trace.c 117 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 <generated/utsrelease.h>
16
17
#include <linux/stacktrace.h>
#include <linux/writeback.h>
18
19
#include <linux/kallsyms.h>
#include <linux/seq_file.h>
20
#include <linux/notifier.h>
21
#include <linux/irqflags.h>
22
#include <linux/debugfs.h>
23
#include <linux/pagemap.h>
24
25
26
#include <linux/hardirq.h>
#include <linux/linkage.h>
#include <linux/uaccess.h>
27
#include <linux/kprobes.h>
28
29
30
#include <linux/ftrace.h>
#include <linux/module.h>
#include <linux/percpu.h>
31
#include <linux/splice.h>
32
#include <linux/kdebug.h>
33
#include <linux/string.h>
34
#include <linux/rwsem.h>
35
#include <linux/slab.h>
36
37
#include <linux/ctype.h>
#include <linux/init.h>
38
#include <linux/poll.h>
39
#include <linux/nmi.h>
40
#include <linux/fs.h>
Ingo Molnar's avatar
Ingo Molnar committed
41

42
#include "trace.h"
43
#include "trace_output.h"
44

45
46
47
48
/*
 * 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.
 */
49
int ring_buffer_expanded;
50

51
52
/*
 * We need to change this state when a selftest is running.
53
54
 * A selftest will lurk into the ring-buffer to count the
 * entries inserted during the selftest although some concurrent
55
 * insertions into the ring-buffer such as trace_printk could occurred
56
57
 * at the same time, giving false positive or negative results.
 */
58
static bool __read_mostly tracing_selftest_running;
59

60
61
62
/*
 * If a tracer is running, we do not want to run SELFTEST.
 */
63
bool __read_mostly tracing_selftest_disabled;
64

65
66
67
68
69
70
71
72
73
74
75
76
77
78
/* 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;
}
79
80
81
82
83
84
85

/*
 * 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.
 */
86
static int tracing_disabled = 1;
87

88
DEFINE_PER_CPU(int, ftrace_cpu_disabled);
89

90
cpumask_var_t __read_mostly	tracing_buffer_mask;
91

92
93
94
95
96
97
98
99
100
101
102
/*
 * 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
103
104
105
 * /proc/sys/kernel/ftrace_dump_on_oops
 * Set 1 if you want to dump buffers of all CPUs
 * Set 2 if you want to dump the buffer of the CPU that triggered oops
106
 */
107
108

enum ftrace_dump_mode ftrace_dump_on_oops;
109

110
111
static int tracing_set_tracer(const char *buf);

Li Zefan's avatar
Li Zefan committed
112
113
#define MAX_TRACER_SIZE		100
static char bootup_tracer_buf[MAX_TRACER_SIZE] __initdata;
114
static char *default_bootup_tracer;
115

116
static int __init set_cmdline_ftrace(char *str)
117
{
Li Zefan's avatar
Li Zefan committed
118
	strncpy(bootup_tracer_buf, str, MAX_TRACER_SIZE);
119
	default_bootup_tracer = bootup_tracer_buf;
120
121
	/* We are using ftrace early, expand it */
	ring_buffer_expanded = 1;
122
123
	return 1;
}
124
__setup("ftrace=", set_cmdline_ftrace);
125

126
127
static int __init set_ftrace_dump_on_oops(char *str)
{
128
129
130
131
132
133
134
135
136
137
138
	if (*str++ != '=' || !*str) {
		ftrace_dump_on_oops = DUMP_ALL;
		return 1;
	}

	if (!strcmp("orig_cpu", str)) {
		ftrace_dump_on_oops = DUMP_ORIG;
                return 1;
        }

        return 0;
139
140
}
__setup("ftrace_dump_on_oops", set_ftrace_dump_on_oops);
Steven Rostedt's avatar
Steven Rostedt committed
141

142
unsigned long long ns2usecs(cycle_t nsec)
143
144
145
146
147
148
{
	nsec += 500;
	do_div(nsec, 1000);
	return nsec;
}

Steven Rostedt's avatar
Steven Rostedt committed
149
150
151
152
153
154
155
156
157
158
159
160
/*
 * 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.
 */
161
162
163
164
static struct trace_array	global_trace;

static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu);

165
166
int filter_current_check_discard(struct ring_buffer *buffer,
				 struct ftrace_event_call *call, void *rec,
167
168
				 struct ring_buffer_event *event)
{
169
	return filter_check_discard(call, rec, buffer, event);
170
}
171
EXPORT_SYMBOL_GPL(filter_current_check_discard);
172

173
174
175
176
177
178
179
180
181
182
183
184
185
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;
}
186

Steven Rostedt's avatar
Steven Rostedt committed
187
188
189
190
191
192
193
194
195
196
/*
 * 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.
 */
197
198
static struct trace_array	max_tr;

199
static DEFINE_PER_CPU(struct trace_array_cpu, max_tr_data);
200

Steven Rostedt's avatar
Steven Rostedt committed
201
/* tracer_enabled is used to toggle activation of a tracer */
202
static int			tracer_enabled = 1;
Steven Rostedt's avatar
Steven Rostedt committed
203

204
205
206
207
208
209
210
211
212
213
214
215
216
/**
 * 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
217
/*
218
219
220
 * 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.
221
222
223
224
225
 *
 * 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
226
 */
227
#define TRACE_BUF_SIZE_DEFAULT	1441792UL /* 16384 * 88 (sizeof(entry)) */
228

229
static unsigned long		trace_buf_size = TRACE_BUF_SIZE_DEFAULT;
230

Steven Rostedt's avatar
Steven Rostedt committed
231
/* trace_types holds a link list of available tracers. */
232
static struct tracer		*trace_types __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
233
234

/* current_trace points to the tracer that is currently active */
235
static struct tracer		*current_trace __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
236
237
238
239

/*
 * trace_types_lock is used to protect the trace_types list.
 */
240
static DEFINE_MUTEX(trace_types_lock);
Steven Rostedt's avatar
Steven Rostedt committed
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
/*
 * serialize the access of the ring buffer
 *
 * ring buffer serializes readers, but it is low level protection.
 * The validity of the events (which returns by ring_buffer_peek() ..etc)
 * are not protected by ring buffer.
 *
 * The content of events may become garbage if we allow other process consumes
 * these events concurrently:
 *   A) the page of the consumed events may become a normal page
 *      (not reader page) in ring buffer, and this page will be rewrited
 *      by events producer.
 *   B) The page of the consumed events may become a page for splice_read,
 *      and this page will be returned to system.
 *
 * These primitives allow multi process access to different cpu ring buffer
 * concurrently.
 *
 * These primitives don't distinguish read-only and read-consume access.
 * Multi read-only access are also serialized.
 */

#ifdef CONFIG_SMP
static DECLARE_RWSEM(all_cpu_access_lock);
static DEFINE_PER_CPU(struct mutex, cpu_access_lock);

static inline void trace_access_lock(int cpu)
{
	if (cpu == TRACE_PIPE_ALL_CPU) {
		/* gain it for accessing the whole ring buffer. */
		down_write(&all_cpu_access_lock);
	} else {
		/* gain it for accessing a cpu ring buffer. */

		/* Firstly block other trace_access_lock(TRACE_PIPE_ALL_CPU). */
		down_read(&all_cpu_access_lock);

		/* Secondly block other access to this @cpu ring buffer. */
		mutex_lock(&per_cpu(cpu_access_lock, cpu));
	}
}

static inline void trace_access_unlock(int cpu)
{
	if (cpu == TRACE_PIPE_ALL_CPU) {
		up_write(&all_cpu_access_lock);
	} else {
		mutex_unlock(&per_cpu(cpu_access_lock, cpu));
		up_read(&all_cpu_access_lock);
	}
}

static inline void trace_access_lock_init(void)
{
	int cpu;

	for_each_possible_cpu(cpu)
		mutex_init(&per_cpu(cpu_access_lock, cpu));
}

#else

static DEFINE_MUTEX(access_lock);

static inline void trace_access_lock(int cpu)
{
	(void)cpu;
	mutex_lock(&access_lock);
}

static inline void trace_access_unlock(int cpu)
{
	(void)cpu;
	mutex_unlock(&access_lock);
}

static inline void trace_access_lock_init(void)
{
}

#endif

Steven Rostedt's avatar
Steven Rostedt committed
324
/* trace_wait is a waitqueue for tasks blocked on trace_poll */
325
326
static DECLARE_WAIT_QUEUE_HEAD(trace_wait);

327
/* trace_flags holds trace_options default values */
328
unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
329
	TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME |
330
331
	TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE |
	TRACE_ITER_IRQ_INFO;
332

333
static int trace_stop_count;
334
static DEFINE_RAW_SPINLOCK(tracing_start_lock);
335

336
337
338
339
340
341
342
static void wakeup_work_handler(struct work_struct *work)
{
	wake_up(&trace_wait);
}

static DECLARE_DELAYED_WORK(wakeup_work, wakeup_work_handler);

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
/**
 * tracing_on - enable tracing buffers
 *
 * This function enables tracing buffers that may have been
 * disabled with tracing_off.
 */
void tracing_on(void)
{
	if (global_trace.buffer)
		ring_buffer_record_on(global_trace.buffer);
	/*
	 * This flag is only looked at when buffers haven't been
	 * allocated yet. We don't really care about the race
	 * between setting this flag and actually turning
	 * on the buffer.
	 */
	global_trace.buffer_disabled = 0;
}
EXPORT_SYMBOL_GPL(tracing_on);

/**
 * tracing_off - turn off tracing buffers
 *
 * This function stops the tracing buffers from recording data.
 * It does not disable any overhead the tracers themselves may
 * be causing. This function simply causes all recording to
 * the ring buffers to fail.
 */
void tracing_off(void)
{
	if (global_trace.buffer)
374
		ring_buffer_record_off(global_trace.buffer);
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
	/*
	 * This flag is only looked at when buffers haven't been
	 * allocated yet. We don't really care about the race
	 * between setting this flag and actually turning
	 * on the buffer.
	 */
	global_trace.buffer_disabled = 1;
}
EXPORT_SYMBOL_GPL(tracing_off);

/**
 * tracing_is_on - show state of ring buffers enabled
 */
int tracing_is_on(void)
{
	if (global_trace.buffer)
		return ring_buffer_record_is_on(global_trace.buffer);
	return !global_trace.buffer_disabled;
}
EXPORT_SYMBOL_GPL(tracing_is_on);

Steven Rostedt's avatar
Steven Rostedt committed
396
397
398
/**
 * trace_wake_up - wake up tasks waiting for trace input
 *
399
400
401
 * Schedules a delayed work to wake up any task that is blocked on the
 * trace_wait queue. These is used with trace_poll for tasks polling the
 * trace.
Steven Rostedt's avatar
Steven Rostedt committed
402
 */
403
404
void trace_wake_up(void)
{
405
	const unsigned long delay = msecs_to_jiffies(2);
406
407
408

	if (trace_flags & TRACE_ITER_BLOCK)
		return;
409
	schedule_delayed_work(&wakeup_work, delay);
410
}
411

412
static int __init set_buf_size(char *str)
413
{
414
	unsigned long buf_size;
415

416
417
	if (!str)
		return 0;
418
	buf_size = memparse(str, &str);
419
	/* nr_entries can not be zero */
420
	if (buf_size == 0)
421
		return 0;
422
	trace_buf_size = buf_size;
423
424
	return 1;
}
425
__setup("trace_buf_size=", set_buf_size);
426

427
428
static int __init set_tracing_thresh(char *str)
{
429
	unsigned long threshold;
430
431
432
433
	int ret;

	if (!str)
		return 0;
434
	ret = strict_strtoul(str, 0, &threshold);
435
436
	if (ret < 0)
		return 0;
437
	tracing_thresh = threshold * 1000;
438
439
440
441
	return 1;
}
__setup("tracing_thresh=", set_tracing_thresh);

Steven Rostedt's avatar
Steven Rostedt committed
442
443
444
445
446
unsigned long nsecs_to_usecs(unsigned long nsecs)
{
	return nsecs / 1000;
}

Steven Rostedt's avatar
Steven Rostedt committed
447
/* These must match the bit postions in trace_iterator_flags */
448
449
450
451
452
static const char *trace_options[] = {
	"print-parent",
	"sym-offset",
	"sym-addr",
	"verbose",
Ingo Molnar's avatar
Ingo Molnar committed
453
	"raw",
454
	"hex",
Ingo Molnar's avatar
Ingo Molnar committed
455
	"bin",
456
	"block",
Ingo Molnar's avatar
Ingo Molnar committed
457
	"stacktrace",
458
	"trace_printk",
459
	"ftrace_preempt",
460
	"branch",
461
	"annotate",
462
	"userstacktrace",
463
	"sym-userobj",
464
	"printk-msg-only",
465
	"context-info",
466
	"latency-format",
467
	"sleep-time",
468
	"graph-time",
469
	"record-cmd",
470
	"overwrite",
471
	"disable_on_free",
472
	"irq-info",
473
474
475
	NULL
};

476
477
478
479
480
481
static struct {
	u64 (*func)(void);
	const char *name;
} trace_clocks[] = {
	{ trace_clock_local,	"local" },
	{ trace_clock_global,	"global" },
482
	{ trace_clock_counter,	"counter" },
483
484
485
486
};

int trace_clock_id;

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
/*
 * 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)) {
564
		if (parser->idx < parser->size - 1)
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
			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;
}

593
594
595
596
597
ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt)
{
	int len;
	int ret;

598
599
600
	if (!cnt)
		return 0;

601
602
603
604
605
606
607
	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);
608
	if (ret == cnt)
609
610
		return -EFAULT;

611
612
	cnt -= ret;

613
	s->readpos += cnt;
614
	return cnt;
615
616
}

617
static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt)
618
619
620
621
622
623
624
625
626
{
	int len;

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

	len = s->len - s->readpos;
	if (cnt > len)
		cnt = len;
627
	memcpy(buf, s->buffer + s->readpos, cnt);
628

629
	s->readpos += cnt;
630
631
632
	return cnt;
}

633
634
635
636
637
638
/*
 * 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.
 *
639
 * This is defined as a arch_spinlock_t in order to help
640
641
642
643
644
645
 * 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.
 */
646
static arch_spinlock_t ftrace_max_lock =
647
	(arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED;
648

649
650
unsigned long __read_mostly	tracing_thresh;

651
652
653
654
655
656
657
658
659
660
661
662
#ifdef CONFIG_TRACER_MAX_TRACE
unsigned long __read_mostly	tracing_max_latency;

/*
 * 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];
663
	struct trace_array_cpu *max_data;
664
665
666
667

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

668
669
670
671
	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;
672

673
	memcpy(max_data->comm, tsk->comm, TASK_COMM_LEN);
674
675
676
677
678
	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;
679
680
681
682
683

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

Steven Rostedt's avatar
Steven Rostedt committed
684
685
686
687
688
689
690
691
692
/**
 * 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
693
void
694
695
update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
696
	struct ring_buffer *buf = tr->buffer;
697

698
699
700
	if (trace_stop_count)
		return;

701
	WARN_ON_ONCE(!irqs_disabled());
702
703
704
705
	if (!current_trace->use_max_tr) {
		WARN_ON_ONCE(1);
		return;
	}
706
	arch_spin_lock(&ftrace_max_lock);
707
708
709
710

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

711
	__update_max_tr(tr, tsk, cpu);
712
	arch_spin_unlock(&ftrace_max_lock);
713
714
715
716
717
718
719
}

/**
 * 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
720
721
 *
 * Flip the trace of a single CPU buffer between the @tr and the max_tr.
722
 */
Ingo Molnar's avatar
Ingo Molnar committed
723
void
724
725
update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
726
	int ret;
727

728
729
730
	if (trace_stop_count)
		return;

731
	WARN_ON_ONCE(!irqs_disabled());
732
733
734
735
736
	if (!current_trace->use_max_tr) {
		WARN_ON_ONCE(1);
		return;
	}

737
	arch_spin_lock(&ftrace_max_lock);
738

739
740
	ret = ring_buffer_swap_cpu(max_tr.buffer, tr->buffer, cpu);

741
742
743
744
745
746
747
748
749
750
751
752
	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");
	}

	WARN_ON_ONCE(ret && ret != -EAGAIN && ret != -EBUSY);
753
754

	__update_max_tr(tr, tsk, cpu);
755
	arch_spin_unlock(&ftrace_max_lock);
756
}
757
#endif /* CONFIG_TRACER_MAX_TRACE */
758

Steven Rostedt's avatar
Steven Rostedt committed
759
760
761
762
763
764
/**
 * register_tracer - register a tracer with the ftrace system.
 * @type - the plugin for the tracer
 *
 * Register a new plugin tracer.
 */
765
766
767
768
769
770
771
772
773
774
int register_tracer(struct tracer *type)
{
	struct tracer *t;
	int ret = 0;

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

775
	if (strlen(type->name) >= MAX_TRACER_SIZE) {
Li Zefan's avatar
Li Zefan committed
776
777
778
779
		pr_info("Tracer has a name longer than %d\n", MAX_TRACER_SIZE);
		return -1;
	}

780
	mutex_lock(&trace_types_lock);
Ingo Molnar's avatar
Ingo Molnar committed
781

782
783
	tracing_selftest_running = true;

784
785
786
	for (t = trace_types; t; t = t->next) {
		if (strcmp(type->name, t->name) == 0) {
			/* already found */
Li Zefan's avatar
Li Zefan committed
787
			pr_info("Tracer %s already registered\n",
788
789
790
791
792
793
				type->name);
			ret = -1;
			goto out;
		}
	}

794
795
796
797
798
799
800
	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;
801
802
803
	if (!type->wait_pipe)
		type->wait_pipe = default_wait_pipe;

804

Steven Rostedt's avatar
Steven Rostedt committed
805
#ifdef CONFIG_FTRACE_STARTUP_TEST
806
	if (type->selftest && !tracing_selftest_disabled) {
Steven Rostedt's avatar
Steven Rostedt committed
807
808
		struct tracer *saved_tracer = current_trace;
		struct trace_array *tr = &global_trace;
809

Steven Rostedt's avatar
Steven Rostedt committed
810
811
812
813
814
815
816
		/*
		 * 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.
		 */
817
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
818

Steven Rostedt's avatar
Steven Rostedt committed
819
		current_trace = type;
820
821
822

		/* If we expanded the buffers, make sure the max is expanded too */
		if (ring_buffer_expanded && type->use_max_tr)
823
824
			ring_buffer_resize(max_tr.buffer, trace_buf_size,
						RING_BUFFER_ALL_CPUS);
825

Steven Rostedt's avatar
Steven Rostedt committed
826
827
828
829
830
831
832
		/* 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");
833
834
			/* Add the warning after printing 'FAILED' */
			WARN_ON(1);
Steven Rostedt's avatar
Steven Rostedt committed
835
836
			goto out;
		}
Steven Rostedt's avatar
Steven Rostedt committed
837
		/* Only reset on passing, to avoid touching corrupted buffers */
838
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
839

840
841
		/* Shrink the max buffer again */
		if (ring_buffer_expanded && type->use_max_tr)
842
843
			ring_buffer_resize(max_tr.buffer, 1,
						RING_BUFFER_ALL_CPUS);
844

Steven Rostedt's avatar
Steven Rostedt committed
845
846
847
848
		printk(KERN_CONT "PASSED\n");
	}
#endif

849
850
	type->next = trace_types;
	trace_types = type;
Steven Rostedt's avatar
Steven Rostedt committed
851

852
 out:
853
	tracing_selftest_running = false;
854
855
	mutex_unlock(&trace_types_lock);

Steven Rostedt's avatar
Steven Rostedt committed
856
857
858
	if (ret || !default_bootup_tracer)
		goto out_unlock;

Li Zefan's avatar
Li Zefan committed
859
	if (strncmp(default_bootup_tracer, type->name, MAX_TRACER_SIZE))
Steven Rostedt's avatar
Steven Rostedt committed
860
861
862
863
864
865
866
867
		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;
868
#ifdef CONFIG_FTRACE_STARTUP_TEST
Steven Rostedt's avatar
Steven Rostedt committed
869
870
	printk(KERN_INFO "Disabling FTRACE selftests due to running tracer '%s'\n",
	       type->name);
871
872
#endif

Steven Rostedt's avatar
Steven Rostedt committed
873
 out_unlock:
874
875
876
877
878
879
880
881
882
883
884
885
	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
886
	pr_info("Tracer %s not registered\n", type->name);
887
888
889
890
	goto out;

 found:
	*t = (*t)->next;
891
892
893
894
895
896
897
898

	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
899
out:
900
901
902
	mutex_unlock(&trace_types_lock);
}

903
904
905
906
907
908
909
910
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();
911
	ring_buffer_reset_cpu(buffer, cpu);
912
913
914
915

	ring_buffer_record_enable(buffer);
}

916
917
void tracing_reset_online_cpus(struct trace_array *tr)
{
918
	struct ring_buffer *buffer = tr->buffer;
919
920
	int cpu;

921
922
923
924
925
	ring_buffer_record_disable(buffer);

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

926
927
928
	tr->time_start = ftrace_now(tr->cpu);

	for_each_online_cpu(cpu)
929
		ring_buffer_reset_cpu(buffer, cpu);
930
931

	ring_buffer_record_enable(buffer);
932
933
}

934
935
936
937
938
939
940
941
942
943
void tracing_reset_current(int cpu)
{
	tracing_reset(&global_trace, cpu);
}

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

944
#define SAVED_CMDLINES 128
945
#define NO_CMDLINE_MAP UINT_MAX
946
947
948
949
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;
950
static arch_spinlock_t trace_cmdline_lock = __ARCH_SPIN_LOCK_UNLOCKED;
951
952

/* temporary disable recording */
953
static atomic_t trace_record_cmdline_disabled __read_mostly;
954
955
956

static void trace_init_cmdlines(void)
{
957
958
	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));
959
960
961
	cmdline_idx = 0;
}

962
963
964
965
966
int is_tracing_stopped(void)
{
	return trace_stop_count;
}

967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
/**
 * 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();
}

982
983
984
985
986
987
988
989
990
991
992
993
994
995
/**
 * 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;

996
	raw_spin_lock_irqsave(&tracing_start_lock, flags);
997
998
999
1000
1001
1002
	if (--trace_stop_count) {
		if (trace_stop_count < 0) {
			/* Someone screwed up their debugging */
			WARN_ON_ONCE(1);
			trace_stop_count = 0;
		}
1003
1004
1005
		goto out;
	}

1006
1007
	/* Prevent the buffers from switching */
	arch_spin_lock(&ftrace_max_lock);
1008
1009
1010
1011
1012
1013
1014
1015
1016

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

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

1017
1018
	arch_spin_unlock(&ftrace_max_lock);

1019
1020
	ftrace_start();
 out:
1021
	raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
}

/**
 * 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();
1036
	raw_spin_lock_irqsave(&tracing_start_lock, flags);
1037
1038
1039
	if (trace_stop_count++)
		goto out;

1040
1041
1042
	/* Prevent the buffers from switching */
	arch_spin_lock(&ftrace_max_lock);

1043
1044
1045
1046
1047
1048
1049
1050
	buffer = global_trace.buffer;
	if (buffer)
		ring_buffer_record_disable(buffer);

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

1051
1052
	arch_spin_unlock(&ftrace_max_lock);

1053
 out:
1054
	raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
1055
1056
}

Ingo Molnar's avatar
Ingo Molnar committed
1057
void trace_stop_cmdline_recording(void);
1058

Ingo Molnar's avatar
Ingo Molnar committed
1059
static void trace_save_cmdline(struct task_struct *tsk)
1060
{
1061
	unsigned pid, idx;
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071

	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.
	 */
1072
	if (!arch_spin_trylock(&trace_cmdline_lock))
1073
1074
1075
		return;

	idx = map_pid_to_cmdline[tsk->pid];
1076
	if (idx == NO_CMDLINE_MAP) {
1077
1078
		idx = (cmdline_idx + 1) % SAVED_CMDLINES;

1079
1080
1081
1082
1083
1084
1085
1086
1087
		/*
		 * 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;
1088

1089
		map_cmdline_to_pid[idx] = tsk->pid;
1090
1091
1092
1093
1094
1095
1096
		map_pid_to_cmdline[tsk->pid] = idx;

		cmdline_idx = idx;
	}

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

1097
	arch_spin_unlock(&trace_cmdline_lock);
1098
1099
}

1100
void trace_find_cmdline(int pid, char comm[])
1101
1102
1103
{
	unsigned map;

1104
1105
1106
1107
	if (!pid) {
		strcpy(comm, "<idle>");
		return;
	}
1108

1109
1110
1111
1112
1113
	if (WARN_ON_ONCE(pid < 0)) {
		strcpy(comm, "<XXX>");
		return;
	}

1114
1115
1116
1117
	if (pid > PID_MAX_DEFAULT) {
		strcpy(comm, "<...>");
		return;
	}
1118

1119
	preempt_disable();
1120
	arch_spin_lock(&trace_cmdline_lock);
1121
	map = map_pid_to_cmdline[pid];
1122
1123
1124
1125
	if (map != NO_CMDLINE_MAP)
		strcpy(comm, saved_cmdlines[map]);
	else
		strcpy(comm, "<...>");
1126

1127
	arch_spin_unlock(&trace_cmdline_lock);
1128
	preempt_enable();
1129
1130
}

Ingo Molnar's avatar
Ingo Molnar committed
1131
void tracing_record_cmdline(struct task_struct *tsk)
1132
{
1133
1134
	if (atomic_read(&trace_record_cmdline_disabled) || !tracer_enabled ||
	    !tracing_is_on())
1135
1136
1137
1138
1139
		return;

	trace_save_cmdline(tsk);
}

1140
void
1141
1142
tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
			     int pc)
1143
1144
1145
{
	struct task_struct *tsk = current;

1146
1147
	entry->preempt_count		= pc & 0xff;
	entry->pid			= (tsk) ? tsk->pid : 0;
1148
	entry->padding			= 0;
1149
	entry->flags =
1150
#ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT
1151
		(irqs_disabled_flags(flags) ? TRACE_FLAG_IRQS_OFF : 0) |
1152
1153
1154
#else
		TRACE_FLAG_IRQS_NOSUPPORT |
#endif
1155
1156
1157
1158
		((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
		((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
		(need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
}
1159
EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
1160

1161
1162
1163
1164
1165
struct ring_buffer_event *
trace_buffer_lock_reserve(struct ring_buffer *buffer,
			  int type,
			  unsigned long len,
			  unsigned long flags, int pc)
1166
1167
1168
{
	struct ring_buffer_event *event;

1169
	event = ring_buffer_lock_reserve(buffer, len);
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
	if (event != NULL) {
		struct trace_entry *ent = ring_buffer_event_data(event);

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

	return event;
}

1180
1181
1182
1183
1184
static inline void
__trace_buffer_unlock_commit(struct ring_buffer *buffer,
			     struct ring_buffer_event *event,
			     unsigned long flags, int pc,
			     int wake)
1185
{
1186
	ring_buffer_unlock_commit(buffer, event);
1187

1188
1189
	ftrace_trace_stack(buffer, flags, 6, pc);
	ftrace_trace_userstack(buffer, flags, pc);
1190
1191
1192
1193
1194

	if (wake)
		trace_wake_up();
}

1195
1196
1197
void trace_buffer_unlock_commit(struct ring_buffer *buffer,
				struct ring_buffer_event *event,
				unsigned long flags, int pc)
1198
{
1199
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1200
1201
}

1202
struct ring_buffer_event *
1203
1204
trace_current_buffer_lock_reserve(struct ring_buffer **current_rb,
				  int type, unsigned long len,
1205
1206
				  unsigned long flags, int pc)
{
1207
1208
	*current_rb = global_trace.buffer;
	return trace_buffer_lock_reserve(*current_rb,
1209
1210
					 type, len, flags, pc);
}
1211
EXPORT_SYMBOL_GPL(trace_current_buffer_lock_reserve);
1212

1213
1214
void trace_current_buffer_unlock_commit(struct ring_buffer *buffer,
					struct ring_buffer_event *event,
1215
1216
					unsigned long flags, int pc)
{
1217
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1218
}
1219
EXPORT_SYMBOL_GPL(trace_current_buffer_unlock_commit);
1220

1221
1222
1223
void trace_nowake_buffer_unlock_commit(struct ring_buffer *buffer,
				       struct ring_buffer_event *event,
				       unsigned long flags, int pc)
1224
{
1225
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 0);
1226
}
1227
EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit);
1228

1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
void trace_nowake_buffer_unlock_commit_regs(struct ring_buffer *buffer,
					    struct ring_buffer_event *event,
					    unsigned long flags, int pc,
					    struct pt_regs *regs)
{
	ring_buffer_unlock_commit(buffer, event);

	ftrace_trace_stack_regs(buffer, flags, 0, pc, regs);
	ftrace_trace_userstack(buffer, flags, pc);
}
EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit_regs);

1241
1242
void trace_current_buffer_discard_commit(struct ring_buffer *buffer,
					 struct ring_buffer_event *event)
1243
{
1244
	ring_buffer_discard_commit(buffer, event);
1245
}
1246
EXPORT_SYMBOL_GPL(trace_current_buffer_discard_commit);
1247

Ingo Molnar's avatar
Ingo Molnar committed
1248
void
1249
trace_function(struct trace_array *tr,
1250
1251
	       unsigned long ip, unsigned long parent_ip, unsigned long flags,
	       int pc)
1252
{
1253
	struct ftrace_event_call *call = &event_function;
1254
	struct ring_buffer *buffer = tr->buffer;
1255
	struct ring_buffer_event *event;
1256
	struct ftrace_entry *entry;