check_by_ssh: fix child process leak on timeouts

When check_by_ssh runs into a timeout it simply exits keeping all child processes running.
Simply adopting the kill loop from runcmd_timeout_alarm_handler() fixes this.

Signed-off-by: Sven Nierlein <sven@nierlein.de>

1 files changed, 13 insertions, 0 deletions

diff --git a/lib/utils_cmd.h b/lib/utils_cmd.h
index ebaf15be..6f3aeb81 100644
--- a/lib/utils_cmd.h
+++ b/lib/utils_cmd.h
@@ -32,4 +32,17 @@ void cmd_init (void);
 #define CMD_NO_ARRAYS 0x01   /* don't populate arrays at all */
 #define CMD_NO_ASSOC 0x02    /* output.line won't point to buf */
 
+/* This variable must be global, since there's no way the caller
+ * can forcibly slay a dead or ungainly running program otherwise.
+ * Multithreading apps and plugins can initialize it (via CMD_INIT)
+ * in an async safe manner PRIOR to calling cmd_run() or cmd_run_array()
+ * for the first time.
+ *
+ * The check for initialized values is atomic and can
+ * occur in any number of threads simultaneously. */
+static pid_t *_cmd_pids = NULL;
+
+RETSIGTYPE timeout_alarm_handler (int);
+
+
 #endif /* _UTILS_CMD_ */