Re: [RFC 1/3] pidfd: allow pidfd_open() on non-thread-group leaders
From: Oleg Nesterov <oleg@redhat.com>
Date: 2023-11-30 17:41:03
Also in:
lkml
Hi Tycho, I can't really read this patch now, possibly I am wrong, but... On 11/30, Tycho Andersen wrote:
quoted hunk ↗ jump to hunk
@@ -263,16 +263,25 @@ void release_task(struct task_struct *p) */ zap_leader = 0; leader = p->group_leader; - if (leader != p && thread_group_empty(leader) - && leader->exit_state == EXIT_ZOMBIE) { - /* - * If we were the last child thread and the leader has - * exited already, and the leader's parent ignores SIGCHLD, - * then we are the one who should release the leader. - */ - zap_leader = do_notify_parent(leader, leader->exit_signal); - if (zap_leader) - leader->exit_state = EXIT_DEAD; + if (leader != p) { + if (thread_group_empty(leader) + && leader->exit_state == EXIT_ZOMBIE) { + /* + * If we were the last child thread and the leader has + * exited already, and the leader's parent ignores SIGCHLD, + * then we are the one who should release the leader. + */ + zap_leader = do_notify_parent(leader, + leader->exit_signal); + if (zap_leader) + leader->exit_state = EXIT_DEAD; + } else { + /* + * wake up pidfd pollers anyway, they want to know this + * thread is dying. + */ + wake_up_all(&thread_pid->wait_pidfd);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ somehow I can't believe this is a good change after a quick glance ;) I think that wake_up_all(wait_pidfd) should have a single caller, do_notify_pidfd(). This probably means it should be shiftef from do_notify_parent() to exit_notify(), I am not sure... No? Oleg.