Re: [PATCH v3 1/1] pidfd: implement PIDFD_THREAD flag for pidfd_open()

From: Tycho Andersen
Date: Wed Jan 31 2024 - 12:24:20 EST


Hi Oleg,

On Wed, Jan 31, 2024 at 02:26:02PM +0100, Oleg Nesterov wrote:
> With this flag:
>
> - pidfd_open() doesn't require that the target task must be
> a thread-group leader
>
> - pidfd_poll() succeeds when the task exits and becomes a
> zombie (iow, passes exit_notify()), even if it is a leader
> and thread-group is not empty.
>
> This means that the behaviour of pidfd_poll(PIDFD_THREAD,
> pid-of-group-leader) is not well defined if it races with
> exec() from its sub-thread; pidfd_poll() can succeed or not
> depending on whether pidfd_task_exited() is called before
> or after exchange_tids().
>
> Perhaps we can improve this behaviour later, pidfd_poll()
> can probably take sig->group_exec_task into account. But
> this doesn't really differ from the case when the leader
> exits before other threads (so pidfd_poll() succeeds) and
> then another thread execs and pidfd_poll() will block again.

I don't have a strong opinion here; leaving it "undefined" for now is
fine with me.

> @@ -2173,7 +2195,9 @@ static int __pidfd_prepare(struct pid *pid, unsigned int flags, struct file **re
> */
> int pidfd_prepare(struct pid *pid, unsigned int flags, struct file **ret)
> {
> - if (!pid || !pid_has_task(pid, PIDTYPE_TGID))
> + bool thread = flags & PIDFD_THREAD;
> +
> + if (!pid || !pid_has_task(pid, thread ? PIDTYPE_PID : PIDTYPE_TGID));

Small typo here, trailing ;. When I fix that, tests pass for me.

Assuming that's fixed up:

Reviewed-by: Tycho Andersen <tandersen@xxxxxxxxxxx>
Tested-by: Tycho Andersen <tandersen@xxxxxxxxxxx>

Thanks for your help!

Tycho