cgroup: avoid the unnecessary list_add(dying_tasks) in cgroup_exit()

Hello,

Sorry for delay and for confusing you in our previous discussion here https://lore.kernel.org/all/[email protected]/

No, cgroup_exit() can't rely on group_dead, this is racy.

And no, we can't shift css_set_skip_task_iters/etc from cgroup_release() to cgroup_exit(), an execing sub-thread can change the group leader.

Let me at least send the simple patch which looks "obviously good" to me.

I would really like to remove the usage of signal->live in cgroup.c, but so far I do not see a simple solution.

With or without this change cgroup.procs can be empty but cgroup.threads is not. But at least the exiting sub-threads which have already passed atomic_dec_and_test() should call cgroup_exit() "soon".

Oleg.

cgroup_exit() needs to do this only if the exiting task is a leader and it is not the last live thread. The patch doesn't use delay_group_leader(), atomic_read(signal->live) matches the code css_task_iter_advance() more.

cgroup_release() can now check list_empty(task->cg_list) before it takes css_set_lock and calls ss_set_skip_task_iters().

---
 kernel/cgroup/cgroup.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

index e32b6972c478..0cbee8118c95 100644
@@ -6686,8 +6686,10 @@ void cgroup_exit(struct task_struct *tsk)
 	WARN_ON_ONCE(list_empty(&tsk->cg_list));
 	cset = task_css_set(tsk);
 	css_set_move_task(tsk, cset, NULL, false);
-	list_add_tail(&tsk->cg_list, &cset->dying_tasks);
 	cset->nr_tasks--;
+	/* matches the signal->live check in css_task_iter_advance() */
+	if (thread_group_leader(tsk) && atomic_read(&tsk->signal->live))
+		list_add_tail(&tsk->cg_list, &cset->dying_tasks);
 
 	if (dl_task(tsk))
 		dec_dl_tasks_cs(tsk);

@@ -6714,10 +6716,12 @@ void cgroup_release(struct task_struct *task)
 		ss->release(task);
 	} while_each_subsys_mask();
 
-	spin_lock_irq(&css_set_lock);
-	css_set_skip_task_iters(task_css_set(task), task);
-	list_del_init(&task->cg_list);
-	spin_unlock_irq(&css_set_lock);
+	if (!list_empty(&task->cg_list)) {
+		spin_lock_irq(&css_set_lock);
+		css_set_skip_task_iters(task_css_set(task), task);
+		list_del_init(&task->cg_list);
+		spin_unlock_irq(&css_set_lock);
+	}
 }
 
 void cgroup_free(struct task_struct *task)

2.25.1.362.g51ebf55

Applied to cgroup/for-6.11.

Thanks.

tejun