sched: rt-group: make rt groups scheduling configurable

Make the rt group scheduler compile time configurable.
Keep it experimental for now.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

include/linux/cgroup_subsys.h

@@ -25,7 +25,7 @@ SUBSYS(ns)
 
 /* */
 
-#ifdef CONFIG_FAIR_CGROUP_SCHED
+#ifdef CONFIG_CGROUP_SCHED
 SUBSYS(cpu_cgroup)
 #endif
 

include/linux/sched.h

@@ -590,7 +590,7 @@ struct user_struct {
 	struct hlist_node uidhash_node;
 	uid_t uid;
 
-#ifdef CONFIG_FAIR_USER_SCHED
+#ifdef CONFIG_USER_SCHED
 	struct task_group *tg;
 #ifdef CONFIG_SYSFS
 	struct kobject kobj;
@@ -973,7 +973,7 @@ struct sched_rt_entity {
 	unsigned long timeout;
 	int nr_cpus_allowed;
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_RT_GROUP_SCHED
 	struct sched_rt_entity	*parent;
 	/* rq on which this entity is (to be) queued: */
 	struct rt_rq		*rt_rq;
@@ -2027,19 +2027,22 @@ extern int sched_mc_power_savings, sched_smt_power_savings;
 
 extern void normalize_rt_tasks(void);
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_GROUP_SCHED
 
 extern struct task_group init_task_group;
 
 extern struct task_group *sched_create_group(void);
 extern void sched_destroy_group(struct task_group *tg);
 extern void sched_move_task(struct task_struct *tsk);
+#ifdef CONFIG_FAIR_GROUP_SCHED
 extern int sched_group_set_shares(struct task_group *tg, unsigned long shares);
 extern unsigned long sched_group_shares(struct task_group *tg);
+#endif
+#ifdef CONFIG_RT_GROUP_SCHED
 extern int sched_group_set_rt_runtime(struct task_group *tg,
 				      long rt_runtime_us);
 extern long sched_group_rt_runtime(struct task_group *tg);
-
+#endif
 #endif
 
 #ifdef CONFIG_TASK_XACCT

init/Kconfig

@@ -311,25 +311,36 @@ config CPUSETS
 
 	  Say N if unsure.
 
-config FAIR_GROUP_SCHED
-	bool "Fair group CPU scheduler"
+config GROUP_SCHED
+	bool "Group CPU scheduler"
 	default y
 	help
 	  This feature lets CPU scheduler recognize task groups and control CPU
 	  bandwidth allocation to such task groups.
 
+config FAIR_GROUP_SCHED
+	bool "Group scheduling for SCHED_OTHER"
+	depends on GROUP_SCHED
+	default y
+
+config RT_GROUP_SCHED
+	bool "Group scheduling for SCHED_RR/FIFO"
+	depends on EXPERIMENTAL
+	depends on GROUP_SCHED
+	default n
+
 choice
-	depends on FAIR_GROUP_SCHED
+	depends on GROUP_SCHED
 	prompt "Basis for grouping tasks"
-	default FAIR_USER_SCHED
+	default USER_SCHED
 
-config FAIR_USER_SCHED
+config USER_SCHED
 	bool "user id"
 	help
 	  This option will choose userid as the basis for grouping
 	  tasks, thus providing equal CPU bandwidth to each user.
 
-config FAIR_CGROUP_SCHED
+config CGROUP_SCHED
 	bool "Control groups"
  	depends on CGROUPS
  	help

kernel/sched_rt.c

@@ -55,7 +55,7 @@ static inline int on_rt_rq(struct sched_rt_entity *rt_se)
 	return !list_empty(&rt_se->run_list);
 }
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_RT_GROUP_SCHED
 
 static inline u64 sched_rt_runtime(struct rt_rq *rt_rq)
 {
@@ -177,7 +177,7 @@ static inline int rt_rq_throttled(struct rt_rq *rt_rq)
 
 static inline int rt_se_prio(struct sched_rt_entity *rt_se)
 {
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_RT_GROUP_SCHED
 	struct rt_rq *rt_rq = group_rt_rq(rt_se);
 
 	if (rt_rq)
@@ -269,7 +269,7 @@ void inc_rt_tasks(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 {
 	WARN_ON(!rt_prio(rt_se_prio(rt_se)));
 	rt_rq->rt_nr_running++;
-#if defined CONFIG_SMP || defined CONFIG_FAIR_GROUP_SCHED
+#if defined CONFIG_SMP || defined CONFIG_RT_GROUP_SCHED
 	if (rt_se_prio(rt_se) < rt_rq->highest_prio)
 		rt_rq->highest_prio = rt_se_prio(rt_se);
 #endif
@@ -281,7 +281,7 @@ void inc_rt_tasks(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 
 	update_rt_migration(rq_of_rt_rq(rt_rq));
 #endif
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_RT_GROUP_SCHED
 	if (rt_se_boosted(rt_se))
 		rt_rq->rt_nr_boosted++;
 #endif
@@ -293,7 +293,7 @@ void dec_rt_tasks(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 	WARN_ON(!rt_prio(rt_se_prio(rt_se)));
 	WARN_ON(!rt_rq->rt_nr_running);
 	rt_rq->rt_nr_running--;
-#if defined CONFIG_SMP || defined CONFIG_FAIR_GROUP_SCHED
+#if defined CONFIG_SMP || defined CONFIG_RT_GROUP_SCHED
 	if (rt_rq->rt_nr_running) {
 		struct rt_prio_array *array;
 
@@ -315,7 +315,7 @@ void dec_rt_tasks(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
 
 	update_rt_migration(rq_of_rt_rq(rt_rq));
 #endif /* CONFIG_SMP */
-#ifdef CONFIG_FAIR_GROUP_SCHED
+#ifdef CONFIG_RT_GROUP_SCHED
 	if (rt_se_boosted(rt_se))
 		rt_rq->rt_nr_boosted--;
 

kernel/user.c

@@ -57,7 +57,7 @@ struct user_struct root_user = {
 	.uid_keyring	= &root_user_keyring,
 	.session_keyring = &root_session_keyring,
 #endif
-#ifdef CONFIG_FAIR_USER_SCHED
+#ifdef CONFIG_USER_SCHED
 	.tg		= &init_task_group,
 #endif
 };
@@ -90,7 +90,7 @@ static struct user_struct *uid_hash_find(uid_t uid, struct hlist_head *hashent)
 	return NULL;
 }
 
-#ifdef CONFIG_FAIR_USER_SCHED
+#ifdef CONFIG_USER_SCHED
 
 static void sched_destroy_user(struct user_struct *up)
 {
@@ -113,15 +113,15 @@ static void sched_switch_user(struct task_struct *p)
 	sched_move_task(p);
 }
 
-#else	/* CONFIG_FAIR_USER_SCHED */
+#else	/* CONFIG_USER_SCHED */
 
 static void sched_destroy_user(struct user_struct *up) { }
 static int sched_create_user(struct user_struct *up) { return 0; }
 static void sched_switch_user(struct task_struct *p) { }
 
-#endif	/* CONFIG_FAIR_USER_SCHED */
+#endif	/* CONFIG_USER_SCHED */
 
-#if defined(CONFIG_FAIR_USER_SCHED) && defined(CONFIG_SYSFS)
+#if defined(CONFIG_USER_SCHED) && defined(CONFIG_SYSFS)
 
 static struct kset *uids_kset; /* represents the /sys/kernel/uids/ directory */
 static DEFINE_MUTEX(uids_mutex);
@@ -137,6 +137,7 @@ static inline void uids_mutex_unlock(void)
 }
 
 /* uid directory attributes */
+#ifdef CONFIG_FAIR_GROUP_SCHED
 static ssize_t cpu_shares_show(struct kobject *kobj,
 			       struct kobj_attribute *attr,
 			       char *buf)
@@ -163,7 +164,9 @@ static ssize_t cpu_shares_store(struct kobject *kobj,
 
 static struct kobj_attribute cpu_share_attr =
 	__ATTR(cpu_share, 0644, cpu_shares_show, cpu_shares_store);
+#endif
 
+#ifdef CONFIG_RT_GROUP_SCHED
 static ssize_t cpu_rt_runtime_show(struct kobject *kobj,
 				   struct kobj_attribute *attr,
 				   char *buf)
@@ -190,11 +193,16 @@ static ssize_t cpu_rt_runtime_store(struct kobject *kobj,
 
 static struct kobj_attribute cpu_rt_runtime_attr =
 	__ATTR(cpu_rt_runtime, 0644, cpu_rt_runtime_show, cpu_rt_runtime_store);
+#endif
 
 /* default attributes per uid directory */
 static struct attribute *uids_attributes[] = {
+#ifdef CONFIG_FAIR_GROUP_SCHED
 	&cpu_share_attr.attr,
+#endif
+#ifdef CONFIG_RT_GROUP_SCHED
 	&cpu_rt_runtime_attr.attr,
+#endif
 	NULL
 };
 
@@ -297,7 +305,7 @@ static inline void free_user(struct user_struct *up, unsigned long flags)
 	schedule_work(&up->work);
 }
 
-#else	/* CONFIG_FAIR_USER_SCHED && CONFIG_SYSFS */
+#else	/* CONFIG_USER_SCHED && CONFIG_SYSFS */
 
 int uids_sysfs_init(void) { return 0; }
 static inline int uids_user_create(struct user_struct *up) { return 0; }
@@ -401,7 +409,7 @@ struct user_struct * alloc_uid(struct user_namespace *ns, uid_t uid)
 		spin_lock_irq(&uidhash_lock);
 		up = uid_hash_find(uid, hashent);
 		if (up) {
-			/* This case is not possible when CONFIG_FAIR_USER_SCHED
+			/* This case is not possible when CONFIG_USER_SCHED
 			 * is defined, since we serialize alloc_uid() using
 			 * uids_mutex. Hence no need to call
 			 * sched_destroy_user() or remove_user_sysfs_dir().