1 files changed, 99 insertions, 93 deletions

diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 1652f2bb54b7..e14358178849 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Pressure stall information for CPU, memory and IO
  *
@@ -34,13 +35,19 @@
  * delayed on that resource such that nobody is advancing and the CPU
  * goes idle. This leaves both workload and CPU unproductive.
  *
- * Naturally, the FULL state doesn't exist for the CPU resource at the
- * system level, but exist at the cgroup level, means all non-idle tasks
- * in a cgroup are delayed on the CPU resource which used by others outside
- * of the cgroup or throttled by the cgroup cpu.max configuration.
- *
  *	SOME = nr_delayed_tasks != 0
- *	FULL = nr_delayed_tasks != 0 && nr_running_tasks == 0
+ *	FULL = nr_delayed_tasks != 0 && nr_productive_tasks == 0
+ *
+ * What it means for a task to be productive is defined differently
+ * for each resource. For IO, productive means a running task. For
+ * memory, productive means a running task that isn't a reclaimer. For
+ * CPU, productive means an oncpu task.
+ *
+ * Naturally, the FULL state doesn't exist for the CPU resource at the
+ * system level, but exist at the cgroup level. At the cgroup level,
+ * FULL means all non-idle tasks in the cgroup are delayed on the CPU
+ * resource which is being used by others outside of the cgroup or
+ * throttled by the cgroup cpu.max configuration.
  *
  * The percentage of wallclock time spent in those compound stall
  * states gives pressure numbers between 0 and 100 for each resource,
@@ -81,13 +88,13 @@
  *
  *	threads = min(nr_nonidle_tasks, nr_cpus)
  *	   SOME = min(nr_delayed_tasks / threads, 1)
- *	   FULL = (threads - min(nr_running_tasks, threads)) / threads
+ *	   FULL = (threads - min(nr_productive_tasks, threads)) / threads
  *
  * For the 257 number crunchers on 256 CPUs, this yields:
  *
  *	threads = min(257, 256)
  *	   SOME = min(1 / 256, 1)             = 0.4%
- *	   FULL = (256 - min(257, 256)) / 256 = 0%
+ *	   FULL = (256 - min(256, 256)) / 256 = 0%
  *
  * For the 1 out of 4 memory-delayed tasks, this yields:
  *
@@ -112,7 +119,7 @@
  * For each runqueue, we track:
  *
  *	   tSOME[cpu] = time(nr_delayed_tasks[cpu] != 0)
- *	   tFULL[cpu] = time(nr_delayed_tasks[cpu] && !nr_running_tasks[cpu])
+ *	   tFULL[cpu] = time(nr_delayed_tasks[cpu] && !nr_productive_tasks[cpu])
  *	tNONIDLE[cpu] = time(nr_nonidle_tasks[cpu] != 0)
  *
  * and then periodically aggregate:
@@ -233,7 +240,8 @@ static bool test_state(unsigned int *tasks, enum psi_states state)
 	case PSI_MEM_SOME:
 		return unlikely(tasks[NR_MEMSTALL]);
 	case PSI_MEM_FULL:
-		return unlikely(tasks[NR_MEMSTALL] && !tasks[NR_RUNNING]);
+		return unlikely(tasks[NR_MEMSTALL] &&
+			tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING]);
 	case PSI_CPU_SOME:
 		return unlikely(tasks[NR_RUNNING] > tasks[NR_ONCPU]);
 	case PSI_CPU_FULL:
@@ -710,10 +718,11 @@ static void psi_group_change(struct psi_group *group, int cpu,
 		if (groupc->tasks[t]) {
 			groupc->tasks[t]--;
 		} else if (!psi_bug) {
-			printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u] clear=%x set=%x\n",
+			printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u %u] clear=%x set=%x\n",
 					cpu, t, groupc->tasks[0],
 					groupc->tasks[1], groupc->tasks[2],
-					groupc->tasks[3], clear, set);
+					groupc->tasks[3], groupc->tasks[4],
+					clear, set);
 			psi_bug = 1;
 		}
 	}
@@ -833,7 +842,6 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next,
 		/*
 		 * When switching between tasks that have an identical
 		 * runtime state, the cgroup that contains both tasks
-		 * runtime state, the cgroup that contains both tasks
 		 * we reach the first common ancestor. Iterate @next's
 		 * ancestors only until we encounter @prev's ONCPU.
 		 */
@@ -854,12 +862,15 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next,
 		int clear = TSK_ONCPU, set = 0;
 
 		/*
-		 * When we're going to sleep, psi_dequeue() lets us handle
-		 * TSK_RUNNING and TSK_IOWAIT here, where we can combine it
-		 * with TSK_ONCPU and save walking common ancestors twice.
+		 * When we're going to sleep, psi_dequeue() lets us
+		 * handle TSK_RUNNING, TSK_MEMSTALL_RUNNING and
+		 * TSK_IOWAIT here, where we can combine it with
+		 * TSK_ONCPU and save walking common ancestors twice.
 		 */
 		if (sleep) {
 			clear |= TSK_RUNNING;
+			if (prev->in_memstall)
+				clear |= TSK_MEMSTALL_RUNNING;
 			if (prev->in_iowait)
 				set |= TSK_IOWAIT;
 		}
@@ -908,7 +919,7 @@ void psi_memstall_enter(unsigned long *flags)
 	rq = this_rq_lock_irq(&rf);
 
 	current->in_memstall = 1;
-	psi_task_change(current, 0, TSK_MEMSTALL);
+	psi_task_change(current, 0, TSK_MEMSTALL | TSK_MEMSTALL_RUNNING);
 
 	rq_unlock_irq(rq, &rf);
 }
@@ -937,7 +948,7 @@ void psi_memstall_leave(unsigned long *flags)
 	rq = this_rq_lock_irq(&rf);
 
 	current->in_memstall = 0;
-	psi_task_change(current, TSK_MEMSTALL, 0);
+	psi_task_change(current, TSK_MEMSTALL | TSK_MEMSTALL_RUNNING, 0);
 
 	rq_unlock_irq(rq, &rf);
 }
@@ -1071,44 +1082,6 @@ int psi_show(struct seq_file *m, struct psi_group *group, enum psi_res res)
 	return 0;
 }
 
-static int psi_io_show(struct seq_file *m, void *v)
-{
-	return psi_show(m, &psi_system, PSI_IO);
-}
-
-static int psi_memory_show(struct seq_file *m, void *v)
-{
-	return psi_show(m, &psi_system, PSI_MEM);
-}
-
-static int psi_cpu_show(struct seq_file *m, void *v)
-{
-	return psi_show(m, &psi_system, PSI_CPU);
-}
-
-static int psi_open(struct file *file, int (*psi_show)(struct seq_file *, void *))
-{
-	if (file->f_mode & FMODE_WRITE && !capable(CAP_SYS_RESOURCE))
-		return -EPERM;
-
-	return single_open(file, psi_show, NULL);
-}
-
-static int psi_io_open(struct inode *inode, struct file *file)
-{
-	return psi_open(file, psi_io_show);
-}
-
-static int psi_memory_open(struct inode *inode, struct file *file)
-{
-	return psi_open(file, psi_memory_show);
-}
-
-static int psi_cpu_open(struct inode *inode, struct file *file)
-{
-	return psi_open(file, psi_cpu_show);
-}
-
 struct psi_trigger *psi_trigger_create(struct psi_group *group,
 			char *buf, size_t nbytes, enum psi_res res)
 {
@@ -1151,7 +1124,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group,
 	t->event = 0;
 	t->last_event_time = 0;
 	init_waitqueue_head(&t->event_wait);
-	kref_init(&t->refcount);
 
 	mutex_lock(&group->trigger_lock);
 
@@ -1180,15 +1152,19 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group,
 	return t;
 }
 
-static void psi_trigger_destroy(struct kref *ref)
+void psi_trigger_destroy(struct psi_trigger *t)
 {
-	struct psi_trigger *t = container_of(ref, struct psi_trigger, refcount);
-	struct psi_group *group = t->group;
+	struct psi_group *group;
 	struct task_struct *task_to_destroy = NULL;
 
-	if (static_branch_likely(&psi_disabled))
+	/*
+	 * We do not check psi_disabled since it might have been disabled after
+	 * the trigger got created.
+	 */
+	if (!t)
 		return;
 
+	group = t->group;
 	/*
 	 * Wakeup waiters to stop polling. Can happen if cgroup is deleted
 	 * from under a polling process.
@@ -1224,9 +1200,9 @@ static void psi_trigger_destroy(struct kref *ref)
 	mutex_unlock(&group->trigger_lock);
 
 	/*
-	 * Wait for both *trigger_ptr from psi_trigger_replace and
-	 * poll_task RCUs to complete their read-side critical sections
-	 * before destroying the trigger and optionally the poll_task
+	 * Wait for psi_schedule_poll_work RCU to complete its read-side
+	 * critical section before destroying the trigger and optionally the
+	 * poll_task.
 	 */
 	synchronize_rcu();
 	/*
@@ -1243,18 +1219,6 @@ static void psi_trigger_destroy(struct kref *ref)
 	kfree(t);
 }
 
-void psi_trigger_replace(void **trigger_ptr, struct psi_trigger *new)
-{
-	struct psi_trigger *old = *trigger_ptr;
-
-	if (static_branch_likely(&psi_disabled))
-		return;
-
-	rcu_assign_pointer(*trigger_ptr, new);
-	if (old)
-		kref_put(&old->refcount, psi_trigger_destroy);
-}
-
 __poll_t psi_trigger_poll(void **trigger_ptr,
 				struct file *file, poll_table *wait)
 {
@@ -1264,27 +1228,57 @@ __poll_t psi_trigger_poll(void **trigger_ptr,
 	if (static_branch_likely(&psi_disabled))
 		return DEFAULT_POLLMASK | EPOLLERR | EPOLLPRI;
 
-	rcu_read_lock();
-
-	t = rcu_dereference(*(void __rcu __force **)trigger_ptr);
-	if (!t) {
-		rcu_read_unlock();
+	t = smp_load_acquire(trigger_ptr);
+	if (!t)
 		return DEFAULT_POLLMASK | EPOLLERR | EPOLLPRI;
-	}
-	kref_get(&t->refcount);
-
-	rcu_read_unlock();
 
 	poll_wait(file, &t->event_wait, wait);
 
 	if (cmpxchg(&t->event, 1, 0) == 1)
 		ret |= EPOLLPRI;
 
-	kref_put(&t->refcount, psi_trigger_destroy);
-
 	return ret;
 }
 
+#ifdef CONFIG_PROC_FS
+static int psi_io_show(struct seq_file *m, void *v)
+{
+	return psi_show(m, &psi_system, PSI_IO);
+}
+
+static int psi_memory_show(struct seq_file *m, void *v)
+{
+	return psi_show(m, &psi_system, PSI_MEM);
+}
+
+static int psi_cpu_show(struct seq_file *m, void *v)
+{
+	return psi_show(m, &psi_system, PSI_CPU);
+}
+
+static int psi_open(struct file *file, int (*psi_show)(struct seq_file *, void *))
+{
+	if (file->f_mode & FMODE_WRITE && !capable(CAP_SYS_RESOURCE))
+		return -EPERM;
+
+	return single_open(file, psi_show, NULL);
+}
+
+static int psi_io_open(struct inode *inode, struct file *file)
+{
+	return psi_open(file, psi_io_show);
+}
+
+static int psi_memory_open(struct inode *inode, struct file *file)
+{
+	return psi_open(file, psi_memory_show);
+}
+
+static int psi_cpu_open(struct inode *inode, struct file *file)
+{
+	return psi_open(file, psi_cpu_show);
+}
+
 static ssize_t psi_write(struct file *file, const char __user *user_buf,
 			 size_t nbytes, enum psi_res res)
 {
@@ -1305,14 +1299,24 @@ static ssize_t psi_write(struct file *file, const char __user *user_buf,
 
 	buf[buf_size - 1] = '\0';
 
-	new = psi_trigger_create(&psi_system, buf, nbytes, res);
-	if (IS_ERR(new))
-		return PTR_ERR(new);
-
 	seq = file->private_data;
+
 	/* Take seq->lock to protect seq->private from concurrent writes */
 	mutex_lock(&seq->lock);
-	psi_trigger_replace(&seq->private, new);
+
+	/* Allow only one trigger per file descriptor */
+	if (seq->private) {
+		mutex_unlock(&seq->lock);
+		return -EBUSY;
+	}
+
+	new = psi_trigger_create(&psi_system, buf, nbytes, res);
+	if (IS_ERR(new)) {
+		mutex_unlock(&seq->lock);
+		return PTR_ERR(new);
+	}
+
+	smp_store_release(&seq->private, new);
 	mutex_unlock(&seq->lock);
 
 	return nbytes;
@@ -1347,7 +1351,7 @@ static int psi_fop_release(struct inode *inode, struct file *file)
 {
 	struct seq_file *seq = file->private_data;
 
-	psi_trigger_replace(&seq->private, NULL);
+	psi_trigger_destroy(seq->private);
 	return single_release(inode, file);
 }
 
@@ -1389,3 +1393,5 @@ static int __init psi_proc_init(void)
 	return 0;
 }
 module_init(psi_proc_init);
+
+#endif /* CONFIG_PROC_FS */