Merge tag 'cgroup-for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup

Pull cgroup updates from Tejun Heo:

 - Defer task cgroup unlink until after the dying task's final context
   switch so that controllers see the cgroup properly populated until
   the task is truly gone

 - cpuset cleanups and simplifications.

   Enforce that domain isolated CPUs stay in root or isolated partitions
   and fail if isolated+nohz_full would leave no housekeeping CPU. Fix
   sched/deadline root domain handling during CPU hot-unplug and race
   for tasks in attaching cpusets

 - Misc fixes including memory reclaim protection documentation and
   selftest KTAP conformance

* tag 'cgroup-for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup: (21 commits)
  cpuset: Treat cpusets in attaching as populated
  sched/deadline: Walk up cpuset hierarchy to decide root domain when hot-unplug
  cgroup/cpuset: Introduce cpuset_cpus_allowed_locked()
  docs: cgroup: No special handling of unpopulated memcgs
  docs: cgroup: Note about sibling relative reclaim protection
  docs: cgroup: Explain reclaim protection target
  selftests/cgroup: conform test to KTAP format output
  cpuset: remove need_rebuild_sched_domains
  cpuset: remove global remote_children list
  cpuset: simplify node setting on error
  cgroup: include missing header for struct irq_work
  cgroup: Fix sleeping from invalid context warning on PREEMPT_RT
  cgroup/cpuset: Globally track isolated_cpus update
  cgroup/cpuset: Ensure domain isolated CPUs stay in root or isolated partition
  cgroup/cpuset: Move up prstate_housekeeping_conflict() helper
  cgroup/cpuset: Fail if isolated and nohz_full don't leave any housekeeping
  cgroup/cpuset: Rename update_unbound_workqueue_cpumask() to update_isolation_cpumasks()
  cgroup: Defer task cgroup unlink until after the task is done switching out
  cgroup: Move dying_tasks cleanup from cgroup_task_release() to cgroup_task_free()
  cgroup: Rename cgroup lifecycle hooks to cgroup_task_*()
  ...

1 files changed, 76 insertions, 15 deletions

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index ae1eb7a85eb4..fa08ea288737 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -60,6 +60,7 @@
 #include <linux/sched/deadline.h>
 #include <linux/psi.h>
 #include <linux/nstree.h>
+#include <linux/irq_work.h>
 #include <net/sock.h>
 
 #define CREATE_TRACE_POINTS
@@ -287,6 +288,7 @@ static void kill_css(struct cgroup_subsys_state *css);
 static int cgroup_addrm_files(struct cgroup_subsys_state *css,
 			      struct cgroup *cgrp, struct cftype cfts[],
 			      bool is_add);
+static void cgroup_rt_init(void);
 
 #ifdef CONFIG_DEBUG_CGROUP_REF
 #define CGROUP_REF_FN_ATTRS	noinline
@@ -941,7 +943,8 @@ static void css_set_move_task(struct task_struct *task,
 		/*
 		 * We are synchronized through cgroup_threadgroup_rwsem
 		 * against PF_EXITING setting such that we can't race
-		 * against cgroup_exit()/cgroup_free() dropping the css_set.
+		 * against cgroup_task_dead()/cgroup_task_free() dropping
+		 * the css_set.
 		 */
 		WARN_ON_ONCE(task->flags & PF_EXITING);
 
@@ -6354,6 +6357,7 @@ int __init cgroup_init(void)
 	BUG_ON(ss_rstat_init(NULL));
 
 	get_user_ns(init_cgroup_ns.user_ns);
+	cgroup_rt_init();
 
 	cgroup_lock();
 
@@ -6967,19 +6971,29 @@ void cgroup_post_fork(struct task_struct *child,
 }
 
 /**
- * cgroup_exit - detach cgroup from exiting task
+ * cgroup_task_exit - detach cgroup from exiting task
  * @tsk: pointer to task_struct of exiting process
  *
  * Description: Detach cgroup from @tsk.
  *
  */
-void cgroup_exit(struct task_struct *tsk)
+void cgroup_task_exit(struct task_struct *tsk)
 {
 	struct cgroup_subsys *ss;
-	struct css_set *cset;
 	int i;
 
-	spin_lock_irq(&css_set_lock);
+	/* see cgroup_post_fork() for details */
+	do_each_subsys_mask(ss, i, have_exit_callback) {
+		ss->exit(tsk);
+	} while_each_subsys_mask();
+}
+
+static void do_cgroup_task_dead(struct task_struct *tsk)
+{
+	struct css_set *cset;
+	unsigned long flags;
+
+	spin_lock_irqsave(&css_set_lock, flags);
 
 	WARN_ON_ONCE(list_empty(&tsk->cg_list));
 	cset = task_css_set(tsk);
@@ -6997,15 +7011,61 @@ void cgroup_exit(struct task_struct *tsk)
 		     test_bit(CGRP_FREEZE, &task_dfl_cgroup(tsk)->flags)))
 		cgroup_update_frozen(task_dfl_cgroup(tsk));
 
-	spin_unlock_irq(&css_set_lock);
+	spin_unlock_irqrestore(&css_set_lock, flags);
+}
 
-	/* see cgroup_post_fork() for details */
-	do_each_subsys_mask(ss, i, have_exit_callback) {
-		ss->exit(tsk);
-	} while_each_subsys_mask();
+#ifdef CONFIG_PREEMPT_RT
+/*
+ * cgroup_task_dead() is called from finish_task_switch() which doesn't allow
+ * scheduling even in RT. As the task_dead path requires grabbing css_set_lock,
+ * this lead to sleeping in the invalid context warning bug. css_set_lock is too
+ * big to become a raw_spinlock. The task_dead path doesn't need to run
+ * synchronously but can't be delayed indefinitely either as the dead task pins
+ * the cgroup and task_struct can be pinned indefinitely. Bounce through lazy
+ * irq_work to allow batching while ensuring timely completion.
+ */
+static DEFINE_PER_CPU(struct llist_head, cgrp_dead_tasks);
+static DEFINE_PER_CPU(struct irq_work, cgrp_dead_tasks_iwork);
+
+static void cgrp_dead_tasks_iwork_fn(struct irq_work *iwork)
+{
+	struct llist_node *lnode;
+	struct task_struct *task, *next;
+
+	lnode = llist_del_all(this_cpu_ptr(&cgrp_dead_tasks));
+	llist_for_each_entry_safe(task, next, lnode, cg_dead_lnode) {
+		do_cgroup_task_dead(task);
+		put_task_struct(task);
+	}
+}
+
+static void __init cgroup_rt_init(void)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu) {
+		init_llist_head(per_cpu_ptr(&cgrp_dead_tasks, cpu));
+		per_cpu(cgrp_dead_tasks_iwork, cpu) =
+			IRQ_WORK_INIT_LAZY(cgrp_dead_tasks_iwork_fn);
+	}
+}
+
+void cgroup_task_dead(struct task_struct *task)
+{
+	get_task_struct(task);
+	llist_add(&task->cg_dead_lnode, this_cpu_ptr(&cgrp_dead_tasks));
+	irq_work_queue(this_cpu_ptr(&cgrp_dead_tasks_iwork));
 }
+#else	/* CONFIG_PREEMPT_RT */
+static void __init cgroup_rt_init(void) {}
 
-void cgroup_release(struct task_struct *task)
+void cgroup_task_dead(struct task_struct *task)
+{
+	do_cgroup_task_dead(task);
+}
+#endif	/* CONFIG_PREEMPT_RT */
+
+void cgroup_task_release(struct task_struct *task)
 {
 	struct cgroup_subsys *ss;
 	int ssid;
@@ -7013,6 +7073,11 @@ void cgroup_release(struct task_struct *task)
 	do_each_subsys_mask(ss, ssid, have_release_callback) {
 		ss->release(task);
 	} while_each_subsys_mask();
+}
+
+void cgroup_task_free(struct task_struct *task)
+{
+	struct css_set *cset = task_css_set(task);
 
 	if (!list_empty(&task->cg_list)) {
 		spin_lock_irq(&css_set_lock);
@@ -7020,11 +7085,7 @@ void cgroup_release(struct task_struct *task)
 		list_del_init(&task->cg_list);
 		spin_unlock_irq(&css_set_lock);
 	}
-}
 
-void cgroup_free(struct task_struct *task)
-{
-	struct css_set *cset = task_css_set(task);
 	put_css_set(cset);
 }