ctdb-mutex: Avoid corner case where helper is already reparented to init

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12113

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit 1f942ec36c98127e25d7aa9b8370a4f621b9d362)

2 files changed, 15 insertions, 1 deletions

diff --git a/ctdb/doc/cluster_mutex_helper.txt b/ctdb/doc/cluster_mutex_helper.txt
index 0fc3a50dc7c..20c8eb2b51d 100644
--- a/ctdb/doc/cluster_mutex_helper.txt
+++ b/ctdb/doc/cluster_mutex_helper.txt
@@ -64,7 +64,8 @@ Valid status codes are:
 If a 0 status code is sent then it the helper should periodically
 check if the (original) parent processes still exists while awaiting
 termination.  If the parent process disappears then the helper should
-release the mutex and exit.  This avoids stale mutexes.
+release the mutex and exit.  This avoids stale mutexes.  Note that a
+helper should never wait for parent process ID 1!
 
 If a non-0 status code is sent then the helper can exit immediately.
 However, if the helper does not exit then it must terminate if it
diff --git a/ctdb/server/ctdb_mutex_fcntl_helper.c b/ctdb/server/ctdb_mutex_fcntl_helper.c
index 06c2205d496..87358beb234 100644
--- a/ctdb/server/ctdb_mutex_fcntl_helper.c
+++ b/ctdb/server/ctdb_mutex_fcntl_helper.c
@@ -82,6 +82,19 @@ int main(int argc, char *argv[])
 	}
 
 	ppid = getppid();
+
+	if (ppid == 1) {
+		/* The original parent is gone and the process has
+		 * been reparented to init.  This can happen if the
+		 * helper is started just as the parent is killed
+		 * during shutdown.  The error message doesn't need to
+		 * be stellar, since there won't be anything around to
+		 * capture and log it...
+		 */
+		fprintf(stderr, "%s: PPID == 1\n", progname);
+		exit(1);
+	}
+
 	file = argv[1];
 
 	result = fcntl_lock(file, &fd);