core: reduce the number of stalled PIDs from the watched processes list when possible

Some PIDs can remain in the watched list even though their processes have
exited since a long time. It can easily happen if the main process of a forking
service manages to spawn a child before the control process exits for example.

However when a pid is about to be mapped to a unit by calling unit_watch_pid(),
the caller usually knows if the pid should belong to this unit exclusively: if
we just forked() off a child, then we can be sure that its PID is otherwise
unused. In this case we take this opportunity to remove any stalled PIDs from
the watched process list.

If we learnt about a PID in any other form (for example via PID file, via
searching, MAINPID= and so on), then we can't assume anything.

1 files changed, 8 insertions, 8 deletions

diff --git a/src/core/service.c b/src/core/service.c
index 4682e86f8e..883ca7bbfe 100644
--- a/src/core/service.c
+++ b/src/core/service.c
@@ -990,7 +990,7 @@ static int service_load_pid_file(Service *s, bool may_warn) {
         if (r < 0)
                 return r;
 
-        r = unit_watch_pid(UNIT(s), pid);
+        r = unit_watch_pid(UNIT(s), pid, false);
         if (r < 0) /* FIXME: we need to do something here */
                 return log_unit_warning_errno(UNIT(s), r, "Failed to watch PID "PID_FMT" for service: %m", pid);
 
@@ -1020,7 +1020,7 @@ static void service_search_main_pid(Service *s) {
         if (service_set_main_pid(s, pid) < 0)
                 return;
 
-        r = unit_watch_pid(UNIT(s), pid);
+        r = unit_watch_pid(UNIT(s), pid, false);
         if (r < 0)
                 /* FIXME: we need to do something here */
                 log_unit_warning_errno(UNIT(s), r, "Failed to watch PID "PID_FMT" from: %m", pid);
@@ -1154,7 +1154,7 @@ static int service_coldplug(Unit *u) {
                     SERVICE_RUNNING, SERVICE_RELOAD,
                     SERVICE_STOP, SERVICE_STOP_WATCHDOG, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST,
                     SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL))) {
-                r = unit_watch_pid(UNIT(s), s->main_pid);
+                r = unit_watch_pid(UNIT(s), s->main_pid, false);
                 if (r < 0)
                         return r;
         }
@@ -1166,7 +1166,7 @@ static int service_coldplug(Unit *u) {
                    SERVICE_RELOAD,
                    SERVICE_STOP, SERVICE_STOP_WATCHDOG, SERVICE_STOP_SIGTERM, SERVICE_STOP_SIGKILL, SERVICE_STOP_POST,
                    SERVICE_FINAL_SIGTERM, SERVICE_FINAL_SIGKILL)) {
-                r = unit_watch_pid(UNIT(s), s->control_pid);
+                r = unit_watch_pid(UNIT(s), s->control_pid, false);
                 if (r < 0)
                         return r;
         }
@@ -1566,8 +1566,8 @@ static int service_spawn(
         s->exec_fd_event_source = TAKE_PTR(exec_fd_source);
         s->exec_fd_hot = false;
 
-        r = unit_watch_pid(UNIT(s), pid);
-        if (r < 0) /* FIXME: we need to do something here */
+        r = unit_watch_pid(UNIT(s), pid, true);
+        if (r < 0)
                 return r;
 
         *_pid = pid;
@@ -3705,7 +3705,7 @@ static void service_notify_message(
                         if (r > 0) {
                                 service_set_main_pid(s, new_main_pid);
 
-                                r = unit_watch_pid(UNIT(s), new_main_pid);
+                                r = unit_watch_pid(UNIT(s), new_main_pid, false);
                                 if (r < 0)
                                         log_unit_warning_errno(UNIT(s), r, "Failed to watch new main PID "PID_FMT" for service: %m", new_main_pid);
 
@@ -3923,7 +3923,7 @@ static void service_bus_name_owner_change(
                         log_unit_debug(u, "D-Bus name %s is now owned by process " PID_FMT, name, pid);
 
                         service_set_main_pid(s, pid);
-                        unit_watch_pid(UNIT(s), pid);
+                        unit_watch_pid(UNIT(s), pid, false);
                 }
         }
 }