trace.c 100 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
/*
 * ring buffer based function tracer
 *
 * Copyright (C) 2007-2008 Steven Rostedt <srostedt@redhat.com>
 * Copyright (C) 2008 Ingo Molnar <mingo@redhat.com>
 *
 * Originally taken from the RT patch by:
 *    Arnaldo Carvalho de Melo <acme@redhat.com>
 *
 * Based on code from the latency_tracer, that is:
 *  Copyright (C) 2004-2006 Ingo Molnar
 *  Copyright (C) 2004 William Lee Irwin III
 */
14
#include <linux/ring_buffer.h>
15
#include <linux/utsrelease.h>
16
17
#include <linux/stacktrace.h>
#include <linux/writeback.h>
18
19
#include <linux/kallsyms.h>
#include <linux/seq_file.h>
Alexey Dobriyan's avatar
Alexey Dobriyan committed
20
#include <linux/smp_lock.h>
21
#include <linux/notifier.h>
22
#include <linux/irqflags.h>
23
#include <linux/debugfs.h>
24
#include <linux/pagemap.h>
25
26
27
#include <linux/hardirq.h>
#include <linux/linkage.h>
#include <linux/uaccess.h>
28
#include <linux/kprobes.h>
29
30
31
#include <linux/ftrace.h>
#include <linux/module.h>
#include <linux/percpu.h>
32
#include <linux/splice.h>
33
#include <linux/kdebug.h>
34
#include <linux/string.h>
35
36
#include <linux/ctype.h>
#include <linux/init.h>
37
#include <linux/poll.h>
38
39
#include <linux/gfp.h>
#include <linux/fs.h>
Ingo Molnar's avatar
Ingo Molnar committed
40

41
#include "trace.h"
42
#include "trace_output.h"
43

44
45
#define TRACE_BUFFER_FLAGS	(RB_FL_OVERWRITE)

46
47
48
49
/*
 * On boot up, the ring buffer is set to the minimum size, so that
 * we do not waste memory on systems that are not using tracing.
 */
50
int ring_buffer_expanded;
51

52
53
/*
 * We need to change this state when a selftest is running.
54
55
 * A selftest will lurk into the ring-buffer to count the
 * entries inserted during the selftest although some concurrent
56
 * insertions into the ring-buffer such as trace_printk could occurred
57
58
 * at the same time, giving false positive or negative results.
 */
59
static bool __read_mostly tracing_selftest_running;
60

61
62
63
/*
 * If a tracer is running, we do not want to run SELFTEST.
 */
64
bool __read_mostly tracing_selftest_disabled;
65

66
67
68
69
70
71
72
73
74
75
76
77
78
79
/* For tracers that don't implement custom flags */
static struct tracer_opt dummy_tracer_opt[] = {
	{ }
};

static struct tracer_flags dummy_tracer_flags = {
	.val = 0,
	.opts = dummy_tracer_opt
};

static int dummy_set_flag(u32 old_flags, u32 bit, int set)
{
	return 0;
}
80
81
82
83
84
85
86

/*
 * Kill all tracing for good (never come back).
 * It is initialized to 1 but will turn to zero if the initialization
 * of the tracer is successful. But that is the only place that sets
 * this back to zero.
 */
87
static int tracing_disabled = 1;
88

89
DEFINE_PER_CPU(local_t, ftrace_cpu_disabled);
90
91
92
93
94
95
96
97
98
99
100
101
102

static inline void ftrace_disable_cpu(void)
{
	preempt_disable();
	local_inc(&__get_cpu_var(ftrace_cpu_disabled));
}

static inline void ftrace_enable_cpu(void)
{
	local_dec(&__get_cpu_var(ftrace_cpu_disabled));
	preempt_enable();
}

103
static cpumask_var_t __read_mostly	tracing_buffer_mask;
104

105
106
107
/* Define which cpu buffers are currently read in trace_pipe */
static cpumask_var_t			tracing_reader_cpumask;

108
#define for_each_tracing_cpu(cpu)	\
109
	for_each_cpu(cpu, tracing_buffer_mask)
110

111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
/*
 * ftrace_dump_on_oops - variable to dump ftrace buffer on oops
 *
 * If there is an oops (or kernel panic) and the ftrace_dump_on_oops
 * is set, then ftrace_dump is called. This will output the contents
 * of the ftrace buffers to the console.  This is very useful for
 * capturing traces that lead to crashes and outputing it to a
 * serial console.
 *
 * It is default off, but you can enable it with either specifying
 * "ftrace_dump_on_oops" in the kernel command line, or setting
 * /proc/sys/kernel/ftrace_dump_on_oops to true.
 */
int ftrace_dump_on_oops;

126
127
static int tracing_set_tracer(const char *buf);

Li Zefan's avatar
Li Zefan committed
128
129
#define MAX_TRACER_SIZE		100
static char bootup_tracer_buf[MAX_TRACER_SIZE] __initdata;
130
static char *default_bootup_tracer;
131

132
static int __init set_cmdline_ftrace(char *str)
133
{
Li Zefan's avatar
Li Zefan committed
134
	strncpy(bootup_tracer_buf, str, MAX_TRACER_SIZE);
135
	default_bootup_tracer = bootup_tracer_buf;
136
137
	/* We are using ftrace early, expand it */
	ring_buffer_expanded = 1;
138
139
	return 1;
}
140
__setup("ftrace=", set_cmdline_ftrace);
141

142
143
144
145
146
147
static int __init set_ftrace_dump_on_oops(char *str)
{
	ftrace_dump_on_oops = 1;
	return 1;
}
__setup("ftrace_dump_on_oops", set_ftrace_dump_on_oops);
Steven Rostedt's avatar
Steven Rostedt committed
148

149
unsigned long long ns2usecs(cycle_t nsec)
150
151
152
153
154
155
{
	nsec += 500;
	do_div(nsec, 1000);
	return nsec;
}

Steven Rostedt's avatar
Steven Rostedt committed
156
157
158
159
160
161
162
163
164
165
166
167
/*
 * The global_trace is the descriptor that holds the tracing
 * buffers for the live tracing. For each CPU, it contains
 * a link list of pages that will store trace entries. The
 * page descriptor of the pages in the memory is used to hold
 * the link list by linking the lru item in the page descriptor
 * to each of the pages in the buffer per CPU.
 *
 * For each active CPU there is a data field that holds the
 * pages for the buffer for that CPU. Each CPU has the same number
 * of pages allocated for its buffer.
 */
168
169
170
171
static struct trace_array	global_trace;

static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu);

172
173
int filter_current_check_discard(struct ring_buffer *buffer,
				 struct ftrace_event_call *call, void *rec,
174
175
				 struct ring_buffer_event *event)
{
176
	return filter_check_discard(call, rec, buffer, event);
177
}
178
EXPORT_SYMBOL_GPL(filter_current_check_discard);
179

180
181
182
183
184
185
186
187
188
189
190
191
192
cycle_t ftrace_now(int cpu)
{
	u64 ts;

	/* Early boot up does not have a buffer yet */
	if (!global_trace.buffer)
		return trace_clock_local();

	ts = ring_buffer_time_stamp(global_trace.buffer, cpu);
	ring_buffer_normalize_time_stamp(global_trace.buffer, cpu, &ts);

	return ts;
}
193

Steven Rostedt's avatar
Steven Rostedt committed
194
195
196
197
198
199
200
201
202
203
/*
 * The max_tr is used to snapshot the global_trace when a maximum
 * latency is reached. Some tracers will use this to store a maximum
 * trace while it continues examining live traces.
 *
 * The buffers for the max_tr are set up the same as the global_trace.
 * When a snapshot is taken, the link list of the max_tr is swapped
 * with the link list of the global_trace and the buffers are reset for
 * the global_trace so the tracing can continue.
 */
204
205
206
207
static struct trace_array	max_tr;

static DEFINE_PER_CPU(struct trace_array_cpu, max_data);

Steven Rostedt's avatar
Steven Rostedt committed
208
/* tracer_enabled is used to toggle activation of a tracer */
209
static int			tracer_enabled = 1;
Steven Rostedt's avatar
Steven Rostedt committed
210

211
212
213
214
215
216
217
218
219
220
221
222
223
/**
 * tracing_is_enabled - return tracer_enabled status
 *
 * This function is used by other tracers to know the status
 * of the tracer_enabled flag.  Tracers may use this function
 * to know if it should enable their features when starting
 * up. See irqsoff tracer for an example (start_irqsoff_tracer).
 */
int tracing_is_enabled(void)
{
	return tracer_enabled;
}

Steven Rostedt's avatar
Steven Rostedt committed
224
/*
225
226
227
 * trace_buf_size is the size in bytes that is allocated
 * for a buffer. Note, the number of bytes is always rounded
 * to page size.
228
229
230
231
232
 *
 * This number is purposely set to a low number of 16384.
 * If the dump on oops happens, it will be much appreciated
 * to not have to wait for all that output. Anyway this can be
 * boot time and run time configurable.
Steven Rostedt's avatar
Steven Rostedt committed
233
 */
234
#define TRACE_BUF_SIZE_DEFAULT	1441792UL /* 16384 * 88 (sizeof(entry)) */
235

236
static unsigned long		trace_buf_size = TRACE_BUF_SIZE_DEFAULT;
237

Steven Rostedt's avatar
Steven Rostedt committed
238
/* trace_types holds a link list of available tracers. */
239
static struct tracer		*trace_types __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
240
241

/* current_trace points to the tracer that is currently active */
242
static struct tracer		*current_trace __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
243
244
245
246
247
248
249

/*
 * trace_types_lock is used to protect the trace_types list.
 * This lock is also used to keep user access serialized.
 * Accesses from userspace will grab this lock while userspace
 * activities happen inside the kernel.
 */
250
static DEFINE_MUTEX(trace_types_lock);
Steven Rostedt's avatar
Steven Rostedt committed
251
252

/* trace_wait is a waitqueue for tasks blocked on trace_poll */
253
254
static DECLARE_WAIT_QUEUE_HEAD(trace_wait);

255
/* trace_flags holds trace_options default values */
256
unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
257
258
	TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME |
	TRACE_ITER_GRAPH_TIME;
259

260
261
262
static int trace_stop_count;
static DEFINE_SPINLOCK(tracing_start_lock);

Steven Rostedt's avatar
Steven Rostedt committed
263
264
265
266
267
268
/**
 * trace_wake_up - wake up tasks waiting for trace input
 *
 * Simply wakes up any task that is blocked on the trace_wait
 * queue. These is used with trace_poll for tasks polling the trace.
 */
269
270
void trace_wake_up(void)
{
271
272
273
274
	int cpu;

	if (trace_flags & TRACE_ITER_BLOCK)
		return;
Ingo Molnar's avatar
Ingo Molnar committed
275
276
277
278
	/*
	 * The runqueue_is_locked() can fail, but this is the best we
	 * have for now:
	 */
279
280
	cpu = get_cpu();
	if (!runqueue_is_locked(cpu))
281
		wake_up(&trace_wait);
282
	put_cpu();
283
}
284

285
static int __init set_buf_size(char *str)
286
{
287
	unsigned long buf_size;
288

289
290
	if (!str)
		return 0;
291
	buf_size = memparse(str, &str);
292
	/* nr_entries can not be zero */
293
	if (buf_size == 0)
294
		return 0;
295
	trace_buf_size = buf_size;
296
297
	return 1;
}
298
__setup("trace_buf_size=", set_buf_size);
299

Steven Rostedt's avatar
Steven Rostedt committed
300
301
302
303
304
unsigned long nsecs_to_usecs(unsigned long nsecs)
{
	return nsecs / 1000;
}

Steven Rostedt's avatar
Steven Rostedt committed
305
/* These must match the bit postions in trace_iterator_flags */
306
307
308
309
310
static const char *trace_options[] = {
	"print-parent",
	"sym-offset",
	"sym-addr",
	"verbose",
Ingo Molnar's avatar
Ingo Molnar committed
311
	"raw",
312
	"hex",
Ingo Molnar's avatar
Ingo Molnar committed
313
	"bin",
314
	"block",
Ingo Molnar's avatar
Ingo Molnar committed
315
	"stacktrace",
Ingo Molnar's avatar
Ingo Molnar committed
316
	"sched-tree",
317
	"trace_printk",
318
	"ftrace_preempt",
319
	"branch",
320
	"annotate",
321
	"userstacktrace",
322
	"sym-userobj",
323
	"printk-msg-only",
324
	"context-info",
325
	"latency-format",
326
	"sleep-time",
327
	"graph-time",
328
329
330
	NULL
};

331
332
333
334
335
336
337
338
339
340
static struct {
	u64 (*func)(void);
	const char *name;
} trace_clocks[] = {
	{ trace_clock_local,	"local" },
	{ trace_clock_global,	"global" },
};

int trace_clock_id;

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
/*
 * trace_parser_get_init - gets the buffer for trace parser
 */
int trace_parser_get_init(struct trace_parser *parser, int size)
{
	memset(parser, 0, sizeof(*parser));

	parser->buffer = kmalloc(size, GFP_KERNEL);
	if (!parser->buffer)
		return 1;

	parser->size = size;
	return 0;
}

/*
 * trace_parser_put - frees the buffer for trace parser
 */
void trace_parser_put(struct trace_parser *parser)
{
	kfree(parser->buffer);
}

/*
 * trace_get_user - reads the user input string separated by  space
 * (matched by isspace(ch))
 *
 * For each string found the 'struct trace_parser' is updated,
 * and the function returns.
 *
 * Returns number of bytes read.
 *
 * See kernel/trace/trace.h for 'struct trace_parser' details.
 */
int trace_get_user(struct trace_parser *parser, const char __user *ubuf,
	size_t cnt, loff_t *ppos)
{
	char ch;
	size_t read = 0;
	ssize_t ret;

	if (!*ppos)
		trace_parser_clear(parser);

	ret = get_user(ch, ubuf++);
	if (ret)
		goto out;

	read++;
	cnt--;

	/*
	 * The parser is not finished with the last write,
	 * continue reading the user input without skipping spaces.
	 */
	if (!parser->cont) {
		/* skip white space */
		while (cnt && isspace(ch)) {
			ret = get_user(ch, ubuf++);
			if (ret)
				goto out;
			read++;
			cnt--;
		}

		/* only spaces were written */
		if (isspace(ch)) {
			*ppos += read;
			ret = read;
			goto out;
		}

		parser->idx = 0;
	}

	/* read the non-space input */
	while (cnt && !isspace(ch)) {
418
		if (parser->idx < parser->size - 1)
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
			parser->buffer[parser->idx++] = ch;
		else {
			ret = -EINVAL;
			goto out;
		}
		ret = get_user(ch, ubuf++);
		if (ret)
			goto out;
		read++;
		cnt--;
	}

	/* We either got finished input or we have to wait for another call. */
	if (isspace(ch)) {
		parser->buffer[parser->idx] = 0;
		parser->cont = false;
	} else {
		parser->cont = true;
		parser->buffer[parser->idx++] = ch;
	}

	*ppos += read;
	ret = read;

out:
	return ret;
}

447
448
449
450
451
ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt)
{
	int len;
	int ret;

452
453
454
	if (!cnt)
		return 0;

455
456
457
458
459
460
461
	if (s->len <= s->readpos)
		return -EBUSY;

	len = s->len - s->readpos;
	if (cnt > len)
		cnt = len;
	ret = copy_to_user(ubuf, s->buffer + s->readpos, cnt);
462
	if (ret == cnt)
463
464
		return -EFAULT;

465
466
	cnt -= ret;

467
	s->readpos += cnt;
468
	return cnt;
469
470
}

471
static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt)
472
473
474
475
476
477
478
479
480
481
482
483
484
485
{
	int len;
	void *ret;

	if (s->len <= s->readpos)
		return -EBUSY;

	len = s->len - s->readpos;
	if (cnt > len)
		cnt = len;
	ret = memcpy(buf, s->buffer + s->readpos, cnt);
	if (!ret)
		return -EFAULT;

486
	s->readpos += cnt;
487
488
489
	return cnt;
}

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
/*
 * ftrace_max_lock is used to protect the swapping of buffers
 * when taking a max snapshot. The buffers themselves are
 * protected by per_cpu spinlocks. But the action of the swap
 * needs its own lock.
 *
 * This is defined as a raw_spinlock_t in order to help
 * with performance when lockdep debugging is enabled.
 *
 * It is also used in other places outside the update_max_tr
 * so it needs to be defined outside of the
 * CONFIG_TRACER_MAX_TRACE.
 */
static raw_spinlock_t ftrace_max_lock =
	(raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;

#ifdef CONFIG_TRACER_MAX_TRACE
unsigned long __read_mostly	tracing_max_latency;
unsigned long __read_mostly	tracing_thresh;

/*
 * Copy the new maximum trace into the separate maximum-trace
 * structure. (this way the maximum trace is permanently saved,
 * for later retrieval via /sys/kernel/debug/tracing/latency_trace)
 */
static void
__update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
	struct trace_array_cpu *data = tr->data[cpu];
519
	struct trace_array_cpu *max_data = tr->data[cpu];
520
521
522
523

	max_tr.cpu = cpu;
	max_tr.time_start = data->preempt_timestamp;

524
525
526
527
	max_data = max_tr.data[cpu];
	max_data->saved_latency = tracing_max_latency;
	max_data->critical_start = data->critical_start;
	max_data->critical_end = data->critical_end;
528
529

	memcpy(data->comm, tsk->comm, TASK_COMM_LEN);
530
531
532
533
534
	max_data->pid = tsk->pid;
	max_data->uid = task_uid(tsk);
	max_data->nice = tsk->static_prio - 20 - MAX_RT_PRIO;
	max_data->policy = tsk->policy;
	max_data->rt_priority = tsk->rt_priority;
535
536
537
538
539

	/* record this tasks comm */
	tracing_record_cmdline(tsk);
}

Steven Rostedt's avatar
Steven Rostedt committed
540
541
542
543
544
545
546
547
548
/**
 * update_max_tr - snapshot all trace buffers from global_trace to max_tr
 * @tr: tracer
 * @tsk: the task with the latency
 * @cpu: The cpu that initiated the trace.
 *
 * Flip the buffers between the @tr and the max_tr and record information
 * about which task was the cause of this latency.
 */
Ingo Molnar's avatar
Ingo Molnar committed
549
void
550
551
update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
552
	struct ring_buffer *buf = tr->buffer;
553

554
555
556
	if (trace_stop_count)
		return;

557
	WARN_ON_ONCE(!irqs_disabled());
558
	__raw_spin_lock(&ftrace_max_lock);
559
560
561
562

	tr->buffer = max_tr.buffer;
	max_tr.buffer = buf;

563
	__update_max_tr(tr, tsk, cpu);
564
	__raw_spin_unlock(&ftrace_max_lock);
565
566
567
568
569
570
571
}

/**
 * update_max_tr_single - only copy one trace over, and reset the rest
 * @tr - tracer
 * @tsk - task with the latency
 * @cpu - the cpu of the buffer to copy.
Steven Rostedt's avatar
Steven Rostedt committed
572
573
 *
 * Flip the trace of a single CPU buffer between the @tr and the max_tr.
574
 */
Ingo Molnar's avatar
Ingo Molnar committed
575
void
576
577
update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
578
	int ret;
579

580
581
582
	if (trace_stop_count)
		return;

583
	WARN_ON_ONCE(!irqs_disabled());
584
	__raw_spin_lock(&ftrace_max_lock);
585

586
587
	ftrace_disable_cpu();

588
589
	ret = ring_buffer_swap_cpu(max_tr.buffer, tr->buffer, cpu);

590
591
592
593
594
595
596
597
598
599
600
	if (ret == -EBUSY) {
		/*
		 * We failed to swap the buffer due to a commit taking
		 * place on this CPU. We fail to record, but we reset
		 * the max trace buffer (no one writes directly to it)
		 * and flag that it failed.
		 */
		trace_array_printk(&max_tr, _THIS_IP_,
			"Failed to swap buffers due to commit in progress\n");
	}

601
602
	ftrace_enable_cpu();

603
	WARN_ON_ONCE(ret && ret != -EAGAIN && ret != -EBUSY);
604
605

	__update_max_tr(tr, tsk, cpu);
606
	__raw_spin_unlock(&ftrace_max_lock);
607
}
608
#endif /* CONFIG_TRACER_MAX_TRACE */
609

Steven Rostedt's avatar
Steven Rostedt committed
610
611
612
613
614
615
/**
 * register_tracer - register a tracer with the ftrace system.
 * @type - the plugin for the tracer
 *
 * Register a new plugin tracer.
 */
616
int register_tracer(struct tracer *type)
617
618
__releases(kernel_lock)
__acquires(kernel_lock)
619
620
621
622
623
624
625
626
627
{
	struct tracer *t;
	int ret = 0;

	if (!type->name) {
		pr_info("Tracer must have a name\n");
		return -1;
	}

Li Zefan's avatar
Li Zefan committed
628
629
630
631
632
	if (strlen(type->name) > MAX_TRACER_SIZE) {
		pr_info("Tracer has a name longer than %d\n", MAX_TRACER_SIZE);
		return -1;
	}

Ingo Molnar's avatar
Ingo Molnar committed
633
634
635
636
637
638
639
	/*
	 * When this gets called we hold the BKL which means that
	 * preemption is disabled. Various trace selftests however
	 * need to disable and enable preemption for successful tests.
	 * So we drop the BKL here and grab it after the tests again.
	 */
	unlock_kernel();
640
	mutex_lock(&trace_types_lock);
Ingo Molnar's avatar
Ingo Molnar committed
641

642
643
	tracing_selftest_running = true;

644
645
646
	for (t = trace_types; t; t = t->next) {
		if (strcmp(type->name, t->name) == 0) {
			/* already found */
Li Zefan's avatar
Li Zefan committed
647
			pr_info("Tracer %s already registered\n",
648
649
650
651
652
653
				type->name);
			ret = -1;
			goto out;
		}
	}

654
655
656
657
658
659
660
	if (!type->set_flag)
		type->set_flag = &dummy_set_flag;
	if (!type->flags)
		type->flags = &dummy_tracer_flags;
	else
		if (!type->flags->opts)
			type->flags->opts = dummy_tracer_opt;
661
662
663
	if (!type->wait_pipe)
		type->wait_pipe = default_wait_pipe;

664

Steven Rostedt's avatar
Steven Rostedt committed
665
#ifdef CONFIG_FTRACE_STARTUP_TEST
666
	if (type->selftest && !tracing_selftest_disabled) {
Steven Rostedt's avatar
Steven Rostedt committed
667
668
		struct tracer *saved_tracer = current_trace;
		struct trace_array *tr = &global_trace;
669

Steven Rostedt's avatar
Steven Rostedt committed
670
671
672
673
674
675
676
		/*
		 * Run a selftest on this tracer.
		 * Here we reset the trace buffer, and set the current
		 * tracer to be this tracer. The tracer can then run some
		 * internal tracing to verify that everything is in order.
		 * If we fail, we do not register this tracer.
		 */
677
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
678

Steven Rostedt's avatar
Steven Rostedt committed
679
680
681
682
683
684
685
686
687
688
		current_trace = type;
		/* the test is responsible for initializing and enabling */
		pr_info("Testing tracer %s: ", type->name);
		ret = type->selftest(type, tr);
		/* the test is responsible for resetting too */
		current_trace = saved_tracer;
		if (ret) {
			printk(KERN_CONT "FAILED!\n");
			goto out;
		}
Steven Rostedt's avatar
Steven Rostedt committed
689
		/* Only reset on passing, to avoid touching corrupted buffers */
690
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
691

Steven Rostedt's avatar
Steven Rostedt committed
692
693
694
695
		printk(KERN_CONT "PASSED\n");
	}
#endif

696
697
	type->next = trace_types;
	trace_types = type;
Steven Rostedt's avatar
Steven Rostedt committed
698

699
 out:
700
	tracing_selftest_running = false;
701
702
	mutex_unlock(&trace_types_lock);

Steven Rostedt's avatar
Steven Rostedt committed
703
704
705
	if (ret || !default_bootup_tracer)
		goto out_unlock;

Li Zefan's avatar
Li Zefan committed
706
	if (strncmp(default_bootup_tracer, type->name, MAX_TRACER_SIZE))
Steven Rostedt's avatar
Steven Rostedt committed
707
708
709
710
711
712
713
714
		goto out_unlock;

	printk(KERN_INFO "Starting tracer '%s'\n", type->name);
	/* Do we want this tracer to start on bootup? */
	tracing_set_tracer(type->name);
	default_bootup_tracer = NULL;
	/* disable other selftests, since this will break it. */
	tracing_selftest_disabled = 1;
715
#ifdef CONFIG_FTRACE_STARTUP_TEST
Steven Rostedt's avatar
Steven Rostedt committed
716
717
	printk(KERN_INFO "Disabling FTRACE selftests due to running tracer '%s'\n",
	       type->name);
718
719
#endif

Steven Rostedt's avatar
Steven Rostedt committed
720
 out_unlock:
721
	lock_kernel();
722
723
724
725
726
727
728
729
730
731
732
733
	return ret;
}

void unregister_tracer(struct tracer *type)
{
	struct tracer **t;

	mutex_lock(&trace_types_lock);
	for (t = &trace_types; *t; t = &(*t)->next) {
		if (*t == type)
			goto found;
	}
Li Zefan's avatar
Li Zefan committed
734
	pr_info("Tracer %s not registered\n", type->name);
735
736
737
738
	goto out;

 found:
	*t = (*t)->next;
739
740
741
742
743
744
745
746

	if (type == current_trace && tracer_enabled) {
		tracer_enabled = 0;
		tracing_stop();
		if (current_trace->stop)
			current_trace->stop(&global_trace);
		current_trace = &nop_trace;
	}
Li Zefan's avatar
Li Zefan committed
747
out:
748
749
750
	mutex_unlock(&trace_types_lock);
}

751
static void __tracing_reset(struct trace_array *tr, int cpu)
752
{
753
	ftrace_disable_cpu();
754
	ring_buffer_reset_cpu(tr->buffer, cpu);
755
	ftrace_enable_cpu();
756
757
}

758
759
760
761
762
763
764
765
766
767
768
769
770
void tracing_reset(struct trace_array *tr, int cpu)
{
	struct ring_buffer *buffer = tr->buffer;

	ring_buffer_record_disable(buffer);

	/* Make sure all commits have finished */
	synchronize_sched();
	__tracing_reset(tr, cpu);

	ring_buffer_record_enable(buffer);
}

771
772
void tracing_reset_online_cpus(struct trace_array *tr)
{
773
	struct ring_buffer *buffer = tr->buffer;
774
775
	int cpu;

776
777
778
779
780
	ring_buffer_record_disable(buffer);

	/* Make sure all commits have finished */
	synchronize_sched();

781
782
783
	tr->time_start = ftrace_now(tr->cpu);

	for_each_online_cpu(cpu)
784
		__tracing_reset(tr, cpu);
785
786

	ring_buffer_record_enable(buffer);
787
788
}

789
790
791
792
793
794
795
796
797
798
void tracing_reset_current(int cpu)
{
	tracing_reset(&global_trace, cpu);
}

void tracing_reset_current_online_cpus(void)
{
	tracing_reset_online_cpus(&global_trace);
}

799
#define SAVED_CMDLINES 128
800
#define NO_CMDLINE_MAP UINT_MAX
801
802
803
804
static unsigned map_pid_to_cmdline[PID_MAX_DEFAULT+1];
static unsigned map_cmdline_to_pid[SAVED_CMDLINES];
static char saved_cmdlines[SAVED_CMDLINES][TASK_COMM_LEN];
static int cmdline_idx;
805
static raw_spinlock_t trace_cmdline_lock = __RAW_SPIN_LOCK_UNLOCKED;
806
807

/* temporary disable recording */
808
static atomic_t trace_record_cmdline_disabled __read_mostly;
809
810
811

static void trace_init_cmdlines(void)
{
812
813
	memset(&map_pid_to_cmdline, NO_CMDLINE_MAP, sizeof(map_pid_to_cmdline));
	memset(&map_cmdline_to_pid, NO_CMDLINE_MAP, sizeof(map_cmdline_to_pid));
814
815
816
	cmdline_idx = 0;
}

817
818
819
820
821
int is_tracing_stopped(void)
{
	return trace_stop_count;
}

822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
/**
 * ftrace_off_permanent - disable all ftrace code permanently
 *
 * This should only be called when a serious anomally has
 * been detected.  This will turn off the function tracing,
 * ring buffers, and other tracing utilites. It takes no
 * locks and can be called from any context.
 */
void ftrace_off_permanent(void)
{
	tracing_disabled = 1;
	ftrace_stop();
	tracing_off_permanent();
}

837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
/**
 * tracing_start - quick start of the tracer
 *
 * If tracing is enabled but was stopped by tracing_stop,
 * this will start the tracer back up.
 */
void tracing_start(void)
{
	struct ring_buffer *buffer;
	unsigned long flags;

	if (tracing_disabled)
		return;

	spin_lock_irqsave(&tracing_start_lock, flags);
852
853
854
855
856
857
	if (--trace_stop_count) {
		if (trace_stop_count < 0) {
			/* Someone screwed up their debugging */
			WARN_ON_ONCE(1);
			trace_stop_count = 0;
		}
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
		goto out;
	}


	buffer = global_trace.buffer;
	if (buffer)
		ring_buffer_record_enable(buffer);

	buffer = max_tr.buffer;
	if (buffer)
		ring_buffer_record_enable(buffer);

	ftrace_start();
 out:
	spin_unlock_irqrestore(&tracing_start_lock, flags);
}

/**
 * tracing_stop - quick stop of the tracer
 *
 * Light weight way to stop tracing. Use in conjunction with
 * tracing_start.
 */
void tracing_stop(void)
{
	struct ring_buffer *buffer;
	unsigned long flags;

	ftrace_stop();
	spin_lock_irqsave(&tracing_start_lock, flags);
	if (trace_stop_count++)
		goto out;

	buffer = global_trace.buffer;
	if (buffer)
		ring_buffer_record_disable(buffer);

	buffer = max_tr.buffer;
	if (buffer)
		ring_buffer_record_disable(buffer);

 out:
	spin_unlock_irqrestore(&tracing_start_lock, flags);
}

Ingo Molnar's avatar
Ingo Molnar committed
903
void trace_stop_cmdline_recording(void);
904

Ingo Molnar's avatar
Ingo Molnar committed
905
static void trace_save_cmdline(struct task_struct *tsk)
906
{
907
	unsigned pid, idx;
908
909
910
911
912
913
914
915
916
917

	if (!tsk->pid || unlikely(tsk->pid > PID_MAX_DEFAULT))
		return;

	/*
	 * It's not the end of the world if we don't get
	 * the lock, but we also don't want to spin
	 * nor do we want to disable interrupts,
	 * so if we miss here, then better luck next time.
	 */
918
	if (!__raw_spin_trylock(&trace_cmdline_lock))
919
920
921
		return;

	idx = map_pid_to_cmdline[tsk->pid];
922
	if (idx == NO_CMDLINE_MAP) {
923
924
		idx = (cmdline_idx + 1) % SAVED_CMDLINES;

925
926
927
928
929
930
931
932
933
		/*
		 * Check whether the cmdline buffer at idx has a pid
		 * mapped. We are going to overwrite that entry so we
		 * need to clear the map_pid_to_cmdline. Otherwise we
		 * would read the new comm for the old pid.
		 */
		pid = map_cmdline_to_pid[idx];
		if (pid != NO_CMDLINE_MAP)
			map_pid_to_cmdline[pid] = NO_CMDLINE_MAP;
934

935
		map_cmdline_to_pid[idx] = tsk->pid;
936
937
938
939
940
941
942
		map_pid_to_cmdline[tsk->pid] = idx;

		cmdline_idx = idx;
	}

	memcpy(&saved_cmdlines[idx], tsk->comm, TASK_COMM_LEN);

943
	__raw_spin_unlock(&trace_cmdline_lock);
944
945
}

946
void trace_find_cmdline(int pid, char comm[])
947
948
949
{
	unsigned map;

950
951
952
953
	if (!pid) {
		strcpy(comm, "<idle>");
		return;
	}
954

955
956
957
958
	if (pid > PID_MAX_DEFAULT) {
		strcpy(comm, "<...>");
		return;
	}
959

960
	preempt_disable();
961
	__raw_spin_lock(&trace_cmdline_lock);
962
	map = map_pid_to_cmdline[pid];
963
964
965
966
	if (map != NO_CMDLINE_MAP)
		strcpy(comm, saved_cmdlines[map]);
	else
		strcpy(comm, "<...>");
967

968
	__raw_spin_unlock(&trace_cmdline_lock);
969
	preempt_enable();
970
971
}

Ingo Molnar's avatar
Ingo Molnar committed
972
void tracing_record_cmdline(struct task_struct *tsk)
973
{
974
975
	if (atomic_read(&trace_record_cmdline_disabled) || !tracer_enabled ||
	    !tracing_is_on())
976
977
978
979
980
		return;

	trace_save_cmdline(tsk);
}

981
void
982
983
tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
			     int pc)
984
985
986
{
	struct task_struct *tsk = current;

987
988
	entry->preempt_count		= pc & 0xff;
	entry->pid			= (tsk) ? tsk->pid : 0;
989
	entry->lock_depth		= (tsk) ? tsk->lock_depth : 0;
990
	entry->flags =
991
#ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT
992
		(irqs_disabled_flags(flags) ? TRACE_FLAG_IRQS_OFF : 0) |
993
994
995
#else
		TRACE_FLAG_IRQS_NOSUPPORT |
#endif
996
997
998
999
		((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
		((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
		(need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
}
1000
EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
1001

1002
1003
1004
1005
1006
struct ring_buffer_event *
trace_buffer_lock_reserve(struct ring_buffer *buffer,
			  int type,
			  unsigned long len,
			  unsigned long flags, int pc)
1007
1008
1009
{
	struct ring_buffer_event *event;

1010
	event = ring_buffer_lock_reserve(buffer, len);
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
	if (event != NULL) {
		struct trace_entry *ent = ring_buffer_event_data(event);

		tracing_generic_entry_update(ent, flags, pc);
		ent->type = type;
	}

	return event;
}

1021
1022
1023
1024
1025
static inline void
__trace_buffer_unlock_commit(struct ring_buffer *buffer,
			     struct ring_buffer_event *event,
			     unsigned long flags, int pc,
			     int wake)
1026
{
1027
	ring_buffer_unlock_commit(buffer, event);
1028

1029
1030
	ftrace_trace_stack(buffer, flags, 6, pc);
	ftrace_trace_userstack(buffer, flags, pc);
1031
1032
1033
1034
1035

	if (wake)
		trace_wake_up();
}

1036
1037
1038
void trace_buffer_unlock_commit(struct ring_buffer *buffer,
				struct ring_buffer_event *event,
				unsigned long flags, int pc)
1039
{
1040
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1041
1042
}

1043
struct ring_buffer_event *
1044
1045
trace_current_buffer_lock_reserve(struct ring_buffer **current_rb,
				  int type, unsigned long len,
1046
1047
				  unsigned long flags, int pc)
{
1048
1049
	*current_rb = global_trace.buffer;
	return trace_buffer_lock_reserve(*current_rb,
1050
1051
					 type, len, flags, pc);
}
1052
EXPORT_SYMBOL_GPL(trace_current_buffer_lock_reserve);
1053

1054
1055
void trace_current_buffer_unlock_commit(struct ring_buffer *buffer,
					struct ring_buffer_event *event,
1056
1057
					unsigned long flags, int pc)
{
1058
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1059
}
1060
EXPORT_SYMBOL_GPL(trace_current_buffer_unlock_commit);
1061

1062
1063
1064
void trace_nowake_buffer_unlock_commit(struct ring_buffer *buffer,
				       struct ring_buffer_event *event,
				       unsigned long flags, int pc)
1065
{
1066
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 0);
1067
}
1068
EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit);
1069

1070
1071
void trace_current_buffer_discard_commit(struct ring_buffer *buffer,
					 struct ring_buffer_event *event)
1072
{
1073
	ring_buffer_discard_commit(buffer, event);
1074
}
1075
EXPORT_SYMBOL_GPL(trace_current_buffer_discard_commit);
1076

Ingo Molnar's avatar
Ingo Molnar committed
1077
void
1078
trace_function(struct trace_array *tr,
1079
1080
	       unsigned long ip, unsigned long parent_ip, unsigned long flags,
	       int pc)
1081
{
1082
	struct ftrace_event_call *call = &event_function;
1083
	struct ring_buffer *buffer = tr->buffer;
1084
	struct ring_buffer_event *event;
1085
	struct ftrace_entry *entry;
1086

1087
1088
1089
1090
	/* If we are reading the ring buffer, don't trace */
	if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled))))
		return;

1091
	event = trace_buffer_lock_reserve(buffer, TRACE_FN, sizeof(*entry),
1092
					  flags, pc);
1093
1094
1095
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
1096
1097
	entry->ip			= ip;
	entry->parent_ip		= parent_ip;
1098

1099
1100
	if (!filter_check_discard(call, entry, buffer, event))
		ring_buffer_unlock_commit(buffer, event);
1101
1102
}

Ingo Molnar's avatar
Ingo Molnar committed
1103
void
Ingo Molnar's avatar
Ingo Molnar committed
1104
ftrace(struct trace_array *tr, struct trace_array_cpu *data,
1105
1106
       unsigned long ip, unsigned long parent_ip, unsigned long flags,
       int pc)
Ingo Molnar's avatar
Ingo Molnar committed
1107
1108
{
	if (likely(!atomic_read(&data->disabled)))
1109
		trace_function(tr, ip, parent_ip, flags, pc);
Ingo Molnar's avatar
Ingo Molnar committed
1110
1111
}

1112
#ifdef CONFIG_STACKTRACE
1113
static void __ftrace_trace_stack(struct ring_buffer *buffer,
1114
1115
				 unsigned long flags,
				 int skip, int pc)
Ingo Molnar's avatar
Ingo Molnar committed
1116
{
1117
	struct ftrace_event_call *call = &event_kernel_stack;
1118
	struct ring_buffer_event *event;
1119
	struct stack_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
1120
1121
	struct stack_trace trace;

1122
	event = trace_buffer_lock_reserve(buffer, TRACE_STACK,
1123
					  sizeof(*entry), flags, pc);
1124
1125
1126
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
1127
	memset(&entry->caller, 0, sizeof(entry->caller));
Ingo Molnar's avatar
Ingo Molnar committed
1128
1129
1130
1131

	trace.nr_entries	= 0;
	trace.max_entries	= FTRACE_STACK_ENTRIES;
	trace.skip		= skip;
1132
	trace.entries		= entry->caller;
Ingo Molnar's avatar
Ingo Molnar committed
1133
1134

	save_stack_trace(&trace);
1135
1136
	if (!filter_check_discard(call, entry, buffer, event))
		ring_buffer_unlock_commit(buffer, event);
Ingo Molnar's avatar
Ingo Molnar committed
1137
1138
}

1139
1140
void ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags,
			int skip, int pc)
1141
1142
1143
1144
{
	if (!(trace_flags & TRACE_ITER_STACKTRACE))
		return;

1145
	__ftrace_trace_stack(buffer, flags, skip, pc);
1146
1147
}

1148
1149
void __trace_stack(struct trace_array *tr, unsigned long flags, int skip,
		   int pc)
1150
{
1151
	__ftrace_trace_stack(tr->buffer, flags, skip, pc);
1152
1153
}

1154
1155
void
ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc)
1156
{
1157
	struct ftrace_event_call *call = &event_user_stack;
1158
	struct ring_buffer_event *event;
1159
1160
1161
1162
1163
1164
	struct userstack_entry *entry;
	struct stack_trace trace;

	if (!(trace_flags & TRACE_ITER_USERSTACKTRACE))
		return;

1165
	event = trace_buffer_lock_reserve(buffer, TRACE_USER_STACK,
1166
					  sizeof(*entry), flags, pc);
1167
1168
1169
1170
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);

1171
	entry->tgid		= current->tgid;
1172
1173
1174
1175
1176
1177
1178
1179
	memset(&entry->caller, 0, sizeof(entry->caller));

	trace.nr_entries	= 0;
	trace.max_entries	= FTRACE_STACK_ENTRIES;
	trace.skip		= 0;
	trace.entries		= entry->caller;

	save_stack_trace_user(&trace);
1180
1181
	if (!filter_check_discard(call, entry, buffer, event))
		ring_buffer_unlock_commit(buffer, event);
1182
1183
}

1184
1185
#ifdef UNUSED
static void __trace_userstack(struct trace_array *tr, unsigned long flags)
1186
{
1187
	ftrace_trace_userstack(tr, flags, preempt_count());
1188
}
1189
#endif /* UNUSED */
1190

1191
1192
#endif /* CONFIG_STACKTRACE */

1193
static void
1194
ftrace_trace_special(void *__tr,
1195
1196
		     unsigned long arg1, unsigned long arg2, unsigned long arg3,
		     int pc)
Ingo Molnar's avatar
Ingo Molnar committed
1197
{
1198