state: fix race condition when reading cgroup

by the time crun attempts to read from the cgroup, systemd might have already cleaned it up. When using systemd, on ENOENT state reports the container as "stopped" instead of an error. Closes: containers/podman#7148 Signed-off-by: Giuseppe Scrivano <[email protected]>
giuseppe · Aug 27, 2020 · ed9c3e6 · ed9c3e6
1 parent 8bd4f3e
commit ed9c3e6
Showing 1 changed file with 11 additions and 1 deletion.
diff --git a/src/libcrun/container.c b/src/libcrun/container.c
@@ -2194,7 +2194,17 @@ libcrun_get_container_state_string (const char *id, libcrun_container_status_t *
 
       ret = libcrun_cgroup_is_container_paused (status->cgroup_path, cgroup_mode, &paused, err);
       if (UNLIKELY (ret < 0))
-        return ret;
+        {
+          /* The cgroup might have been cleaned up by systemd by the time we try to read it, so ignore ENOENT.  */
+          if (status->systemd_cgroup && crun_error_get_errno (err) == ENOENT)
+            {
+              crun_error_release (err);
+              *container_status = "stopped";
+              return 0;
+            }
+
+          return ret;
+        }
     }
 
   if (! *running)