Comment 6 for bug 728531

Here is a comment from the Upstart changelog which attempts to explain the scenario leading to the out-of-memory condition:

        * init/job.c: job_new(): Fix for nasty OOM scenario when an attempt is made
          to start a chroot job with same name as running single-instance non-chroot
          job. Previously, the path for a chrooted Job was set from the (nul) instance
          name of its parent JobClass. However, since that job instance name already
          existed outside the chroot and was already D-Bus registered,
          nih_dbus_object_new() (which allocates storage and is called by
          job_register()) returned NULL and kept doing so due to being called
          within NIH_MUST(). This resulted in OOM due to an as-yet unidentified
          D-Bus bug (possibly a leak in find_subtree_recurse()).