7 @@ func (c *Container) shareFiles(m Mount, idx int, hostSharedDir, guestSharedDir s
 	} else {
 		// These mounts are created in the shared dir
 		mountDest := filepath.Join(hostSharedDir, c.sandbox.id, filename)
-		if err := bindMount(c.ctx, m.Source, mountDest, false); err != nil {
+		if err := bindMount(c.ctx, m.Source, mountDest, false, "private"); err != nil {
 			return "", false, err
 		}
 		// Save HostPath mount value into the mount list of the container.
diff --git a/virtcontainers/fc.go b/virtcontainers/fc.go
index 653eae1824..3ebb00c839 100644
--- a/virtcontainers/fc.go
+++ b/virtcontainers/fc.go
@@ -180,46 +180,6 @@ func (fc *firecracker) trace(name string) (opentracing.Span, context.Context) {
 	return span, ctx
 }
 
-// bindMount bind mounts a source in to a destination. This will
-// do some bookkeeping:
-// * evaluate all symlinks
-// * ensure the source exists
-// * recursively create the destination
-func (fc *firecracker) bindMount(ctx context.Context, source, destination string, readonly bool) error {
-	span, _ := trace(ctx, "bindMount")
-	defer span.Finish()
-
-	if source == "" {
-		return fmt.Errorf("source must be specified")
-	}
-	if destination == "" {
-		return fmt.Errorf("destination must be specified")
-	}
-
-	absSource, err := filepath.EvalSymlinks(source)
-	if err != nil {
-		return fmt.Errorf("Could not resolve symlink for source %v", source)
-	}
-
-	if err := ensureDestinationExists(absSource, destination); err != nil {
-		return fmt.Errorf("Could not create destination mount point %v: %v", destination, err)
-	}
-
-	fc.Logger().WithFields(logrus.Fields{"src": absSource, "dst": destination}).Debug("Bind mounting resource")
-
-	if err := syscall.Mount(absSource, destination, "bind", syscall.MS_BIND|syscall.MS_SLAVE, ""); err != nil {
-		return fmt.Errorf("Could not bind mount %v to %v: %v", absSource, destination, err)
-	}
-
-	// For readonly bind mounts, we need to remount with the readonly flag.
-	// This is needed as only very recent versions of libmount/util-linux support "bind,ro"
-	if readonly {
-		return syscall.Mount(absSource, destination, "bind", uintptr(syscall.MS_BIND|syscall.MS_SLAVE|syscall.MS_REMOUNT|syscall.MS_RDONLY), "")
-	}
-
-	return nil
-}
-
 // For firecracker this call only sets the internal structure up.
 // The sandbox will be created and started through startSandbox().
 func (fc *firecracker) createSandbox(ctx context.Context, id string, networkNS NetworkNamespace, hypervisorConfig *HypervisorConfig, stateful bool) error {
@@ -543,7 +503,7 @@ func (fc *firecracker) fcJailResource(src, dst string) (string, error) {
 			src, dst)
 	}
 	jailedLocation := filepath.Join(fc.jailerRoot, dst)
-	if err := fc.bindMount(context.Background(), src, jailedLocation, false); err != nil {
+	if err := bindMount(context.Background(), src, jailedLocation, false, "slave"); err != nil {
 		fc.Logger().WithField("bindMount failed", err).Error()
 		return "", err
 	}
diff --git a/virtcontainers/mount.go b/virtcontainers/mount.go
index 9987a9b177..93951ac4fd 100644
--- a/virtcontainers/mount.go
+++ b/virtcontainers/mount.go
@@ -31,6 +31,13 @@ var rootfsDir = "rootfs"
 
 var systemMountPrefixes = []string{"/proc", "/sys"}
 
+var propagationTypes = map[string]uintptr{
+	"shared":  syscall.MS_SHARED,
+	"private": syscall.MS_PRIVATE,
+	"slave":   syscall.MS_SLAVE,
+	"ubind":   syscall.MS_UNBINDABLE,
+}
+
 func isSystemMount(m string) bool {
 	for _, p := range systemMountPrefixes {
 		if m == p || strings.HasPrefix(m, p+"/") {
@@ -260,7 +267,8 @@ const mountPerm = os.FileMode(0755)
 // * evaluate all symlinks
 // * ensure the source exists
 // * recursively create the destination
-func bindMount(ctx context.Context, source, destination string, readonly bool) error {
+// pgtypes stands for propagation types, which are shared, private, slave, and ubind.
+func bindMount(ctx context.Context, source, destination string, readonly bool, pgtypes string) error {
 	span, _ := trace(ctx, "bindMount")
 	defer span.Finish()
 
@@ -284,8 +292,12 @@ func bindMount(ctx context.Context, source, destination string, readonly bool) e
 		return fmt.Errorf("Could not bind mount %v to %v: %v", absSource, destination, err)
 	}
 
-	if err := syscall.Mount("none", destination, "", syscall.MS_PRIVATE, ""); err != nil {
-		return fmt.Errorf("Could not make mount point %v private: %v", destination, err)
+	if pgtype, exist := propagationTypes[pgtypes]; exist {
+		if err := syscall.Mount("none", destination, "", pgtype, ""); err != nil {
+			return fmt.Errorf("Could not make mount point %v %s: %v", destination, pgtypes, err)
+		}
+	} else {
+		return fmt.Errorf("Wrong propagation type %s", pgtypes)
 	}
 
 	// For readonly bind mounts, we need to remount with the readonly flag.
@@ -305,7 +317,7 @@ func bindMountContainerRootfs(ctx context.Context, sharedDir, sandboxID, cID, cR
 
 	rootfsDest := filepath.Join(sharedDir, sandboxID, cID, rootfsDir)
 
-	return bindMount(ctx, cRootFs, rootfsDest, readonly)
+	return bindMount(ctx, cRootFs, rootfsDest, readonly, "private")
 }
 
 // Mount describes a container mount.