[PATCH 05/19] sched: Core-wide rq->lock

[Peter Zijlstra]

Introduce the basic infrastructure to have a core wide rq->lock.

This relies on the rq->__lock order being in increasing CPU number. It
is also constrained to SMT8 per lockdep (and SMT256 per preempt_count).

Luckily SMT8 is the max supported SMT count for Linux (Mips, Sparc and
Power are known to have this).

Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
---
 kernel/Kconfig.preempt |    6 ++
 kernel/sched/core.c    |  139 +++++++++++++++++++++++++++++++++++++++++++++++++
 kernel/sched/sched.h   |   37 +++++++++++++
 3 files changed, 182 insertions(+)

--- a/kernel/Kconfig.preempt
+++ b/kernel/Kconfig.preempt
@@ -99,3 +99,9 @@ config PREEMPT_DYNAMIC
 
 	  Interesting if you want the same pre-built kernel should be used for
 	  both Server and Desktop workloads.
+
+config SCHED_CORE
+	bool "Core Scheduling for SMT"
+	default y
+	depends on SCHED_SMT
+
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -84,6 +84,103 @@ unsigned int sysctl_sched_rt_period = 10
 
 __read_mostly int scheduler_running;
 
+#ifdef CONFIG_SCHED_CORE
+
+DEFINE_STATIC_KEY_FALSE(__sched_core_enabled);
+
+/*
+ * Magic required such that:
+ *
+ *	raw_spin_rq_lock(rq);
+ *	...
+ *	raw_spin_rq_unlock(rq);
+ *
+ * ends up locking and unlocking the _same_ lock, and all CPUs
+ * always agree on what rq has what lock.
+ *
+ * XXX entirely possible to selectively enable cores, don't bother for now.
+ */
+
+static DEFINE_MUTEX(sched_core_mutex);
+static int sched_core_count;
+static struct cpumask sched_core_mask;
+
+static void __sched_core_flip(bool enabled)
+{
+	int cpu, t, i;
+
+	cpus_read_lock();
+
+	/*
+	 * Toggle the online cores, one by one.
+	 */
+	cpumask_copy(&sched_core_mask, cpu_online_mask);
+	for_each_cpu(cpu, &sched_core_mask) {
+		const struct cpumask *smt_mask = cpu_smt_mask(cpu);
+
+		i = 0;
+		local_irq_disable();
+		for_each_cpu(t, smt_mask) {
+			/* supports up to SMT8 */
+			raw_spin_lock_nested(&cpu_rq(t)->__lock, i++);
+		}
+
+		for_each_cpu(t, smt_mask)
+			cpu_rq(t)->core_enabled = enabled;
+
+		for_each_cpu(t, smt_mask)
+			raw_spin_unlock(&cpu_rq(t)->__lock);
+		local_irq_enable();
+
+		cpumask_andnot(&sched_core_mask, &sched_core_mask, smt_mask);
+	}
+
+	/*
+	 * Toggle the offline CPUs.
+	 */
+	cpumask_copy(&sched_core_mask, cpu_possible_mask);
+	cpumask_andnot(&sched_core_mask, &sched_core_mask, cpu_online_mask);
+
+	for_each_cpu(cpu, &sched_core_mask)
+		cpu_rq(cpu)->core_enabled = enabled;
+
+	cpus_read_unlock();
+}
+
+static void __sched_core_enable(void)
+{
+	// XXX verify there are no cookie tasks (yet)
+
+	static_branch_enable(&__sched_core_enabled);
+	__sched_core_flip(true);
+}
+
+static void __sched_core_disable(void)
+{
+	// XXX verify there are no cookie tasks (left)
+
+	__sched_core_flip(false);
+	static_branch_disable(&__sched_core_enabled);
+}
+
+void sched_core_get(void)
+{
+	mutex_lock(&sched_core_mutex);
+	if (!sched_core_count++)
+		__sched_core_enable();
+	mutex_unlock(&sched_core_mutex);
+}
+
+void sched_core_put(void)
+{
+	mutex_lock(&sched_core_mutex);
+	if (!--sched_core_count)
+		__sched_core_disable();
+	mutex_unlock(&sched_core_mutex);
+}
+
+#endif /* CONFIG_SCHED_CORE */
+
 /*
  * part of the period that we allow rt tasks to run in us.
  * default: 0.95s
@@ -5042,6 +5139,40 @@ pick_next_task(struct rq *rq, struct tas
 	BUG();
 }
 
+#ifdef CONFIG_SCHED_CORE
+
+static inline void sched_core_cpu_starting(unsigned int cpu)
+{
+	const struct cpumask *smt_mask = cpu_smt_mask(cpu);
+	struct rq *rq, *core_rq = NULL;
+	int i;
+
+	core_rq = cpu_rq(cpu)->core;
+
+	if (!core_rq) {
+		for_each_cpu(i, smt_mask) {
+			rq = cpu_rq(i);
+			if (rq->core && rq->core == rq)
+				core_rq = rq;
+		}
+
+		if (!core_rq)
+			core_rq = cpu_rq(cpu);
+
+		for_each_cpu(i, smt_mask) {
+			rq = cpu_rq(i);
+
+			WARN_ON_ONCE(rq->core && rq->core != core_rq);
+			rq->core = core_rq;
+		}
+	}
+}
+#else /* !CONFIG_SCHED_CORE */
+
+static inline void sched_core_cpu_starting(unsigned int cpu) {}
+
+#endif /* CONFIG_SCHED_CORE */
+
 /*
  * __schedule() is the main scheduler function.
  *
@@ -8006,6 +8137,7 @@ static void sched_rq_cpu_starting(unsign
 
 int sched_cpu_starting(unsigned int cpu)
 {
+	sched_core_cpu_starting(cpu);
 	sched_rq_cpu_starting(cpu);
 	sched_tick_start(cpu);
 	return 0;
@@ -8290,6 +8424,11 @@ void __init sched_init(void)
 #endif /* CONFIG_SMP */
 		hrtick_rq_init(rq);
 		atomic_set(&rq->nr_iowait, 0);
+
+#ifdef CONFIG_SCHED_CORE
+		rq->core = NULL;
+		rq->core_enabled = 0;
+#endif
 	}
 
 	set_load_weight(&init_task, false);
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1075,6 +1075,12 @@ struct rq {
 #endif
 	unsigned int		push_busy;
 	struct cpu_stop_work	push_work;
+
+#ifdef CONFIG_SCHED_CORE
+	/* per rq */
+	struct rq		*core;
+	unsigned int		core_enabled;
+#endif
 };
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
@@ -1113,6 +1119,35 @@ static inline bool is_migration_disabled
 #endif
 }
 
+#ifdef CONFIG_SCHED_CORE
+
+DECLARE_STATIC_KEY_FALSE(__sched_core_enabled);
+
+static inline bool sched_core_enabled(struct rq *rq)
+{
+	return static_branch_unlikely(&__sched_core_enabled) && rq->core_enabled;
+}
+
+static inline bool sched_core_disabled(void)
+{
+	return !static_branch_unlikely(&__sched_core_enabled);
+}
+
+static inline raw_spinlock_t *rq_lockp(struct rq *rq)
+{
+	if (sched_core_enabled(rq))
+		return &rq->core->__lock;
+
+	return &rq->__lock;
+}
+
+#else /* !CONFIG_SCHED_CORE */
+
+static inline bool sched_core_enabled(struct rq *rq)
+{
+	return false;
+}
+
 static inline bool sched_core_disabled(void)
 {
 	return true;
@@ -1123,6 +1158,8 @@ static inline raw_spinlock_t *rq_lockp(s
 	return &rq->__lock;
 }
 
+#endif /* CONFIG_SCHED_CORE */
+
 static inline void lockdep_assert_rq_held(struct rq *rq)
 {
 	lockdep_assert_held(rq_lockp(rq));