summaryrefslogtreecommitdiffstats
path: root/kernel/rcu/rcuperf.c
blob: d38ab08a3fe7eb5d92caf5b0c4d590299d912d90 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
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
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
/*
 * Read-Copy Update module-based performance-test facility
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, you can access it online at
 * http://www.gnu.org/licenses/gpl-2.0.html.
 *
 * Copyright (C) IBM Corporation, 2015
 *
 * Authors: Paul E. McKenney <paulmck@us.ibm.com>
 */
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/kthread.h>
#include <linux/err.h>
#include <linux/spinlock.h>
#include <linux/smp.h>
#include <linux/rcupdate.h>
#include <linux/interrupt.h>
#include <linux/sched.h>
#include <linux/atomic.h>
#include <linux/bitops.h>
#include <linux/completion.h>
#include <linux/moduleparam.h>
#include <linux/percpu.h>
#include <linux/notifier.h>
#include <linux/reboot.h>
#include <linux/freezer.h>
#include <linux/cpu.h>
#include <linux/delay.h>
#include <linux/stat.h>
#include <linux/srcu.h>
#include <linux/slab.h>
#include <asm/byteorder.h>
#include <linux/torture.h>
#include <linux/vmalloc.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Paul E. McKenney <paulmck@linux.vnet.ibm.com>");

#define PERF_FLAG "-perf:"
#define PERFOUT_STRING(s) \
	pr_alert("%s" PERF_FLAG s "\n", perf_type)
#define VERBOSE_PERFOUT_STRING(s) \
	do { if (verbose) pr_alert("%s" PERF_FLAG " %s\n", perf_type, s); } while (0)
#define VERBOSE_PERFOUT_ERRSTRING(s) \
	do { if (verbose) pr_alert("%s" PERF_FLAG "!!! %s\n", perf_type, s); } while (0)

torture_param(bool, gp_exp, false, "Use expedited GP wait primitives");
torture_param(int, holdoff, 10, "Holdoff time before test start (s)");
torture_param(int, nreaders, -1, "Number of RCU reader threads");
torture_param(int, nwriters, -1, "Number of RCU updater threads");
torture_param(bool, shutdown, false, "Shutdown at end of performance tests.");
torture_param(bool, verbose, true, "Enable verbose debugging printk()s");

static char *perf_type = "rcu";
module_param(perf_type, charp, 0444);
MODULE_PARM_DESC(perf_type, "Type of RCU to performance-test (rcu, rcu_bh, ...)");

static int nrealreaders;
static int nrealwriters;
static struct task_struct **writer_tasks;
static struct task_struct **reader_tasks;
static struct task_struct *shutdown_task;

static u64 **writer_durations;
static int *writer_n_durations;
static atomic_t n_rcu_perf_reader_started;
static atomic_t n_rcu_perf_writer_started;
static atomic_t n_rcu_perf_writer_finished;
static wait_queue_head_t shutdown_wq;
static u64 t_rcu_perf_writer_started;
static u64 t_rcu_perf_writer_finished;
static unsigned long b_rcu_perf_writer_started;
static unsigned long b_rcu_perf_writer_finished;

static int rcu_perf_writer_state;
#define RTWS_INIT		0
#define RTWS_EXP_SYNC		1
#define RTWS_SYNC		2
#define RTWS_IDLE		2
#define RTWS_STOPPING		3

#define MAX_MEAS 10000
#define MIN_MEAS 100

static int perf_runnable = IS_ENABLED(MODULE);
module_param(perf_runnable, int, 0444);
MODULE_PARM_DESC(perf_runnable, "Start rcuperf at boot");

/*
 * Operations vector for selecting different types of tests.
 */

struct rcu_perf_ops {
	int ptype;
	void (*init)(void);
	void (*cleanup)(void);
	int (*readlock)(void);
	void (*readunlock)(int idx);
	unsigned long (*started)(void);
	unsigned long (*completed)(void);
	unsigned long (*exp_completed)(void);
	void (*sync)(void);
	void (*exp_sync)(void);
	const char *name;
};

static struct rcu_perf_ops *cur_ops;

/*
 * Definitions for rcu perf testing.
 */

static int rcu_perf_read_lock(void) __acquires(RCU)
{
	rcu_read_lock();
	return 0;
}

static void rcu_perf_read_unlock(int idx) __releases(RCU)
{
	rcu_read_unlock();
}

static unsigned long __maybe_unused rcu_no_completed(void)
{
	return 0;
}

static void rcu_sync_perf_init(void)
{
}

static struct rcu_perf_ops rcu_ops = {
	.ptype		= RCU_FLAVOR,
	.init		= rcu_sync_perf_init,
	.readlock	= rcu_perf_read_lock,
	.readunlock	= rcu_perf_read_unlock,
	.started	= rcu_batches_started,
	.completed	= rcu_batches_completed,
	.exp_completed	= rcu_exp_batches_completed,
	.sync		= synchronize_rcu,
	.exp_sync	= synchronize_rcu_expedited,
	.name		= "rcu"
};

/*
 * Definitions for rcu_bh perf testing.
 */

static int rcu_bh_perf_read_lock(void) __acquires(RCU_BH)
{
	rcu_read_lock_bh();
	return 0;
}

static void rcu_bh_perf_read_unlock(int idx) __releases(RCU_BH)
{
	rcu_read_unlock_bh();
}

static struct rcu_perf_ops rcu_bh_ops = {
	.ptype		= RCU_BH_FLAVOR,
	.init		= rcu_sync_perf_init,
	.readlock	= rcu_bh_perf_read_lock,
	.readunlock	= rcu_bh_perf_read_unlock,
	.started	= rcu_batches_started_bh,
	.completed	= rcu_batches_completed_bh,
	.exp_completed	= rcu_exp_batches_completed_sched,
	.sync		= synchronize_rcu_bh,
	.exp_sync	= synchronize_rcu_bh_expedited,
	.name		= "rcu_bh"
};

/*
 * Definitions for srcu perf testing.
 */

DEFINE_STATIC_SRCU(srcu_ctl_perf);
static struct srcu_struct *srcu_ctlp = &srcu_ctl_perf;

static int srcu_perf_read_lock(void) __acquires(srcu_ctlp)
{
	return srcu_read_lock(srcu_ctlp);
}

static void srcu_perf_read_unlock(int idx) __releases(srcu_ctlp)
{
	srcu_read_unlock(srcu_ctlp, idx);
}

static unsigned long srcu_perf_completed(void)
{
	return srcu_batches_completed(srcu_ctlp);
}

static void srcu_perf_synchronize(void)
{
	synchronize_srcu(srcu_ctlp);
}

static void srcu_perf_synchronize_expedited(void)
{
	synchronize_srcu_expedited(srcu_ctlp);
}

static struct rcu_perf_ops srcu_ops = {
	.ptype		= SRCU_FLAVOR,
	.init		= rcu_sync_perf_init,
	.readlock	= srcu_perf_read_lock,
	.readunlock	= srcu_perf_read_unlock,
	.started	= NULL,
	.completed	= srcu_perf_completed,
	.exp_completed	= srcu_perf_completed,
	.sync		= srcu_perf_synchronize,
	.exp_sync	= srcu_perf_synchronize_expedited,
	.name		= "srcu"
};

/*
 * Definitions for sched perf testing.
 */

static int sched_perf_read_lock(void)
{
	preempt_disable();
	return 0;
}

static void sched_perf_read_unlock(int idx)
{
	preempt_enable();
}

static struct rcu_perf_ops sched_ops = {
	.ptype		= RCU_SCHED_FLAVOR,
	.init		= rcu_sync_perf_init,
	.readlock	= sched_perf_read_lock,
	.readunlock	= sched_perf_read_unlock,
	.started	= rcu_batches_started_sched,
	.completed	= rcu_batches_completed_sched,
	.exp_completed	= rcu_exp_batches_completed_sched,
	.sync		= synchronize_sched,
	.exp_sync	= synchronize_sched_expedited,
	.name		= "sched"
};

#ifdef CONFIG_TASKS_RCU

/*
 * Definitions for RCU-tasks perf testing.
 */

static int tasks_perf_read_lock(void)
{
	return 0;
}

static void tasks_perf_read_unlock(int idx)
{
}

static struct rcu_perf_ops tasks_ops = {
	.ptype		= RCU_TASKS_FLAVOR,
	.init		= rcu_sync_perf_init,
	.readlock	= tasks_perf_read_lock,
	.readunlock	= tasks_perf_read_unlock,
	.started	= rcu_no_completed,
	.completed	= rcu_no_completed,
	.sync		= synchronize_rcu_tasks,
	.exp_sync	= synchronize_rcu_tasks,
	.name		= "tasks"
};

#define RCUPERF_TASKS_OPS &tasks_ops,

static bool __maybe_unused torturing_tasks(void)
{
	return cur_ops == &tasks_ops;
}

#else /* #ifdef CONFIG_TASKS_RCU */

#define RCUPERF_TASKS_OPS

static bool __maybe_unused torturing_tasks(void)
{
	return false;
}

#endif /* #else #ifdef CONFIG_TASKS_RCU */

/*
 * If performance tests complete, wait for shutdown to commence.
 */
static void rcu_perf_wait_shutdown(void)
{
	cond_resched_rcu_qs();
	if (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters)
		return;
	while (!torture_must_stop())
		schedule_timeout_uninterruptible(1);
}

/*
 * RCU perf reader kthread.  Repeatedly does empty RCU read-side
 * critical section, minimizing update-side interference.
 */
static int
rcu_perf_reader(void *arg)
{
	unsigned long flags;
	int idx;
	long me = (long)arg;

	VERBOSE_PERFOUT_STRING("rcu_perf_reader task started");
	set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids));
	set_user_nice(current, MAX_NICE);
	atomic_inc(&n_rcu_perf_reader_started);

	do {
		local_irq_save(flags);
		idx = cur_ops->readlock();
		cur_ops->readunlock(idx);
		local_irq_restore(flags);
		rcu_perf_wait_shutdown();
	} while (!torture_must_stop());
	torture_kthread_stopping("rcu_perf_reader");
	return 0;
}

/*
 * RCU perf writer kthread.  Repeatedly does a grace period.
 */
static int
rcu_perf_writer(void *arg)
{
	int i = 0;
	int i_max;
	long me = (long)arg;
	struct sched_param sp;
	bool started = false, done = false, alldone = false;
	u64 t;
	u64 *wdp;
	u64 *wdpp = writer_durations[me];

	VERBOSE_PERFOUT_STRING("rcu_perf_writer task started");
	WARN_ON(!wdpp);
	set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids));
	sp.sched_priority = 1;
	sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);

	if (holdoff)
		schedule_timeout_uninterruptible(holdoff * HZ);

	t = ktime_get_mono_fast_ns();
	if (atomic_inc_return(&n_rcu_perf_writer_started) >= nrealwriters) {
		t_rcu_perf_writer_started = t;
		if (gp_exp) {
			b_rcu_perf_writer_started =
				cur_ops->exp_completed() / 2;
		} else {
			b_rcu_perf_writer_started =
				cur_ops->completed();
		}
	}

	do {
		wdp = &wdpp[i];
		*wdp = ktime_get_mono_fast_ns();
		if (gp_exp) {
			rcu_perf_writer_state = RTWS_EXP_SYNC;
			cur_ops->exp_sync();
		} else {
			rcu_perf_writer_state = RTWS_SYNC;
			cur_ops->sync();
		}
		rcu_perf_writer_state = RTWS_IDLE;
		t = ktime_get_mono_fast_ns();
		*wdp = t - *wdp;
		i_max = i;
		if (!started &&
		    atomic_read(&n_rcu_perf_writer_started) >= nrealwriters)
			started = true;
		if (!done && i >= MIN_MEAS) {
			done = true;
			sp.sched_priority = 0;
			sched_setscheduler_nocheck(current,
						   SCHED_NORMAL, &sp);
			pr_alert("%s" PERF_FLAG
				 "rcu_perf_writer %ld has %d measurements\n",
				 perf_type, me, MIN_MEAS);
			if (atomic_inc_return(&n_rcu_perf_writer_finished) >=
			    nrealwriters) {
				schedule_timeout_interruptible(10);
				rcu_ftrace_dump(DUMP_ALL);
				PERFOUT_STRING("Test complete");
				t_rcu_perf_writer_finished = t;
				if (gp_exp) {
					b_rcu_perf_writer_finished =
						cur_ops->exp_completed() / 2;
				} else {
					b_rcu_perf_writer_finished =
						cur_ops->completed();
				}
				if (shutdown) {
					smp_mb(); /* Assign before wake. */
					wake_up(&shutdown_wq);
				}
			}
		}
		if (done && !alldone &&
		    atomic_read(&n_rcu_perf_writer_finished) >= nrealwriters)
			alldone = true;
		if (started && !alldone && i < MAX_MEAS - 1)
			i++;
		rcu_perf_wait_shutdown();
	} while (!torture_must_stop());
	rcu_perf_writer_state = RTWS_STOPPING;
	writer_n_durations[me] = i_max;
	torture_kthread_stopping("rcu_perf_writer");
	return 0;
}

static inline void
rcu_perf_print_module_parms(struct rcu_perf_ops *cur_ops, const char *tag)
{
	pr_alert("%s" PERF_FLAG
		 "--- %s: nreaders=%d nwriters=%d verbose=%d shutdown=%d\n",
		 perf_type, tag, nrealreaders, nrealwriters, verbose, shutdown);
}

static void
rcu_perf_cleanup(void)
{
	int i;
	int j;
	int ngps = 0;
	u64 *wdp;
	u64 *wdpp;

	if (torture_cleanup_begin())
		return;

	if (reader_tasks) {
		for (i = 0; i < nrealreaders; i++)
			torture_stop_kthread(rcu_perf_reader,
					     reader_tasks[i]);
		kfree(reader_tasks);
	}

	if (writer_tasks) {
		for (i = 0; i < nrealwriters; i++) {
			torture_stop_kthread(rcu_perf_writer,
					     writer_tasks[i]);
			if (!writer_n_durations)
				continue;
			j = writer_n_durations[i];
			pr_alert("%s%s writer %d gps: %d\n",
				 perf_type, PERF_FLAG, i, j);
			ngps += j;
		}
		pr_alert("%s%s start: %llu end: %llu duration: %llu gps: %d batches: %ld\n",
			 perf_type, PERF_FLAG,
			 t_rcu_perf_writer_started, t_rcu_perf_writer_finished,
			 t_rcu_perf_writer_finished -
			 t_rcu_perf_writer_started,
			 ngps,
			 b_rcu_perf_writer_finished -
			 b_rcu_perf_writer_started);
		for (i = 0; i < nrealwriters; i++) {
			if (!writer_durations)
				break;
			if (!writer_n_durations)
				continue;
			wdpp = writer_durations[i];
			if (!wdpp)
				continue;
			for (j = 0; j <= writer_n_durations[i]; j++) {
				wdp = &wdpp[j];
				pr_alert("%s%s %4d writer-duration: %5d %llu\n",
					perf_type, PERF_FLAG,
					i, j, *wdp);
				if (j % 100 == 0)
					schedule_timeout_uninterruptible(1);
			}
			kfree(writer_durations[i]);
		}
		kfree(writer_tasks);
		kfree(writer_durations);
		kfree(writer_n_durations);
	}

	/* Do flavor-specific cleanup operations.  */
	if (cur_ops->cleanup != NULL)
		cur_ops->cleanup();

	torture_cleanup_end();
}

/*
 * Return the number if non-negative.  If -1, the number of CPUs.
 * If less than -1, that much less than the number of CPUs, but
 * at least one.
 */
static int compute_real(int n)
{
	int nr;

	if (n >= 0) {
		nr = n;
	} else {
		nr = num_online_cpus() + 1 + n;
		if (nr <= 0)
			nr = 1;
	}
	return nr;
}

/*
 * RCU perf shutdown kthread.  Just waits to be awakened, then shuts
 * down system.
 */
static int
rcu_perf_shutdown(void *arg)
{
	do {
		wait_event(shutdown_wq,
			   atomic_read(&n_rcu_perf_writer_finished) >=
			   nrealwriters);
	} while (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters);
	smp_mb(); /* Wake before output. */
	rcu_perf_cleanup();
	kernel_power_off();
	return -EINVAL;
}

static int __init
rcu_perf_init(void)
{
	long i;
	int firsterr = 0;
	static struct rcu_perf_ops *perf_ops[] = {
		&rcu_ops, &rcu_bh_ops, &srcu_ops, &sched_ops,
		RCUPERF_TASKS_OPS
	};

	if (!torture_init_begin(perf_type, verbose, &perf_runnable))
		return -EBUSY;

	/* Process args and tell the world that the perf'er is on the job. */
	for (i = 0; i < ARRAY_SIZE(perf_ops); i++) {
		cur_ops = perf_ops[i];
		if (strcmp(perf_type, cur_ops->name) == 0)
			break;
	}
	if (i == ARRAY_SIZE(perf_ops)) {
		pr_alert("rcu-perf: invalid perf type: \"%s\"\n",
			 perf_type);
		pr_alert("rcu-perf types:");
		for (i = 0; i < ARRAY_SIZE(perf_ops); i++)
			pr_alert(" %s", perf_ops[i]->name);
		pr_alert("\n");
		firsterr = -EINVAL;
		goto unwind;
	}
	if (cur_ops->init)
		cur_ops->init();

	nrealwriters = compute_real(nwriters);
	nrealreaders = compute_real(nreaders);
	atomic_set(&n_rcu_perf_reader_started, 0);
	atomic_set(&n_rcu_perf_writer_started, 0);
	atomic_set(&n_rcu_perf_writer_finished, 0);
	rcu_perf_print_module_parms(cur_ops, "Start of test");

	/* Start up the kthreads. */

	if (shutdown) {
		init_waitqueue_head(&shutdown_wq);
		firsterr = torture_create_kthread(rcu_perf_shutdown, NULL,
						  shutdown_task);
		if (firsterr)
			goto unwind;
		schedule_timeout_uninterruptible(1);
	}
	reader_tasks = kcalloc(nrealreaders, sizeof(reader_tasks[0]),
			       GFP_KERNEL);
	if (reader_tasks == NULL) {
		VERBOSE_PERFOUT_ERRSTRING("out of memory");
		firsterr = -ENOMEM;
		goto unwind;
	}
	for (i = 0; i < nrealreaders; i++) {
		firsterr = torture_create_kthread(rcu_perf_reader, (void *)i,
						  reader_tasks[i]);
		if (firsterr)
			goto unwind;
	}
	while (atomic_read(&n_rcu_perf_reader_started) < nrealreaders)
		schedule_timeout_uninterruptible(1);
	writer_tasks = kcalloc(nrealwriters, sizeof(reader_tasks[0]),
			       GFP_KERNEL);
	writer_durations = kcalloc(nrealwriters, sizeof(*writer_durations),
				   GFP_KERNEL);
	writer_n_durations =
		kcalloc(nrealwriters, sizeof(*writer_n_durations),
			GFP_KERNEL);
	if (!writer_tasks || !writer_durations || !writer_n_durations) {
		VERBOSE_PERFOUT_ERRSTRING("out of memory");
		firsterr = -ENOMEM;
		goto unwind;
	}
	if (rcu_gp_is_expedited() && !rcu_gp_is_normal() && !gp_exp) {
		VERBOSE_PERFOUT_ERRSTRING("All grace periods expedited, no normal ones to measure!");
		firsterr = -EINVAL;
		goto unwind;
	}
	if (rcu_gp_is_normal() && gp_exp) {
		VERBOSE_PERFOUT_ERRSTRING("All grace periods normal, no expedited ones to measure!");
		firsterr = -EINVAL;
		goto unwind;
	}
	for (i = 0; i < nrealwriters; i++) {
		writer_durations[i] =
			kcalloc(MAX_MEAS, sizeof(*writer_durations[i]),
				GFP_KERNEL);
		if (!writer_durations[i]) {
			firsterr = -ENOMEM;
			goto unwind;
		}
		firsterr = torture_create_kthread(rcu_perf_writer, (void *)i,
						  writer_tasks[i]);
		if (firsterr)
			goto unwind;
	}
	torture_init_end();
	return 0;

unwind:
	torture_init_end();
	rcu_perf_cleanup();
	return firsterr;
}

module_init(rcu_perf_init);
module_exit(rcu_perf_cleanup);