ptrace: do_notify_parent_cldstop: fix the wrong ->nsproxy usage
authorOleg Nesterov <oleg@redhat.com>
Wed, 17 Jun 2009 23:27:35 +0000 (16:27 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 18 Jun 2009 20:03:52 +0000 (13:03 -0700)
If the non-traced sub-thread calls do_notify_parent_cldstop(), we send the
notification to group_leader->real_parent and we report group_leader's
pid.

But, if group_leader is traced we use the wrong ->parent->nsproxy->pid_ns,
the tracer and parent can live in different namespaces.  Change the code
to use "parent" instead of tsk->parent.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Roland McGrath <roland@redhat.com>
Acked-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
kernel/signal.c

index 09ccc1c0e1f88216eb626c99557958bd5d311f9a..ccf1ceedaebef64b137553e2e82250dbfe6a7707 100644 (file)
@@ -1499,7 +1499,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk, int why)
         * see comment in do_notify_parent() abot the following 3 lines
         */
        rcu_read_lock();
-       info.si_pid = task_pid_nr_ns(tsk, tsk->parent->nsproxy->pid_ns);
+       info.si_pid = task_pid_nr_ns(tsk, parent->nsproxy->pid_ns);
        info.si_uid = __task_cred(tsk)->uid;
        rcu_read_unlock();