net: workaround a problem when iptables can't open /run/xtables.lock

Starting with iptables 1.6.2, we have to use the --wait option,
but it doesn't work properly with userns, because in this case,
we don't have enough rights to open /run/xtables.lock.

(00.174703)      1: 	Running iptables-restore -w for iptables-restore -w Fatal: can't open lock file /run/xtables.lock: Permission denied
(00.192058)      1: Error (criu/util.c:842): exited, status=4
(00.192080)      1: Error (criu/net.c:1738): iptables-restore -w failed
(00.192088)      1: Error (criu/net.c:2389): Can't create net_ns
(00.192131)      1: Error (criu/util.c:1567): Can't wait or bad status: errno=0, status=65280

This patch workarounds this problem by mounting tmpfs into /run.
Net namespaces are restored in a separate process, so we can create a
new mount namespace and create new mounts.

https://github.com/checkpoint-restore/criu/issues/469

Cc: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Cc: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>

criu/net.c

@@ -1965,6 +1965,48 @@ out:
 	return ret;
 }
 
+/*
+ * iptables-restore is executed from a target userns and it may have not enough
+ * rights to open /run/xtables.lock. Here we try to workaround this problem.
+ */
+static int prepare_xtable_lock()
+{
+	int fd;
+
+	fd = open("/run/xtables.lock", O_RDONLY);
+	if (fd >= 0) {
+		close(fd);
+		return 0;
+	}
+
+	/*
+	 * __prepare_net_namespaces is executed in a separate process,
+	 * so a mount namespace can be changed.
+	 */
+	if (unshare(CLONE_NEWNS)) {
+		pr_perror("Unable to create a mount namespace");
+		return -1;
+	}
+
+	if (mount(NULL, "/",  NULL, MS_SLAVE | MS_REC, NULL)) {
+		pr_perror("Unable to conver mounts to slave mounts");
+		return -1;
+	}
+	/*
+	 * /run/xtables.lock may not exist, so we can't just bind-mount a file
+	 * over it.
+	 * A new mount will not be propagated to the host mount namespace,
+	 * because we are in another userns.
+	 */
+
+	if (mount("criu-xtable-lock", "/run", "tmpfs", 0, NULL)) {
+		pr_perror("Unable to mount tmpfs into /run");
+		return -1;
+	}
+
+	return 0;
+}
+
 static inline int restore_iptables(int pid)
 {
 	int ret = -1;
@@ -2311,6 +2353,9 @@ static int __prepare_net_namespaces(void *unused)
 	struct ns_id *nsid;
 	int root_ns;
 
+	if (prepare_xtable_lock())
+		return -1;
+
 	root_ns = open_proc(PROC_SELF, "ns/net");
 	if (root_ns < 0)
 		return -1;