[linux-block.git] / kernel / sched_debug.c

/*
 * kernel/time/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>

/*
 * 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(long long nsec)
{
	if (nsec < 0) {
		nsec = -nsec;
		do_div(nsec, 1000000);
		return -nsec;
	}
	do_div(nsec, 1000000);

	return nsec;
}

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

	return do_div(nsec, 1000000);
}

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

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, p->pid,
		SPLIT_NS(p->se.vruntime),
		(long long)(p->nvcsw + p->nivcsw),
		p->prio);
#ifdef CONFIG_SCHEDSTATS
	SEQ_printf(m, "%15Ld.%06ld %15Ld.%06ld %15Ld.%06ld\n",
		SPLIT_NS(p->se.vruntime),
		SPLIT_NS(p->se.sum_exec_runtime),
		SPLIT_NS(p->se.sum_sleep_runtime));
#else
	SEQ_printf(m, "%15Ld %15Ld %15Ld.%06ld %15Ld.%06ld %15Ld.%06ld\n",
		0LL, 0LL, 0LL, 0L, 0LL, 0L, 0LL, 0L);
#endif
}

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

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

	read_lock_irq(&tasklist_lock);

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

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

	read_unlock_irq(&tasklist_lock);
}

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 = &per_cpu(runqueues, cpu);
	struct sched_entity *last;
	unsigned long flags;

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

	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "exec_clock",
			SPLIT_NS(cfs_rq->exec_clock));

	spin_lock_irqsave(&rq->lock, flags);
	if (cfs_rq->rb_leftmost)
		MIN_vruntime = (__pick_next_entity(cfs_rq))->vruntime;
	last = __pick_last_entity(cfs_rq);
	if (last)
		max_vruntime = last->vruntime;
	min_vruntime = rq->cfs.min_vruntime;
	rq0_min_vruntime = per_cpu(runqueues, 0).cfs.min_vruntime;
	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));
}

static void print_cpu(struct seq_file *m, int cpu)
{
	struct rq *rq = &per_cpu(runqueues, cpu);

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

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

#define P(x) \
	SEQ_printf(m, "  .%-30s: %Ld\n", #x, (long long)(rq->x))
#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);
	SEQ_printf(m, "  .%-30s: %lu\n", "jiffies", jiffies);
	PN(next_balance);
	P(curr->pid);
	PN(clock);
	PN(idle_clock);
	PN(prev_clock_raw);
	P(clock_warps);
	P(clock_overflows);
	P(clock_deep_idle_events);
	PN(clock_max_delta);
	P(cpu_load[0]);
	P(cpu_load[1]);
	P(cpu_load[2]);
	P(cpu_load[3]);
	P(cpu_load[4]);
#undef P
#undef PN

	print_cfs_stats(m, cpu);

	print_rq(m, rq, cpu);
}

static int sched_debug_show(struct seq_file *m, void *v)
{
	u64 now = ktime_to_ns(ktime_get());
	int cpu;

	SEQ_printf(m, "Sched Debug Version: v0.05-v20, %s %.*s\n",
		init_utsname()->release,
		(int)strcspn(init_utsname()->version, " "),
		init_utsname()->version);

	SEQ_printf(m, "now at %Lu.%06ld msecs\n", SPLIT_NS(now));

	for_each_online_cpu(cpu)
		print_cpu(m, cpu);

	SEQ_printf(m, "\n");

	return 0;
}

static void sysrq_sched_debug_show(void)
{
	sched_debug_show(NULL, NULL);
}

static int sched_debug_open(struct inode *inode, struct file *filp)
{
	return single_open(filp, sched_debug_show, NULL);
}

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

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

	pe = create_proc_entry("sched_debug", 0644, NULL);
	if (!pe)
		return -ENOMEM;

	pe->proc_fops = &sched_debug_fops;

	return 0;
}

__initcall(init_sched_debug_procfs);

void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
{
	unsigned long flags;
	int num_threads = 1;

	rcu_read_lock();
	if (lock_task_sighand(p, &flags)) {
		num_threads = atomic_read(&p->signal->count);
		unlock_task_sighand(p, &flags);
	}
	rcu_read_unlock();

	SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, p->pid, num_threads);
	SEQ_printf(m, "----------------------------------------------\n");
#define P(F) \
	SEQ_printf(m, "%-25s:%20Ld\n", #F, (long long)p->F)
#define PN(F) \
	SEQ_printf(m, "%-25s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)p->F))

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

#ifdef CONFIG_SCHEDSTATS
	PN(se.wait_start);
	PN(se.sleep_start);
	PN(se.block_start);
	PN(se.sleep_max);
	PN(se.block_max);
	PN(se.exec_max);
	PN(se.slice_max);
	PN(se.wait_max);
#endif
	SEQ_printf(m, "%-25s:%20Ld\n",
		   "nr_switches", (long long)(p->nvcsw + p->nivcsw));
	P(se.load.weight);
	P(policy);
	P(prio);
#undef P
#undef PN

	{
		u64 t0, t1;

		t0 = sched_clock();
		t1 = sched_clock();
		SEQ_printf(m, "%-25s:%20Ld\n",
			   "clock-delta", (long long)(t1-t0));
	}
}

void proc_sched_set_task(struct task_struct *p)
{
#ifdef CONFIG_SCHEDSTATS
	p->se.sleep_max			= 0;
	p->se.block_max			= 0;
	p->se.exec_max			= 0;
	p->se.slice_max			= 0;
	p->se.wait_max			= 0;
#endif
	p->se.sum_exec_runtime		= 0;
	p->se.prev_sum_exec_runtime	= 0;
}
Commit	Line	Data
43ae34cb IM	1	/*
	2	* kernel/time/sched_debug.c
	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
	19	/*
	20	* This allows printing both to /proc/sched_debug and
	21	* to the console
	22	*/
	23	#define SEQ_printf(m, x...) \
	24	do { \
	25	if (m) \
	26	seq_printf(m, x); \
	27	else \
	28	printk(x); \
	29	} while (0)
	30
ef83a571 IM	31	/*
	32	* Ease the printing of nsec fields:
	33	*/
	34	static long long nsec_high(long long nsec)
	35	{
	36	if (nsec < 0) {
	37	nsec = -nsec;
	38	do_div(nsec, 1000000);
	39	return -nsec;
	40	}
	41	do_div(nsec, 1000000);
	42
	43	return nsec;
	44	}
	45
	46	static unsigned long nsec_low(long long nsec)
	47	{
	48	if (nsec < 0)
	49	nsec = -nsec;
	50
	51	return do_div(nsec, 1000000);
	52	}
	53
	54	#define SPLIT_NS(x) nsec_high(x), nsec_low(x)
	55
43ae34cb	56	static void
a48da48b	57	print_task(struct seq_file m, struct rq rq, struct task_struct *p)
43ae34cb IM	58	{
	59	if (rq->curr == p)
	60	SEQ_printf(m, "R");
	61	else
	62	SEQ_printf(m, " ");
	63
ef83a571	64	SEQ_printf(m, "%15s %5d %9Ld.%06ld %9Ld %5d ",
43ae34cb	65	p->comm, p->pid,
ef83a571	66	SPLIT_NS(p->se.vruntime),
43ae34cb	67	(long long)(p->nvcsw + p->nivcsw),
6f605d83	68	p->prio);
6cfb0d5d	69	#ifdef CONFIG_SCHEDSTATS
ef83a571 IM	70	SEQ_printf(m, "%15Ld.%06ld %15Ld.%06ld %15Ld.%06ld\n",
	71	SPLIT_NS(p->se.vruntime),
	72	SPLIT_NS(p->se.sum_exec_runtime),
	73	SPLIT_NS(p->se.sum_sleep_runtime));
6cfb0d5d	74	#else
ef83a571 IM	75	SEQ_printf(m, "%15Ld %15Ld %15Ld.%06ld %15Ld.%06ld %15Ld.%06ld\n",
ef83a571 IM	76	0LL, 0LL, 0LL, 0L, 0LL, 0L, 0LL, 0L);
6cfb0d5d	77	#endif
43ae34cb IM	78	}
43ae34cb IM	79
a48da48b	80	static void print_rq(struct seq_file m, struct rq rq, int rq_cpu)
43ae34cb IM	81	{
	82	struct task_struct g, p;
	83
	84	SEQ_printf(m,
	85	"\nrunnable tasks:\n"
bbdba7c0 IM	86	" task PID tree-key switches prio"
bbdba7c0 IM	87	" exec-runtime sum-exec sum-sleep\n"
1a75b94f IM	88	"------------------------------------------------------"
1a75b94f IM	89	"------------------------------------------------");
43ae34cb IM	90
	91	read_lock_irq(&tasklist_lock);
	92
	93	do_each_thread(g, p) {
	94	if (!p->se.on_rq \|\| task_cpu(p) != rq_cpu)
	95	continue;
	96
a48da48b	97	print_task(m, rq, p);
43ae34cb IM	98	} while_each_thread(g, p);
	99
	100	read_unlock_irq(&tasklist_lock);
	101	}
	102
5cef9eca	103	void print_cfs_rq(struct seq_file m, int cpu, struct cfs_rq cfs_rq)
43ae34cb	104	{
86d9560c IM	105	s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
86d9560c IM	106	spread, rq0_min_vruntime, spread0;
67e12eac IM	107	struct rq *rq = &per_cpu(runqueues, cpu);
	108	struct sched_entity *last;
	109	unsigned long flags;
	110
5167e75f	111	SEQ_printf(m, "\ncfs_rq\n");
43ae34cb	112
ef83a571 IM	113	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "exec_clock",
ef83a571 IM	114	SPLIT_NS(cfs_rq->exec_clock));
67e12eac IM	115
	116	spin_lock_irqsave(&rq->lock, flags);
	117	if (cfs_rq->rb_leftmost)
	118	MIN_vruntime = (__pick_next_entity(cfs_rq))->vruntime;
	119	last = __pick_last_entity(cfs_rq);
	120	if (last)
	121	max_vruntime = last->vruntime;
86d9560c IM	122	min_vruntime = rq->cfs.min_vruntime;
86d9560c IM	123	rq0_min_vruntime = per_cpu(runqueues, 0).cfs.min_vruntime;
67e12eac	124	spin_unlock_irqrestore(&rq->lock, flags);
ef83a571 IM	125	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "MIN_vruntime",
	126	SPLIT_NS(MIN_vruntime));
	127	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "min_vruntime",
	128	SPLIT_NS(min_vruntime));
	129	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "max_vruntime",
	130	SPLIT_NS(max_vruntime));
67e12eac	131	spread = max_vruntime - MIN_vruntime;
ef83a571 IM	132	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "spread",
ef83a571 IM	133	SPLIT_NS(spread));
86d9560c	134	spread0 = min_vruntime - rq0_min_vruntime;
ef83a571 IM	135	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", "spread0",
ef83a571 IM	136	SPLIT_NS(spread0));
43ae34cb IM	137	}
43ae34cb IM	138
a48da48b	139	static void print_cpu(struct seq_file *m, int cpu)
43ae34cb IM	140	{
	141	struct rq *rq = &per_cpu(runqueues, cpu);
	142
	143	#ifdef CONFIG_X86
	144	{
	145	unsigned int freq = cpu_khz ? : 1;
	146
	147	SEQ_printf(m, "\ncpu#%d, %u.%03u MHz\n",
	148	cpu, freq / 1000, (freq % 1000));
	149	}
	150	#else
	151	SEQ_printf(m, "\ncpu#%d\n", cpu);
	152	#endif
	153
	154	#define P(x) \
	155	SEQ_printf(m, " .%-30s: %Ld\n", #x, (long long)(rq->x))
ef83a571 IM	156	#define PN(x) \
ef83a571 IM	157	SEQ_printf(m, " .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rq->x))
43ae34cb IM	158
	159	P(nr_running);
	160	SEQ_printf(m, " .%-30s: %lu\n", "load",
495eca49	161	rq->load.weight);
43ae34cb IM	162	P(nr_switches);
	163	P(nr_load_updates);
	164	P(nr_uninterruptible);
	165	SEQ_printf(m, " .%-30s: %lu\n", "jiffies", jiffies);
ef83a571	166	PN(next_balance);
43ae34cb	167	P(curr->pid);
ef83a571 IM	168	PN(clock);
	169	PN(idle_clock);
	170	PN(prev_clock_raw);
43ae34cb IM	171	P(clock_warps);
43ae34cb IM	172	P(clock_overflows);
2aa44d05	173	P(clock_deep_idle_events);
ef83a571	174	PN(clock_max_delta);
43ae34cb IM	175	P(cpu_load[0]);
	176	P(cpu_load[1]);
	177	P(cpu_load[2]);
	178	P(cpu_load[3]);
	179	P(cpu_load[4]);
	180	#undef P
ef83a571	181	#undef PN
43ae34cb	182
5cef9eca	183	print_cfs_stats(m, cpu);
43ae34cb	184
a48da48b	185	print_rq(m, rq, cpu);
43ae34cb IM	186	}
	187
	188	static int sched_debug_show(struct seq_file m, void v)
	189	{
	190	u64 now = ktime_to_ns(ktime_get());
	191	int cpu;
	192
6cfb0d5d	193	SEQ_printf(m, "Sched Debug Version: v0.05-v20, %s %.*s\n",
43ae34cb IM	194	init_utsname()->release,
	195	(int)strcspn(init_utsname()->version, " "),
	196	init_utsname()->version);
	197
ef83a571	198	SEQ_printf(m, "now at %Lu.%06ld msecs\n", SPLIT_NS(now));
43ae34cb IM	199
43ae34cb IM	200	for_each_online_cpu(cpu)
a48da48b	201	print_cpu(m, cpu);
43ae34cb IM	202
	203	SEQ_printf(m, "\n");
	204
	205	return 0;
	206	}
	207
f3373461	208	static void sysrq_sched_debug_show(void)
43ae34cb IM	209	{
	210	sched_debug_show(NULL, NULL);
	211	}
	212
	213	static int sched_debug_open(struct inode inode, struct file filp)
	214	{
	215	return single_open(filp, sched_debug_show, NULL);
	216	}
	217
	218	static struct file_operations sched_debug_fops = {
	219	.open = sched_debug_open,
	220	.read = seq_read,
	221	.llseek = seq_lseek,
5ea473a1	222	.release = single_release,
43ae34cb IM	223	};
	224
	225	static int __init init_sched_debug_procfs(void)
	226	{
	227	struct proc_dir_entry *pe;
	228
	229	pe = create_proc_entry("sched_debug", 0644, NULL);
	230	if (!pe)
	231	return -ENOMEM;
	232
	233	pe->proc_fops = &sched_debug_fops;
	234
	235	return 0;
	236	}
	237
	238	__initcall(init_sched_debug_procfs);
	239
	240	void proc_sched_show_task(struct task_struct p, struct seq_file m)
	241	{
	242	unsigned long flags;
	243	int num_threads = 1;
	244
	245	rcu_read_lock();
	246	if (lock_task_sighand(p, &flags)) {
	247	num_threads = atomic_read(&p->signal->count);
	248	unlock_task_sighand(p, &flags);
	249	}
	250	rcu_read_unlock();
	251
	252	SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, p->pid, num_threads);
	253	SEQ_printf(m, "----------------------------------------------\n");
	254	#define P(F) \
	255	SEQ_printf(m, "%-25s:%20Ld\n", #F, (long long)p->F)
ef83a571 IM	256	#define PN(F) \
ef83a571 IM	257	SEQ_printf(m, "%-25s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)p->F))
43ae34cb	258
ef83a571 IM	259	PN(se.exec_start);
	260	PN(se.vruntime);
	261	PN(se.sum_exec_runtime);
6cfb0d5d IM	262
6cfb0d5d IM	263	#ifdef CONFIG_SCHEDSTATS
ef83a571 IM	264	PN(se.wait_start);
	265	PN(se.sleep_start);
	266	PN(se.block_start);
	267	PN(se.sleep_max);
	268	PN(se.block_max);
	269	PN(se.exec_max);
	270	PN(se.slice_max);
	271	PN(se.wait_max);
6cfb0d5d	272	#endif
43ae34cb IM	273	SEQ_printf(m, "%-25s:%20Ld\n",
	274	"nr_switches", (long long)(p->nvcsw + p->nivcsw));
	275	P(se.load.weight);
	276	P(policy);
	277	P(prio);
	278	#undef P
ef83a571	279	#undef PN
43ae34cb IM	280
	281	{
	282	u64 t0, t1;
	283
	284	t0 = sched_clock();
	285	t1 = sched_clock();
	286	SEQ_printf(m, "%-25s:%20Ld\n",
	287	"clock-delta", (long long)(t1-t0));
	288	}
	289	}
	290
	291	void proc_sched_set_task(struct task_struct *p)
	292	{
6cfb0d5d	293	#ifdef CONFIG_SCHEDSTATS
a4b29ba2 IM	294	p->se.sleep_max = 0;
	295	p->se.block_max = 0;
	296	p->se.exec_max = 0;
eba1ed4b	297	p->se.slice_max = 0;
a4b29ba2	298	p->se.wait_max = 0;
6cfb0d5d	299	#endif
a4b29ba2	300	p->se.sum_exec_runtime = 0;
2491b2b8	301	p->se.prev_sum_exec_runtime = 0;
43ae34cb	302	}