Merge branch 'for-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup

Pull cgroup updates from Tejun Heo:

 - cgroup.kill is added which implements atomic killing of the whole
   subtree.

   Down the line, this should be able to replace the multiple userland
   implementations of "keep killing till empty".

 - PSI can now be turned off at boot time to avoid overhead for
   configurations which don't care about PSI.

* 'for-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cgroup: make per-cgroup pressure stall tracking configurable
  cgroup: Fix kernel-doc
  cgroup: inline cgroup_task_freeze()
  tests/cgroup: test cgroup.kill
  tests/cgroup: move cg_wait_for(), cg_prepare_for_wait()
  tests/cgroup: use cgroup.kill in cg_killall()
  docs/cgroup: add entry for cgroup.kill
  cgroup: introduce cgroup.kill

3 files changed, 182 insertions, 30 deletions

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 9cc8c3a686b1..ea5b0f01d2b3 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -209,6 +209,22 @@ struct cgroup_namespace init_cgroup_ns = {
 static struct file_system_type cgroup2_fs_type;
 static struct cftype cgroup_base_files[];
 
+/* cgroup optional features */
+enum cgroup_opt_features {
+#ifdef CONFIG_PSI
+	OPT_FEATURE_PRESSURE,
+#endif
+	OPT_FEATURE_COUNT
+};
+
+static const char *cgroup_opt_feature_names[OPT_FEATURE_COUNT] = {
+#ifdef CONFIG_PSI
+	"pressure",
+#endif
+};
+
+static u16 cgroup_feature_disable_mask __read_mostly;
+
 static int cgroup_apply_control(struct cgroup *cgrp);
 static void cgroup_finalize_control(struct cgroup *cgrp, int ret);
 static void css_task_iter_skip(struct css_task_iter *it,
@@ -2390,7 +2406,7 @@ struct task_struct *cgroup_taskset_next(struct cgroup_taskset *tset,
 }
 
 /**
- * cgroup_taskset_migrate - migrate a taskset
+ * cgroup_migrate_execute - migrate a taskset
  * @mgctx: migration context
  *
  * Migrate tasks in @mgctx as setup by migration preparation functions.
@@ -3632,6 +3648,18 @@ static void cgroup_pressure_release(struct kernfs_open_file *of)
 {
 	psi_trigger_replace(&of->priv, NULL);
 }
+
+bool cgroup_psi_enabled(void)
+{
+	return (cgroup_feature_disable_mask & (1 << OPT_FEATURE_PRESSURE)) == 0;
+}
+
+#else /* CONFIG_PSI */
+bool cgroup_psi_enabled(void)
+{
+	return false;
+}
+
 #endif /* CONFIG_PSI */
 
 static int cgroup_freeze_show(struct seq_file *seq, void *v)
@@ -3668,6 +3696,80 @@ static ssize_t cgroup_freeze_write(struct kernfs_open_file *of,
 	return nbytes;
 }
 
+static void __cgroup_kill(struct cgroup *cgrp)
+{
+	struct css_task_iter it;
+	struct task_struct *task;
+
+	lockdep_assert_held(&cgroup_mutex);
+
+	spin_lock_irq(&css_set_lock);
+	set_bit(CGRP_KILL, &cgrp->flags);
+	spin_unlock_irq(&css_set_lock);
+
+	css_task_iter_start(&cgrp->self, CSS_TASK_ITER_PROCS | CSS_TASK_ITER_THREADED, &it);
+	while ((task = css_task_iter_next(&it))) {
+		/* Ignore kernel threads here. */
+		if (task->flags & PF_KTHREAD)
+			continue;
+
+		/* Skip tasks that are already dying. */
+		if (__fatal_signal_pending(task))
+			continue;
+
+		send_sig(SIGKILL, task, 0);
+	}
+	css_task_iter_end(&it);
+
+	spin_lock_irq(&css_set_lock);
+	clear_bit(CGRP_KILL, &cgrp->flags);
+	spin_unlock_irq(&css_set_lock);
+}
+
+static void cgroup_kill(struct cgroup *cgrp)
+{
+	struct cgroup_subsys_state *css;
+	struct cgroup *dsct;
+
+	lockdep_assert_held(&cgroup_mutex);
+
+	cgroup_for_each_live_descendant_pre(dsct, css, cgrp)
+		__cgroup_kill(dsct);
+}
+
+static ssize_t cgroup_kill_write(struct kernfs_open_file *of, char *buf,
+				 size_t nbytes, loff_t off)
+{
+	ssize_t ret = 0;
+	int kill;
+	struct cgroup *cgrp;
+
+	ret = kstrtoint(strstrip(buf), 0, &kill);
+	if (ret)
+		return ret;
+
+	if (kill != 1)
+		return -ERANGE;
+
+	cgrp = cgroup_kn_lock_live(of->kn, false);
+	if (!cgrp)
+		return -ENOENT;
+
+	/*
+	 * Killing is a process directed operation, i.e. the whole thread-group
+	 * is taken down so act like we do for cgroup.procs and only make this
+	 * writable in non-threaded cgroups.
+	 */
+	if (cgroup_is_threaded(cgrp))
+		ret = -EOPNOTSUPP;
+	else
+		cgroup_kill(cgrp);
+
+	cgroup_kn_unlock(of->kn);
+
+	return ret ?: nbytes;
+}
+
 static int cgroup_file_open(struct kernfs_open_file *of)
 {
 	struct cftype *cft = of_cft(of);
@@ -3882,6 +3984,8 @@ static int cgroup_addrm_files(struct cgroup_subsys_state *css,
 restart:
 	for (cft = cfts; cft != cft_end && cft->name[0] != '\0'; cft++) {
 		/* does cft->flags tell us to skip this file on @cgrp? */
+		if ((cft->flags & CFTYPE_PRESSURE) && !cgroup_psi_enabled())
+			continue;
 		if ((cft->flags & __CFTYPE_ONLY_ON_DFL) && !cgroup_on_dfl(cgrp))
 			continue;
 		if ((cft->flags & __CFTYPE_NOT_ON_DFL) && cgroup_on_dfl(cgrp))
@@ -3959,6 +4063,9 @@ static int cgroup_init_cftypes(struct cgroup_subsys *ss, struct cftype *cfts)
 
 		WARN_ON(cft->ss || cft->kf_ops);
 
+		if ((cft->flags & CFTYPE_PRESSURE) && !cgroup_psi_enabled())
+			continue;
+
 		if (cft->seq_start)
 			kf_ops = &cgroup_kf_ops;
 		else
@@ -4861,12 +4968,18 @@ static struct cftype cgroup_base_files[] = {
 		.write = cgroup_freeze_write,
 	},
 	{
+		.name = "cgroup.kill",
+		.flags = CFTYPE_NOT_ON_ROOT,
+		.write = cgroup_kill_write,
+	},
+	{
 		.name = "cpu.stat",
 		.seq_show = cpu_stat_show,
 	},
 #ifdef CONFIG_PSI
 	{
 		.name = "io.pressure",
+		.flags = CFTYPE_PRESSURE,
 		.seq_show = cgroup_io_pressure_show,
 		.write = cgroup_io_pressure_write,
 		.poll = cgroup_pressure_poll,
@@ -4874,6 +4987,7 @@ static struct cftype cgroup_base_files[] = {
 	},
 	{
 		.name = "memory.pressure",
+		.flags = CFTYPE_PRESSURE,
 		.seq_show = cgroup_memory_pressure_show,
 		.write = cgroup_memory_pressure_write,
 		.poll = cgroup_pressure_poll,
@@ -4881,6 +4995,7 @@ static struct cftype cgroup_base_files[] = {
 	},
 	{
 		.name = "cpu.pressure",
+		.flags = CFTYPE_PRESSURE,
 		.seq_show = cgroup_cpu_pressure_show,
 		.write = cgroup_cpu_pressure_write,
 		.poll = cgroup_pressure_poll,
@@ -6080,6 +6195,8 @@ void cgroup_post_fork(struct task_struct *child,
 		      struct kernel_clone_args *kargs)
 	__releases(&cgroup_threadgroup_rwsem) __releases(&cgroup_mutex)
 {
+	unsigned long cgrp_flags = 0;
+	bool kill = false;
 	struct cgroup_subsys *ss;
 	struct css_set *cset;
 	int i;
@@ -6091,6 +6208,11 @@ void cgroup_post_fork(struct task_struct *child,
 
 	/* init tasks are special, only link regular threads */
 	if (likely(child->pid)) {
+		if (kargs->cgrp)
+			cgrp_flags = kargs->cgrp->flags;
+		else
+			cgrp_flags = cset->dfl_cgrp->flags;
+
 		WARN_ON_ONCE(!list_empty(&child->cg_list));
 		cset->nr_tasks++;
 		css_set_move_task(child, NULL, cset, false);
@@ -6099,23 +6221,32 @@ void cgroup_post_fork(struct task_struct *child,
 		cset = NULL;
 	}
 
-	/*
-	 * If the cgroup has to be frozen, the new task has too.  Let's set
-	 * the JOBCTL_TRAP_FREEZE jobctl bit to get the task into the
-	 * frozen state.
-	 */
-	if (unlikely(cgroup_task_freeze(child))) {
-		spin_lock(&child->sighand->siglock);
-		WARN_ON_ONCE(child->frozen);
-		child->jobctl |= JOBCTL_TRAP_FREEZE;
-		spin_unlock(&child->sighand->siglock);
+	if (!(child->flags & PF_KTHREAD)) {
+		if (unlikely(test_bit(CGRP_FREEZE, &cgrp_flags))) {
+			/*
+			 * If the cgroup has to be frozen, the new task has
+			 * too. Let's set the JOBCTL_TRAP_FREEZE jobctl bit to
+			 * get the task into the frozen state.
+			 */
+			spin_lock(&child->sighand->siglock);
+			WARN_ON_ONCE(child->frozen);
+			child->jobctl |= JOBCTL_TRAP_FREEZE;
+			spin_unlock(&child->sighand->siglock);
+
+			/*
+			 * Calling cgroup_update_frozen() isn't required here,
+			 * because it will be called anyway a bit later from
+			 * do_freezer_trap(). So we avoid cgroup's transient
+			 * switch from the frozen state and back.
+			 */
+		}
 
 		/*
-		 * Calling cgroup_update_frozen() isn't required here,
-		 * because it will be called anyway a bit later from
-		 * do_freezer_trap(). So we avoid cgroup's transient switch
-		 * from the frozen state and back.
+		 * If the cgroup is to be killed notice it now and take the
+		 * child down right after we finished preparing it for
+		 * userspace.
 		 */
+		kill = test_bit(CGRP_KILL, &cgrp_flags);
 	}
 
 	spin_unlock_irq(&css_set_lock);
@@ -6138,6 +6269,10 @@ void cgroup_post_fork(struct task_struct *child,
 		put_css_set(rcset);
 	}
 
+	/* Cgroup has to be killed so take down child immediately. */
+	if (unlikely(kill))
+		do_send_sig_info(SIGKILL, SEND_SIG_NOINFO, child, PIDTYPE_TGID);
+
 	cgroup_css_set_put_fork(kargs);
 }
 
@@ -6163,7 +6298,8 @@ void cgroup_exit(struct task_struct *tsk)
 	cset->nr_tasks--;
 
 	WARN_ON_ONCE(cgroup_task_frozen(tsk));
-	if (unlikely(cgroup_task_freeze(tsk)))
+	if (unlikely(!(tsk->flags & PF_KTHREAD) &&
+		     test_bit(CGRP_FREEZE, &task_dfl_cgroup(tsk)->flags)))
 		cgroup_update_frozen(task_dfl_cgroup(tsk));
 
 	spin_unlock_irq(&css_set_lock);
@@ -6214,6 +6350,15 @@ static int __init cgroup_disable(char *str)
 			pr_info("Disabling %s control group subsystem\n",
 				ss->name);
 		}
+
+		for (i = 0; i < OPT_FEATURE_COUNT; i++) {
+			if (strcmp(token, cgroup_opt_feature_names[i]))
+				continue;
+			cgroup_feature_disable_mask |= 1 << i;
+			pr_info("Disabling %s control group feature\n",
+				cgroup_opt_feature_names[i]);
+			break;
+		}
 	}
 	return 1;
 }
@@ -6512,6 +6657,9 @@ static ssize_t show_delegatable_files(struct cftype *files, char *buf,
 		if (!(cft->flags & CFTYPE_NS_DELEGATABLE))
 			continue;
 
+		if ((cft->flags & CFTYPE_PRESSURE) && !cgroup_psi_enabled())
+			continue;
+
 		if (prefix)
 			ret += snprintf(buf + ret, size - ret, "%s.", prefix);
 
diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index cee265cb535c..7f0e58917432 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -220,7 +220,7 @@ void cgroup_rstat_flush_irqsafe(struct cgroup *cgrp)
 }
 
 /**
- * cgroup_rstat_flush_begin - flush stats in @cgrp's subtree and hold
+ * cgroup_rstat_flush_hold - flush stats in @cgrp's subtree and hold
  * @cgrp: target cgroup
  *
  * Flush stats in @cgrp's subtree and prevent further flushes.  Must be
diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 58b36d17a09a..1652f2bb54b7 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -148,6 +148,7 @@
 static int psi_bug __read_mostly;
 
 DEFINE_STATIC_KEY_FALSE(psi_disabled);
+DEFINE_STATIC_KEY_TRUE(psi_cgroups_enabled);
 
 #ifdef CONFIG_PSI_DEFAULT_DISABLED
 static bool psi_enable;
@@ -215,6 +216,9 @@ void __init psi_init(void)
 		return;
 	}
 
+	if (!cgroup_psi_enabled())
+		static_branch_disable(&psi_cgroups_enabled);
+
 	psi_period = jiffies_to_nsecs(PSI_FREQ);
 	group_init(&psi_system);
 }
@@ -748,23 +752,23 @@ static void psi_group_change(struct psi_group *group, int cpu,
 
 static struct psi_group *iterate_groups(struct task_struct *task, void **iter)
 {
+	if (*iter == &psi_system)
+		return NULL;
+
 #ifdef CONFIG_CGROUPS
-	struct cgroup *cgroup = NULL;
+	if (static_branch_likely(&psi_cgroups_enabled)) {
+		struct cgroup *cgroup = NULL;
 
-	if (!*iter)
-		cgroup = task->cgroups->dfl_cgrp;
-	else if (*iter == &psi_system)
-		return NULL;
-	else
-		cgroup = cgroup_parent(*iter);
+		if (!*iter)
+			cgroup = task->cgroups->dfl_cgrp;
+		else
+			cgroup = cgroup_parent(*iter);
 
-	if (cgroup && cgroup_parent(cgroup)) {
-		*iter = cgroup;
-		return cgroup_psi(cgroup);
+		if (cgroup && cgroup_parent(cgroup)) {
+			*iter = cgroup;
+			return cgroup_psi(cgroup);
+		}
 	}
-#else
-	if (*iter)
-		return NULL;
 #endif
 	*iter = &psi_system;
 	return &psi_system;