Re: [PATCH RFC v1] pidfd: fix a race in setting exit_state for pidfd polling
From: Christian Brauner
Date: Fri Jul 19 2019 - 12:53:57 EST
On July 19, 2019 6:51:20 PM GMT+02:00, Joel Fernandes <joelaf@xxxxxxxxxx> wrote:
>On Fri, Jul 19, 2019 at 12:27 PM Christian Brauner
>> On Fri, Jul 19, 2019 at 06:14:05PM +0200, Oleg Nesterov wrote:
>> > it seems that I missed something else...
>> > On 07/17, Joel Fernandes (Google) wrote:
>> > >
>> > > @@ -1156,10 +1157,11 @@ static int wait_task_zombie(struct
>wait_opts *wo, struct task_struct *p)
>> > > ptrace_unlink(p);
>> > >
>> > > /* If parent wants a zombie, don't release it now */
>> > > - state = EXIT_ZOMBIE;
>> > > + p->exit_state = EXIT_ZOMBIE;
>> > > if (do_notify_parent(p, p->exit_signal))
>> > > - state = EXIT_DEAD;
>> > > - p->exit_state = state;
>> > > + p->exit_state = EXIT_DEAD;
>> > > +
>> > > + state = p->exit_state;
>> > > write_unlock_irq(&tasklist_lock);
>> > why do you think we also need to change wait_task_zombie() ?
>> > pidfd_poll() only needs the exit_state != 0 check, we know that it
>> > is not zero at this point. Why do we need to change exit_state
>> > do_notify_parent() ?
>> Oh, because of?:
>> * Move the task's state to DEAD/TRACE, only one thread can
>> state = (ptrace_reparented(p) && thread_group_leader(p)) ?
>> EXIT_TRACE : EXIT_DEAD;
>> if (cmpxchg(&p->exit_state, EXIT_ZOMBIE, state) !=
>> return 0;
>> So exit_state will definitely be set in this scenario. Good point.
>Agreed. Christian, do you mind dropping this hunk from the patch or do
>you want me to resend the patch with the hunk dropped?
Yeah, no problem. :)