Caj Larsson <c...@omnicloud.com> writes:
> Hello,
>
> We are setting up a container using a CLONE_NEWNS linux namespace.
> Previously we used the 3.4.4 kernel, which worked fine. After I
> upgraded i also experienced races in netlink, which has been resolved
> by placing a monitor around the namespace setup. When we upgraded to
> Linux 3.8.0 however our init processes does not get reaped when the
> namespace is killed and lingers as zombie process under the global
> init.
>
> The init has multiple threads when running and two remain in the
> zombie. One of them is hung in zap_pid_ns_processes() and has been set
> uninterruptible. The other one, which has Tgid=PID, is still in
> do_exit().
Doh.
Thank you for the detailed bug report it appears I goofed, and failed
to account for a multi-threaded init.
Will you please verify that the following patch fixes it for you.
Signed-off-by: "Eric W. Biederman" <ebied...@xmission.com>
---
diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c
index c1c3dc1..72b7722 100644
--- a/kernel/pid_namespace.c
+++ b/kernel/pid_namespace.c
@@ -181,6 +181,7 @@ void zap_pid_ns_processes(struct pid_namespace *pid_ns)
int nr;
int rc;
struct task_struct *task, *me = current;
+ int init_pids = task_pid_vnr(me) == 1 ? 1 : 2;
/* Don't allow any more processes into the pid namespace */
disable_pid_allocation(pid_ns);
@@ -230,7 +231,7 @@ void zap_pid_ns_processes(struct pid_namespace *pid_ns)
*/
for (;;) {
set_current_state(TASK_UNINTERRUPTIBLE);
- if (pid_ns->nr_hashed == 1)
+ if (pid_ns->nr_hashed == init_pids)
break;
schedule();
}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
