1 files changed, 203 insertions, 126 deletions

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 373ff5f55884..a80a73909dc2 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -1064,6 +1064,23 @@ update_stats_curr_start(struct cfs_rq *cfs_rq, struct sched_entity *se)
  * Scheduling class queueing methods:
  */
 
+static inline bool is_core_idle(int cpu)
+{
+#ifdef CONFIG_SCHED_SMT
+	int sibling;
+
+	for_each_cpu(sibling, cpu_smt_mask(cpu)) {
+		if (cpu == sibling)
+			continue;
+
+		if (!idle_cpu(sibling))
+			return false;
+	}
+#endif
+
+	return true;
+}
+
 #ifdef CONFIG_NUMA
 #define NUMA_IMBALANCE_MIN 2
 
@@ -1700,23 +1717,6 @@ struct numa_stats {
 	int idle_cpu;
 };
 
-static inline bool is_core_idle(int cpu)
-{
-#ifdef CONFIG_SCHED_SMT
-	int sibling;
-
-	for_each_cpu(sibling, cpu_smt_mask(cpu)) {
-		if (cpu == sibling)
-			continue;
-
-		if (!idle_cpu(sibling))
-			return false;
-	}
-#endif
-
-	return true;
-}
-
 struct task_numa_env {
 	struct task_struct *p;
 
@@ -5577,6 +5577,14 @@ static void __cfsb_csd_unthrottle(void *arg)
 	rq_lock(rq, &rf);
 
 	/*
+	 * Iterating over the list can trigger several call to
+	 * update_rq_clock() in unthrottle_cfs_rq().
+	 * Do it once and skip the potential next ones.
+	 */
+	update_rq_clock(rq);
+	rq_clock_start_loop_update(rq);
+
+	/*
 	 * Since we hold rq lock we're safe from concurrent manipulation of
 	 * the CSD list. However, this RCU critical section annotates the
 	 * fact that we pair with sched_free_group_rcu(), so that we cannot
@@ -5595,6 +5603,7 @@ static void __cfsb_csd_unthrottle(void *arg)
 
 	rcu_read_unlock();
 
+	rq_clock_stop_loop_update(rq);
 	rq_unlock(rq, &rf);
 }
 
@@ -6115,6 +6124,13 @@ static void __maybe_unused unthrottle_offline_cfs_rqs(struct rq *rq)
 
 	lockdep_assert_rq_held(rq);
 
+	/*
+	 * The rq clock has already been updated in the
+	 * set_rq_offline(), so we should skip updating
+	 * the rq clock again in unthrottle_cfs_rq().
+	 */
+	rq_clock_start_loop_update(rq);
+
 	rcu_read_lock();
 	list_for_each_entry_rcu(tg, &task_groups, list) {
 		struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
@@ -6137,6 +6153,8 @@ static void __maybe_unused unthrottle_offline_cfs_rqs(struct rq *rq)
 			unthrottle_cfs_rq(cfs_rq);
 	}
 	rcu_read_unlock();
+
+	rq_clock_stop_loop_update(rq);
 }
 
 #else /* CONFIG_CFS_BANDWIDTH */
@@ -7202,14 +7220,58 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target)
 	return target;
 }
 
-/*
- * Predicts what cpu_util(@cpu) would return if @p was removed from @cpu
- * (@dst_cpu = -1) or migrated to @dst_cpu.
- */
-static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu)
+/**
+ * cpu_util() - Estimates the amount of CPU capacity used by CFS tasks.
+ * @cpu: the CPU to get the utilization for
+ * @p: task for which the CPU utilization should be predicted or NULL
+ * @dst_cpu: CPU @p migrates to, -1 if @p moves from @cpu or @p == NULL
+ * @boost: 1 to enable boosting, otherwise 0
+ *
+ * The unit of the return value must be the same as the one of CPU capacity
+ * so that CPU utilization can be compared with CPU capacity.
+ *
+ * CPU utilization is the sum of running time of runnable tasks plus the
+ * recent utilization of currently non-runnable tasks on that CPU.
+ * It represents the amount of CPU capacity currently used by CFS tasks in
+ * the range [0..max CPU capacity] with max CPU capacity being the CPU
+ * capacity at f_max.
+ *
+ * The estimated CPU utilization is defined as the maximum between CPU
+ * utilization and sum of the estimated utilization of the currently
+ * runnable tasks on that CPU. It preserves a utilization "snapshot" of
+ * previously-executed tasks, which helps better deduce how busy a CPU will
+ * be when a long-sleeping task wakes up. The contribution to CPU utilization
+ * of such a task would be significantly decayed at this point of time.
+ *
+ * Boosted CPU utilization is defined as max(CPU runnable, CPU utilization).
+ * CPU contention for CFS tasks can be detected by CPU runnable > CPU
+ * utilization. Boosting is implemented in cpu_util() so that internal
+ * users (e.g. EAS) can use it next to external users (e.g. schedutil),
+ * latter via cpu_util_cfs_boost().
+ *
+ * CPU utilization can be higher than the current CPU capacity
+ * (f_curr/f_max * max CPU capacity) or even the max CPU capacity because
+ * of rounding errors as well as task migrations or wakeups of new tasks.
+ * CPU utilization has to be capped to fit into the [0..max CPU capacity]
+ * range. Otherwise a group of CPUs (CPU0 util = 121% + CPU1 util = 80%)
+ * could be seen as over-utilized even though CPU1 has 20% of spare CPU
+ * capacity. CPU utilization is allowed to overshoot current CPU capacity
+ * though since this is useful for predicting the CPU capacity required
+ * after task migrations (scheduler-driven DVFS).
+ *
+ * Return: (Boosted) (estimated) utilization for the specified CPU.
+ */
+static unsigned long
+cpu_util(int cpu, struct task_struct *p, int dst_cpu, int boost)
 {
 	struct cfs_rq *cfs_rq = &cpu_rq(cpu)->cfs;
 	unsigned long util = READ_ONCE(cfs_rq->avg.util_avg);
+	unsigned long runnable;
+
+	if (boost) {
+		runnable = READ_ONCE(cfs_rq->avg.runnable_avg);
+		util = max(util, runnable);
+	}
 
 	/*
 	 * If @dst_cpu is -1 or @p migrates from @cpu to @dst_cpu remove its
@@ -7217,9 +7279,9 @@ static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu)
 	 * contribution. In all the other cases @cpu is not impacted by the
 	 * migration so its util_avg is already correct.
 	 */
-	if (task_cpu(p) == cpu && dst_cpu != cpu)
+	if (p && task_cpu(p) == cpu && dst_cpu != cpu)
 		lsub_positive(&util, task_util(p));
-	else if (task_cpu(p) != cpu && dst_cpu == cpu)
+	else if (p && task_cpu(p) != cpu && dst_cpu == cpu)
 		util += task_util(p);
 
 	if (sched_feat(UTIL_EST)) {
@@ -7227,6 +7289,9 @@ static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu)
 
 		util_est = READ_ONCE(cfs_rq->avg.util_est.enqueued);
 
+		if (boost)
+			util_est = max(util_est, runnable);
+
 		/*
 		 * During wake-up @p isn't enqueued yet and doesn't contribute
 		 * to any cpu_rq(cpu)->cfs.avg.util_est.enqueued.
@@ -7255,7 +7320,7 @@ static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu)
 		 */
 		if (dst_cpu == cpu)
 			util_est += _task_util_est(p);
-		else if (unlikely(task_on_rq_queued(p) || current == p))
+		else if (p && unlikely(task_on_rq_queued(p) || current == p))
 			lsub_positive(&util_est, _task_util_est(p));
 
 		util = max(util, util_est);
@@ -7264,6 +7329,16 @@ static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu)
 	return min(util, capacity_orig_of(cpu));
 }
 
+unsigned long cpu_util_cfs(int cpu)
+{
+	return cpu_util(cpu, NULL, -1, 0);
+}
+
+unsigned long cpu_util_cfs_boost(int cpu)
+{
+	return cpu_util(cpu, NULL, -1, 1);
+}
+
 /*
  * cpu_util_without: compute cpu utilization without any contributions from *p
  * @cpu: the CPU which utilization is requested
@@ -7281,9 +7356,9 @@ static unsigned long cpu_util_without(int cpu, struct task_struct *p)
 {
 	/* Task has no contribution or is new */
 	if (cpu != task_cpu(p) || !READ_ONCE(p->se.avg.last_update_time))
-		return cpu_util_cfs(cpu);
+		p = NULL;
 
-	return cpu_util_next(cpu, p, -1);
+	return cpu_util(cpu, p, -1, 0);
 }
 
 /*
@@ -7330,7 +7405,7 @@ static inline void eenv_task_busy_time(struct energy_env *eenv,
  * cpu_capacity.
  *
  * The contribution of the task @p for which we want to estimate the
- * energy cost is removed (by cpu_util_next()) and must be calculated
+ * energy cost is removed (by cpu_util()) and must be calculated
  * separately (see eenv_task_busy_time). This ensures:
  *
  *   - A stable PD utilization, no matter which CPU of that PD we want to place
@@ -7351,7 +7426,7 @@ static inline void eenv_pd_busy_time(struct energy_env *eenv,
 	int cpu;
 
 	for_each_cpu(cpu, pd_cpus) {
-		unsigned long util = cpu_util_next(cpu, p, -1);
+		unsigned long util = cpu_util(cpu, p, -1, 0);
 
 		busy_time += effective_cpu_util(cpu, util, ENERGY_UTIL, NULL);
 	}
@@ -7375,8 +7450,8 @@ eenv_pd_max_util(struct energy_env *eenv, struct cpumask *pd_cpus,
 
 	for_each_cpu(cpu, pd_cpus) {
 		struct task_struct *tsk = (cpu == dst_cpu) ? p : NULL;
-		unsigned long util = cpu_util_next(cpu, p, dst_cpu);
-		unsigned long cpu_util;
+		unsigned long util = cpu_util(cpu, p, dst_cpu, 1);
+		unsigned long eff_util;
 
 		/*
 		 * Performance domain frequency: utilization clamping
@@ -7385,8 +7460,8 @@ eenv_pd_max_util(struct energy_env *eenv, struct cpumask *pd_cpus,
 		 * NOTE: in case RT tasks are running, by default the
 		 * FREQUENCY_UTIL's utilization can be max OPP.
 		 */
-		cpu_util = effective_cpu_util(cpu, util, FREQUENCY_UTIL, tsk);
-		max_util = max(max_util, cpu_util);
+		eff_util = effective_cpu_util(cpu, util, FREQUENCY_UTIL, tsk);
+		max_util = max(max_util, eff_util);
 	}
 
 	return min(max_util, eenv->cpu_cap);
@@ -7521,7 +7596,7 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu)
 			if (!cpumask_test_cpu(cpu, p->cpus_ptr))
 				continue;
 
-			util = cpu_util_next(cpu, p, cpu);
+			util = cpu_util(cpu, p, cpu, 0);
 			cpu_cap = capacity_of(cpu);
 
 			/*
@@ -9331,96 +9406,61 @@ group_type group_classify(unsigned int imbalance_pct,
 }
 
 /**
- * asym_smt_can_pull_tasks - Check whether the load balancing CPU can pull tasks
- * @dst_cpu:	Destination CPU of the load balancing
+ * sched_use_asym_prio - Check whether asym_packing priority must be used
+ * @sd:		The scheduling domain of the load balancing
+ * @cpu:	A CPU
+ *
+ * Always use CPU priority when balancing load between SMT siblings. When
+ * balancing load between cores, it is not sufficient that @cpu is idle. Only
+ * use CPU priority if the whole core is idle.
+ *
+ * Returns: True if the priority of @cpu must be followed. False otherwise.
+ */
+static bool sched_use_asym_prio(struct sched_domain *sd, int cpu)
+{
+	if (!sched_smt_active())
+		return true;
+
+	return sd->flags & SD_SHARE_CPUCAPACITY || is_core_idle(cpu);
+}
+
+/**
+ * sched_asym - Check if the destination CPU can do asym_packing load balance
+ * @env:	The load balancing environment
  * @sds:	Load-balancing data with statistics of the local group
  * @sgs:	Load-balancing statistics of the candidate busiest group
- * @sg:		The candidate busiest group
+ * @group:	The candidate busiest group
  *
- * Check the state of the SMT siblings of both @sds::local and @sg and decide
- * if @dst_cpu can pull tasks.
+ * @env::dst_cpu can do asym_packing if it has higher priority than the
+ * preferred CPU of @group.
  *
- * If @dst_cpu does not have SMT siblings, it can pull tasks if two or more of
- * the SMT siblings of @sg are busy. If only one CPU in @sg is busy, pull tasks
- * only if @dst_cpu has higher priority.
+ * SMT is a special case. If we are balancing load between cores, @env::dst_cpu
+ * can do asym_packing balance only if all its SMT siblings are idle. Also, it
+ * can only do it if @group is an SMT group and has exactly on busy CPU. Larger
+ * imbalances in the number of CPUS are dealt with in find_busiest_group().
  *
- * If both @dst_cpu and @sg have SMT siblings, and @sg has exactly one more
- * busy CPU than @sds::local, let @dst_cpu pull tasks if it has higher priority.
- * Bigger imbalances in the number of busy CPUs will be dealt with in
- * update_sd_pick_busiest().
+ * If we are balancing load within an SMT core, or at DIE domain level, always
+ * proceed.
  *
- * If @sg does not have SMT siblings, only pull tasks if all of the SMT siblings
- * of @dst_cpu are idle and @sg has lower priority.
- *
- * Return: true if @dst_cpu can pull tasks, false otherwise.
+ * Return: true if @env::dst_cpu can do with asym_packing load balance. False
+ * otherwise.
  */
-static bool asym_smt_can_pull_tasks(int dst_cpu, struct sd_lb_stats *sds,
-				    struct sg_lb_stats *sgs,
-				    struct sched_group *sg)
+static inline bool
+sched_asym(struct lb_env *env, struct sd_lb_stats *sds,  struct sg_lb_stats *sgs,
+	   struct sched_group *group)
 {
-#ifdef CONFIG_SCHED_SMT
-	bool local_is_smt, sg_is_smt;
-	int sg_busy_cpus;
-
-	local_is_smt = sds->local->flags & SD_SHARE_CPUCAPACITY;
-	sg_is_smt = sg->flags & SD_SHARE_CPUCAPACITY;
-
-	sg_busy_cpus = sgs->group_weight - sgs->idle_cpus;
-
-	if (!local_is_smt) {
-		/*
-		 * If we are here, @dst_cpu is idle and does not have SMT
-		 * siblings. Pull tasks if candidate group has two or more
-		 * busy CPUs.
-		 */
-		if (sg_busy_cpus >= 2) /* implies sg_is_smt */
-			return true;
-
-		/*
-		 * @dst_cpu does not have SMT siblings. @sg may have SMT
-		 * siblings and only one is busy. In such case, @dst_cpu
-		 * can help if it has higher priority and is idle (i.e.,
-		 * it has no running tasks).
-		 */
-		return sched_asym_prefer(dst_cpu, sg->asym_prefer_cpu);
-	}
-
-	/* @dst_cpu has SMT siblings. */
-
-	if (sg_is_smt) {
-		int local_busy_cpus = sds->local->group_weight -
-				      sds->local_stat.idle_cpus;
-		int busy_cpus_delta = sg_busy_cpus - local_busy_cpus;
-
-		if (busy_cpus_delta == 1)
-			return sched_asym_prefer(dst_cpu, sg->asym_prefer_cpu);
-
+	/* Ensure that the whole local core is idle, if applicable. */
+	if (!sched_use_asym_prio(env->sd, env->dst_cpu))
 		return false;
-	}
 
 	/*
-	 * @sg does not have SMT siblings. Ensure that @sds::local does not end
-	 * up with more than one busy SMT sibling and only pull tasks if there
-	 * are not busy CPUs (i.e., no CPU has running tasks).
+	 * CPU priorities does not make sense for SMT cores with more than one
+	 * busy sibling.
 	 */
-	if (!sds->local_stat.sum_nr_running)
-		return sched_asym_prefer(dst_cpu, sg->asym_prefer_cpu);
-
-	return false;
-#else
-	/* Always return false so that callers deal with non-SMT cases. */
-	return false;
-#endif
-}
-
-static inline bool
-sched_asym(struct lb_env *env, struct sd_lb_stats *sds,  struct sg_lb_stats *sgs,
-	   struct sched_group *group)
-{
-	/* Only do SMT checks if either local or candidate have SMT siblings */
-	if ((sds->local->flags & SD_SHARE_CPUCAPACITY) ||
-	    (group->flags & SD_SHARE_CPUCAPACITY))
-		return asym_smt_can_pull_tasks(env->dst_cpu, sds, sgs, group);
+	if (group->flags & SD_SHARE_CPUCAPACITY) {
+		if (sgs->group_weight - sgs->idle_cpus != 1)
+			return false;
+	}
 
 	return sched_asym_prefer(env->dst_cpu, group->asym_prefer_cpu);
 }
@@ -9610,10 +9650,22 @@ static bool update_sd_pick_busiest(struct lb_env *env,
 		 * contention when accessing shared HW resources.
 		 *
 		 * XXX for now avg_load is not computed and always 0 so we
-		 * select the 1st one.
+		 * select the 1st one, except if @sg is composed of SMT
+		 * siblings.
 		 */
-		if (sgs->avg_load <= busiest->avg_load)
+
+		if (sgs->avg_load < busiest->avg_load)
 			return false;
+
+		if (sgs->avg_load == busiest->avg_load) {
+			/*
+			 * SMT sched groups need more help than non-SMT groups.
+			 * If @sg happens to also be SMT, either choice is good.
+			 */
+			if (sds->busiest->flags & SD_SHARE_CPUCAPACITY)
+				return false;
+		}
+
 		break;
 
 	case group_has_spare:
@@ -10088,7 +10140,6 @@ static void update_idle_cpu_scan(struct lb_env *env,
 
 static inline void update_sd_lb_stats(struct lb_env *env, struct sd_lb_stats *sds)
 {
-	struct sched_domain *child = env->sd->child;
 	struct sched_group *sg = env->sd->groups;
 	struct sg_lb_stats *local = &sds->local_stat;
 	struct sg_lb_stats tmp_sgs;
@@ -10129,8 +10180,13 @@ next_group:
 		sg = sg->next;
 	} while (sg != env->sd->groups);
 
-	/* Tag domain that child domain prefers tasks go to siblings first */
-	sds->prefer_sibling = child && child->flags & SD_PREFER_SIBLING;
+	/*
+	 * Indicate that the child domain of the busiest group prefers tasks
+	 * go to a child's sibling domains first. NB the flags of a sched group
+	 * are those of the child domain.
+	 */
+	if (sds->busiest)
+		sds->prefer_sibling = !!(sds->busiest->flags & SD_PREFER_SIBLING);
 
 
 	if (env->sd->flags & SD_NUMA)
@@ -10440,7 +10496,10 @@ static struct sched_group *find_busiest_group(struct lb_env *env)
 			goto out_balanced;
 	}
 
-	/* Try to move all excess tasks to child's sibling domain */
+	/*
+	 * Try to move all excess tasks to a sibling domain of the busiest
+	 * group's child domain.
+	 */
 	if (sds.prefer_sibling && local->group_type == group_has_spare &&
 	    busiest->sum_nr_running > local->sum_nr_running + 1)
 		goto force_balance;
@@ -10542,8 +10601,15 @@ static struct rq *find_busiest_queue(struct lb_env *env,
 		    nr_running == 1)
 			continue;
 
-		/* Make sure we only pull tasks from a CPU of lower priority */
+		/*
+		 * Make sure we only pull tasks from a CPU of lower priority
+		 * when balancing between SMT siblings.
+		 *
+		 * If balancing between cores, let lower priority CPUs help
+		 * SMT cores with more than one busy sibling.
+		 */
 		if ((env->sd->flags & SD_ASYM_PACKING) &&
+		    sched_use_asym_prio(env->sd, i) &&
 		    sched_asym_prefer(i, env->dst_cpu) &&
 		    nr_running == 1)
 			continue;
@@ -10581,7 +10647,7 @@ static struct rq *find_busiest_queue(struct lb_env *env,
 			break;
 
 		case migrate_util:
-			util = cpu_util_cfs(i);
+			util = cpu_util_cfs_boost(i);
 
 			/*
 			 * Don't try to pull utilization from a CPU with one
@@ -10632,12 +10698,19 @@ static inline bool
 asym_active_balance(struct lb_env *env)
 {
 	/*
-	 * ASYM_PACKING needs to force migrate tasks from busy but
-	 * lower priority CPUs in order to pack all tasks in the
-	 * highest priority CPUs.
+	 * ASYM_PACKING needs to force migrate tasks from busy but lower
+	 * priority CPUs in order to pack all tasks in the highest priority
+	 * CPUs. When done between cores, do it only if the whole core if the
+	 * whole core is idle.
+	 *
+	 * If @env::src_cpu is an SMT core with busy siblings, let
+	 * the lower priority @env::dst_cpu help it. Do not follow
+	 * CPU priority.
 	 */
 	return env->idle != CPU_NOT_IDLE && (env->sd->flags & SD_ASYM_PACKING) &&
-	       sched_asym_prefer(env->dst_cpu, env->src_cpu);
+	       sched_use_asym_prio(env->sd, env->dst_cpu) &&
+	       (sched_asym_prefer(env->dst_cpu, env->src_cpu) ||
+		!sched_use_asym_prio(env->sd, env->src_cpu));
 }
 
 static inline bool
@@ -10744,7 +10817,7 @@ static int load_balance(int this_cpu, struct rq *this_rq,
 		.sd		= sd,
 		.dst_cpu	= this_cpu,
 		.dst_rq		= this_rq,
-		.dst_grpmask    = sched_group_span(sd->groups),
+		.dst_grpmask    = group_balance_mask(sd->groups),
 		.idle		= idle,
 		.loop_break	= SCHED_NR_MIGRATE_BREAK,
 		.cpus		= cpus,
@@ -11371,9 +11444,13 @@ static void nohz_balancer_kick(struct rq *rq)
 		 * When ASYM_PACKING; see if there's a more preferred CPU
 		 * currently idle; in which case, kick the ILB to move tasks
 		 * around.
+		 *
+		 * When balancing betwen cores, all the SMT siblings of the
+		 * preferred CPU must be idle.
 		 */
 		for_each_cpu_and(i, sched_domain_span(sd), nohz.idle_cpus_mask) {
-			if (sched_asym_prefer(i, cpu)) {
+			if (sched_use_asym_prio(sd, i) &&
+			    sched_asym_prefer(i, cpu)) {
 				flags = NOHZ_STATS_KICK | NOHZ_BALANCE_KICK;
 				goto unlock;
 			}