debug.c 17.8 KB
Newer Older
I
Ingo Molnar 已提交
1
/*
2
 * kernel/sched/debug.c
I
Ingo Molnar 已提交
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
 *
 * 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>
I
Ingo Molnar 已提交
18
#include <linux/mempolicy.h>
19
#include <linux/debugfs.h>
I
Ingo Molnar 已提交
20

21 22
#include "sched.h"

23 24
static DEFINE_SPINLOCK(sched_debug_lock);

I
Ingo Molnar 已提交
25 26 27 28 29 30 31 32 33 34 35 36
/*
 * 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)

I
Ingo Molnar 已提交
37 38 39
/*
 * Ease the printing of nsec fields:
 */
I
Ingo Molnar 已提交
40
static long long nsec_high(unsigned long long nsec)
I
Ingo Molnar 已提交
41
{
I
Ingo Molnar 已提交
42
	if ((long long)nsec < 0) {
I
Ingo Molnar 已提交
43 44 45 46 47 48 49 50 51
		nsec = -nsec;
		do_div(nsec, 1000000);
		return -nsec;
	}
	do_div(nsec, 1000000);

	return nsec;
}

I
Ingo Molnar 已提交
52
static unsigned long nsec_low(unsigned long long nsec)
I
Ingo Molnar 已提交
53
{
I
Ingo Molnar 已提交
54
	if ((long long)nsec < 0)
I
Ingo Molnar 已提交
55 56 57 58 59 60 61
		nsec = -nsec;

	return do_div(nsec, 1000000);
}

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

62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191
#define SCHED_FEAT(name, enabled)	\
	#name ,

static const char * const sched_feat_names[] = {
#include "features.h"
};

#undef SCHED_FEAT

static int sched_feat_show(struct seq_file *m, void *v)
{
	int i;

	for (i = 0; i < __SCHED_FEAT_NR; i++) {
		if (!(sysctl_sched_features & (1UL << i)))
			seq_puts(m, "NO_");
		seq_printf(m, "%s ", sched_feat_names[i]);
	}
	seq_puts(m, "\n");

	return 0;
}

#ifdef HAVE_JUMP_LABEL

#define jump_label_key__true  STATIC_KEY_INIT_TRUE
#define jump_label_key__false STATIC_KEY_INIT_FALSE

#define SCHED_FEAT(name, enabled)	\
	jump_label_key__##enabled ,

struct static_key sched_feat_keys[__SCHED_FEAT_NR] = {
#include "features.h"
};

#undef SCHED_FEAT

static void sched_feat_disable(int i)
{
	static_key_disable(&sched_feat_keys[i]);
}

static void sched_feat_enable(int i)
{
	static_key_enable(&sched_feat_keys[i]);
}
#else
static void sched_feat_disable(int i) { };
static void sched_feat_enable(int i) { };
#endif /* HAVE_JUMP_LABEL */

static int sched_feat_set(char *cmp)
{
	int i;
	int neg = 0;

	if (strncmp(cmp, "NO_", 3) == 0) {
		neg = 1;
		cmp += 3;
	}

	for (i = 0; i < __SCHED_FEAT_NR; i++) {
		if (strcmp(cmp, sched_feat_names[i]) == 0) {
			if (neg) {
				sysctl_sched_features &= ~(1UL << i);
				sched_feat_disable(i);
			} else {
				sysctl_sched_features |= (1UL << i);
				sched_feat_enable(i);
			}
			break;
		}
	}

	return i;
}

static ssize_t
sched_feat_write(struct file *filp, const char __user *ubuf,
		size_t cnt, loff_t *ppos)
{
	char buf[64];
	char *cmp;
	int i;
	struct inode *inode;

	if (cnt > 63)
		cnt = 63;

	if (copy_from_user(&buf, ubuf, cnt))
		return -EFAULT;

	buf[cnt] = 0;
	cmp = strstrip(buf);

	/* Ensure the static_key remains in a consistent state */
	inode = file_inode(filp);
	inode_lock(inode);
	i = sched_feat_set(cmp);
	inode_unlock(inode);
	if (i == __SCHED_FEAT_NR)
		return -EINVAL;

	*ppos += cnt;

	return cnt;
}

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

static const struct file_operations sched_feat_fops = {
	.open		= sched_feat_open,
	.write		= sched_feat_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= single_release,
};

static __init int sched_init_debug(void)
{
	debugfs_create_file("sched_features", 0644, NULL, NULL,
			&sched_feat_fops);

	return 0;
}
late_initcall(sched_init_debug);

192
#ifdef CONFIG_FAIR_GROUP_SCHED
193
static void print_cfs_group_stats(struct seq_file *m, int cpu, struct task_group *tg)
194 195 196 197 198 199 200 201
{
	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))

Y
Yuyang Du 已提交
202
	if (!se)
203 204
		return;

205 206 207 208
	PN(se->exec_start);
	PN(se->vruntime);
	PN(se->sum_exec_runtime);
#ifdef CONFIG_SCHEDSTATS
209 210 211 212 213 214 215 216 217 218 219 220
	if (schedstat_enabled()) {
		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);
	}
221 222
#endif
	P(se->load.weight);
223
#ifdef CONFIG_SMP
224 225
	P(se->avg.load_avg);
	P(se->avg.util_avg);
226
#endif
227 228 229 230 231
#undef PN
#undef P
}
#endif

232 233 234 235 236
#ifdef CONFIG_CGROUP_SCHED
static char group_path[PATH_MAX];

static char *task_group_path(struct task_group *tg)
{
237 238 239
	if (autogroup_path(tg, group_path, PATH_MAX))
		return group_path;

T
Tejun Heo 已提交
240
	return cgroup_path(tg->css.cgroup, group_path, PATH_MAX);
241 242 243
}
#endif

I
Ingo Molnar 已提交
244
static void
245
print_task(struct seq_file *m, struct rq *rq, struct task_struct *p)
I
Ingo Molnar 已提交
246 247 248 249 250 251
{
	if (rq->curr == p)
		SEQ_printf(m, "R");
	else
		SEQ_printf(m, " ");

I
Ingo Molnar 已提交
252
	SEQ_printf(m, "%15s %5d %9Ld.%06ld %9Ld %5d ",
253
		p->comm, task_pid_nr(p),
I
Ingo Molnar 已提交
254
		SPLIT_NS(p->se.vruntime),
I
Ingo Molnar 已提交
255
		(long long)(p->nvcsw + p->nivcsw),
256
		p->prio);
I
Ingo Molnar 已提交
257
#ifdef CONFIG_SCHEDSTATS
258 259 260 261 262 263
	if (schedstat_enabled()) {
		SEQ_printf(m, "%9Ld.%06ld %9Ld.%06ld %9Ld.%06ld",
			SPLIT_NS(p->se.statistics.wait_sum),
			SPLIT_NS(p->se.sum_exec_runtime),
			SPLIT_NS(p->se.statistics.sum_sleep_runtime));
	}
I
Ingo Molnar 已提交
264
#else
265 266 267 268
	SEQ_printf(m, "%9Ld.%06ld %9Ld.%06ld %9Ld.%06ld",
		0LL, 0L,
		SPLIT_NS(p->se.sum_exec_runtime),
		0LL, 0L);
I
Ingo Molnar 已提交
269
#endif
I
Ingo Molnar 已提交
270
#ifdef CONFIG_NUMA_BALANCING
271
	SEQ_printf(m, " %d %d", task_node(p), task_numa_group_id(p));
I
Ingo Molnar 已提交
272
#endif
273 274 275
#ifdef CONFIG_CGROUP_SCHED
	SEQ_printf(m, " %s", task_group_path(task_group(p)));
#endif
276 277

	SEQ_printf(m, "\n");
I
Ingo Molnar 已提交
278 279
}

280
static void print_rq(struct seq_file *m, struct rq *rq, int rq_cpu)
I
Ingo Molnar 已提交
281 282 283 284 285
{
	struct task_struct *g, *p;

	SEQ_printf(m,
	"\nrunnable tasks:\n"
286
	"            task   PID         tree-key  switches  prio"
287
	"     wait-time             sum-exec        sum-sleep\n"
288
	"------------------------------------------------------"
289
	"----------------------------------------------------\n");
I
Ingo Molnar 已提交
290

291
	rcu_read_lock();
292
	for_each_process_thread(g, p) {
I
Ingo Molnar 已提交
293
		if (task_cpu(p) != rq_cpu)
I
Ingo Molnar 已提交
294 295
			continue;

296
		print_task(m, rq, p);
297
	}
298
	rcu_read_unlock();
I
Ingo Molnar 已提交
299 300
}

301
void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
I
Ingo Molnar 已提交
302
{
I
Ingo Molnar 已提交
303 304
	s64 MIN_vruntime = -1, min_vruntime, max_vruntime = -1,
		spread, rq0_min_vruntime, spread0;
305
	struct rq *rq = cpu_rq(cpu);
I
Ingo Molnar 已提交
306 307 308
	struct sched_entity *last;
	unsigned long flags;

309 310 311
#ifdef CONFIG_FAIR_GROUP_SCHED
	SEQ_printf(m, "\ncfs_rq[%d]:%s\n", cpu, task_group_path(cfs_rq->tg));
#else
312
	SEQ_printf(m, "\ncfs_rq[%d]:\n", cpu);
313
#endif
I
Ingo Molnar 已提交
314 315
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "exec_clock",
			SPLIT_NS(cfs_rq->exec_clock));
I
Ingo Molnar 已提交
316

317
	raw_spin_lock_irqsave(&rq->lock, flags);
I
Ingo Molnar 已提交
318
	if (cfs_rq->rb_leftmost)
319
		MIN_vruntime = (__pick_first_entity(cfs_rq))->vruntime;
I
Ingo Molnar 已提交
320 321 322
	last = __pick_last_entity(cfs_rq);
	if (last)
		max_vruntime = last->vruntime;
P
Peter Zijlstra 已提交
323
	min_vruntime = cfs_rq->min_vruntime;
324
	rq0_min_vruntime = cpu_rq(0)->cfs.min_vruntime;
325
	raw_spin_unlock_irqrestore(&rq->lock, flags);
I
Ingo Molnar 已提交
326 327 328 329 330 331
	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));
I
Ingo Molnar 已提交
332
	spread = max_vruntime - MIN_vruntime;
I
Ingo Molnar 已提交
333 334
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "spread",
			SPLIT_NS(spread));
I
Ingo Molnar 已提交
335
	spread0 = min_vruntime - rq0_min_vruntime;
I
Ingo Molnar 已提交
336 337
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", "spread0",
			SPLIT_NS(spread0));
P
Peter Zijlstra 已提交
338
	SEQ_printf(m, "  .%-30s: %d\n", "nr_spread_over",
P
Peter Zijlstra 已提交
339
			cfs_rq->nr_spread_over);
340
	SEQ_printf(m, "  .%-30s: %d\n", "nr_running", cfs_rq->nr_running);
P
Peter Zijlstra 已提交
341
	SEQ_printf(m, "  .%-30s: %ld\n", "load", cfs_rq->load.weight);
342
#ifdef CONFIG_SMP
343 344
	SEQ_printf(m, "  .%-30s: %lu\n", "load_avg",
			cfs_rq->avg.load_avg);
345 346
	SEQ_printf(m, "  .%-30s: %lu\n", "runnable_load_avg",
			cfs_rq->runnable_load_avg);
347 348 349 350 351 352
	SEQ_printf(m, "  .%-30s: %lu\n", "util_avg",
			cfs_rq->avg.util_avg);
	SEQ_printf(m, "  .%-30s: %ld\n", "removed_load_avg",
			atomic_long_read(&cfs_rq->removed_load_avg));
	SEQ_printf(m, "  .%-30s: %ld\n", "removed_util_avg",
			atomic_long_read(&cfs_rq->removed_util_avg));
353
#ifdef CONFIG_FAIR_GROUP_SCHED
354 355
	SEQ_printf(m, "  .%-30s: %lu\n", "tg_load_avg_contrib",
			cfs_rq->tg_load_avg_contrib);
356 357
	SEQ_printf(m, "  .%-30s: %ld\n", "tg_load_avg",
			atomic_long_read(&cfs_rq->tg->load_avg));
358
#endif
359
#endif
360 361 362 363 364 365
#ifdef CONFIG_CFS_BANDWIDTH
	SEQ_printf(m, "  .%-30s: %d\n", "throttled",
			cfs_rq->throttled);
	SEQ_printf(m, "  .%-30s: %d\n", "throttle_count",
			cfs_rq->throttle_count);
#endif
P
Peter Zijlstra 已提交
366

367
#ifdef CONFIG_FAIR_GROUP_SCHED
368
	print_cfs_group_stats(m, cpu, cfs_rq->tg);
369
#endif
I
Ingo Molnar 已提交
370 371
}

372 373
void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq)
{
374 375 376
#ifdef CONFIG_RT_GROUP_SCHED
	SEQ_printf(m, "\nrt_rq[%d]:%s\n", cpu, task_group_path(rt_rq->tg));
#else
377
	SEQ_printf(m, "\nrt_rq[%d]:\n", cpu);
378
#endif
379 380 381 382 383 384 385 386 387 388 389 390 391 392 393

#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
}

394 395 396 397 398 399
void print_dl_rq(struct seq_file *m, int cpu, struct dl_rq *dl_rq)
{
	SEQ_printf(m, "\ndl_rq[%d]:\n", cpu);
	SEQ_printf(m, "  .%-30s: %ld\n", "dl_nr_running", dl_rq->dl_nr_running);
}

400 401
extern __read_mostly int sched_clock_running;

402
static void print_cpu(struct seq_file *m, int cpu)
I
Ingo Molnar 已提交
403
{
404
	struct rq *rq = cpu_rq(cpu);
405
	unsigned long flags;
I
Ingo Molnar 已提交
406 407 408 409 410

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

411
		SEQ_printf(m, "cpu#%d, %u.%03u MHz\n",
I
Ingo Molnar 已提交
412 413 414
			   cpu, freq / 1000, (freq % 1000));
	}
#else
415
	SEQ_printf(m, "cpu#%d\n", cpu);
I
Ingo Molnar 已提交
416 417
#endif

418 419 420 421 422 423 424 425
#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)

I
Ingo Molnar 已提交
426 427
#define PN(x) \
	SEQ_printf(m, "  .%-30s: %Ld.%06ld\n", #x, SPLIT_NS(rq->x))
I
Ingo Molnar 已提交
428 429 430

	P(nr_running);
	SEQ_printf(m, "  .%-30s: %lu\n", "load",
431
		   rq->load.weight);
I
Ingo Molnar 已提交
432 433 434
	P(nr_switches);
	P(nr_load_updates);
	P(nr_uninterruptible);
I
Ingo Molnar 已提交
435
	PN(next_balance);
436
	SEQ_printf(m, "  .%-30s: %ld\n", "curr->pid", (long)(task_pid_nr(rq->curr)));
I
Ingo Molnar 已提交
437
	PN(clock);
438
	PN(clock_task);
I
Ingo Molnar 已提交
439 440 441 442 443 444
	P(cpu_load[0]);
	P(cpu_load[1]);
	P(cpu_load[2]);
	P(cpu_load[3]);
	P(cpu_load[4]);
#undef P
I
Ingo Molnar 已提交
445
#undef PN
I
Ingo Molnar 已提交
446

P
Peter Zijlstra 已提交
447 448
#ifdef CONFIG_SCHEDSTATS
#define P(n) SEQ_printf(m, "  .%-30s: %d\n", #n, rq->n);
M
Mike Galbraith 已提交
449
#define P64(n) SEQ_printf(m, "  .%-30s: %Ld\n", #n, rq->n);
P
Peter Zijlstra 已提交
450

M
Mike Galbraith 已提交
451 452
#ifdef CONFIG_SMP
	P64(avg_idle);
453
	P64(max_idle_balance_cost);
M
Mike Galbraith 已提交
454
#endif
P
Peter Zijlstra 已提交
455

456 457 458 459 460 461 462
	if (schedstat_enabled()) {
		P(yld_count);
		P(sched_count);
		P(sched_goidle);
		P(ttwu_count);
		P(ttwu_local);
	}
P
Peter Zijlstra 已提交
463 464

#undef P
465
#undef P64
P
Peter Zijlstra 已提交
466
#endif
467
	spin_lock_irqsave(&sched_debug_lock, flags);
468
	print_cfs_stats(m, cpu);
469
	print_rt_stats(m, cpu);
470
	print_dl_stats(m, cpu);
I
Ingo Molnar 已提交
471

472
	print_rq(m, rq, cpu);
473
	spin_unlock_irqrestore(&sched_debug_lock, flags);
474
	SEQ_printf(m, "\n");
I
Ingo Molnar 已提交
475 476
}

477 478 479 480 481 482
static const char *sched_tunable_scaling_names[] = {
	"none",
	"logaritmic",
	"linear"
};

483
static void sched_debug_header(struct seq_file *m)
I
Ingo Molnar 已提交
484
{
485 486
	u64 ktime, sched_clk, cpu_clk;
	unsigned long flags;
I
Ingo Molnar 已提交
487

488 489 490 491 492 493
	local_irq_save(flags);
	ktime = ktime_to_ns(ktime_get());
	sched_clk = sched_clock();
	cpu_clk = local_clock();
	local_irq_restore(flags);

I
Ingo Molnar 已提交
494
	SEQ_printf(m, "Sched Debug Version: v0.11, %s %.*s\n",
I
Ingo Molnar 已提交
495 496 497 498
		init_utsname()->release,
		(int)strcspn(init_utsname()->version, " "),
		init_utsname()->version);

499 500 501 502 503 504 505 506 507
#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
508
	P(sched_clock_stable());
509 510 511 512 513 514
#endif
#undef PN
#undef P

	SEQ_printf(m, "\n");
	SEQ_printf(m, "sysctl_sched\n");
I
Ingo Molnar 已提交
515

I
Ingo Molnar 已提交
516
#define P(x) \
517
	SEQ_printf(m, "  .%-40s: %Ld\n", #x, (long long)(x))
I
Ingo Molnar 已提交
518
#define PN(x) \
519
	SEQ_printf(m, "  .%-40s: %Ld.%06ld\n", #x, SPLIT_NS(x))
I
Ingo Molnar 已提交
520
	PN(sysctl_sched_latency);
521
	PN(sysctl_sched_min_granularity);
I
Ingo Molnar 已提交
522
	PN(sysctl_sched_wakeup_granularity);
523
	P(sysctl_sched_child_runs_first);
I
Ingo Molnar 已提交
524 525 526 527
	P(sysctl_sched_features);
#undef PN
#undef P

528 529
	SEQ_printf(m, "  .%-40s: %d (%s)\n",
		"sysctl_sched_tunable_scaling",
530 531
		sysctl_sched_tunable_scaling,
		sched_tunable_scaling_names[sysctl_sched_tunable_scaling]);
532 533
	SEQ_printf(m, "\n");
}
534

535 536 537
static int sched_debug_show(struct seq_file *m, void *v)
{
	int cpu = (unsigned long)(v - 2);
I
Ingo Molnar 已提交
538

539 540 541 542
	if (cpu != -1)
		print_cpu(m, cpu);
	else
		sched_debug_header(m);
I
Ingo Molnar 已提交
543 544 545 546

	return 0;
}

547
void sysrq_sched_debug_show(void)
I
Ingo Molnar 已提交
548
{
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606
	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;
I
Ingo Molnar 已提交
607 608 609 610
}

static int sched_debug_open(struct inode *inode, struct file *filp)
{
611 612 613 614 615
	int ret = 0;

	ret = seq_open(filp, &sched_debug_sops);

	return ret;
I
Ingo Molnar 已提交
616 617
}

618
static const struct file_operations sched_debug_fops = {
I
Ingo Molnar 已提交
619 620 621
	.open		= sched_debug_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
622
	.release	= sched_debug_release,
I
Ingo Molnar 已提交
623 624 625 626 627 628
};

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

629
	pe = proc_create("sched_debug", 0444, NULL, &sched_debug_fops);
I
Ingo Molnar 已提交
630 631 632 633 634 635 636
	if (!pe)
		return -ENOMEM;
	return 0;
}

__initcall(init_sched_debug_procfs);

I
Ingo Molnar 已提交
637 638 639 640 641 642 643 644 645 646
#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))


647 648 649 650 651 652 653 654 655 656 657
#ifdef CONFIG_NUMA_BALANCING
void print_numa_stats(struct seq_file *m, int node, unsigned long tsf,
		unsigned long tpf, unsigned long gsf, unsigned long gpf)
{
	SEQ_printf(m, "numa_faults node=%d ", node);
	SEQ_printf(m, "task_private=%lu task_shared=%lu ", tsf, tpf);
	SEQ_printf(m, "group_private=%lu group_shared=%lu\n", gsf, gpf);
}
#endif


I
Ingo Molnar 已提交
658 659 660 661 662 663 664 665 666 667 668 669 670 671 672
static void sched_show_numa(struct task_struct *p, struct seq_file *m)
{
#ifdef CONFIG_NUMA_BALANCING
	struct mempolicy *pol;

	if (p->mm)
		P(mm->numa_scan_seq);

	task_lock(p);
	pol = p->mempolicy;
	if (pol && !(pol->flags & MPOL_F_MORON))
		pol = NULL;
	mpol_get(pol);
	task_unlock(p);

673 674 675 676 677 678
	P(numa_pages_migrated);
	P(numa_preferred_nid);
	P(total_numa_faults);
	SEQ_printf(m, "current_node=%d, numa_group_id=%d\n",
			task_node(p), task_numa_group_id(p));
	show_numa_stats(p, m);
I
Ingo Molnar 已提交
679 680 681 682
	mpol_put(pol);
#endif
}

I
Ingo Molnar 已提交
683 684
void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
{
685
	unsigned long nr_switches;
I
Ingo Molnar 已提交
686

687
	SEQ_printf(m, "%s (%d, #threads: %d)\n", p->comm, task_pid_nr(p),
688
						get_nr_threads(p));
689
	SEQ_printf(m,
690 691
		"---------------------------------------------------------"
		"----------\n");
692
#define __P(F) \
693
	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)F)
I
Ingo Molnar 已提交
694
#define P(F) \
695
	SEQ_printf(m, "%-45s:%21Ld\n", #F, (long long)p->F)
696
#define __PN(F) \
697
	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)F))
I
Ingo Molnar 已提交
698
#define PN(F) \
699
	SEQ_printf(m, "%-45s:%14Ld.%06ld\n", #F, SPLIT_NS((long long)p->F))
I
Ingo Molnar 已提交
700

I
Ingo Molnar 已提交
701 702 703
	PN(se.exec_start);
	PN(se.vruntime);
	PN(se.sum_exec_runtime);
I
Ingo Molnar 已提交
704

705 706
	nr_switches = p->nvcsw + p->nivcsw;

I
Ingo Molnar 已提交
707
#ifdef CONFIG_SCHEDSTATS
708 709
	P(se.nr_migrations);

710
	if (schedstat_enabled()) {
711 712
		u64 avg_atom, avg_per_cpu;

713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
		PN(se.statistics.sum_sleep_runtime);
		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.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);

741 742
		avg_atom = p->se.sum_exec_runtime;
		if (nr_switches)
743
			avg_atom = div64_ul(avg_atom, nr_switches);
744 745 746 747
		else
			avg_atom = -1LL;

		avg_per_cpu = p->se.sum_exec_runtime;
748
		if (p->se.nr_migrations) {
R
Roman Zippel 已提交
749 750
			avg_per_cpu = div64_u64(avg_per_cpu,
						p->se.nr_migrations);
751
		} else {
752
			avg_per_cpu = -1LL;
753
		}
754 755 756 757

		__PN(avg_atom);
		__PN(avg_per_cpu);
	}
I
Ingo Molnar 已提交
758
#endif
759
	__P(nr_switches);
760
	SEQ_printf(m, "%-45s:%21Ld\n",
761
		   "nr_voluntary_switches", (long long)p->nvcsw);
762
	SEQ_printf(m, "%-45s:%21Ld\n",
763 764
		   "nr_involuntary_switches", (long long)p->nivcsw);

I
Ingo Molnar 已提交
765
	P(se.load.weight);
766
#ifdef CONFIG_SMP
767 768 769 770 771
	P(se.avg.load_sum);
	P(se.avg.util_sum);
	P(se.avg.load_avg);
	P(se.avg.util_avg);
	P(se.avg.last_update_time);
772
#endif
I
Ingo Molnar 已提交
773 774
	P(policy);
	P(prio);
I
Ingo Molnar 已提交
775
#undef PN
776 777 778
#undef __PN
#undef P
#undef __P
I
Ingo Molnar 已提交
779 780

	{
781
		unsigned int this_cpu = raw_smp_processor_id();
I
Ingo Molnar 已提交
782 783
		u64 t0, t1;

784 785
		t0 = cpu_clock(this_cpu);
		t1 = cpu_clock(this_cpu);
786
		SEQ_printf(m, "%-45s:%21Ld\n",
I
Ingo Molnar 已提交
787 788
			   "clock-delta", (long long)(t1-t0));
	}
I
Ingo Molnar 已提交
789 790

	sched_show_numa(p, m);
I
Ingo Molnar 已提交
791 792 793 794
}

void proc_sched_set_task(struct task_struct *p)
{
I
Ingo Molnar 已提交
795
#ifdef CONFIG_SCHEDSTATS
796
	memset(&p->se.statistics, 0, sizeof(p->se.statistics));
I
Ingo Molnar 已提交
797
#endif
I
Ingo Molnar 已提交
798
}