Merge git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched

* git://git.kernel.org/pub/scm/linux/kernel/git/mingo/linux-2.6-sched:
  sched: tweak the sched_runtime_limit tunable
  sched: skip updating rq's next_balance under null SD
  sched: fix broken SMT/MC optimizations
  sched: accounting regression since rc1
  sched: fix sysctl directory permissions
  sched: sched_clock_idle_[sleep|wakeup]_event()

2 files changed, 53 insertions, 18 deletions

diff --git a/kernel/sched.c b/kernel/sched.c
index 45e17b83b7f..96e9b82246d 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -262,7 +262,8 @@ struct rq {
 	s64 clock_max_delta;
 
 	unsigned int clock_warps, clock_overflows;
-	unsigned int clock_unstable_events;
+	u64 idle_clock;
+	unsigned int clock_deep_idle_events;
 	u64 tick_timestamp;
 
 	atomic_t nr_iowait;
@@ -556,18 +557,40 @@ static inline struct rq *this_rq_lock(void)
 }
 
 /*
- * CPU frequency is/was unstable - start new by setting prev_clock_raw:
+ * We are going deep-idle (irqs are disabled):
  */
-void sched_clock_unstable_event(void)
+void sched_clock_idle_sleep_event(void)
 {
-	unsigned long flags;
-	struct rq *rq;
+	struct rq *rq = cpu_rq(smp_processor_id());
 
-	rq = task_rq_lock(current, &flags);
-	rq->prev_clock_raw = sched_clock();
-	rq->clock_unstable_events++;
-	task_rq_unlock(rq, &flags);
+	spin_lock(&rq->lock);
+	__update_rq_clock(rq);
+	spin_unlock(&rq->lock);
+	rq->clock_deep_idle_events++;
 }
+EXPORT_SYMBOL_GPL(sched_clock_idle_sleep_event);
+
+/*
+ * We just idled delta nanoseconds (called with irqs disabled):
+ */
+void sched_clock_idle_wakeup_event(u64 delta_ns)
+{
+	struct rq *rq = cpu_rq(smp_processor_id());
+	u64 now = sched_clock();
+
+	rq->idle_clock += delta_ns;
+	/*
+	 * Override the previous timestamp and ignore all
+	 * sched_clock() deltas that occured while we idled,
+	 * and use the PM-provided delta_ns to advance the
+	 * rq clock:
+	 */
+	spin_lock(&rq->lock);
+	rq->prev_clock_raw = now;
+	rq->clock += delta_ns;
+	spin_unlock(&rq->lock);
+}
+EXPORT_SYMBOL_GPL(sched_clock_idle_wakeup_event);
 
 /*
  * resched_task - mark a task 'to be rescheduled now'.
@@ -2494,7 +2517,7 @@ group_next:
 	 * a think about bumping its value to force at least one task to be
 	 * moved
 	 */
-	if (*imbalance + SCHED_LOAD_SCALE_FUZZ < busiest_load_per_task/2) {
+	if (*imbalance + SCHED_LOAD_SCALE_FUZZ < busiest_load_per_task) {
 		unsigned long tmp, pwr_now, pwr_move;
 		unsigned int imbn;
 
@@ -3020,6 +3043,7 @@ static inline void rebalance_domains(int cpu, enum cpu_idle_type idle)
 	struct sched_domain *sd;
 	/* Earliest time when we have to do rebalance again */
 	unsigned long next_balance = jiffies + 60*HZ;
+	int update_next_balance = 0;
 
 	for_each_domain(cpu, sd) {
 		if (!(sd->flags & SD_LOAD_BALANCE))
@@ -3056,8 +3080,10 @@ static inline void rebalance_domains(int cpu, enum cpu_idle_type idle)
 		if (sd->flags & SD_SERIALIZE)
 			spin_unlock(&balancing);
 out:
-		if (time_after(next_balance, sd->last_balance + interval))
+		if (time_after(next_balance, sd->last_balance + interval)) {
 			next_balance = sd->last_balance + interval;
+			update_next_balance = 1;
+		}
 
 		/*
 		 * Stop the load balance at this level. There is another
@@ -3067,7 +3093,14 @@ out:
 		if (!balance)
 			break;
 	}
-	rq->next_balance = next_balance;
+
+	/*
+	 * next_balance will be updated only when there is a need.
+	 * When the cpu is attached to null domain for ex, it will not be
+	 * updated.
+	 */
+	if (likely(update_next_balance))
+		rq->next_balance = next_balance;
 }
 
 /*
@@ -4890,7 +4923,7 @@ static inline void sched_init_granularity(void)
 	if (sysctl_sched_granularity > gran_limit)
 		sysctl_sched_granularity = gran_limit;
 
-	sysctl_sched_runtime_limit = sysctl_sched_granularity * 4;
+	sysctl_sched_runtime_limit = sysctl_sched_granularity * 8;
 	sysctl_sched_wakeup_granularity = sysctl_sched_granularity / 2;
 }
 
@@ -5234,15 +5267,16 @@ static void migrate_dead_tasks(unsigned int dead_cpu)
 static struct ctl_table sd_ctl_dir[] = {
 	{
 		.procname	= "sched_domain",
-		.mode		= 0755,
+		.mode		= 0555,
 	},
 	{0,},
 };
 
 static struct ctl_table sd_ctl_root[] = {
 	{
+		.ctl_name	= CTL_KERN,
 		.procname	= "kernel",
-		.mode		= 0755,
+		.mode		= 0555,
 		.child		= sd_ctl_dir,
 	},
 	{0,},
@@ -5318,7 +5352,7 @@ static ctl_table *sd_alloc_ctl_cpu_table(int cpu)
 	for_each_domain(cpu, sd) {
 		snprintf(buf, 32, "domain%d", i);
 		entry->procname = kstrdup(buf, GFP_KERNEL);
-		entry->mode = 0755;
+		entry->mode = 0555;
 		entry->child = sd_alloc_ctl_domain_table(sd);
 		entry++;
 		i++;
@@ -5338,7 +5372,7 @@ static void init_sched_domain_sysctl(void)
 	for (i = 0; i < cpu_num; i++, entry++) {
 		snprintf(buf, 32, "cpu%d", i);
 		entry->procname = kstrdup(buf, GFP_KERNEL);
-		entry->mode = 0755;
+		entry->mode = 0555;
 		entry->child = sd_alloc_ctl_cpu_table(i);
 	}
 	sd_sysctl_header = register_sysctl_table(sd_ctl_root);
diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
index 87e524762b8..ab18f45f2ab 100644
--- a/kernel/sched_debug.c
+++ b/kernel/sched_debug.c
@@ -154,10 +154,11 @@ static void print_cpu(struct seq_file *m, int cpu)
 	P(next_balance);
 	P(curr->pid);
 	P(clock);
+	P(idle_clock);
 	P(prev_clock_raw);
 	P(clock_warps);
 	P(clock_overflows);
-	P(clock_unstable_events);
+	P(clock_deep_idle_events);
 	P(clock_max_delta);
 	P(cpu_load[0]);
 	P(cpu_load[1]);