Messages in this thread | | | Date | Fri, 16 May 2008 03:33:04 -0700 | From | Andrew Morton <> | Subject | Re: [PATCH 10/10] Bsdacct: account dying tasks in all relevant namespaces |
| |
On Fri, 16 May 2008 14:11:04 +0400 Pavel Emelyanov <xemul@openvz.org> wrote:
> > >> + * handles process accounting for an exiting task > >> + */ > >> +void acct_process(void) > >> +{ > >> + struct pid_namespace *ns; > >> + > >> + for (ns = task_active_pid_ns(current); ns != NULL; ns = ns->parent) > >> + acct_process_in_ns(ns); > >> +} > > > > Is it really safe to walk this list locklessly? > > Well, yes - I wrote it in the comment to the 10th patch: > > "ns->parent access if safe lockless, since current it still > alive and holds its namespace, which in turn holds its parent."
oic.
> Should I also add this as a comment here?
It wouldn't hurt...
| |