[linux-2.6-block.git] / kernel / sched / debug.c

/*
 * kernel/sched/debug.c
 *
 * Print the CFS rbtree
 *
 * Copyright(C) 2007, Red Hat, Inc., Ingo Molnar
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

#include <linux/proc_fs.h>
#include <linux/sched.h>
#include <linux/seq_file.h>
#include <linux/kallsyms.h>
#include <linux/utsname.h>

#include "sched.h"

static DEFINE_SPINLOCK(sched_debug_lock);

/*
 * This allows printing both to /proc/sched_debug and
 * to the console
 */
#define SEQ_printf(m, x...)			\
 do {						\
	if (m)					\
		seq_printf(m, x);		\
	else					\
		printk(x);			\
 } while (0)

/*
 * Ease the printing of nsec fields:
 */
static long long nsec_high(unsigned long long nsec)
{
	if ((long long)nsec < 0) {
		nsec = -nsec;
		do_div(nsec, 1000000);
		return -nsec;
	}
	do_div(nsec, 1000000);

	return nsec;
}

static unsigned long nsec_low(unsigned long long nsec)
{
	if ((long long)nsec < 0)
		nsec = -nsec;

	return do_div(nsec, 1000000);
}

#define SPLIT_NS(x) nsec_high(x), nsec_low(x)

#ifdef CONFIG_FAIR_GROUP_SCHED
static void print_cfs_group_stats(struct seq_file *m, int cpu, struct task_group *tg)
{
	struct sched_entity *se = tg->se[cpu];

#define P(F) \
	SEQ_printf(m, "  .%-30s: %lld\n", #F, (long long)F)
#define PN(F) \
	SEQ_printf(m, "  .%-30s: %lld.%06ld\n", #F, SPLIT_NS((long long)F))

	if (!se) {
		struct sched_avg *avg = &cpu_rq(cpu)->avg;
		P(avg->runnable_avg_sum);
		P(avg->runnable_avg_period);
		return;
	}


	PN(se->exec_start);
	PN(se->vruntime);
	PN(se->sum_exec_runtime);
#ifdef CONFIG_SCHEDSTATS
	PN(se->statistics.wait_start);
	PN(se->statistics.sleep_start);
	PN(se->statistics.block_start);
	PN(se->statistics.sleep_max);
	PN(se->statistics.block_max);
	PN(se->statistics.exec_max);
	PN(se->statistics.slice_max);
	PN(se->statistics.wait_max);
	PN(se->statistics.wait_sum);
	P(se->statistics.wait_count);
#endif
	P(se->load.weight);
#ifdef CONFIG_SMP
	P(se->avg.runnable_avg_sum);
	P(se->avg.runnable_avg_period);
	P(se->avg.load_avg_contrib);
	P(se->avg.decay_count);
#endif
#undef PN
#undef P
}
#endif

#ifdef CONFIG_CGROUP_SCHED
static char group_path[PATH_MAX];

static char *task_group_path(struct task_group *tg)
{
	if (autogroup_path(tg, group_path, PATH_MAX))
		return group_path;

	cgroup_path(tg->css.cgroup, group_path, PATH_MAX);
	return group_path;
}
#endif

static void
print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
{
	if (rq->curr == p)
		SEQ_printf(m, "R");
	else
		SEQ_printf(m, " ");

	SEQ_printf(m, "%15s %5d %9Ld.%06ld %9Ld %5d ",
		p->comm, task_pid_nr(p),
		SPLIT_NS(p->se.vruntime),
		(long long)(p->nvcsw + p->nivcsw),
		p->prio);
#ifdef CONFIG_SCHEDSTATS
	SEQ_printf(m, "%9Ld.%06ld %9Ld.%06ld %9Ld.%06ld",
		SPLIT_NS(p->se.vruntime),
		SPLIT_NS(p->se.sum_exec_runtime),
		SPLIT_NS(p->se.statistics.sum_sleep_runtime));
#else
	SEQ_printf(m, "%15Ld %15Ld %15Ld.%06ld %15Ld.%06ld %15Ld.%06ld",
		0LL, 0LL, 0LL, 0L, 0LL, 0L, 0LL, 0L);
#endif
#ifdef CONFIG_CGROUP_SCHED
	SEQ_printf(m, " %s", task_group_path(task_group(p)));
#endif

	SEQ_printf(m, "\n");
}

static void print_rq(struct seq_file *m, struct rq *rq, int rq_cpu)
{
	struct task_struct *g, *p;
	unsigned long flags;

	SEQ_printf(m,
	"\nrunnable tasks:\n"
	"            task   PID         tree-key  switches  prio"
	"     exec-runtime         sum-exec        sum-sleep\n"
	"------------------------------------------------------"
	"----------------------------------------------------\n");

	read_lock_irqsave(&tasklist_lock, flags);

	do_each_thread(g, p) {
		if (!p->on_rq || task_cpu(p) != rq_cpu)
			continue;

		print_task(m, rq, p);
	} while_each_thread(g, p);

	read_unlock_irqrestore(&tasklist_lock, flags);
}

void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
{
	s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
		spread, rq0_min_vruntime, spread0;
	struct rq *rq = cpu_rq(cpu);
	struct sched_entity *last;
	unsigned long flags;

#ifdef CONFIG_FAIR_GROUP_SCHED
	SEQ_printf(m, "\ncfs_rq[%d]:%s\n", cpu, task_group_path(cfs_rq->tg));
#else
	SEQ_printf(m, "\ncfs_rq[%d]:\n", cpu);
#endif
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "exec_clock",
			SPLIT_NS(cfs_rq->exec_clock));

	raw_spin_lock_irqsave(&rq->lock, flags);
	if (cfs_rq->rb_leftmost)
		MIN_vruntime = (__pick_first_entity(cfs_rq))->vruntime;
	last = __pick_last_entity(cfs_rq);
	if (last)
		max_vruntime = last->vruntime;
	min_vruntime = cfs_rq->min_vruntime;
	rq0_min_vruntime = cpu_rq(0)->cfs.min_vruntime;
	raw_spin_unlock_irqrestore(&rq->lock, flags);
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "MIN_vruntime",
			SPLIT_NS(MIN_vruntime));
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "min_vruntime",
			SPLIT_NS(min_vruntime));
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "max_vruntime",
			SPLIT_NS(max_vruntime));
	spread = max_vruntime - MIN_vruntime;
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "spread",
			SPLIT_NS(spread));
	spread0 = min_vruntime - rq0_min_vruntime;
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "spread0",
			SPLIT_NS(spread0));
	SEQ_printf(m, "  .%-30s: %d\n", "nr_spread_over",
			cfs_rq->nr_spread_over);
	SEQ_printf(m, "  .%-30s: %d\n", "nr_running", cfs_rq->nr_running);
	SEQ_printf(m, "  .%-30s: %ld\n", "load", cfs_rq->load.weight);
#ifdef CONFIG_SMP
	SEQ_printf(m, "  .%-30s: %ld\n", "runnable_load_avg",
			cfs_rq->runnable_load_avg);
	SEQ_printf(m, "  .%-30s: %ld\n", "blocked_load_avg",
			cfs_rq->blocked_load_avg);
#ifdef CONFIG_FAIR_GROUP_SCHED
	SEQ_printf(m, "  .%-30s: %ld\n", "tg_load_contrib",
			cfs_rq->tg_load_contrib);
	SEQ_printf(m, "  .%-30s: %d\n", "tg_runnable_contrib",
			cfs_rq->tg_runnable_contrib);
	SEQ_printf(m, "  .%-30s: %ld\n", "tg_load_avg",
			atomic_long_read(&cfs_rq->tg->load_avg));
	SEQ_printf(m, "  .%-30s: %d\n", "tg->runnable_avg",
			atomic_read(&cfs_rq->tg->runnable_avg));
#endif
#endif

#ifdef CONFIG_FAIR_GROUP_SCHED
	print_cfs_group_stats(m, cpu, cfs_rq->tg);
#endif
}

void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq)
{
#ifdef CONFIG_RT_GROUP_SCHED
	SEQ_printf(m, "\nrt_rq[%d]:%s\n", cpu, task_group_path(rt_rq->tg));
#else
	SEQ_printf(m, "\nrt_rq[%d]:\n", cpu);
#endif

#define P(x) \
	SEQ_printf(m, "  .%-30s: %Ld\n", #x, (long long)(rt_rq->x))
#define PN(x) \
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rt_rq->x))

	P(rt_nr_running);
	P(rt_throttled);
	PN(rt_time);
	PN(rt_runtime);

#undef PN
#undef P
}

extern __read_mostly int sched_clock_running;

static void print_cpu(struct seq_file *m, int cpu)
{
	struct rq *rq = cpu_rq(cpu);
	unsigned long flags;

#ifdef CONFIG_X86
	{
		unsigned int freq = cpu_khz ? : 1;

		SEQ_printf(m, "cpu#%d, %u.%03u MHz\n",
			   cpu, freq / 1000, (freq % 1000));
	}
#else
	SEQ_printf(m, "cpu#%d\n", cpu);
#endif

#define P(x)								\
do {									\
	if (sizeof(rq->x) == 4)						\
		SEQ_printf(m, "  .%-30s: %ld\n", #x, (long)(rq->x));	\
	else								\
		SEQ_printf(m, "  .%-30s: %Ld\n", #x, (long long)(rq->x));\
} while (0)

#define PN(x) \
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rq->x))

	P(nr_running);
	SEQ_printf(m, "  .%-30s: %lu\n", "load",
		   rq->load.weight);
	P(nr_switches);
	P(nr_load_updates);
	P(nr_uninterruptible);
	PN(next_balance);
	SEQ_printf(m, "  .%-30s: %ld\n", "curr->pid", (long)(task_pid_nr(rq->curr)));
	PN(clock);
	P(cpu_load[0]);
	P(cpu_load[1]);
	P(cpu_load[2]);
	P(cpu_load[3]);
	P(cpu_load[4]);
#undef P
#undef PN

#ifdef CONFIG_SCHEDSTATS
#define P(n) SEQ_printf(m, "  .%-30s: %d\n", #n, rq->n);
#define P64(n) SEQ_printf(m, "  .%-30s: %Ld\n", #n, rq->n);

	P(yld_count);

	P(sched_count);
	P(sched_goidle);
#ifdef CONFIG_SMP
	P64(avg_idle);
#endif

	P(ttwu_count);
	P(ttwu_local);

#undef P
#undef P64
#endif
	spin_lock_irqsave(&sched_debug_lock, flags);
	print_cfs_stats(m, cpu);
	print_rt_stats(m, cpu);

	rcu_read_lock();
	print_rq(m, rq, cpu);
	rcu_read_unlock();
	spin_unlock_irqrestore(&sched_debug_lock, flags);
	SEQ_printf(m, "\n");
}

static const char *sched_tunable_scaling_names[] = {
	"none",
	"logaritmic",
	"linear"
};

static void sched_debug_header(struct seq_file *m)
{
	u64 ktime, sched_clk, cpu_clk;
	unsigned long flags;

	local_irq_save(flags);
	ktime = ktime_to_ns(ktime_get());
	sched_clk = sched_clock();
	cpu_clk = local_clock();
	local_irq_restore(flags);

	SEQ_printf(m, "Sched Debug Version: v0.10, %s %.*s\n",
		init_utsname()->release,
		(int)strcspn(init_utsname()->version, " "),
		init_utsname()->version);

#define P(x) \
	SEQ_printf(m, "%-40s: %Ld\n", #x, (long long)(x))
#define PN(x) \
	SEQ_printf(m, "%-40s: %Ld.%06ld\n", #x, SPLIT_NS(x))
	PN(ktime);
	PN(sched_clk);
	PN(cpu_clk);
	P(jiffies);
#ifdef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK
	P(sched_clock_stable);
#endif
#undef PN
#undef P

	SEQ_printf(m, "\n");
	SEQ_printf(m, "sysctl_sched\n");

#define P(x) \
	SEQ_printf(m, "  .%-40s: %Ld\n", #x, (long long)(x))
#define PN(x) \
	SEQ_printf(m, "  .%-40s: %Ld.%06ld\n", #x, SPLIT_NS(x))
	PN(sysctl_sched_latency);
	PN(sysctl_sched_min_granularity);
	PN(sysctl_sched_wakeup_granularity);
	P(sysctl_sched_child_runs_first);
	P(sysctl_sched_features);
#undef PN
#undef P

	SEQ_printf(m, "  .%-40s: %d (%s)\n",
		"sysctl_sched_tunable_scaling",
		sysctl_sched_tunable_scaling,
		sched_tunable_scaling_names[sysctl_sched_tunable_scaling]);
	SEQ_printf(m, "\n");
}

static int sched_debug_show(struct seq_file *m, void *v)
{
	int cpu = (unsigned long)(v - 2);

	if (cpu != -1)
		print_cpu(m, cpu);
	else
		sched_debug_header(m);

	return 0;
}

void sysrq_sched_debug_show(void)
{
	int cpu;

	sched_debug_header(NULL);
	for_each_online_cpu(cpu)
		print_cpu(NULL, cpu);

}

/*
 * This itererator needs some explanation.
 * It returns 1 for the header position.
 * This means 2 is cpu 0.
 * In a hotplugged system some cpus, including cpu 0, may be missing so we have
 * to use cpumask_* to iterate over the cpus.
 */
static void *sched_debug_start(struct seq_file *file, loff_t *offset)
{
	unsigned long n = *offset;

	if (n == 0)
		return (void *) 1;

	n--;

	if (n > 0)
		n = cpumask_next(n - 1, cpu_online_mask);
	else
		n = cpumask_first(cpu_online_mask);

	*offset = n + 1;

	if (n < nr_cpu_ids)
		return (void *)(unsigned long)(n + 2);
	return NULL;
}

static void *sched_debug_next(struct seq_file *file, void *data, loff_t *offset)
{
	(*offset)++;
	return sched_debug_start(file, offset);
}

static void sched_debug_stop(struct seq_file *file, void *data)
{
}

static const struct seq_operations sched_debug_sops = {
	.start = sched_debug_start,
	.next = sched_debug_next,
	.stop = sched_debug_stop,
	.show = sched_debug_show,
};

static int sched_debug_release(struct inode *inode, struct file *file)
{
	seq_release(inode, file);

	return 0;
}

static int sched_debug_open(struct inode *inode, struct file *filp)
{
	int ret = 0;

	ret = seq_open(filp, &sched_debug_sops);

	return ret;
}

static const struct file_operations sched_debug_fops = {
	.open		= sched_debug_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= sched_debug_release,
};

static int __init init_sched_debug_procfs(void)
{
	struct proc_dir_entry *pe;

	pe = proc_create("sched_debug", 0444, NULL, &sched_debug_fops);
	if (!pe)
		return -ENOMEM;
	return 0;
}

__initcall(init_sched_debug_procfs);

void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
{
	unsigned long nr_switches;

	SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, task_pid_nr(p),
						get_nr_threads(p));
	SEQ_printf(m,
		"---------------------------------------------------------"
		"----------\n");
#define __P(F) \
	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)F)
#define P(F) \
	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)p->F)
#define __PN(F) \
	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)F))
#define PN(F) \
	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)p->F))

	PN(se.exec_start);
	PN(se.vruntime);
	PN(se.sum_exec_runtime);

	nr_switches = p->nvcsw + p->nivcsw;

#ifdef CONFIG_SCHEDSTATS
	PN(se.statistics.wait_start);
	PN(se.statistics.sleep_start);
	PN(se.statistics.block_start);
	PN(se.statistics.sleep_max);
	PN(se.statistics.block_max);
	PN(se.statistics.exec_max);
	PN(se.statistics.slice_max);
	PN(se.statistics.wait_max);
	PN(se.statistics.wait_sum);
	P(se.statistics.wait_count);
	PN(se.statistics.iowait_sum);
	P(se.statistics.iowait_count);
	P(se.nr_migrations);
	P(se.statistics.nr_migrations_cold);
	P(se.statistics.nr_failed_migrations_affine);
	P(se.statistics.nr_failed_migrations_running);
	P(se.statistics.nr_failed_migrations_hot);
	P(se.statistics.nr_forced_migrations);
	P(se.statistics.nr_wakeups);
	P(se.statistics.nr_wakeups_sync);
	P(se.statistics.nr_wakeups_migrate);
	P(se.statistics.nr_wakeups_local);
	P(se.statistics.nr_wakeups_remote);
	P(se.statistics.nr_wakeups_affine);
	P(se.statistics.nr_wakeups_affine_attempts);
	P(se.statistics.nr_wakeups_passive);
	P(se.statistics.nr_wakeups_idle);

	{
		u64 avg_atom, avg_per_cpu;

		avg_atom = p->se.sum_exec_runtime;
		if (nr_switches)
			do_div(avg_atom, nr_switches);
		else
			avg_atom = -1LL;

		avg_per_cpu = p->se.sum_exec_runtime;
		if (p->se.nr_migrations) {
			avg_per_cpu = div64_u64(avg_per_cpu,
						p->se.nr_migrations);
		} else {
			avg_per_cpu = -1LL;
		}

		__PN(avg_atom);
		__PN(avg_per_cpu);
	}
#endif
	__P(nr_switches);
	SEQ_printf(m, "%-45s:%21Ld\n",
		   "nr_voluntary_switches", (long long)p->nvcsw);
	SEQ_printf(m, "%-45s:%21Ld\n",
		   "nr_involuntary_switches", (long long)p->nivcsw);

	P(se.load.weight);
#ifdef CONFIG_SMP
	P(se.avg.runnable_avg_sum);
	P(se.avg.runnable_avg_period);
	P(se.avg.load_avg_contrib);
	P(se.avg.decay_count);
#endif
	P(policy);
	P(prio);
#undef PN
#undef __PN
#undef P
#undef __P

	{
		unsigned int this_cpu = raw_smp_processor_id();
		u64 t0, t1;

		t0 = cpu_clock(this_cpu);
		t1 = cpu_clock(this_cpu);
		SEQ_printf(m, "%-45s:%21Ld\n",
			   "clock-delta", (long long)(t1-t0));
	}
}

void proc_sched_set_task(struct task_struct *p)
{
#ifdef CONFIG_SCHEDSTATS
	memset(&p->se.statistics, 0, sizeof(p->se.statistics));
#endif
}
Commit	Line	Data
43ae34cb	1	/*
391e43da	2	* kernel/sched/debug.c
43ae34cb IM	3	*
	4	* Print the CFS rbtree
	5	*
	6	* Copyright(C) 2007, Red Hat, Inc., Ingo Molnar
	7	*
	8	* This program is free software; you can redistribute it and/or modify
	9	* it under the terms of the GNU General Public License version 2 as
	10	* published by the Free Software Foundation.
	11	*/
	12
	13	#include <linux/proc_fs.h>
	14	#include <linux/sched.h>
	15	#include <linux/seq_file.h>
	16	#include <linux/kallsyms.h>
	17	#include <linux/utsname.h>
	18
029632fb PZ	19	#include "sched.h"
029632fb PZ	20
efe25c2c BR	21	static DEFINE_SPINLOCK(sched_debug_lock);
efe25c2c BR	22
43ae34cb IM	23	/*
	24	* This allows printing both to /proc/sched_debug and
	25	* to the console
	26	*/
	27	#define SEQ_printf(m, x...) \
	28	do { \
	29	if (m) \
	30	seq_printf(m, x); \
	31	else \
	32	printk(x); \
	33	} while (0)
	34
ef83a571 IM	35	/*
	36	* Ease the printing of nsec fields:
	37	*/
90b2628f	38	static long long nsec_high(unsigned long long nsec)
ef83a571	39	{
90b2628f	40	if ((long long)nsec < 0) {
ef83a571 IM	41	nsec = -nsec;
	42	do_div(nsec, 1000000);
	43	return -nsec;
	44	}
	45	do_div(nsec, 1000000);
	46
	47	return nsec;
	48	}
	49
90b2628f	50	static unsigned long nsec_low(unsigned long long nsec)
ef83a571	51	{
90b2628f	52	if ((long long)nsec < 0)
ef83a571 IM	53	nsec = -nsec;
	54
	55	return do_div(nsec, 1000000);
	56	}
	57
	58	#define SPLIT_NS(x) nsec_high(x), nsec_low(x)
	59
ff9b48c3	60	#ifdef CONFIG_FAIR_GROUP_SCHED
5091faa4	61	static void print_cfs_group_stats(struct seq_file m, int cpu, struct task_group tg)
ff9b48c3 BR	62	{
ff9b48c3 BR	63	struct sched_entity *se = tg->se[cpu];
ff9b48c3 BR	64
	65	#define P(F) \
	66	SEQ_printf(m, " .%-30s: %lld\n", #F, (long long)F)
	67	#define PN(F) \
	68	SEQ_printf(m, " .%-30s: %lld.%06ld\n", #F, SPLIT_NS((long long)F))
	69
18bf2805 BS	70	if (!se) {
	71	struct sched_avg *avg = &cpu_rq(cpu)->avg;
	72	P(avg->runnable_avg_sum);
	73	P(avg->runnable_avg_period);
	74	return;
	75	}
	76
	77
ff9b48c3 BR	78	PN(se->exec_start);
	79	PN(se->vruntime);
	80	PN(se->sum_exec_runtime);
	81	#ifdef CONFIG_SCHEDSTATS
41acab88 LDM	82	PN(se->statistics.wait_start);
	83	PN(se->statistics.sleep_start);
	84	PN(se->statistics.block_start);
	85	PN(se->statistics.sleep_max);
	86	PN(se->statistics.block_max);
	87	PN(se->statistics.exec_max);
	88	PN(se->statistics.slice_max);
	89	PN(se->statistics.wait_max);
	90	PN(se->statistics.wait_sum);
	91	P(se->statistics.wait_count);
ff9b48c3 BR	92	#endif
ff9b48c3 BR	93	P(se->load.weight);
9d85f21c PT	94	#ifdef CONFIG_SMP
	95	P(se->avg.runnable_avg_sum);
	96	P(se->avg.runnable_avg_period);
2dac754e	97	P(se->avg.load_avg_contrib);
9ee474f5	98	P(se->avg.decay_count);
9d85f21c	99	#endif
ff9b48c3 BR	100	#undef PN
	101	#undef P
	102	}
	103	#endif
	104
efe25c2c BR	105	#ifdef CONFIG_CGROUP_SCHED
	106	static char group_path[PATH_MAX];
	107
	108	static char task_group_path(struct task_group tg)
	109	{
8ecedd7a BR	110	if (autogroup_path(tg, group_path, PATH_MAX))
	111	return group_path;
	112
efe25c2c BR	113	cgroup_path(tg->css.cgroup, group_path, PATH_MAX);
	114	return group_path;
	115	}
	116	#endif
	117
43ae34cb	118	static void
a48da48b	119	print_task(struct seq_file m, struct rq rq, struct task_struct *p)
43ae34cb IM	120	{
	121	if (rq->curr == p)
	122	SEQ_printf(m, "R");
	123	else
	124	SEQ_printf(m, " ");
	125
ef83a571	126	SEQ_printf(m, "%15s %5d %9Ld.%06ld %9Ld %5d ",
fc840914	127	p->comm, task_pid_nr(p),
ef83a571	128	SPLIT_NS(p->se.vruntime),
43ae34cb	129	(long long)(p->nvcsw + p->nivcsw),
6f605d83	130	p->prio);
6cfb0d5d	131	#ifdef CONFIG_SCHEDSTATS
d19ca308	132	SEQ_printf(m, "%9Ld.%06ld %9Ld.%06ld %9Ld.%06ld",
ef83a571 IM	133	SPLIT_NS(p->se.vruntime),
ef83a571 IM	134	SPLIT_NS(p->se.sum_exec_runtime),
41acab88	135	SPLIT_NS(p->se.statistics.sum_sleep_runtime));
6cfb0d5d	136	#else
d19ca308	137	SEQ_printf(m, "%15Ld %15Ld %15Ld.%06ld %15Ld.%06ld %15Ld.%06ld",
ef83a571	138	0LL, 0LL, 0LL, 0L, 0LL, 0L, 0LL, 0L);
6cfb0d5d	139	#endif
efe25c2c BR	140	#ifdef CONFIG_CGROUP_SCHED
	141	SEQ_printf(m, " %s", task_group_path(task_group(p)));
	142	#endif
d19ca308	143
d19ca308	144	SEQ_printf(m, "\n");
43ae34cb IM	145	}
43ae34cb IM	146
a48da48b	147	static void print_rq(struct seq_file m, struct rq rq, int rq_cpu)
43ae34cb IM	148	{
43ae34cb IM	149	struct task_struct g, p;
ab63a633	150	unsigned long flags;
43ae34cb IM	151
	152	SEQ_printf(m,
	153	"\nrunnable tasks:\n"
c86da3a3 MG	154	" task PID tree-key switches prio"
c86da3a3 MG	155	" exec-runtime sum-exec sum-sleep\n"
1a75b94f	156	"------------------------------------------------------"
c86da3a3	157	"----------------------------------------------------\n");
43ae34cb	158
ab63a633	159	read_lock_irqsave(&tasklist_lock, flags);
43ae34cb IM	160
43ae34cb IM	161	do_each_thread(g, p) {
fd2f4419	162	if (!p->on_rq \|\| task_cpu(p) != rq_cpu)
43ae34cb IM	163	continue;
43ae34cb IM	164
a48da48b	165	print_task(m, rq, p);
43ae34cb IM	166	} while_each_thread(g, p);
43ae34cb IM	167
ab63a633	168	read_unlock_irqrestore(&tasklist_lock, flags);
43ae34cb IM	169	}
43ae34cb IM	170
5cef9eca	171	void print_cfs_rq(struct seq_file m, int cpu, struct cfs_rq cfs_rq)
43ae34cb	172	{
86d9560c IM	173	s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
86d9560c IM	174	spread, rq0_min_vruntime, spread0;
348ec61e	175	struct rq *rq = cpu_rq(cpu);
67e12eac IM	176	struct sched_entity *last;
	177	unsigned long flags;
	178
efe25c2c BR	179	#ifdef CONFIG_FAIR_GROUP_SCHED
	180	SEQ_printf(m, "\ncfs_rq[%d]:%s\n", cpu, task_group_path(cfs_rq->tg));
	181	#else
ada18de2	182	SEQ_printf(m, "\ncfs_rq[%d]:\n", cpu);
efe25c2c	183	#endif
ef83a571 IM	184	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "exec_clock",
ef83a571 IM	185	SPLIT_NS(cfs_rq->exec_clock));
67e12eac	186
05fa785c	187	raw_spin_lock_irqsave(&rq->lock, flags);
67e12eac	188	if (cfs_rq->rb_leftmost)
ac53db59	189	MIN_vruntime = (__pick_first_entity(cfs_rq))->vruntime;
67e12eac IM	190	last = __pick_last_entity(cfs_rq);
	191	if (last)
	192	max_vruntime = last->vruntime;
5ac5c4d6	193	min_vruntime = cfs_rq->min_vruntime;
348ec61e	194	rq0_min_vruntime = cpu_rq(0)->cfs.min_vruntime;
05fa785c	195	raw_spin_unlock_irqrestore(&rq->lock, flags);
ef83a571 IM	196	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "MIN_vruntime",
	197	SPLIT_NS(MIN_vruntime));
	198	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "min_vruntime",
	199	SPLIT_NS(min_vruntime));
	200	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "max_vruntime",
	201	SPLIT_NS(max_vruntime));
67e12eac	202	spread = max_vruntime - MIN_vruntime;
ef83a571 IM	203	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "spread",
ef83a571 IM	204	SPLIT_NS(spread));
86d9560c	205	spread0 = min_vruntime - rq0_min_vruntime;
ef83a571 IM	206	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "spread0",
ef83a571 IM	207	SPLIT_NS(spread0));
5ac5c4d6	208	SEQ_printf(m, " .%-30s: %d\n", "nr_spread_over",
ddc97297	209	cfs_rq->nr_spread_over);
c82513e5	210	SEQ_printf(m, " .%-30s: %d\n", "nr_running", cfs_rq->nr_running);
2069dd75	211	SEQ_printf(m, " .%-30s: %ld\n", "load", cfs_rq->load.weight);
c09595f6	212	#ifdef CONFIG_SMP
72a4cf20	213	SEQ_printf(m, " .%-30s: %ld\n", "runnable_load_avg",
2dac754e	214	cfs_rq->runnable_load_avg);
72a4cf20	215	SEQ_printf(m, " .%-30s: %ld\n", "blocked_load_avg",
9ee474f5	216	cfs_rq->blocked_load_avg);
333bb864	217	#ifdef CONFIG_FAIR_GROUP_SCHED
bf5b986e	218	SEQ_printf(m, " .%-30s: %ld\n", "tg_load_contrib",
c566e8e9	219	cfs_rq->tg_load_contrib);
bb17f655 PT	220	SEQ_printf(m, " .%-30s: %d\n", "tg_runnable_contrib",
bb17f655 PT	221	cfs_rq->tg_runnable_contrib);
333bb864 AS	222	SEQ_printf(m, " .%-30s: %ld\n", "tg_load_avg",
333bb864 AS	223	atomic_long_read(&cfs_rq->tg->load_avg));
bb17f655 PT	224	SEQ_printf(m, " .%-30s: %d\n", "tg->runnable_avg",
bb17f655 PT	225	atomic_read(&cfs_rq->tg->runnable_avg));
c09595f6	226	#endif
333bb864	227	#endif
2069dd75	228
333bb864	229	#ifdef CONFIG_FAIR_GROUP_SCHED
ff9b48c3	230	print_cfs_group_stats(m, cpu, cfs_rq->tg);
c09595f6	231	#endif
43ae34cb IM	232	}
43ae34cb IM	233
ada18de2 PZ	234	void print_rt_rq(struct seq_file m, int cpu, struct rt_rq rt_rq)
ada18de2 PZ	235	{
efe25c2c BR	236	#ifdef CONFIG_RT_GROUP_SCHED
	237	SEQ_printf(m, "\nrt_rq[%d]:%s\n", cpu, task_group_path(rt_rq->tg));
	238	#else
ada18de2	239	SEQ_printf(m, "\nrt_rq[%d]:\n", cpu);
efe25c2c	240	#endif
ada18de2 PZ	241
	242	#define P(x) \
	243	SEQ_printf(m, " .%-30s: %Ld\n", #x, (long long)(rt_rq->x))
	244	#define PN(x) \
	245	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rt_rq->x))
	246
	247	P(rt_nr_running);
	248	P(rt_throttled);
	249	PN(rt_time);
	250	PN(rt_runtime);
	251
	252	#undef PN
	253	#undef P
	254	}
	255
5bb6b1ea PZ	256	extern __read_mostly int sched_clock_running;
5bb6b1ea PZ	257
a48da48b	258	static void print_cpu(struct seq_file *m, int cpu)
43ae34cb	259	{
348ec61e	260	struct rq *rq = cpu_rq(cpu);
efe25c2c	261	unsigned long flags;
43ae34cb IM	262
	263	#ifdef CONFIG_X86
	264	{
	265	unsigned int freq = cpu_khz ? : 1;
	266
bbbfeac9	267	SEQ_printf(m, "cpu#%d, %u.%03u MHz\n",
43ae34cb IM	268	cpu, freq / 1000, (freq % 1000));
	269	}
	270	#else
bbbfeac9	271	SEQ_printf(m, "cpu#%d\n", cpu);
43ae34cb IM	272	#endif
43ae34cb IM	273
13e099d2 PZ	274	#define P(x) \
	275	do { \
	276	if (sizeof(rq->x) == 4) \
	277	SEQ_printf(m, " .%-30s: %ld\n", #x, (long)(rq->x)); \
	278	else \
	279	SEQ_printf(m, " .%-30s: %Ld\n", #x, (long long)(rq->x));\
	280	} while (0)
	281
ef83a571 IM	282	#define PN(x) \
ef83a571 IM	283	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rq->x))
43ae34cb IM	284
	285	P(nr_running);
	286	SEQ_printf(m, " .%-30s: %lu\n", "load",
495eca49	287	rq->load.weight);
43ae34cb IM	288	P(nr_switches);
	289	P(nr_load_updates);
	290	P(nr_uninterruptible);
ef83a571	291	PN(next_balance);
fc840914	292	SEQ_printf(m, " .%-30s: %ld\n", "curr->pid", (long)(task_pid_nr(rq->curr)));
ef83a571	293	PN(clock);
43ae34cb IM	294	P(cpu_load[0]);
	295	P(cpu_load[1]);
	296	P(cpu_load[2]);
	297	P(cpu_load[3]);
	298	P(cpu_load[4]);
	299	#undef P
ef83a571	300	#undef PN
43ae34cb	301
5ac5c4d6 PZ	302	#ifdef CONFIG_SCHEDSTATS
5ac5c4d6 PZ	303	#define P(n) SEQ_printf(m, " .%-30s: %d\n", #n, rq->n);
1b9508f6	304	#define P64(n) SEQ_printf(m, " .%-30s: %Ld\n", #n, rq->n);
5ac5c4d6	305
5ac5c4d6 PZ	306	P(yld_count);
5ac5c4d6 PZ	307
5ac5c4d6 PZ	308	P(sched_count);
5ac5c4d6 PZ	309	P(sched_goidle);
1b9508f6 MG	310	#ifdef CONFIG_SMP
	311	P64(avg_idle);
	312	#endif
5ac5c4d6 PZ	313
	314	P(ttwu_count);
	315	P(ttwu_local);
	316
5ac5c4d6	317	#undef P
fce20979	318	#undef P64
5ac5c4d6	319	#endif
efe25c2c	320	spin_lock_irqsave(&sched_debug_lock, flags);
5cef9eca	321	print_cfs_stats(m, cpu);
ada18de2	322	print_rt_stats(m, cpu);
43ae34cb	323
efe25c2c	324	rcu_read_lock();
a48da48b	325	print_rq(m, rq, cpu);
efe25c2c BR	326	rcu_read_unlock();
efe25c2c BR	327	spin_unlock_irqrestore(&sched_debug_lock, flags);
bbbfeac9	328	SEQ_printf(m, "\n");
43ae34cb IM	329	}
43ae34cb IM	330
1983a922 CE	331	static const char *sched_tunable_scaling_names[] = {
	332	"none",
	333	"logaritmic",
	334	"linear"
	335	};
	336
bbbfeac9	337	static void sched_debug_header(struct seq_file *m)
43ae34cb	338	{
5bb6b1ea PZ	339	u64 ktime, sched_clk, cpu_clk;
5bb6b1ea PZ	340	unsigned long flags;
43ae34cb	341
5bb6b1ea PZ	342	local_irq_save(flags);
	343	ktime = ktime_to_ns(ktime_get());
	344	sched_clk = sched_clock();
	345	cpu_clk = local_clock();
	346	local_irq_restore(flags);
	347
	348	SEQ_printf(m, "Sched Debug Version: v0.10, %s %.*s\n",
43ae34cb IM	349	init_utsname()->release,
	350	(int)strcspn(init_utsname()->version, " "),
	351	init_utsname()->version);
	352
5bb6b1ea PZ	353	#define P(x) \
	354	SEQ_printf(m, "%-40s: %Ld\n", #x, (long long)(x))
	355	#define PN(x) \
	356	SEQ_printf(m, "%-40s: %Ld.%06ld\n", #x, SPLIT_NS(x))
	357	PN(ktime);
	358	PN(sched_clk);
	359	PN(cpu_clk);
	360	P(jiffies);
	361	#ifdef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK
	362	P(sched_clock_stable);
	363	#endif
	364	#undef PN
	365	#undef P
	366
	367	SEQ_printf(m, "\n");
	368	SEQ_printf(m, "sysctl_sched\n");
43ae34cb	369
1aa4731e	370	#define P(x) \
d822cece	371	SEQ_printf(m, " .%-40s: %Ld\n", #x, (long long)(x))
1aa4731e	372	#define PN(x) \
d822cece	373	SEQ_printf(m, " .%-40s: %Ld.%06ld\n", #x, SPLIT_NS(x))
1aa4731e	374	PN(sysctl_sched_latency);
b2be5e96	375	PN(sysctl_sched_min_granularity);
1aa4731e	376	PN(sysctl_sched_wakeup_granularity);
eebef746	377	P(sysctl_sched_child_runs_first);
1aa4731e IM	378	P(sysctl_sched_features);
	379	#undef PN
	380	#undef P
	381
bbbfeac9 NZ	382	SEQ_printf(m, " .%-40s: %d (%s)\n",
bbbfeac9 NZ	383	"sysctl_sched_tunable_scaling",
1983a922 CE	384	sysctl_sched_tunable_scaling,
1983a922 CE	385	sched_tunable_scaling_names[sysctl_sched_tunable_scaling]);
bbbfeac9 NZ	386	SEQ_printf(m, "\n");
bbbfeac9 NZ	387	}
1983a922	388
bbbfeac9 NZ	389	static int sched_debug_show(struct seq_file m, void v)
	390	{
	391	int cpu = (unsigned long)(v - 2);
43ae34cb	392
bbbfeac9 NZ	393	if (cpu != -1)
	394	print_cpu(m, cpu);
	395	else
	396	sched_debug_header(m);
43ae34cb IM	397
	398	return 0;
	399	}
	400
029632fb	401	void sysrq_sched_debug_show(void)
43ae34cb	402	{
bbbfeac9 NZ	403	int cpu;
	404
	405	sched_debug_header(NULL);
	406	for_each_online_cpu(cpu)
	407	print_cpu(NULL, cpu);
	408
	409	}
	410
	411	/*
	412	* This itererator needs some explanation.
	413	* It returns 1 for the header position.
	414	* This means 2 is cpu 0.
	415	* In a hotplugged system some cpus, including cpu 0, may be missing so we have
	416	* to use cpumask_* to iterate over the cpus.
	417	*/
	418	static void sched_debug_start(struct seq_file file, loff_t *offset)
	419	{
	420	unsigned long n = *offset;
	421
	422	if (n == 0)
	423	return (void *) 1;
	424
	425	n--;
	426
	427	if (n > 0)
	428	n = cpumask_next(n - 1, cpu_online_mask);
	429	else
	430	n = cpumask_first(cpu_online_mask);
	431
	432	*offset = n + 1;
	433
	434	if (n < nr_cpu_ids)
	435	return (void *)(unsigned long)(n + 2);
	436	return NULL;
	437	}
	438
	439	static void sched_debug_next(struct seq_file file, void data, loff_t offset)
	440	{
	441	(*offset)++;
	442	return sched_debug_start(file, offset);
	443	}
	444
	445	static void sched_debug_stop(struct seq_file file, void data)
	446	{
	447	}
	448
	449	static const struct seq_operations sched_debug_sops = {
	450	.start = sched_debug_start,
	451	.next = sched_debug_next,
	452	.stop = sched_debug_stop,
	453	.show = sched_debug_show,
	454	};
	455
	456	static int sched_debug_release(struct inode inode, struct file file)
	457	{
	458	seq_release(inode, file);
	459
	460	return 0;
43ae34cb IM	461	}
	462
	463	static int sched_debug_open(struct inode inode, struct file filp)
	464	{
bbbfeac9 NZ	465	int ret = 0;
	466
	467	ret = seq_open(filp, &sched_debug_sops);
	468
	469	return ret;
43ae34cb IM	470	}
43ae34cb IM	471
0dbee3a6	472	static const struct file_operations sched_debug_fops = {
43ae34cb IM	473	.open = sched_debug_open,
	474	.read = seq_read,
	475	.llseek = seq_lseek,
bbbfeac9	476	.release = sched_debug_release,
43ae34cb IM	477	};
	478
	479	static int __init init_sched_debug_procfs(void)
	480	{
	481	struct proc_dir_entry *pe;
	482
a9cf4ddb	483	pe = proc_create("sched_debug", 0444, NULL, &sched_debug_fops);
43ae34cb IM	484	if (!pe)
43ae34cb IM	485	return -ENOMEM;
43ae34cb IM	486	return 0;
	487	}
	488
	489	__initcall(init_sched_debug_procfs);
	490
	491	void proc_sched_show_task(struct task_struct p, struct seq_file m)
	492	{
cc367732	493	unsigned long nr_switches;
43ae34cb	494
fc840914	495	SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, task_pid_nr(p),
5089a976	496	get_nr_threads(p));
2d92f227	497	SEQ_printf(m,
add332a1 KB	498	"---------------------------------------------------------"
add332a1 KB	499	"----------\n");
cc367732	500	#define __P(F) \
add332a1	501	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)F)
43ae34cb	502	#define P(F) \
add332a1	503	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)p->F)
cc367732	504	#define __PN(F) \
add332a1	505	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)F))
ef83a571	506	#define PN(F) \
add332a1	507	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)p->F))
43ae34cb	508
ef83a571 IM	509	PN(se.exec_start);
	510	PN(se.vruntime);
	511	PN(se.sum_exec_runtime);
6cfb0d5d	512
cc367732 IM	513	nr_switches = p->nvcsw + p->nivcsw;
cc367732 IM	514
6cfb0d5d	515	#ifdef CONFIG_SCHEDSTATS
41acab88 LDM	516	PN(se.statistics.wait_start);
	517	PN(se.statistics.sleep_start);
	518	PN(se.statistics.block_start);
	519	PN(se.statistics.sleep_max);
	520	PN(se.statistics.block_max);
	521	PN(se.statistics.exec_max);
	522	PN(se.statistics.slice_max);
	523	PN(se.statistics.wait_max);
	524	PN(se.statistics.wait_sum);
	525	P(se.statistics.wait_count);
	526	PN(se.statistics.iowait_sum);
	527	P(se.statistics.iowait_count);
cc367732	528	P(se.nr_migrations);
41acab88 LDM	529	P(se.statistics.nr_migrations_cold);
	530	P(se.statistics.nr_failed_migrations_affine);
	531	P(se.statistics.nr_failed_migrations_running);
	532	P(se.statistics.nr_failed_migrations_hot);
	533	P(se.statistics.nr_forced_migrations);
	534	P(se.statistics.nr_wakeups);
	535	P(se.statistics.nr_wakeups_sync);
	536	P(se.statistics.nr_wakeups_migrate);
	537	P(se.statistics.nr_wakeups_local);
	538	P(se.statistics.nr_wakeups_remote);
	539	P(se.statistics.nr_wakeups_affine);
	540	P(se.statistics.nr_wakeups_affine_attempts);
	541	P(se.statistics.nr_wakeups_passive);
	542	P(se.statistics.nr_wakeups_idle);
cc367732 IM	543
	544	{
	545	u64 avg_atom, avg_per_cpu;
	546
	547	avg_atom = p->se.sum_exec_runtime;
	548	if (nr_switches)
	549	do_div(avg_atom, nr_switches);
	550	else
	551	avg_atom = -1LL;
	552
	553	avg_per_cpu = p->se.sum_exec_runtime;
c1a89740	554	if (p->se.nr_migrations) {
6f6d6a1a RZ	555	avg_per_cpu = div64_u64(avg_per_cpu,
6f6d6a1a RZ	556	p->se.nr_migrations);
c1a89740	557	} else {
cc367732	558	avg_per_cpu = -1LL;
c1a89740	559	}
cc367732 IM	560
	561	__PN(avg_atom);
	562	__PN(avg_per_cpu);
	563	}
6cfb0d5d	564	#endif
cc367732	565	__P(nr_switches);
add332a1	566	SEQ_printf(m, "%-45s:%21Ld\n",
cc367732	567	"nr_voluntary_switches", (long long)p->nvcsw);
add332a1	568	SEQ_printf(m, "%-45s:%21Ld\n",
cc367732 IM	569	"nr_involuntary_switches", (long long)p->nivcsw);
cc367732 IM	570
43ae34cb	571	P(se.load.weight);
333bb864	572	#ifdef CONFIG_SMP
939fd731 KB	573	P(se.avg.runnable_avg_sum);
	574	P(se.avg.runnable_avg_period);
	575	P(se.avg.load_avg_contrib);
	576	P(se.avg.decay_count);
	577	#endif
43ae34cb IM	578	P(policy);
43ae34cb IM	579	P(prio);
ef83a571	580	#undef PN
cc367732 IM	581	#undef __PN
	582	#undef P
	583	#undef __P
43ae34cb IM	584
43ae34cb IM	585	{
29d7b90c	586	unsigned int this_cpu = raw_smp_processor_id();
43ae34cb IM	587	u64 t0, t1;
43ae34cb IM	588
29d7b90c IM	589	t0 = cpu_clock(this_cpu);
29d7b90c IM	590	t1 = cpu_clock(this_cpu);
add332a1	591	SEQ_printf(m, "%-45s:%21Ld\n",
43ae34cb IM	592	"clock-delta", (long long)(t1-t0));
	593	}
	594	}
	595
	596	void proc_sched_set_task(struct task_struct *p)
	597	{
6cfb0d5d	598	#ifdef CONFIG_SCHEDSTATS
41acab88	599	memset(&p->se.statistics, 0, sizeof(p->se.statistics));
6cfb0d5d	600	#endif
43ae34cb	601	}