core/cgroup: upgrade log level when we fail to rescope a pid

keszybz · keszybz · commit 7a2ba4078731 · 2021-06-30T21:55:16.000+02:00
See https://bugzilla.redhat.com/show_bug.cgi?id=1973058 again: systemd[1779]: Started Application launched by gnome-session-binary. systemd[1779]: app-glib-liveinst\x2dsetup-1897.scope: Failed to add PIDs to scope's control group: No such process systemd[1779]: app-glib-liveinst\x2dsetup-1897.scope: Failed with result 'resources'. systemd[1779]: Failed to start Application launched by gnome-session-binary. systemd[1779]: app-glib-xdg\x2duser\x2ddirs-1900.scope: Failed to add PIDs to scope's control group: No such process systemd[1779]: app-glib-xdg\x2duser\x2ddirs-1900.scope: Failed with result 'resources'. systemd[1779]: Failed to start Application launched by gnome-session-binary. systemd[1779]: app-gnome-gsettings\x2ddata\x2dconvert-1902.scope: Failed to add PIDs to scope's control group: No such process systemd[1779]: app-gnome-gsettings\x2ddata\x2dconvert-1902.scope: Failed with result 'resources'. systemd[1779]: Failed to start Application launched by gnome-session-binary. Since we don't show the PID anywhere, it can be quite hard to figure out what is going on. There may be logs from the pid above or below in the log, but we have no PID number to identify them. So let's upgrade the log from unit_attach_pids_to_cgroup() to tell us precisely which PIDs and why couldn't be handled.
diff --git a/src/core/cgroup.c b/src/core/cgroup.c
@@ -2164,19 +2164,23 @@ int unit_attach_pids_to_cgroup(Unit *u, Set *pids, const char *suffix_path) {
                 /* First, attach the PID to the main cgroup hierarchy */
                 q = cg_attach(SYSTEMD_CGROUP_CONTROLLER, p, pid);
                 if (q < 0) {
-                        log_unit_debug_errno(u, q, "Couldn't move process " PID_FMT " to requested cgroup '%s': %m", pid, p);
+                        bool again = MANAGER_IS_USER(u->manager) && ERRNO_IS_PRIVILEGE(q);
 
-                        if (MANAGER_IS_USER(u->manager) && ERRNO_IS_PRIVILEGE(q)) {
+                        log_unit_full_errno(u, again ? LOG_DEBUG : LOG_INFO,  q,
+                                            "Couldn't move process "PID_FMT" to%s requested cgroup '%s': %m",
+                                            pid, again ? " directly" : "", p);
+
+                        if (again) {
                                 int z;
 
-                                /* If we are in a user instance, and we can't move the process ourselves due to
-                                 * permission problems, let's ask the system instance about it instead. Since it's more
-                                 * privileged it might be able to move the process across the leaves of a subtree who's
-                                 * top node is not owned by us. */
+                                /* If we are in a user instance, and we can't move the process ourselves due
+                                 * to permission problems, let's ask the system instance about it instead.
+                                 * Since it's more privileged it might be able to move the process across the
+                                 * leaves of a subtree whose top node is not owned by us. */
 
                                 z = unit_attach_pid_to_cgroup_via_bus(u, pid, suffix_path);
                                 if (z < 0)
-                                        log_unit_debug_errno(u, z, "Couldn't move process " PID_FMT " to requested cgroup '%s' via the system bus either: %m", pid, p);
+                                        log_unit_info_errno(u, z, "Couldn't move process "PID_FMT" to requested cgroup '%s' (directly or via the system bus): %m", pid, p);
                                 else
                                         continue; /* When the bus thing worked via the bus we are fully done for this PID. */
                         }
