trace.c 103 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>
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
#include <linux/rwsem.h>
36
#include <linux/slab.h>
37
38
#include <linux/ctype.h>
#include <linux/init.h>
39
#include <linux/poll.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
#define TRACE_BUFFER_FLAGS	(RB_FL_OVERWRITE)

47
48
49
50
/*
 * 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.
 */
51
int ring_buffer_expanded;
52

53
54
/*
 * We need to change this state when a selftest is running.
55
56
 * A selftest will lurk into the ring-buffer to count the
 * entries inserted during the selftest although some concurrent
57
 * insertions into the ring-buffer such as trace_printk could occurred
58
59
 * at the same time, giving false positive or negative results.
 */
60
static bool __read_mostly tracing_selftest_running;
61

62
63
64
/*
 * If a tracer is running, we do not want to run SELFTEST.
 */
65
bool __read_mostly tracing_selftest_disabled;
66

67
68
69
70
71
72
73
74
75
76
77
78
79
80
/* 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;
}
81
82
83
84
85
86
87

/*
 * 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.
 */
88
static int tracing_disabled = 1;
89

90
DEFINE_PER_CPU(int, ftrace_cpu_disabled);
91
92
93
94

static inline void ftrace_disable_cpu(void)
{
	preempt_disable();
95
	__this_cpu_inc(ftrace_cpu_disabled);
96
97
98
99
}

static inline void ftrace_enable_cpu(void)
{
100
	__this_cpu_dec(ftrace_cpu_disabled);
101
102
103
	preempt_enable();
}

104
static cpumask_var_t __read_mostly	tracing_buffer_mask;
105
106

#define for_each_tracing_cpu(cpu)	\
107
	for_each_cpu(cpu, tracing_buffer_mask)
108

109
110
111
112
113
114
115
116
117
118
119
/*
 * 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
120
121
122
 * /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
123
 */
124
125

enum ftrace_dump_mode ftrace_dump_on_oops;
126

127
128
static int tracing_set_tracer(const char *buf);

Li Zefan's avatar
Li Zefan committed
129
130
#define MAX_TRACER_SIZE		100
static char bootup_tracer_buf[MAX_TRACER_SIZE] __initdata;
131
static char *default_bootup_tracer;
132

133
static int __init set_cmdline_ftrace(char *str)
134
{
Li Zefan's avatar
Li Zefan committed
135
	strncpy(bootup_tracer_buf, str, MAX_TRACER_SIZE);
136
	default_bootup_tracer = bootup_tracer_buf;
137
138
	/* We are using ftrace early, expand it */
	ring_buffer_expanded = 1;
139
140
	return 1;
}
141
__setup("ftrace=", set_cmdline_ftrace);
142

143
144
static int __init set_ftrace_dump_on_oops(char *str)
{
145
146
147
148
149
150
151
152
153
154
155
	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;
156
157
}
__setup("ftrace_dump_on_oops", set_ftrace_dump_on_oops);
Steven Rostedt's avatar
Steven Rostedt committed
158

159
unsigned long long ns2usecs(cycle_t nsec)
160
161
162
163
164
165
{
	nsec += 500;
	do_div(nsec, 1000);
	return nsec;
}

Steven Rostedt's avatar
Steven Rostedt committed
166
167
168
169
170
171
172
173
174
175
176
177
/*
 * 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.
 */
178
179
180
181
static struct trace_array	global_trace;

static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu);

182
183
int filter_current_check_discard(struct ring_buffer *buffer,
				 struct ftrace_event_call *call, void *rec,
184
185
				 struct ring_buffer_event *event)
{
186
	return filter_check_discard(call, rec, buffer, event);
187
}
188
EXPORT_SYMBOL_GPL(filter_current_check_discard);
189

190
191
192
193
194
195
196
197
198
199
200
201
202
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;
}
203

Steven Rostedt's avatar
Steven Rostedt committed
204
205
206
207
208
209
210
211
212
213
/*
 * 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.
 */
214
215
static struct trace_array	max_tr;

216
static DEFINE_PER_CPU(struct trace_array_cpu, max_tr_data);
217

Steven Rostedt's avatar
Steven Rostedt committed
218
/* tracer_enabled is used to toggle activation of a tracer */
219
static int			tracer_enabled = 1;
Steven Rostedt's avatar
Steven Rostedt committed
220

221
222
223
224
225
226
227
228
229
230
231
232
233
/**
 * 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
234
/*
235
236
237
 * 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.
238
239
240
241
242
 *
 * 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
243
 */
244
#define TRACE_BUF_SIZE_DEFAULT	1441792UL /* 16384 * 88 (sizeof(entry)) */
245

246
static unsigned long		trace_buf_size = TRACE_BUF_SIZE_DEFAULT;
247

Steven Rostedt's avatar
Steven Rostedt committed
248
/* trace_types holds a link list of available tracers. */
249
static struct tracer		*trace_types __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
250
251

/* current_trace points to the tracer that is currently active */
252
static struct tracer		*current_trace __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
253
254
255
256

/*
 * trace_types_lock is used to protect the trace_types list.
 */
257
static DEFINE_MUTEX(trace_types_lock);
Steven Rostedt's avatar
Steven Rostedt committed
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
/*
 * 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
341
/* trace_wait is a waitqueue for tasks blocked on trace_poll */
342
343
static DECLARE_WAIT_QUEUE_HEAD(trace_wait);

344
/* trace_flags holds trace_options default values */
345
unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
346
347
	TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME |
	TRACE_ITER_GRAPH_TIME;
348

349
350
351
static int trace_stop_count;
static DEFINE_SPINLOCK(tracing_start_lock);

Steven Rostedt's avatar
Steven Rostedt committed
352
353
354
355
356
357
/**
 * 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.
 */
358
359
void trace_wake_up(void)
{
360
361
362
363
	int cpu;

	if (trace_flags & TRACE_ITER_BLOCK)
		return;
Ingo Molnar's avatar
Ingo Molnar committed
364
365
366
367
	/*
	 * The runqueue_is_locked() can fail, but this is the best we
	 * have for now:
	 */
368
369
	cpu = get_cpu();
	if (!runqueue_is_locked(cpu))
370
		wake_up(&trace_wait);
371
	put_cpu();
372
}
373

374
static int __init set_buf_size(char *str)
375
{
376
	unsigned long buf_size;
377

378
379
	if (!str)
		return 0;
380
	buf_size = memparse(str, &str);
381
	/* nr_entries can not be zero */
382
	if (buf_size == 0)
383
		return 0;
384
	trace_buf_size = buf_size;
385
386
	return 1;
}
387
__setup("trace_buf_size=", set_buf_size);
388

389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
static int __init set_tracing_thresh(char *str)
{
	unsigned long threshhold;
	int ret;

	if (!str)
		return 0;
	ret = strict_strtoul(str, 0, &threshhold);
	if (ret < 0)
		return 0;
	tracing_thresh = threshhold * 1000;
	return 1;
}
__setup("tracing_thresh=", set_tracing_thresh);

Steven Rostedt's avatar
Steven Rostedt committed
404
405
406
407
408
unsigned long nsecs_to_usecs(unsigned long nsecs)
{
	return nsecs / 1000;
}

Steven Rostedt's avatar
Steven Rostedt committed
409
/* These must match the bit postions in trace_iterator_flags */
410
411
412
413
414
static const char *trace_options[] = {
	"print-parent",
	"sym-offset",
	"sym-addr",
	"verbose",
Ingo Molnar's avatar
Ingo Molnar committed
415
	"raw",
416
	"hex",
Ingo Molnar's avatar
Ingo Molnar committed
417
	"bin",
418
	"block",
Ingo Molnar's avatar
Ingo Molnar committed
419
	"stacktrace",
420
	"trace_printk",
421
	"ftrace_preempt",
422
	"branch",
423
	"annotate",
424
	"userstacktrace",
425
	"sym-userobj",
426
	"printk-msg-only",
427
	"context-info",
428
	"latency-format",
429
	"sleep-time",
430
	"graph-time",
431
432
433
	NULL
};

434
435
436
437
438
439
440
441
442
443
static struct {
	u64 (*func)(void);
	const char *name;
} trace_clocks[] = {
	{ trace_clock_local,	"local" },
	{ trace_clock_global,	"global" },
};

int trace_clock_id;

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
/*
 * 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)) {
521
		if (parser->idx < parser->size - 1)
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
			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;
}

550
551
552
553
554
ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt)
{
	int len;
	int ret;

555
556
557
	if (!cnt)
		return 0;

558
559
560
561
562
563
564
	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);
565
	if (ret == cnt)
566
567
		return -EFAULT;

568
569
	cnt -= ret;

570
	s->readpos += cnt;
571
	return cnt;
572
573
}

574
static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt)
575
576
577
578
579
580
581
582
583
584
585
586
587
588
{
	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;

589
	s->readpos += cnt;
590
591
592
	return cnt;
}

593
594
595
596
597
598
/*
 * 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.
 *
599
 * This is defined as a arch_spinlock_t in order to help
600
601
602
603
604
605
 * 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.
 */
606
static arch_spinlock_t ftrace_max_lock =
607
	(arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED;
608

609
610
unsigned long __read_mostly	tracing_thresh;

611
612
613
614
615
616
617
618
619
620
621
622
#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];
623
	struct trace_array_cpu *max_data;
624
625
626
627

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

628
629
630
631
	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;
632

633
	memcpy(max_data->comm, tsk->comm, TASK_COMM_LEN);
634
635
636
637
638
	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;
639
640
641
642
643

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

Steven Rostedt's avatar
Steven Rostedt committed
644
645
646
647
648
649
650
651
652
/**
 * 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
653
void
654
655
update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
656
	struct ring_buffer *buf = tr->buffer;
657

658
659
660
	if (trace_stop_count)
		return;

661
	WARN_ON_ONCE(!irqs_disabled());
662
	arch_spin_lock(&ftrace_max_lock);
663
664
665
666

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

667
	__update_max_tr(tr, tsk, cpu);
668
	arch_spin_unlock(&ftrace_max_lock);
669
670
671
672
673
674
675
}

/**
 * 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
676
677
 *
 * Flip the trace of a single CPU buffer between the @tr and the max_tr.
678
 */
Ingo Molnar's avatar
Ingo Molnar committed
679
void
680
681
update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
682
	int ret;
683

684
685
686
	if (trace_stop_count)
		return;

687
	WARN_ON_ONCE(!irqs_disabled());
688
	arch_spin_lock(&ftrace_max_lock);
689

690
691
	ftrace_disable_cpu();

692
693
	ret = ring_buffer_swap_cpu(max_tr.buffer, tr->buffer, cpu);

694
695
696
697
698
699
700
701
702
703
704
	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");
	}

705
706
	ftrace_enable_cpu();

707
	WARN_ON_ONCE(ret && ret != -EAGAIN && ret != -EBUSY);
708
709

	__update_max_tr(tr, tsk, cpu);
710
	arch_spin_unlock(&ftrace_max_lock);
711
}
712
#endif /* CONFIG_TRACER_MAX_TRACE */
713

Steven Rostedt's avatar
Steven Rostedt committed
714
715
716
717
718
719
/**
 * register_tracer - register a tracer with the ftrace system.
 * @type - the plugin for the tracer
 *
 * Register a new plugin tracer.
 */
720
int register_tracer(struct tracer *type)
721
722
__releases(kernel_lock)
__acquires(kernel_lock)
723
724
725
726
727
728
729
730
731
{
	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
732
733
734
735
736
	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
737
738
739
740
741
742
743
	/*
	 * 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();
744
	mutex_lock(&trace_types_lock);
Ingo Molnar's avatar
Ingo Molnar committed
745

746
747
	tracing_selftest_running = true;

748
749
750
	for (t = trace_types; t; t = t->next) {
		if (strcmp(type->name, t->name) == 0) {
			/* already found */
Li Zefan's avatar
Li Zefan committed
751
			pr_info("Tracer %s already registered\n",
752
753
754
755
756
757
				type->name);
			ret = -1;
			goto out;
		}
	}

758
759
760
761
762
763
764
	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;
765
766
767
	if (!type->wait_pipe)
		type->wait_pipe = default_wait_pipe;

768

Steven Rostedt's avatar
Steven Rostedt committed
769
#ifdef CONFIG_FTRACE_STARTUP_TEST
770
	if (type->selftest && !tracing_selftest_disabled) {
Steven Rostedt's avatar
Steven Rostedt committed
771
772
		struct tracer *saved_tracer = current_trace;
		struct trace_array *tr = &global_trace;
773

Steven Rostedt's avatar
Steven Rostedt committed
774
775
776
777
778
779
780
		/*
		 * 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.
		 */
781
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
782

Steven Rostedt's avatar
Steven Rostedt committed
783
784
785
786
787
788
789
790
791
792
		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
793
		/* Only reset on passing, to avoid touching corrupted buffers */
794
		tracing_reset_online_cpus(tr);
Ingo Molnar's avatar
Ingo Molnar committed
795

Steven Rostedt's avatar
Steven Rostedt committed
796
797
798
799
		printk(KERN_CONT "PASSED\n");
	}
#endif

800
801
	type->next = trace_types;
	trace_types = type;
Steven Rostedt's avatar
Steven Rostedt committed
802

803
 out:
804
	tracing_selftest_running = false;
805
806
	mutex_unlock(&trace_types_lock);

Steven Rostedt's avatar
Steven Rostedt committed
807
808
809
	if (ret || !default_bootup_tracer)
		goto out_unlock;

Li Zefan's avatar
Li Zefan committed
810
	if (strncmp(default_bootup_tracer, type->name, MAX_TRACER_SIZE))
Steven Rostedt's avatar
Steven Rostedt committed
811
812
813
814
815
816
817
818
		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;
819
#ifdef CONFIG_FTRACE_STARTUP_TEST
Steven Rostedt's avatar
Steven Rostedt committed
820
821
	printk(KERN_INFO "Disabling FTRACE selftests due to running tracer '%s'\n",
	       type->name);
822
823
#endif

Steven Rostedt's avatar
Steven Rostedt committed
824
 out_unlock:
825
	lock_kernel();
826
827
828
829
830
831
832
833
834
835
836
837
	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
838
	pr_info("Tracer %s not registered\n", type->name);
839
840
841
842
	goto out;

 found:
	*t = (*t)->next;
843
844
845
846
847
848
849
850

	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
851
out:
852
853
854
	mutex_unlock(&trace_types_lock);
}

855
static void __tracing_reset(struct ring_buffer *buffer, int cpu)
856
{
857
	ftrace_disable_cpu();
858
	ring_buffer_reset_cpu(buffer, cpu);
859
	ftrace_enable_cpu();
860
861
}

862
863
864
865
866
867
868
869
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();
870
	__tracing_reset(buffer, cpu);
871
872
873
874

	ring_buffer_record_enable(buffer);
}

875
876
void tracing_reset_online_cpus(struct trace_array *tr)
{
877
	struct ring_buffer *buffer = tr->buffer;
878
879
	int cpu;

880
881
882
883
884
	ring_buffer_record_disable(buffer);

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

885
886
887
	tr->time_start = ftrace_now(tr->cpu);

	for_each_online_cpu(cpu)
888
		__tracing_reset(buffer, cpu);
889
890

	ring_buffer_record_enable(buffer);
891
892
}

893
894
895
896
897
898
899
900
901
902
void tracing_reset_current(int cpu)
{
	tracing_reset(&global_trace, cpu);
}

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

903
#define SAVED_CMDLINES 128
904
#define NO_CMDLINE_MAP UINT_MAX
905
906
907
908
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;
909
static arch_spinlock_t trace_cmdline_lock = __ARCH_SPIN_LOCK_UNLOCKED;
910
911

/* temporary disable recording */
912
static atomic_t trace_record_cmdline_disabled __read_mostly;
913
914
915

static void trace_init_cmdlines(void)
{
916
917
	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));
918
919
920
	cmdline_idx = 0;
}

921
922
923
924
925
int is_tracing_stopped(void)
{
	return trace_stop_count;
}

926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
/**
 * 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();
}

941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
/**
 * 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);
956
957
958
959
960
961
	if (--trace_stop_count) {
		if (trace_stop_count < 0) {
			/* Someone screwed up their debugging */
			WARN_ON_ONCE(1);
			trace_stop_count = 0;
		}
962
963
964
		goto out;
	}

965
966
	/* Prevent the buffers from switching */
	arch_spin_lock(&ftrace_max_lock);
967
968
969
970
971
972
973
974
975

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

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

976
977
	arch_spin_unlock(&ftrace_max_lock);

978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
	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;

999
1000
1001
	/* Prevent the buffers from switching */
	arch_spin_lock(&ftrace_max_lock);

1002
1003
1004
1005
1006
1007
1008
1009
	buffer = global_trace.buffer;
	if (buffer)
		ring_buffer_record_disable(buffer);

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

1010
1011
	arch_spin_unlock(&ftrace_max_lock);

1012
1013
1014
1015
 out:
	spin_unlock_irqrestore(&tracing_start_lock, flags);
}

Ingo Molnar's avatar
Ingo Molnar committed
1016
void trace_stop_cmdline_recording(void);
1017

Ingo Molnar's avatar
Ingo Molnar committed
1018
static void trace_save_cmdline(struct task_struct *tsk)
1019
{
1020
	unsigned pid, idx;
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030

	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.
	 */
1031
	if (!arch_spin_trylock(&trace_cmdline_lock))
1032
1033
1034
		return;

	idx = map_pid_to_cmdline[tsk->pid];
1035
	if (idx == NO_CMDLINE_MAP) {
1036
1037
		idx = (cmdline_idx + 1) % SAVED_CMDLINES;

1038
1039
1040
1041
1042
1043
1044
1045
1046
		/*
		 * 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;
1047

1048
		map_cmdline_to_pid[idx] = tsk->pid;
1049
1050
1051
1052
1053
1054
1055
		map_pid_to_cmdline[tsk->pid] = idx;

		cmdline_idx = idx;
	}

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

1056
	arch_spin_unlock(&trace_cmdline_lock);
1057
1058
}

1059
void trace_find_cmdline(int pid, char comm[])
1060
1061
1062
{
	unsigned map;

1063
1064
1065
1066
	if (!pid) {
		strcpy(comm, "<idle>");
		return;
	}
1067

1068
1069
1070
1071
1072
	if (WARN_ON_ONCE(pid < 0)) {
		strcpy(comm, "<XXX>");
		return;
	}

1073
1074
1075
1076
	if (pid > PID_MAX_DEFAULT) {
		strcpy(comm, "<...>");
		return;
	}
1077

1078
	preempt_disable();
1079
	arch_spin_lock(&trace_cmdline_lock);
1080
	map = map_pid_to_cmdline[pid];
1081
1082
1083
1084
	if (map != NO_CMDLINE_MAP)
		strcpy(comm, saved_cmdlines[map]);
	else
		strcpy(comm, "<...>");
1085

1086
	arch_spin_unlock(&trace_cmdline_lock);
1087
	preempt_enable();
1088
1089
}

Ingo Molnar's avatar
Ingo Molnar committed
1090
void tracing_record_cmdline(struct task_struct *tsk)
1091
{
1092
1093
	if (atomic_read(&trace_record_cmdline_disabled) || !tracer_enabled ||
	    !tracing_is_on())
1094
1095
1096
1097
1098
		return;

	trace_save_cmdline(tsk);
}

1099
void
1100
1101
tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
			     int pc)
1102
1103
1104
{
	struct task_struct *tsk = current;

1105
1106
	entry->preempt_count		= pc & 0xff;
	entry->pid			= (tsk) ? tsk->pid : 0;
1107
	entry->lock_depth		= (tsk) ? tsk->lock_depth : 0;
1108
	entry->flags =
1109
#ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT
1110
		(irqs_disabled_flags(flags) ? TRACE_FLAG_IRQS_OFF : 0) |
1111
1112
1113
#else
		TRACE_FLAG_IRQS_NOSUPPORT |
#endif
1114
1115
1116
1117
		((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
		((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
		(need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
}
1118
EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
1119

1120
1121
1122
1123
1124
struct ring_buffer_event *
trace_buffer_lock_reserve(struct ring_buffer *buffer,
			  int type,
			  unsigned long len,
			  unsigned long flags, int pc)
1125
1126
1127
{
	struct ring_buffer_event *event;

1128
	event = ring_buffer_lock_reserve(buffer, len);
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
	if (event != NULL) {
		struct trace_entry *ent = ring_buffer_event_data(event);

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

	return event;
}

1139
1140
1141
1142
1143
static inline void
__trace_buffer_unlock_commit(struct ring_buffer *buffer,
			     struct ring_buffer_event *event,
			     unsigned long flags, int pc,
			     int wake)
1144
{
1145
	ring_buffer_unlock_commit(buffer, event);
1146

1147
1148
	ftrace_trace_stack(buffer, flags, 6, pc);
	ftrace_trace_userstack(buffer, flags, pc);
1149
1150
1151
1152
1153

	if (wake)
		trace_wake_up();
}

1154
1155
1156
void trace_buffer_unlock_commit(struct ring_buffer *buffer,
				struct ring_buffer_event *event,
				unsigned long flags, int pc)
1157
{
1158
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1159
1160
}

1161
struct ring_buffer_event *
1162
1163
trace_current_buffer_lock_reserve(struct ring_buffer **current_rb,
				  int type, unsigned long len,
1164
1165
				  unsigned long flags, int pc)
{
1166
1167
	*current_rb = global_trace.buffer;
	return trace_buffer_lock_reserve(*current_rb,
1168
1169
					 type, len, flags, pc);
}
1170
EXPORT_SYMBOL_GPL(trace_current_buffer_lock_reserve);
1171

1172
1173
void trace_current_buffer_unlock_commit(struct ring_buffer *buffer,
					struct ring_buffer_event *event,
1174
1175
					unsigned long flags, int pc)
{
1176
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
1177
}
1178
EXPORT_SYMBOL_GPL(trace_current_buffer_unlock_commit);
1179

1180
1181
1182
void trace_nowake_buffer_unlock_commit(struct ring_buffer *buffer,
				       struct ring_buffer_event *event,
				       unsigned long flags, int pc)
1183
{
1184
	__trace_buffer_unlock_commit(buffer, event, flags, pc, 0);
1185
}
1186
EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit);
1187

1188
1189
void trace_current_buffer_discard_commit(struct ring_buffer *buffer,
					 struct ring_buffer_event *event)
1190
{
1191
	ring_buffer_discard_commit(buffer, event);
1192
}
1193
EXPORT_SYMBOL_GPL(trace_current_buffer_discard_commit);
1194

Ingo Molnar's avatar
Ingo Molnar committed
1195
void
1196
trace_function(struct trace_array *tr,
1197
1198
	       unsigned long ip, unsigned long parent_ip, unsigned long flags,
	       int pc)
1199
{
1200
	struct ftrace_event_call *call = &event_function;
1201
	struct ring_buffer *buffer = tr->buffer;
1202
	struct ring_buffer_event *event;
1203
	struct ftrace_entry *entry;
1204

1205
	/* If we are reading the ring buffer, don't trace */
1206
	if (unlikely(__this_cpu_read(ftrace_cpu_disabled)))
1207
1208
		return;

1209
	event = trace_buffer_lock_reserve(buffer, TRACE_FN, sizeof(*entry),
1210
					  flags, pc);
1211
1212
1213
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
1214
1215
	entry->ip			= ip;
	entry->parent_ip		= parent_ip;
1216

1217
1218
	if (!filter_check_discard(call, entry, buffer, event))
		ring_buffer_unlock_commit(buffer, event);
1219
1220
}

Ingo Molnar's avatar
Ingo Molnar committed
1221
void
Ingo Molnar's avatar
Ingo Molnar committed
1222
ftrace(struct trace_array *tr, struct trace_array_cpu *data,
1223
1224
       unsigned long ip, unsigned long parent_ip, unsigned long flags,
       int pc)
Ingo Molnar's avatar
Ingo Molnar committed
1225
1226
{
	if (likely(!atomic_read(&data->disabled)))
1227
		trace_function(tr, ip, parent_ip, flags, pc);
Ingo Molnar's avatar
Ingo Molnar committed
1228
1229
}

1230
#ifdef CONFIG_STACKTRACE
1231
static void __ftrace_trace_stack(struct ring_buffer *buffer,
1232
1233
				 unsigned long flags,
				 int skip, int pc)
Ingo Molnar's avatar
Ingo Molnar committed
1234
{
1235
	struct ftrace_event_call *call = &event_kernel_stack;
1236
	struct ring_buffer_event *event;
1237
	struct stack_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
1238
1239
	struct stack_trace trace;