trace.c 78 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
/*
 * 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
 */
#include <linux/utsrelease.h>
#include <linux/kallsyms.h>
#include <linux/seq_file.h>
17
#include <linux/notifier.h>
18
#include <linux/debugfs.h>
19
#include <linux/pagemap.h>
20
21
22
23
24
25
#include <linux/hardirq.h>
#include <linux/linkage.h>
#include <linux/uaccess.h>
#include <linux/ftrace.h>
#include <linux/module.h>
#include <linux/percpu.h>
26
#include <linux/kdebug.h>
27
28
#include <linux/ctype.h>
#include <linux/init.h>
29
#include <linux/poll.h>
30
31
#include <linux/gfp.h>
#include <linux/fs.h>
32
#include <linux/kprobes.h>
Steven Rostedt's avatar
Steven Rostedt committed
33
#include <linux/writeback.h>
34

Ingo Molnar's avatar
Ingo Molnar committed
35
#include <linux/stacktrace.h>
36
#include <linux/ring_buffer.h>
37
#include <linux/irqflags.h>
Ingo Molnar's avatar
Ingo Molnar committed
38

39
40
#include "trace.h"

41
42
#define TRACE_BUFFER_FLAGS	(RB_FL_OVERWRITE)

43
44
45
unsigned long __read_mostly	tracing_max_latency = (cycle_t)ULONG_MAX;
unsigned long __read_mostly	tracing_thresh;

46
47
48
49
50
51
52
53
54

/*
 * 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.
 */
int tracing_disabled = 1;

55
56
57
58
59
60
61
62
63
64
65
66
67
68
static DEFINE_PER_CPU(local_t, ftrace_cpu_disabled);

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();
}

69
70
71
72
73
static cpumask_t __read_mostly		tracing_buffer_mask;

#define for_each_tracing_cpu(cpu)	\
	for_each_cpu_mask(cpu, tracing_buffer_mask)

74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
/*
 * 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;

89
90
91
92
93
94
95
96
97
static int tracing_set_tracer(char *buf);

static int __init set_ftrace(char *str)
{
	tracing_set_tracer(str);
	return 1;
}
__setup("ftrace", set_ftrace);

98
99
100
101
102
103
104
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);

105
long
106
107
108
109
110
111
112
ns2usecs(cycle_t nsec)
{
	nsec += 500;
	do_div(nsec, 1000);
	return nsec;
}

Ingo Molnar's avatar
Ingo Molnar committed
113
cycle_t ftrace_now(int cpu)
114
{
115
116
117
	u64 ts = ring_buffer_time_stamp(cpu);
	ring_buffer_normalize_time_stamp(cpu, &ts);
	return ts;
118
119
}

Steven Rostedt's avatar
Steven Rostedt committed
120
121
122
123
124
125
126
127
128
129
130
131
/*
 * 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.
 */
132
133
134
135
static struct trace_array	global_trace;

static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu);

Steven Rostedt's avatar
Steven Rostedt committed
136
137
138
139
140
141
142
143
144
145
/*
 * 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.
 */
146
147
148
149
static struct trace_array	max_tr;

static DEFINE_PER_CPU(struct trace_array_cpu, max_data);

Steven Rostedt's avatar
Steven Rostedt committed
150
/* tracer_enabled is used to toggle activation of a tracer */
151
static int			tracer_enabled = 1;
Steven Rostedt's avatar
Steven Rostedt committed
152

153
154
155
156
157
158
159
160
161
162
163
164
165
/**
 * 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;
}

166
167
168
/* function tracing enabled */
int				ftrace_function_enabled;

Steven Rostedt's avatar
Steven Rostedt committed
169
/*
170
171
172
 * 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.
173
174
175
176
177
 *
 * 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
178
 */
179
#define TRACE_BUF_SIZE_DEFAULT	1441792UL /* 16384 * 88 (sizeof(entry)) */
180

181
static unsigned long		trace_buf_size = TRACE_BUF_SIZE_DEFAULT;
182

Steven Rostedt's avatar
Steven Rostedt committed
183
/* trace_types holds a link list of available tracers. */
184
static struct tracer		*trace_types __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
185
186

/* current_trace points to the tracer that is currently active */
187
static struct tracer		*current_trace __read_mostly;
Steven Rostedt's avatar
Steven Rostedt committed
188
189
190
191
192
193

/*
 * max_tracer_type_len is used to simplify the allocating of
 * buffers to read userspace tracer names. We keep track of
 * the longest tracer name registered.
 */
194
195
static int			max_tracer_type_len;

Steven Rostedt's avatar
Steven Rostedt committed
196
197
198
199
200
201
/*
 * 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.
 */
202
static DEFINE_MUTEX(trace_types_lock);
Steven Rostedt's avatar
Steven Rostedt committed
203
204

/* trace_wait is a waitqueue for tasks blocked on trace_poll */
205
206
static DECLARE_WAIT_QUEUE_HEAD(trace_wait);

207
/* trace_flags holds trace_options default values */
208
209
unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
	TRACE_ITER_ANNOTATE;
210

Steven Rostedt's avatar
Steven Rostedt committed
211
212
213
214
215
216
/**
 * 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.
 */
217
218
void trace_wake_up(void)
{
Ingo Molnar's avatar
Ingo Molnar committed
219
220
221
222
223
	/*
	 * The runqueue_is_locked() can fail, but this is the best we
	 * have for now:
	 */
	if (!(trace_flags & TRACE_ITER_BLOCK) && !runqueue_is_locked())
224
225
		wake_up(&trace_wait);
}
226

227
static int __init set_buf_size(char *str)
228
{
229
	unsigned long buf_size;
230
231
	int ret;

232
233
	if (!str)
		return 0;
234
	ret = strict_strtoul(str, 0, &buf_size);
235
	/* nr_entries can not be zero */
236
	if (ret < 0 || buf_size == 0)
237
		return 0;
238
	trace_buf_size = buf_size;
239
240
	return 1;
}
241
__setup("trace_buf_size=", set_buf_size);
242

Steven Rostedt's avatar
Steven Rostedt committed
243
244
245
246
247
unsigned long nsecs_to_usecs(unsigned long nsecs)
{
	return nsecs / 1000;
}

Steven Rostedt's avatar
Steven Rostedt committed
248
/* These must match the bit postions in trace_iterator_flags */
249
250
251
252
253
static const char *trace_options[] = {
	"print-parent",
	"sym-offset",
	"sym-addr",
	"verbose",
Ingo Molnar's avatar
Ingo Molnar committed
254
	"raw",
255
	"hex",
Ingo Molnar's avatar
Ingo Molnar committed
256
	"bin",
257
	"block",
Ingo Molnar's avatar
Ingo Molnar committed
258
	"stacktrace",
Ingo Molnar's avatar
Ingo Molnar committed
259
	"sched-tree",
260
	"ftrace_printk",
261
	"ftrace_preempt",
262
#ifdef CONFIG_BRANCH_TRACER
263
	"branch",
264
#endif
265
	"annotate",
266
267
268
	NULL
};

Steven Rostedt's avatar
Steven Rostedt committed
269
270
271
272
273
274
275
276
277
/*
 * 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.
 */
278
279
static raw_spinlock_t ftrace_max_lock =
	(raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
280
281
282
283
284
285

/*
 * Copy the new maximum trace into the separate maximum-trace
 * structure. (this way the maximum trace is permanently saved,
 * for later retrieval via /debugfs/tracing/latency_trace)
 */
Ingo Molnar's avatar
Ingo Molnar committed
286
static void
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
__update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
	struct trace_array_cpu *data = tr->data[cpu];

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

	data = max_tr.data[cpu];
	data->saved_latency = tracing_max_latency;

	memcpy(data->comm, tsk->comm, TASK_COMM_LEN);
	data->pid = tsk->pid;
	data->uid = tsk->uid;
	data->nice = tsk->static_prio - 20 - MAX_RT_PRIO;
	data->policy = tsk->policy;
	data->rt_priority = tsk->rt_priority;

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

Steven Rostedt's avatar
Steven Rostedt committed
308
309
310
311
312
313
314
315
316
317
318
/**
 * trace_seq_printf - sequence printing of trace information
 * @s: trace sequence descriptor
 * @fmt: printf format string
 *
 * The tracer may use either sequence operations or its own
 * copy to user routines. To simplify formating of a trace
 * trace_seq_printf is used to store strings into a special
 * buffer (@s). Then the output may be either used by
 * the sequencer or pulled into another buffer.
 */
319
int
320
321
322
323
trace_seq_printf(struct trace_seq *s, const char *fmt, ...)
{
	int len = (PAGE_SIZE - 1) - s->len;
	va_list ap;
324
	int ret;
325
326
327
328
329

	if (!len)
		return 0;

	va_start(ap, fmt);
330
	ret = vsnprintf(s->buffer + s->len, len, fmt, ap);
331
332
	va_end(ap);

333
	/* If we can't write it all, don't bother writing anything */
334
	if (ret >= len)
335
336
337
		return 0;

	s->len += ret;
338
339
340
341

	return len;
}

Steven Rostedt's avatar
Steven Rostedt committed
342
343
344
345
346
347
348
349
350
351
/**
 * trace_seq_puts - trace sequence printing of simple string
 * @s: trace sequence descriptor
 * @str: simple string to record
 *
 * The tracer may use either the sequence operations or its own
 * copy to user routines. This function records a simple string
 * into a special buffer (@s) for later retrieval by a sequencer
 * or other mechanism.
 */
Ingo Molnar's avatar
Ingo Molnar committed
352
static int
353
354
355
356
357
trace_seq_puts(struct trace_seq *s, const char *str)
{
	int len = strlen(str);

	if (len > ((PAGE_SIZE - 1) - s->len))
358
		return 0;
359
360
361
362
363
364
365

	memcpy(s->buffer + s->len, str, len);
	s->len += len;

	return len;
}

Ingo Molnar's avatar
Ingo Molnar committed
366
static int
367
368
369
370
371
372
373
374
375
376
trace_seq_putc(struct trace_seq *s, unsigned char c)
{
	if (s->len >= (PAGE_SIZE - 1))
		return 0;

	s->buffer[s->len++] = c;

	return 1;
}

Ingo Molnar's avatar
Ingo Molnar committed
377
static int
Ingo Molnar's avatar
Ingo Molnar committed
378
379
380
381
382
383
384
385
386
387
388
trace_seq_putmem(struct trace_seq *s, void *mem, size_t len)
{
	if (len > ((PAGE_SIZE - 1) - s->len))
		return 0;

	memcpy(s->buffer + s->len, mem, len);
	s->len += len;

	return len;
}

389
390
#define MAX_MEMHEX_BYTES	8
#define HEX_CHARS		(MAX_MEMHEX_BYTES*2 + 1)
391

Ingo Molnar's avatar
Ingo Molnar committed
392
static int
393
394
395
trace_seq_putmem_hex(struct trace_seq *s, void *mem, size_t len)
{
	unsigned char hex[HEX_CHARS];
Thomas Gleixner's avatar
Thomas Gleixner committed
396
	unsigned char *data = mem;
397
398
399
400
401
402
403
	int i, j;

#ifdef __BIG_ENDIAN
	for (i = 0, j = 0; i < len; i++) {
#else
	for (i = len-1, j = 0; i >= 0; i--) {
#endif
404
405
		hex[j++] = hex_asc_hi(data[i]);
		hex[j++] = hex_asc_lo(data[i]);
406
	}
Thomas Gleixner's avatar
Thomas Gleixner committed
407
	hex[j++] = ' ';
408
409
410
411

	return trace_seq_putmem(s, hex, j);
}

Ingo Molnar's avatar
Ingo Molnar committed
412
static void
413
414
415
trace_seq_reset(struct trace_seq *s)
{
	s->len = 0;
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
	s->readpos = 0;
}

ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt)
{
	int len;
	int ret;

	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);
	if (ret)
		return -EFAULT;

	s->readpos += len;
	return cnt;
436
437
}

Ingo Molnar's avatar
Ingo Molnar committed
438
static void
439
440
441
442
443
444
445
446
447
448
trace_print_seq(struct seq_file *m, struct trace_seq *s)
{
	int len = s->len >= PAGE_SIZE ? PAGE_SIZE - 1 : s->len;

	s->buffer[len] = 0;
	seq_puts(m, s->buffer);

	trace_seq_reset(s);
}

Steven Rostedt's avatar
Steven Rostedt committed
449
450
451
452
453
454
455
456
457
/**
 * 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
458
void
459
460
update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
461
	struct ring_buffer *buf = tr->buffer;
462

463
	WARN_ON_ONCE(!irqs_disabled());
464
	__raw_spin_lock(&ftrace_max_lock);
465
466
467
468

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

469
	ftrace_disable_cpu();
470
	ring_buffer_reset(tr->buffer);
471
	ftrace_enable_cpu();
472
473

	__update_max_tr(tr, tsk, cpu);
474
	__raw_spin_unlock(&ftrace_max_lock);
475
476
477
478
479
480
481
}

/**
 * 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
482
483
 *
 * Flip the trace of a single CPU buffer between the @tr and the max_tr.
484
 */
Ingo Molnar's avatar
Ingo Molnar committed
485
void
486
487
update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
{
488
	int ret;
489

490
	WARN_ON_ONCE(!irqs_disabled());
491
	__raw_spin_lock(&ftrace_max_lock);
492

493
494
	ftrace_disable_cpu();

495
496
497
	ring_buffer_reset(max_tr.buffer);
	ret = ring_buffer_swap_cpu(max_tr.buffer, tr->buffer, cpu);

498
499
	ftrace_enable_cpu();

500
	WARN_ON_ONCE(ret);
501
502

	__update_max_tr(tr, tsk, cpu);
503
	__raw_spin_unlock(&ftrace_max_lock);
504
505
}

Steven Rostedt's avatar
Steven Rostedt committed
506
507
508
509
510
511
/**
 * register_tracer - register a tracer with the ftrace system.
 * @type - the plugin for the tracer
 *
 * Register a new plugin tracer.
 */
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
int register_tracer(struct tracer *type)
{
	struct tracer *t;
	int len;
	int ret = 0;

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

	mutex_lock(&trace_types_lock);
	for (t = trace_types; t; t = t->next) {
		if (strcmp(type->name, t->name) == 0) {
			/* already found */
			pr_info("Trace %s already registered\n",
				type->name);
			ret = -1;
			goto out;
		}
	}

Steven Rostedt's avatar
Steven Rostedt committed
534
#ifdef CONFIG_FTRACE_STARTUP_TEST
535
536
537
538
539
540
541
	/*
	 * 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();
Steven Rostedt's avatar
Steven Rostedt committed
542
543
544
545
546
547
548
549
550
551
552
	if (type->selftest) {
		struct tracer *saved_tracer = current_trace;
		struct trace_array *tr = &global_trace;
		int i;
		/*
		 * 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.
		 */
553
		for_each_tracing_cpu(i) {
554
			tracing_reset(tr, i);
Steven Rostedt's avatar
Steven Rostedt committed
555
556
557
558
559
560
561
562
563
564
565
		}
		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
566
		/* Only reset on passing, to avoid touching corrupted buffers */
567
		for_each_tracing_cpu(i) {
568
			tracing_reset(tr, i);
Steven Rostedt's avatar
Steven Rostedt committed
569
		}
Steven Rostedt's avatar
Steven Rostedt committed
570
571
		printk(KERN_CONT "PASSED\n");
	}
572
	lock_kernel();
Steven Rostedt's avatar
Steven Rostedt committed
573
574
#endif

575
576
577
578
579
	type->next = trace_types;
	trace_types = type;
	len = strlen(type->name);
	if (len > max_tracer_type_len)
		max_tracer_type_len = len;
Steven Rostedt's avatar
Steven Rostedt committed
580

581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
 out:
	mutex_unlock(&trace_types_lock);

	return ret;
}

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

	mutex_lock(&trace_types_lock);
	for (t = &trace_types; *t; t = &(*t)->next) {
		if (*t == type)
			goto found;
	}
	pr_info("Trace %s not registered\n", type->name);
	goto out;

 found:
	*t = (*t)->next;
	if (strlen(type->name) != max_tracer_type_len)
		goto out;

	max_tracer_type_len = 0;
	for (t = &trace_types; *t; t = &(*t)->next) {
		len = strlen((*t)->name);
		if (len > max_tracer_type_len)
			max_tracer_type_len = len;
	}
 out:
	mutex_unlock(&trace_types_lock);
}

615
void tracing_reset(struct trace_array *tr, int cpu)
616
{
617
	ftrace_disable_cpu();
618
	ring_buffer_reset_cpu(tr->buffer, cpu);
619
	ftrace_enable_cpu();
620
621
622
623
624
625
626
627
}

#define SAVED_CMDLINES 128
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;
static DEFINE_SPINLOCK(trace_cmdline_lock);
628
629
630

/* temporary disable recording */
atomic_t trace_record_cmdline_disabled __read_mostly;
631
632
633
634
635
636
637
638

static void trace_init_cmdlines(void)
{
	memset(&map_pid_to_cmdline, -1, sizeof(map_pid_to_cmdline));
	memset(&map_cmdline_to_pid, -1, sizeof(map_cmdline_to_pid));
	cmdline_idx = 0;
}

639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
static int trace_stop_count;
static DEFINE_SPINLOCK(tracing_start_lock);

/**
 * 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);
	if (--trace_stop_count)
		goto out;

	if (trace_stop_count < 0) {
		/* Someone screwed up their debugging */
		WARN_ON_ONCE(1);
		trace_stop_count = 0;
		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
709
void trace_stop_cmdline_recording(void);
710

Ingo Molnar's avatar
Ingo Molnar committed
711
static void trace_save_cmdline(struct task_struct *tsk)
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
{
	unsigned map;
	unsigned idx;

	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.
	 */
	if (!spin_trylock(&trace_cmdline_lock))
		return;

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

		map = map_cmdline_to_pid[idx];
		if (map <= PID_MAX_DEFAULT)
			map_pid_to_cmdline[map] = (unsigned)-1;

		map_pid_to_cmdline[tsk->pid] = idx;

		cmdline_idx = idx;
	}

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

	spin_unlock(&trace_cmdline_lock);
}

Ingo Molnar's avatar
Ingo Molnar committed
746
static char *trace_find_cmdline(int pid)
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
{
	char *cmdline = "<...>";
	unsigned map;

	if (!pid)
		return "<idle>";

	if (pid > PID_MAX_DEFAULT)
		goto out;

	map = map_pid_to_cmdline[pid];
	if (map >= SAVED_CMDLINES)
		goto out;

	cmdline = saved_cmdlines[map];

 out:
	return cmdline;
}

Ingo Molnar's avatar
Ingo Molnar committed
767
void tracing_record_cmdline(struct task_struct *tsk)
768
769
770
771
772
773
774
{
	if (atomic_read(&trace_record_cmdline_disabled))
		return;

	trace_save_cmdline(tsk);
}

775
void
776
777
tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
			     int pc)
778
779
780
{
	struct task_struct *tsk = current;

781
782
783
	entry->preempt_count		= pc & 0xff;
	entry->pid			= (tsk) ? tsk->pid : 0;
	entry->flags =
784
#ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT
785
		(irqs_disabled_flags(flags) ? TRACE_FLAG_IRQS_OFF : 0) |
786
787
788
#else
		TRACE_FLAG_IRQS_NOSUPPORT |
#endif
789
790
791
792
793
		((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
		((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
		(need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
}

Ingo Molnar's avatar
Ingo Molnar committed
794
void
795
trace_function(struct trace_array *tr, struct trace_array_cpu *data,
796
797
	       unsigned long ip, unsigned long parent_ip, unsigned long flags,
	       int pc)
798
{
799
	struct ring_buffer_event *event;
800
	struct ftrace_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
801
	unsigned long irq_flags;
802

803
804
805
806
	/* If we are reading the ring buffer, don't trace */
	if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled))))
		return;

807
808
809
810
811
	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
					 &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
812
	tracing_generic_entry_update(&entry->ent, flags, pc);
813
814
815
	entry->ent.type			= TRACE_FN;
	entry->ip			= ip;
	entry->parent_ip		= parent_ip;
816
	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
817
818
}

819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
#ifdef CONFIG_FUNCTION_RET_TRACER
static void __trace_function_return(struct trace_array *tr,
				struct trace_array_cpu *data,
				struct ftrace_retfunc *trace,
				unsigned long flags,
				int pc)
{
	struct ring_buffer_event *event;
	struct ftrace_ret_entry *entry;
	unsigned long irq_flags;

	if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled))))
		return;

	event = ring_buffer_lock_reserve(global_trace.buffer, sizeof(*entry),
					 &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
	tracing_generic_entry_update(&entry->ent, flags, pc);
	entry->ent.type			= TRACE_FN_RET;
	entry->ip			= trace->func;
	entry->parent_ip	= trace->ret;
	entry->rettime		= trace->rettime;
	entry->calltime		= trace->calltime;
	ring_buffer_unlock_commit(global_trace.buffer, event, irq_flags);
}
#endif

Ingo Molnar's avatar
Ingo Molnar committed
848
void
Ingo Molnar's avatar
Ingo Molnar committed
849
ftrace(struct trace_array *tr, struct trace_array_cpu *data,
850
851
       unsigned long ip, unsigned long parent_ip, unsigned long flags,
       int pc)
Ingo Molnar's avatar
Ingo Molnar committed
852
853
{
	if (likely(!atomic_read(&data->disabled)))
854
		trace_function(tr, data, ip, parent_ip, flags, pc);
Ingo Molnar's avatar
Ingo Molnar committed
855
856
}

857
858
859
860
static void ftrace_trace_stack(struct trace_array *tr,
			       struct trace_array_cpu *data,
			       unsigned long flags,
			       int skip, int pc)
Ingo Molnar's avatar
Ingo Molnar committed
861
{
862
#ifdef CONFIG_STACKTRACE
863
	struct ring_buffer_event *event;
864
	struct stack_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
865
	struct stack_trace trace;
866
	unsigned long irq_flags;
Ingo Molnar's avatar
Ingo Molnar committed
867
868
869
870

	if (!(trace_flags & TRACE_ITER_STACKTRACE))
		return;

871
872
873
874
875
	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
					 &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
876
	tracing_generic_entry_update(&entry->ent, flags, pc);
877
	entry->ent.type		= TRACE_STACK;
Ingo Molnar's avatar
Ingo Molnar committed
878

879
	memset(&entry->caller, 0, sizeof(entry->caller));
Ingo Molnar's avatar
Ingo Molnar committed
880
881
882
883

	trace.nr_entries	= 0;
	trace.max_entries	= FTRACE_STACK_ENTRIES;
	trace.skip		= skip;
884
	trace.entries		= entry->caller;
Ingo Molnar's avatar
Ingo Molnar committed
885
886

	save_stack_trace(&trace);
887
	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
888
#endif
Ingo Molnar's avatar
Ingo Molnar committed
889
890
}

891
892
893
894
895
896
897
898
899
900
901
902
void __trace_stack(struct trace_array *tr,
		   struct trace_array_cpu *data,
		   unsigned long flags,
		   int skip)
{
	ftrace_trace_stack(tr, data, flags, skip, preempt_count());
}

static void
ftrace_trace_special(void *__tr, void *__data,
		     unsigned long arg1, unsigned long arg2, unsigned long arg3,
		     int pc)
Ingo Molnar's avatar
Ingo Molnar committed
903
{
904
	struct ring_buffer_event *event;
Ingo Molnar's avatar
Ingo Molnar committed
905
906
	struct trace_array_cpu *data = __data;
	struct trace_array *tr = __tr;
907
	struct special_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
908
909
	unsigned long irq_flags;

910
911
912
913
914
	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
					 &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
915
	tracing_generic_entry_update(&entry->ent, 0, pc);
916
917
918
919
	entry->ent.type			= TRACE_SPECIAL;
	entry->arg1			= arg1;
	entry->arg2			= arg2;
	entry->arg3			= arg3;
920
	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
921
	ftrace_trace_stack(tr, data, irq_flags, 4, pc);
Ingo Molnar's avatar
Ingo Molnar committed
922
923
924
925

	trace_wake_up();
}

926
927
928
929
930
931
932
void
__trace_special(void *__tr, void *__data,
		unsigned long arg1, unsigned long arg2, unsigned long arg3)
{
	ftrace_trace_special(__tr, __data, arg1, arg2, arg3, preempt_count());
}

Ingo Molnar's avatar
Ingo Molnar committed
933
void
934
935
tracing_sched_switch_trace(struct trace_array *tr,
			   struct trace_array_cpu *data,
Ingo Molnar's avatar
Ingo Molnar committed
936
937
			   struct task_struct *prev,
			   struct task_struct *next,
938
			   unsigned long flags, int pc)
939
{
940
	struct ring_buffer_event *event;
941
	struct ctx_switch_entry *entry;
Ingo Molnar's avatar
Ingo Molnar committed
942
	unsigned long irq_flags;
943

944
945
946
947
948
	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
					   &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
949
	tracing_generic_entry_update(&entry->ent, flags, pc);
950
951
952
953
954
955
956
957
	entry->ent.type			= TRACE_CTX;
	entry->prev_pid			= prev->pid;
	entry->prev_prio		= prev->prio;
	entry->prev_state		= prev->state;
	entry->next_pid			= next->pid;
	entry->next_prio		= next->prio;
	entry->next_state		= next->state;
	entry->next_cpu	= task_cpu(next);
958
	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
959
	ftrace_trace_stack(tr, data, flags, 5, pc);
960
961
}

962
963
964
void
tracing_sched_wakeup_trace(struct trace_array *tr,
			   struct trace_array_cpu *data,
Ingo Molnar's avatar
Ingo Molnar committed
965
966
			   struct task_struct *wakee,
			   struct task_struct *curr,
967
			   unsigned long flags, int pc)
968
{
969
	struct ring_buffer_event *event;
970
	struct ctx_switch_entry *entry;
971
972
	unsigned long irq_flags;

973
974
975
976
977
	event = ring_buffer_lock_reserve(tr->buffer, sizeof(*entry),
					   &irq_flags);
	if (!event)
		return;
	entry	= ring_buffer_event_data(event);
978
	tracing_generic_entry_update(&entry->ent, flags, pc);
979
980
981
982
983
984
985
986
	entry->ent.type			= TRACE_WAKE;
	entry->prev_pid			= curr->pid;
	entry->prev_prio		= curr->prio;
	entry->prev_state		= curr->state;
	entry->next_pid			= wakee->pid;
	entry->next_prio		= wakee->prio;
	entry->next_state		= wakee->state;
	entry->next_cpu			= task_cpu(wakee);
987
	ring_buffer_unlock_commit(tr->buffer, event, irq_flags);
988
	ftrace_trace_stack(tr, data, flags, 6, pc);
Ingo Molnar's avatar
Ingo Molnar committed
989
990

	trace_wake_up();
991
992
}

993
994
995
996
997
void
ftrace_special(unsigned long arg1, unsigned long arg2, unsigned long arg3)
{
	struct trace_array *tr = &global_trace;
	struct trace_array_cpu *data;
998
	unsigned long flags;
999
	int cpu;
1000
	int pc;
1001

1002
	if (tracing_disabled)
1003
1004
		return;

1005
	pc = preempt_count();
1006
	local_irq_save(flags);
1007
1008
1009
	cpu = raw_smp_processor_id();
	data = tr->data[cpu];

1010
	if (likely(atomic_inc_return(&data->disabled) == 1))
1011
		ftrace_trace_special(tr, data, arg1, arg2, arg3, pc);
1012

1013
1014
	atomic_dec(&data->disabled);
	local_irq_restore(flags);
1015
1016
}

1017
#ifdef CONFIG_FUNCTION_TRACER
Ingo Molnar's avatar
Ingo Molnar committed
1018
static void
1019
function_trace_call_preempt_only(unsigned long ip, unsigned long parent_ip)
Ingo Molnar's avatar
Ingo Molnar committed
1020
1021
1022
1023
1024
{
	struct trace_array *tr = &global_trace;
	struct trace_array_cpu *data;
	unsigned long flags;
	long disabled;
1025
1026
	int cpu, resched;
	int pc;
Ingo Molnar's avatar
Ingo Molnar committed
1027

1028
	if (unlikely(!ftrace_function_enabled))
Ingo Molnar's avatar
Ingo Molnar committed
1029
1030
		return;

1031
	pc = preempt_count();
1032
	resched = ftrace_preempt_disable();
1033
	local_save_flags(flags);
Ingo Molnar's avatar
Ingo Molnar committed
1034
1035
1036
1037
1038
	cpu = raw_smp_processor_id();
	data = tr->data[cpu];
	disabled = atomic_inc_return(&data->disabled);

	if (likely(disabled == 1))
1039
		trace_function(tr, data, ip, parent_ip, flags, pc);
Ingo Molnar's avatar
Ingo Molnar committed
1040
1041

	atomic_dec(&data->disabled);
1042
	ftrace_preempt_enable(resched);
Ingo Molnar's avatar
Ingo Molnar committed
1043
1044
}

1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
static void
function_trace_call(unsigned long ip, unsigned long parent_ip)
{
	struct trace_array *tr = &global_trace;
	struct trace_array_cpu *data;
	unsigned long flags;
	long disabled;
	int cpu;
	int pc;

	if (unlikely(!ftrace_function_enabled))
		return;

	/*
	 * Need to use raw, since this must be called before the
	 * recursive protection is performed.
	 */
1062
	local_irq_save(flags);
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
	cpu = raw_smp_processor_id();
	data = tr->data[cpu];
	disabled = atomic_inc_return(&data->disabled);

	if (likely(disabled == 1)) {
		pc = preempt_count();
		trace_function(tr, data, ip, parent_ip, flags, pc);
	}

	atomic_dec(&data->disabled);
1073
	local_irq_restore(flags);
1074
1075
}

1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
#ifdef CONFIG_FUNCTION_RET_TRACER
void trace_function_return(struct ftrace_retfunc *trace)
{
	struct trace_array *tr = &global_trace;
	struct trace_array_cpu *data;
	unsigned long flags;
	long disabled;
	int cpu;
	int pc;

	raw_local_irq_save(flags);
	cpu = raw_smp_processor_id();
	data = tr->data[cpu];
	disabled = atomic_inc_return(&data->disabled);
	if (likely(disabled == 1)) {
		pc = preempt_count();
		__trace_function_return(tr, data, trace, flags, pc);
	}
	atomic_dec(&data->disabled);
	raw_local_irq_restore(flags);
}
#endif /* CONFIG_FUNCTION_RET_TRACER */

Ingo Molnar's avatar
Ingo Molnar committed
1099
1100
1101
1102
1103
static struct ftrace_ops trace_ops __read_mostly =
{
	.func = function_trace_call,
};

Ingo Molnar's avatar
Ingo Molnar committed
1104
void tracing_start_function_trace(void)
Ingo Molnar's avatar
Ingo Molnar committed
1105
{
1106
	ftrace_function_enabled = 0;
1107
1108
1109
1110
1111
1112

	if (trace_flags & TRACE_ITER_PREEMPTONLY)
		trace_ops.func = function_trace_call_preempt_only;
	else
		trace_ops.func = function_trace_call;

Ingo Molnar's avatar
Ingo Molnar committed
1113
	register_ftrace_function(&trace_ops);
1114
	ftrace_function_enabled = 1;
Ingo Molnar's avatar
Ingo Molnar committed
1115
1116
}

Ingo Molnar's avatar
Ingo Molnar committed
1117
void tracing_stop_function_trace(void)
Ingo Molnar's avatar
Ingo Molnar committed
1118
{
1119
	ftrace_function_enabled = 0;
Ingo Molnar's avatar
Ingo Molnar committed
1120
1121
1122
1123
	unregister_ftrace_function(&trace_ops);
}
#endif

1124
1125
enum trace_file_type {
	TRACE_FILE_LAT_FMT	= 1,
1126
	TRACE_FILE_ANNOTATE	= 2,
1127
1128
};

1129
1130
static void trace_iterator_increment(struct trace_iterator *iter, int cpu)
{
1131
1132
1133
	/* Don't allow ftrace to trace into the ring buffers */
	ftrace_disable_cpu();

1134
	iter->idx++;
1135
1136
1137
1138
	if (iter->buffer_iter[iter->cpu])
		ring_buffer_read(iter->buffer_iter[iter->cpu], NULL);

	ftrace_enable_cpu();
1139
1140
}

Ingo Molnar's avatar
Ingo Molnar committed
1141
static struct trace_entry *
1142
peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts)
1143
{
1144
1145
	struct ring_buffer_event *event;
	struct ring_buffer_iter *buf_iter = iter->buffer_iter[cpu];
1146

1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
	/* Don't allow ftrace to trace into the ring buffers */
	ftrace_disable_cpu();

	if (buf_iter)
		event = ring_buffer_iter_peek(buf_iter, ts);
	else
		event = ring_buffer_peek(iter->tr->buffer, cpu, ts);

	ftrace_enable_cpu();

1157
	return event ? ring_buffer_event_data(event) : NULL;
1158
}
1159

1160
static struct trace_entry *
1161
__find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts)
1162
{
1163
	struct ring_buffer *buffer = iter->tr->buffer;
1164
	struct trace_entry *ent, *next = NULL;
1165
	u64 next_ts = 0, ts;
1166
1167
1168
	int next_cpu = -1;
	int cpu;

1169
	for_each_tracing_cpu(cpu) {
1170

1171
1172
		if (ring_buffer_empty_cpu(buffer, cpu))
			continue;
1173

1174
		ent = peek_next_entry(iter, cpu, &ts);
1175

Ingo Molnar's avatar
Ingo Molnar committed
1176
1177
1178
		/*
		 * Pick the entry with the smallest timestamp:
		 */
1179
		if (ent && (!next || ts < next_ts)) {
1180
1181
			next = ent;
			next_cpu = cpu;
1182
			next_ts = ts;
1183
1184
1185
1186
1187
1188
		}
	}

	if (ent_cpu)
		*ent_cpu = next_cpu;

1189
1190
1191
	if (ent_ts)
		*ent_ts = next_ts;

1192
1193
1194
	return next;
}

1195
1196
/* Find the next real entry, without updating the iterator itself */
static struct trace_entry *
1197
find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts)
1198
{
1199
	return __find_next_entry(iter, ent_cpu, ent_ts);
1200
1201
1202
1203
1204
}

/* Find the next real entry, and increment the iterator to the next entry */
static void *find_next_entry_inc(struct trace_iterator *iter)
{
1205
	iter->ent = __find_next_entry(iter, &iter->cpu, &iter->ts);
1206

1207
	if (iter->ent)
1208
1209
		trace_iterator_increment(iter, iter->cpu);

1210
	return iter->ent ? iter : NULL;
1211
}
1212

Ingo Molnar's avatar
Ingo Molnar committed
1213
static void trace_consume(struct trace_iterator *iter)
1214
{
1215
1216
	/* Don't allow ftrace to trace into the ring buffers */
	ftrace_disable_cpu();
1217
	ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts);
1218
	ftrace_enable_cpu();
1219
1220
}

Ingo Molnar's avatar
Ingo Molnar committed
1221
static void *s_next(struct seq_file *m, void *v, loff_t *pos)
1222
1223
1224
{
	struct trace_iterator *iter = m->private;
	int i = (int)*pos;
Ingo Molnar's avatar
Ingo Molnar committed
1225
	void *ent;
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250

	(*pos)++;

	/* can't go backwards */
	if (iter->idx > i)
		return NULL;

	if (iter->idx < 0)
		ent = find_next_entry_inc(iter);
	else
		ent = iter;

	while (ent && iter->idx < i)
		ent = find_next_entry_inc(iter);

	iter->pos = *pos;

	return ent;
}

static void *s_start(struct seq_file *m, loff_t *pos)
{
	struct trace_iterator *iter = m->private;
	void *p = NULL;
	loff_t l = 0;
1251
	int cpu;
1252
1253
1254

	mutex_lock(&trace_types_lock);

1255
1256
	if (!current_trace || current_trace != iter->trace) {
		mutex_unlock(&trace_types_lock);
1257
		return NULL;
1258
	}
1259
1260
1261
1262
1263
1264
1265
1266

	atomic_inc(&trace_record_cmdline_disabled);

	if (*pos != iter->pos) {
		iter->ent = NULL;
		iter->cpu = 0;
		iter->idx = -1;

1267
1268
		ftrace_disable_cpu();

1269
1270
		for_each_tracing_cpu(cpu) {
			ring_buffer_iter_reset(iter->buffer_iter[cpu]);
1271
		}
1272

1273
1274
		ftrace_enable_cpu();

1275
1276
1277
1278
		for (p = iter; p && l < *pos; p = s_next(m, p, &l))
			;

	} else {
1279
		l = *pos - 1;
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291