ftrace.c 165.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12
/*
 * Infrastructure for profiling code inserted by 'gcc -pg'.
 *
 * Copyright (C) 2007-2008 Steven Rostedt <srostedt@redhat.com>
 * Copyright (C) 2004-2008 Ingo Molnar <mingo@redhat.com>
 *
 * Originally ported from the -rt patch by:
 *   Copyright (C) 2007 Arnaldo Carvalho de Melo <acme@redhat.com>
 *
 * Based on code in the latency_tracer, that is:
 *
 *  Copyright (C) 2004-2006 Ingo Molnar
13
 *  Copyright (C) 2004 Nadia Yvette Chambers
14 15
 */

16 17
#include <linux/stop_machine.h>
#include <linux/clocksource.h>
18
#include <linux/sched/task.h>
19
#include <linux/kallsyms.h>
20
#include <linux/seq_file.h>
21
#include <linux/suspend.h>
22
#include <linux/tracefs.h>
23
#include <linux/hardirq.h>
I
Ingo Molnar 已提交
24
#include <linux/kthread.h>
25
#include <linux/uaccess.h>
26
#include <linux/bsearch.h>
27
#include <linux/module.h>
I
Ingo Molnar 已提交
28
#include <linux/ftrace.h>
29
#include <linux/sysctl.h>
30
#include <linux/slab.h>
31
#include <linux/ctype.h>
32
#include <linux/sort.h>
33
#include <linux/list.h>
34
#include <linux/hash.h>
35
#include <linux/rcupdate.h>
36

37
#include <trace/events/sched.h>
38

39
#include <asm/sections.h>
40
#include <asm/setup.h>
41

42
#include "trace_output.h"
S
Steven Rostedt 已提交
43
#include "trace_stat.h"
44

45
#define FTRACE_WARN_ON(cond)			\
46 47 48
	({					\
		int ___r = cond;		\
		if (WARN_ON(___r))		\
49
			ftrace_kill();		\
50 51
		___r;				\
	})
52 53

#define FTRACE_WARN_ON_ONCE(cond)		\
54 55 56
	({					\
		int ___r = cond;		\
		if (WARN_ON_ONCE(___r))		\
57
			ftrace_kill();		\
58 59
		___r;				\
	})
60

61 62 63
/* hash bits for specific function selection */
#define FTRACE_HASH_BITS 7
#define FTRACE_FUNC_HASHSIZE (1 << FTRACE_HASH_BITS)
64 65
#define FTRACE_HASH_DEFAULT_BITS 10
#define FTRACE_HASH_MAX_BITS 12
66

67
#ifdef CONFIG_DYNAMIC_FTRACE
68 69 70
#define INIT_OPS_HASH(opsname)	\
	.func_hash		= &opsname.local_hash,			\
	.local_hash.regex_lock	= __MUTEX_INITIALIZER(opsname.local_hash.regex_lock),
71 72 73
#define ASSIGN_OPS_HASH(opsname, val) \
	.func_hash		= val, \
	.local_hash.regex_lock	= __MUTEX_INITIALIZER(opsname.local_hash.regex_lock),
74
#else
75
#define INIT_OPS_HASH(opsname)
76
#define ASSIGN_OPS_HASH(opsname, val)
77 78
#endif

79 80
static struct ftrace_ops ftrace_list_end __read_mostly = {
	.func		= ftrace_stub,
81
	.flags		= FTRACE_OPS_FL_RECURSION_SAFE | FTRACE_OPS_FL_STUB,
82
	INIT_OPS_HASH(ftrace_list_end)
83 84
};

85 86
/* ftrace_enabled is a method to turn ftrace on or off */
int ftrace_enabled __read_mostly;
87
static int last_ftrace_enabled;
88

89 90
/* Current function tracing op */
struct ftrace_ops *function_trace_op __read_mostly = &ftrace_list_end;
91 92
/* What to set function_trace_op to */
static struct ftrace_ops *set_function_trace_op;
93

94
static bool ftrace_pids_enabled(struct ftrace_ops *ops)
95
{
96 97 98 99 100 101 102 103
	struct trace_array *tr;

	if (!(ops->flags & FTRACE_OPS_FL_PID) || !ops->private)
		return false;

	tr = ops->private;

	return tr->function_pids != NULL;
104 105 106 107
}

static void ftrace_update_trampoline(struct ftrace_ops *ops);

108 109 110 111 112 113
/*
 * ftrace_disabled is set when an anomaly is discovered.
 * ftrace_disabled is much stronger than ftrace_enabled.
 */
static int ftrace_disabled __read_mostly;

114
static DEFINE_MUTEX(ftrace_lock);
115

116
static struct ftrace_ops __rcu *ftrace_ops_list __read_mostly = &ftrace_list_end;
117
ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub;
118
static struct ftrace_ops global_ops;
119

120 121
#if ARCH_SUPPORTS_FTRACE_OPS
static void ftrace_ops_list_func(unsigned long ip, unsigned long parent_ip,
122
				 struct ftrace_ops *op, struct pt_regs *regs);
123 124 125 126 127
#else
/* See comment below, where ftrace_ops_list_func is defined */
static void ftrace_ops_no_ops(unsigned long ip, unsigned long parent_ip);
#define ftrace_ops_list_func ((ftrace_func_t)ftrace_ops_no_ops)
#endif
128

129 130
/*
 * Traverse the ftrace_global_list, invoking all entries.  The reason that we
131
 * can use rcu_dereference_raw_notrace() is that elements removed from this list
132
 * are simply leaked, so there is no need to interact with a grace-period
133
 * mechanism.  The rcu_dereference_raw_notrace() calls are needed to handle
134 135 136 137 138
 * concurrent insertions into the ftrace_global_list.
 *
 * Silly Alpha and silly pointer-speculation compiler optimizations!
 */
#define do_for_each_ftrace_op(op, list)			\
139
	op = rcu_dereference_raw_notrace(list);			\
140 141 142 143 144 145
	do

/*
 * Optimized for just a single item in the list (as that is the normal case).
 */
#define while_for_each_ftrace_op(op)				\
146
	while (likely(op = rcu_dereference_raw_notrace((op)->next)) &&	\
147 148
	       unlikely((op) != &ftrace_list_end))

149 150 151 152
static inline void ftrace_ops_init(struct ftrace_ops *ops)
{
#ifdef CONFIG_DYNAMIC_FTRACE
	if (!(ops->flags & FTRACE_OPS_FL_INITIALIZED)) {
153 154
		mutex_init(&ops->local_hash.regex_lock);
		ops->func_hash = &ops->local_hash;
155 156 157 158 159
		ops->flags |= FTRACE_OPS_FL_INITIALIZED;
	}
#endif
}

160
static void ftrace_pid_func(unsigned long ip, unsigned long parent_ip,
161
			    struct ftrace_ops *op, struct pt_regs *regs)
162
{
163 164 165
	struct trace_array *tr = op->private;

	if (tr && this_cpu_read(tr->trace_buffer.data->ftrace_ignore_pid))
166 167
		return;

168
	op->saved_func(ip, parent_ip, op, regs);
169 170
}

171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
static void ftrace_sync(struct work_struct *work)
{
	/*
	 * This function is just a stub to implement a hard force
	 * of synchronize_sched(). This requires synchronizing
	 * tasks even in userspace and idle.
	 *
	 * Yes, function tracing is rude.
	 */
}

static void ftrace_sync_ipi(void *data)
{
	/* Probably not needed, but do it anyway */
	smp_rmb();
}

188 189
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
static void update_function_graph_func(void);
190 191 192 193 194

/* Both enabled by default (can be cleared by function_graph tracer flags */
static bool fgraph_sleep_time = true;
static bool fgraph_graph_time = true;

195 196 197 198
#else
static inline void update_function_graph_func(void) { }
#endif

199 200 201 202

static ftrace_func_t ftrace_ops_get_list_func(struct ftrace_ops *ops)
{
	/*
203
	 * If this is a dynamic, RCU, or per CPU ops, or we force list func,
204 205
	 * then it needs to call the list anyway.
	 */
206 207
	if (ops->flags & (FTRACE_OPS_FL_DYNAMIC | FTRACE_OPS_FL_RCU) ||
	    FTRACE_FORCE_LIST_FUNC)
208 209 210 211 212
		return ftrace_ops_list_func;

	return ftrace_ops_get_func(ops);
}

213 214 215 216
static void update_ftrace_function(void)
{
	ftrace_func_t func;

217 218 219 220 221
	/*
	 * Prepare the ftrace_ops that the arch callback will use.
	 * If there's only one ftrace_ops registered, the ftrace_ops_list
	 * will point to the ops we want.
	 */
222 223
	set_function_trace_op = rcu_dereference_protected(ftrace_ops_list,
						lockdep_is_held(&ftrace_lock));
224 225

	/* If there's no ftrace_ops registered, just call the stub function */
226
	if (set_function_trace_op == &ftrace_list_end) {
227 228
		func = ftrace_stub;

229 230
	/*
	 * If we are at the end of the list and this ops is
231 232
	 * recursion safe and not dynamic and the arch supports passing ops,
	 * then have the mcount trampoline call the function directly.
233
	 */
234 235
	} else if (rcu_dereference_protected(ftrace_ops_list->next,
			lockdep_is_held(&ftrace_lock)) == &ftrace_list_end) {
236
		func = ftrace_ops_get_list_func(ftrace_ops_list);
237

238 239
	} else {
		/* Just use the default ftrace_ops */
240
		set_function_trace_op = &ftrace_list_end;
241
		func = ftrace_ops_list_func;
242
	}
243

244 245
	update_function_graph_func();

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
	/* If there's no change, then do nothing more here */
	if (ftrace_trace_function == func)
		return;

	/*
	 * If we are using the list function, it doesn't care
	 * about the function_trace_ops.
	 */
	if (func == ftrace_ops_list_func) {
		ftrace_trace_function = func;
		/*
		 * Don't even bother setting function_trace_ops,
		 * it would be racy to do so anyway.
		 */
		return;
	}

#ifndef CONFIG_DYNAMIC_FTRACE
	/*
	 * For static tracing, we need to be a bit more careful.
	 * The function change takes affect immediately. Thus,
	 * we need to coorditate the setting of the function_trace_ops
	 * with the setting of the ftrace_trace_function.
	 *
	 * Set the function to the list ops, which will call the
	 * function we want, albeit indirectly, but it handles the
	 * ftrace_ops and doesn't depend on function_trace_op.
	 */
	ftrace_trace_function = ftrace_ops_list_func;
	/*
	 * Make sure all CPUs see this. Yes this is slow, but static
	 * tracing is slow and nasty to have enabled.
	 */
	schedule_on_each_cpu(ftrace_sync);
	/* Now all cpus are using the list ops. */
	function_trace_op = set_function_trace_op;
	/* Make sure the function_trace_op is visible on all CPUs */
	smp_wmb();
	/* Nasty way to force a rmb on all cpus */
	smp_call_function(ftrace_sync_ipi, NULL, 1);
	/* OK, we are all set to update the ftrace_trace_function now! */
#endif /* !CONFIG_DYNAMIC_FTRACE */

289 290 291
	ftrace_trace_function = func;
}

292 293
static void add_ftrace_ops(struct ftrace_ops __rcu **list,
			   struct ftrace_ops *ops)
294
{
295 296
	rcu_assign_pointer(ops->next, *list);

297
	/*
298
	 * We are entering ops into the list but another
299 300
	 * CPU might be walking that list. We need to make sure
	 * the ops->next pointer is valid before another CPU sees
301
	 * the ops pointer included into the list.
302
	 */
303
	rcu_assign_pointer(*list, ops);
304 305
}

306 307
static int remove_ftrace_ops(struct ftrace_ops __rcu **list,
			     struct ftrace_ops *ops)
308 309 310 311
{
	struct ftrace_ops **p;

	/*
312 313
	 * If we are removing the last function, then simply point
	 * to the ftrace_stub.
314
	 */
315 316 317 318
	if (rcu_dereference_protected(*list,
			lockdep_is_held(&ftrace_lock)) == ops &&
	    rcu_dereference_protected(ops->next,
			lockdep_is_held(&ftrace_lock)) == &ftrace_list_end) {
319
		*list = &ftrace_list_end;
S
Steven Rostedt 已提交
320
		return 0;
321 322
	}

323
	for (p = list; *p != &ftrace_list_end; p = &(*p)->next)
324 325 326
		if (*p == ops)
			break;

S
Steven Rostedt 已提交
327 328
	if (*p != ops)
		return -1;
329 330

	*p = (*p)->next;
331 332
	return 0;
}
333

334 335
static void ftrace_update_trampoline(struct ftrace_ops *ops);

336 337
static int __register_ftrace_function(struct ftrace_ops *ops)
{
338 339 340
	if (ops->flags & FTRACE_OPS_FL_DELETED)
		return -EINVAL;

341 342 343
	if (WARN_ON(ops->flags & FTRACE_OPS_FL_ENABLED))
		return -EBUSY;

344
#ifndef CONFIG_DYNAMIC_FTRACE_WITH_REGS
345 346 347 348 349 350 351 352 353 354 355 356 357
	/*
	 * If the ftrace_ops specifies SAVE_REGS, then it only can be used
	 * if the arch supports it, or SAVE_REGS_IF_SUPPORTED is also set.
	 * Setting SAVE_REGS_IF_SUPPORTED makes SAVE_REGS irrelevant.
	 */
	if (ops->flags & FTRACE_OPS_FL_SAVE_REGS &&
	    !(ops->flags & FTRACE_OPS_FL_SAVE_REGS_IF_SUPPORTED))
		return -EINVAL;

	if (ops->flags & FTRACE_OPS_FL_SAVE_REGS_IF_SUPPORTED)
		ops->flags |= FTRACE_OPS_FL_SAVE_REGS;
#endif

358 359 360
	if (!core_kernel_data((unsigned long)ops))
		ops->flags |= FTRACE_OPS_FL_DYNAMIC;

361
	add_ftrace_ops(&ftrace_ops_list, ops);
362

363 364 365
	/* Always save the function, and reset at unregistering */
	ops->saved_func = ops->func;

366
	if (ftrace_pids_enabled(ops))
367 368
		ops->func = ftrace_pid_func;

369 370
	ftrace_update_trampoline(ops);

371 372 373 374 375 376 377 378 379 380
	if (ftrace_enabled)
		update_ftrace_function();

	return 0;
}

static int __unregister_ftrace_function(struct ftrace_ops *ops)
{
	int ret;

381 382 383
	if (WARN_ON(!(ops->flags & FTRACE_OPS_FL_ENABLED)))
		return -EBUSY;

384
	ret = remove_ftrace_ops(&ftrace_ops_list, ops);
385

386 387
	if (ret < 0)
		return ret;
388

389 390
	if (ftrace_enabled)
		update_ftrace_function();
391

392 393
	ops->func = ops->saved_func;

S
Steven Rostedt 已提交
394
	return 0;
395 396
}

397 398
static void ftrace_update_pid_func(void)
{
399 400
	struct ftrace_ops *op;

401
	/* Only do something if we are tracing something */
402
	if (ftrace_trace_function == ftrace_stub)
403
		return;
404

405 406
	do_for_each_ftrace_op(op, ftrace_ops_list) {
		if (op->flags & FTRACE_OPS_FL_PID) {
407 408
			op->func = ftrace_pids_enabled(op) ?
				ftrace_pid_func : op->saved_func;
409 410 411 412
			ftrace_update_trampoline(op);
		}
	} while_for_each_ftrace_op(op);

413
	update_ftrace_function();
414 415
}

416 417 418 419 420
#ifdef CONFIG_FUNCTION_PROFILER
struct ftrace_profile {
	struct hlist_node		node;
	unsigned long			ip;
	unsigned long			counter;
421 422
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
	unsigned long long		time;
423
	unsigned long long		time_squared;
424
#endif
425 426
};

427 428 429 430
struct ftrace_profile_page {
	struct ftrace_profile_page	*next;
	unsigned long			index;
	struct ftrace_profile		records[];
431 432
};

433 434 435 436 437 438 439 440
struct ftrace_profile_stat {
	atomic_t			disabled;
	struct hlist_head		*hash;
	struct ftrace_profile_page	*pages;
	struct ftrace_profile_page	*start;
	struct tracer_stat		stat;
};

441 442
#define PROFILE_RECORDS_SIZE						\
	(PAGE_SIZE - offsetof(struct ftrace_profile_page, records))
443

444 445
#define PROFILES_PER_PAGE					\
	(PROFILE_RECORDS_SIZE / sizeof(struct ftrace_profile))
446

447 448 449
static int ftrace_profile_enabled __read_mostly;

/* ftrace_profile_lock - synchronize the enable and disable of the profiler */
S
Steven Rostedt 已提交
450 451
static DEFINE_MUTEX(ftrace_profile_lock);

452
static DEFINE_PER_CPU(struct ftrace_profile_stat, ftrace_profile_stats);
453

454 455
#define FTRACE_PROFILE_HASH_BITS 10
#define FTRACE_PROFILE_HASH_SIZE (1 << FTRACE_PROFILE_HASH_BITS)
456

S
Steven Rostedt 已提交
457 458 459
static void *
function_stat_next(void *v, int idx)
{
460 461
	struct ftrace_profile *rec = v;
	struct ftrace_profile_page *pg;
S
Steven Rostedt 已提交
462

463
	pg = (struct ftrace_profile_page *)((unsigned long)rec & PAGE_MASK);
S
Steven Rostedt 已提交
464 465

 again:
L
Li Zefan 已提交
466 467 468
	if (idx != 0)
		rec++;

S
Steven Rostedt 已提交
469 470 471 472 473
	if ((void *)rec >= (void *)&pg->records[pg->index]) {
		pg = pg->next;
		if (!pg)
			return NULL;
		rec = &pg->records[0];
474 475
		if (!rec->counter)
			goto again;
S
Steven Rostedt 已提交
476 477 478 479 480 481 482
	}

	return rec;
}

static void *function_stat_start(struct tracer_stat *trace)
{
483 484 485 486 487 488 489
	struct ftrace_profile_stat *stat =
		container_of(trace, struct ftrace_profile_stat, stat);

	if (!stat || !stat->start)
		return NULL;

	return function_stat_next(&stat->start->records[0], 0);
S
Steven Rostedt 已提交
490 491
}

492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
/* function graph compares on total time */
static int function_stat_cmp(void *p1, void *p2)
{
	struct ftrace_profile *a = p1;
	struct ftrace_profile *b = p2;

	if (a->time < b->time)
		return -1;
	if (a->time > b->time)
		return 1;
	else
		return 0;
}
#else
/* not function graph compares against hits */
S
Steven Rostedt 已提交
508 509
static int function_stat_cmp(void *p1, void *p2)
{
510 511
	struct ftrace_profile *a = p1;
	struct ftrace_profile *b = p2;
S
Steven Rostedt 已提交
512 513 514 515 516 517 518 519

	if (a->counter < b->counter)
		return -1;
	if (a->counter > b->counter)
		return 1;
	else
		return 0;
}
520
#endif
S
Steven Rostedt 已提交
521 522 523

static int function_stat_headers(struct seq_file *m)
{
524
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
525 526 527 528
	seq_puts(m, "  Function                               "
		 "Hit    Time            Avg             s^2\n"
		    "  --------                               "
		 "---    ----            ---             ---\n");
529
#else
530 531
	seq_puts(m, "  Function                               Hit\n"
		    "  --------                               ---\n");
532
#endif
S
Steven Rostedt 已提交
533 534 535 536 537
	return 0;
}

static int function_stat_show(struct seq_file *m, void *v)
{
538
	struct ftrace_profile *rec = v;
S
Steven Rostedt 已提交
539
	char str[KSYM_SYMBOL_LEN];
540
	int ret = 0;
541
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
542 543
	static struct trace_seq s;
	unsigned long long avg;
544
	unsigned long long stddev;
545
#endif
546 547 548 549 550 551 552
	mutex_lock(&ftrace_profile_lock);

	/* we raced with function_profile_reset() */
	if (unlikely(rec->counter == 0)) {
		ret = -EBUSY;
		goto out;
	}
S
Steven Rostedt 已提交
553

554 555 556 557 558 559 560
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
	avg = rec->time;
	do_div(avg, rec->counter);
	if (tracing_thresh && (avg < tracing_thresh))
		goto out;
#endif

S
Steven Rostedt 已提交
561
	kallsyms_lookup(rec->ip, NULL, NULL, NULL, str);
562 563 564
	seq_printf(m, "  %-30.30s  %10lu", str, rec->counter);

#ifdef CONFIG_FUNCTION_GRAPH_TRACER
565
	seq_puts(m, "    ");
566

567 568 569 570
	/* Sample standard deviation (s^2) */
	if (rec->counter <= 1)
		stddev = 0;
	else {
571 572 573 574 575 576 577
		/*
		 * Apply Welford's method:
		 * s^2 = 1 / (n * (n-1)) * (n * \Sum (x_i)^2 - (\Sum x_i)^2)
		 */
		stddev = rec->counter * rec->time_squared -
			 rec->time * rec->time;

578 579 580 581
		/*
		 * Divide only 1000 for ns^2 -> us^2 conversion.
		 * trace_print_graph_duration will divide 1000 again.
		 */
582
		do_div(stddev, rec->counter * (rec->counter - 1) * 1000);
583 584
	}

585 586 587 588
	trace_seq_init(&s);
	trace_print_graph_duration(rec->time, &s);
	trace_seq_puts(&s, "    ");
	trace_print_graph_duration(avg, &s);
589 590
	trace_seq_puts(&s, "    ");
	trace_print_graph_duration(stddev, &s);
591 592 593
	trace_print_seq(m, &s);
#endif
	seq_putc(m, '\n');
594 595
out:
	mutex_unlock(&ftrace_profile_lock);
S
Steven Rostedt 已提交
596

597
	return ret;
S
Steven Rostedt 已提交
598 599
}

600
static void ftrace_profile_reset(struct ftrace_profile_stat *stat)
S
Steven Rostedt 已提交
601
{
602
	struct ftrace_profile_page *pg;
S
Steven Rostedt 已提交
603

604
	pg = stat->pages = stat->start;
S
Steven Rostedt 已提交
605

606 607 608 609
	while (pg) {
		memset(pg->records, 0, PROFILE_RECORDS_SIZE);
		pg->index = 0;
		pg = pg->next;
S
Steven Rostedt 已提交
610 611
	}

612
	memset(stat->hash, 0,
613 614
	       FTRACE_PROFILE_HASH_SIZE * sizeof(struct hlist_head));
}
S
Steven Rostedt 已提交
615

616
int ftrace_profile_pages_init(struct ftrace_profile_stat *stat)
617 618
{
	struct ftrace_profile_page *pg;
619 620
	int functions;
	int pages;
621
	int i;
S
Steven Rostedt 已提交
622

623
	/* If we already allocated, do nothing */
624
	if (stat->pages)
625
		return 0;
S
Steven Rostedt 已提交
626

627 628
	stat->pages = (void *)get_zeroed_page(GFP_KERNEL);
	if (!stat->pages)
629
		return -ENOMEM;
S
Steven Rostedt 已提交
630

631 632 633 634 635 636 637 638 639 640 641 642 643
#ifdef CONFIG_DYNAMIC_FTRACE
	functions = ftrace_update_tot_cnt;
#else
	/*
	 * We do not know the number of functions that exist because
	 * dynamic tracing is what counts them. With past experience
	 * we have around 20K functions. That should be more than enough.
	 * It is highly unlikely we will execute every function in
	 * the kernel.
	 */
	functions = 20000;
#endif

644
	pg = stat->start = stat->pages;
S
Steven Rostedt 已提交
645

646 647
	pages = DIV_ROUND_UP(functions, PROFILES_PER_PAGE);

648
	for (i = 1; i < pages; i++) {
649 650
		pg->next = (void *)get_zeroed_page(GFP_KERNEL);
		if (!pg->next)
651
			goto out_free;
652 653 654 655
		pg = pg->next;
	}

	return 0;
656 657 658 659 660 661 662 663 664 665 666 667 668 669

 out_free:
	pg = stat->start;
	while (pg) {
		unsigned long tmp = (unsigned long)pg;

		pg = pg->next;
		free_page(tmp);
	}

	stat->pages = NULL;
	stat->start = NULL;

	return -ENOMEM;
S
Steven Rostedt 已提交
670 671
}

672
static int ftrace_profile_init_cpu(int cpu)
S
Steven Rostedt 已提交
673
{
674
	struct ftrace_profile_stat *stat;
675
	int size;
S
Steven Rostedt 已提交
676

677 678 679
	stat = &per_cpu(ftrace_profile_stats, cpu);

	if (stat->hash) {
680
		/* If the profile is already created, simply reset it */
681
		ftrace_profile_reset(stat);
682 683
		return 0;
	}
S
Steven Rostedt 已提交
684

685 686 687 688 689
	/*
	 * We are profiling all functions, but usually only a few thousand
	 * functions are hit. We'll make a hash of 1024 items.
	 */
	size = FTRACE_PROFILE_HASH_SIZE;
S
Steven Rostedt 已提交
690

K
Kees Cook 已提交
691
	stat->hash = kcalloc(size, sizeof(struct hlist_head), GFP_KERNEL);
692

693
	if (!stat->hash)
694 695
		return -ENOMEM;

696
	/* Preallocate the function profiling pages */
697 698 699
	if (ftrace_profile_pages_init(stat) < 0) {
		kfree(stat->hash);
		stat->hash = NULL;
700 701 702 703
		return -ENOMEM;
	}

	return 0;
S
Steven Rostedt 已提交
704 705
}

706 707 708 709 710
static int ftrace_profile_init(void)
{
	int cpu;
	int ret = 0;

711
	for_each_possible_cpu(cpu) {
712 713 714 715 716 717 718 719
		ret = ftrace_profile_init_cpu(cpu);
		if (ret)
			break;
	}

	return ret;
}

720
/* interrupts must be disabled */
721 722
static struct ftrace_profile *
ftrace_find_profiled_func(struct ftrace_profile_stat *stat, unsigned long ip)
S
Steven Rostedt 已提交
723
{
724
	struct ftrace_profile *rec;
S
Steven Rostedt 已提交
725 726 727
	struct hlist_head *hhd;
	unsigned long key;

728
	key = hash_long(ip, FTRACE_PROFILE_HASH_BITS);
729
	hhd = &stat->hash[key];
S
Steven Rostedt 已提交
730 731 732 733

	if (hlist_empty(hhd))
		return NULL;

734
	hlist_for_each_entry_rcu_notrace(rec, hhd, node) {
S
Steven Rostedt 已提交
735
		if (rec->ip == ip)
736 737 738 739 740 741
			return rec;
	}

	return NULL;
}

742 743
static void ftrace_add_profile(struct ftrace_profile_stat *stat,
			       struct ftrace_profile *rec)
744 745 746
{
	unsigned long key;

747
	key = hash_long(rec->ip, FTRACE_PROFILE_HASH_BITS);
748
	hlist_add_head_rcu(&rec->node, &stat->hash[key]);
749 750
}

751 752 753
/*
 * The memory is already allocated, this simply finds a new record to use.
 */
754
static struct ftrace_profile *
755
ftrace_profile_alloc(struct ftrace_profile_stat *stat, unsigned long ip)
756 757 758
{
	struct ftrace_profile *rec = NULL;

759
	/* prevent recursion (from NMIs) */
760
	if (atomic_inc_return(&stat->disabled) != 1)
761 762 763
		goto out;

	/*
764 765
	 * Try to find the function again since an NMI
	 * could have added it
766
	 */
767
	rec = ftrace_find_profiled_func(stat, ip);
768
	if (rec)
769
		goto out;
770

771 772 773 774
	if (stat->pages->index == PROFILES_PER_PAGE) {
		if (!stat->pages->next)
			goto out;
		stat->pages = stat->pages->next;
S
Steven Rostedt 已提交
775
	}
776

777
	rec = &stat->pages->records[stat->pages->index++];
778
	rec->ip = ip;
779
	ftrace_add_profile(stat, rec);
780

S
Steven Rostedt 已提交
781
 out:
782
	atomic_dec(&stat->disabled);
S
Steven Rostedt 已提交
783 784 785 786 787

	return rec;
}

static void
788
function_profile_call(unsigned long ip, unsigned long parent_ip,
789
		      struct ftrace_ops *ops, struct pt_regs *regs)
S
Steven Rostedt 已提交
790
{
791
	struct ftrace_profile_stat *stat;
792
	struct ftrace_profile *rec;
S
Steven Rostedt 已提交
793 794 795 796 797 798
	unsigned long flags;

	if (!ftrace_profile_enabled)
		return;

	local_irq_save(flags);
799

800
	stat = this_cpu_ptr(&ftrace_profile_stats);
801
	if (!stat->hash || !ftrace_profile_enabled)
802 803 804
		goto out;

	rec = ftrace_find_profiled_func(stat, ip);
805
	if (!rec) {
806
		rec = ftrace_profile_alloc(stat, ip);
807 808 809
		if (!rec)
			goto out;
	}
S
Steven Rostedt 已提交
810 811 812 813 814 815

	rec->counter++;
 out:
	local_irq_restore(flags);
}

816 817 818
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
static int profile_graph_entry(struct ftrace_graph_ent *trace)
{
819 820
	int index = trace->depth;

821
	function_profile_call(trace->func, 0, NULL, NULL);
822

823 824 825 826
	/* If function graph is shutting down, ret_stack can be NULL */
	if (!current->ret_stack)
		return 0;

827 828 829
	if (index >= 0 && index < FTRACE_RETFUNC_DEPTH)
		current->ret_stack[index].subtime = 0;

830 831 832 833 834
	return 1;
}

static void profile_graph_return(struct ftrace_graph_ret *trace)
{
835
	struct ftrace_profile_stat *stat;
836
	unsigned long long calltime;
837
	struct ftrace_profile *rec;
838
	unsigned long flags;
839 840

	local_irq_save(flags);
841
	stat = this_cpu_ptr(&ftrace_profile_stats);
842
	if (!stat->hash || !ftrace_profile_enabled)
843 844
		goto out;

845 846 847 848
	/* If the calltime was zero'd ignore it */
	if (!trace->calltime)
		goto out;

849 850
	calltime = trace->rettime - trace->calltime;

851
	if (!fgraph_graph_time) {
852 853 854 855 856 857 858 859 860 861 862 863 864 865
		int index;

		index = trace->depth;

		/* Append this call time to the parent time to subtract */
		if (index)
			current->ret_stack[index - 1].subtime += calltime;

		if (current->ret_stack[index].subtime < calltime)
			calltime -= current->ret_stack[index].subtime;
		else
			calltime = 0;
	}

866
	rec = ftrace_find_profiled_func(stat, trace->func);
867
	if (rec) {
868
		rec->time += calltime;
869 870
		rec->time_squared += calltime * calltime;
	}
871

872
 out:
873 874 875 876 877 878 879 880 881 882 883 884 885 886
	local_irq_restore(flags);
}

static int register_ftrace_profiler(void)
{
	return register_ftrace_graph(&profile_graph_return,
				     &profile_graph_entry);
}

static void unregister_ftrace_profiler(void)
{
	unregister_ftrace_graph();
}
#else
887
static struct ftrace_ops ftrace_profile_ops __read_mostly = {
888
	.func		= function_profile_call,
889
	.flags		= FTRACE_OPS_FL_RECURSION_SAFE | FTRACE_OPS_FL_INITIALIZED,
890
	INIT_OPS_HASH(ftrace_profile_ops)
S
Steven Rostedt 已提交
891 892
};

893 894 895 896 897 898 899 900 901 902 903
static int register_ftrace_profiler(void)
{
	return register_ftrace_function(&ftrace_profile_ops);
}

static void unregister_ftrace_profiler(void)
{
	unregister_ftrace_function(&ftrace_profile_ops);
}
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */

S
Steven Rostedt 已提交
904 905 906 907 908 909 910
static ssize_t
ftrace_profile_write(struct file *filp, const char __user *ubuf,
		     size_t cnt, loff_t *ppos)
{
	unsigned long val;
	int ret;

911 912
	ret = kstrtoul_from_user(ubuf, cnt, 10, &val);
	if (ret)
S
Steven Rostedt 已提交
913 914 915 916 917 918 919
		return ret;

	val = !!val;

	mutex_lock(&ftrace_profile_lock);
	if (ftrace_profile_enabled ^ val) {
		if (val) {
920 921 922 923 924 925
			ret = ftrace_profile_init();
			if (ret < 0) {
				cnt = ret;
				goto out;
			}

926 927 928 929 930
			ret = register_ftrace_profiler();
			if (ret < 0) {
				cnt = ret;
				goto out;
			}
S
Steven Rostedt 已提交
931 932 933
			ftrace_profile_enabled = 1;
		} else {
			ftrace_profile_enabled = 0;
934 935 936 937
			/*
			 * unregister_ftrace_profiler calls stop_machine
			 * so this acts like an synchronize_sched.
			 */
938
			unregister_ftrace_profiler();
S
Steven Rostedt 已提交
939 940
		}
	}
941
 out:
S
Steven Rostedt 已提交
942 943
	mutex_unlock(&ftrace_profile_lock);

944
	*ppos += cnt;
S
Steven Rostedt 已提交
945 946 947 948

	return cnt;
}

949 950 951 952
static ssize_t
ftrace_profile_read(struct file *filp, char __user *ubuf,
		     size_t cnt, loff_t *ppos)
{
953
	char buf[64];		/* big enough to hold a number */
954 955 956 957 958 959
	int r;

	r = sprintf(buf, "%u\n", ftrace_profile_enabled);
	return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
}

S
Steven Rostedt 已提交
960 961 962 963
static const struct file_operations ftrace_profile_fops = {
	.open		= tracing_open_generic,
	.read		= ftrace_profile_read,
	.write		= ftrace_profile_write,
964
	.llseek		= default_llseek,
S
Steven Rostedt 已提交
965 966
};

967 968
/* used to initialize the real stat files */
static struct tracer_stat function_stats __initdata = {
969 970 971 972 973 974
	.name		= "functions",
	.stat_start	= function_stat_start,
	.stat_next	= function_stat_next,
	.stat_cmp	= function_stat_cmp,
	.stat_headers	= function_stat_headers,
	.stat_show	= function_stat_show
975 976
};

977
static __init void ftrace_profile_tracefs(struct dentry *d_tracer)
S
Steven Rostedt 已提交
978
{
979
	struct ftrace_profile_stat *stat;
S
Steven Rostedt 已提交
980
	struct dentry *entry;
981
	char *name;
S
Steven Rostedt 已提交
982
	int ret;
983 984 985 986 987
	int cpu;

	for_each_possible_cpu(cpu) {
		stat = &per_cpu(ftrace_profile_stats, cpu);

988
		name = kasprintf(GFP_KERNEL, "function%d", cpu);
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008
		if (!name) {
			/*
			 * The files created are permanent, if something happens
			 * we still do not free memory.
			 */
			WARN(1,
			     "Could not allocate stat file for cpu %d\n",
			     cpu);
			return;
		}
		stat->stat = function_stats;
		stat->stat.name = name;
		ret = register_stat_tracer(&stat->stat);
		if (ret) {
			WARN(1,
			     "Could not register function stat for cpu %d\n",
			     cpu);
			kfree(name);
			return;
		}
S
Steven Rostedt 已提交
1009 1010
	}

1011
	entry = tracefs_create_file("function_profile_enabled", 0644,
S
Steven Rostedt 已提交
1012 1013
				    d_tracer, NULL, &ftrace_profile_fops);
	if (!entry)
1014
		pr_warn("Could not create tracefs 'function_profile_enabled' entry\n");
S
Steven Rostedt 已提交
1015 1016 1017
}

#else /* CONFIG_FUNCTION_PROFILER */
1018
static __init void ftrace_profile_tracefs(struct dentry *d_tracer)
S
Steven Rostedt 已提交
1019 1020 1021 1022
{
}
#endif /* CONFIG_FUNCTION_PROFILER */

1023 1024
static struct pid * const ftrace_swapper_pid = &init_struct_pid;

1025 1026 1027 1028 1029 1030
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
static int ftrace_graph_active;
#else
# define ftrace_graph_active 0
#endif

1031 1032
#ifdef CONFIG_DYNAMIC_FTRACE

1033 1034
static struct ftrace_ops *removed_ops;

1035 1036 1037 1038 1039 1040
/*
 * Set when doing a global update, like enabling all recs or disabling them.
 * It is not set when just updating a single ftrace_ops.
 */
static bool update_all_ops;

1041 1042 1043 1044
#ifndef CONFIG_FTRACE_MCOUNT_RECORD
# error Dynamic ftrace depends on MCOUNT_RECORD
#endif

1045 1046 1047 1048 1049
struct ftrace_func_entry {
	struct hlist_node hlist;
	unsigned long ip;
};

1050 1051 1052 1053 1054
struct ftrace_func_probe {
	struct ftrace_probe_ops	*probe_ops;
	struct ftrace_ops	ops;
	struct trace_array	*tr;
	struct list_head	list;
1055
	void			*data;
1056 1057 1058
	int			ref;
};

1059 1060 1061 1062 1063 1064 1065 1066 1067
/*
 * We make these constant because no one should touch them,
 * but they are used as the default "empty hash", to avoid allocating
 * it all the time. These are in a read only section such that if
 * anyone does try to modify it, it will cause an exception.
 */
static const struct hlist_head empty_buckets[1];
static const struct ftrace_hash empty_hash = {
	.buckets = (struct hlist_head *)empty_buckets,
1068
};
1069
#define EMPTY_HASH	((struct ftrace_hash *)&empty_hash)
1070

1071
static struct ftrace_ops global_ops = {
1072 1073 1074 1075 1076
	.func				= ftrace_stub,
	.local_hash.notrace_hash	= EMPTY_HASH,
	.local_hash.filter_hash		= EMPTY_HASH,
	INIT_OPS_HASH(global_ops)
	.flags				= FTRACE_OPS_FL_RECURSION_SAFE |
1077 1078
					  FTRACE_OPS_FL_INITIALIZED |
					  FTRACE_OPS_FL_PID,
1079 1080
};

1081
/*
1082
 * Used by the stack undwinder to know about dynamic ftrace trampolines.
1083
 */
1084
struct ftrace_ops *ftrace_ops_trampoline(unsigned long addr)
1085
{
1086
	struct ftrace_ops *op = NULL;
1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102

	/*
	 * Some of the ops may be dynamically allocated,
	 * they are freed after a synchronize_sched().
	 */
	preempt_disable_notrace();

	do_for_each_ftrace_op(op, ftrace_ops_list) {
		/*
		 * This is to check for dynamically allocated trampolines.
		 * Trampolines that are in kernel text will have
		 * core_kernel_text() return true.
		 */
		if (op->trampoline && op->trampoline_size)
			if (addr >= op->trampoline &&
			    addr < op->trampoline + op->trampoline_size) {
1103 1104
				preempt_enable_notrace();
				return op;
1105 1106 1107 1108
			}
	} while_for_each_ftrace_op(op);
	preempt_enable_notrace();

1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
	return NULL;
}

/*
 * This is used by __kernel_text_address() to return true if the
 * address is on a dynamically allocated trampoline that would
 * not return true for either core_kernel_text() or
 * is_module_text_address().
 */
bool is_ftrace_trampoline(unsigned long addr)
{
	return ftrace_ops_trampoline(addr) != NULL;
1121 1122
}

1123 1124
struct ftrace_page {
	struct ftrace_page	*next;
1125
	struct dyn_ftrace	*records;
1126
	int			index;
1127
	int			size;
1128 1129
};

1130 1131
#define ENTRY_SIZE sizeof(struct dyn_ftrace)
#define ENTRIES_PER_PAGE (PAGE_SIZE / ENTRY_SIZE)
1132 1133 1134 1135 1136 1137 1138

/* estimate from running different kernels */
#define NR_TO_INIT		10000

static struct ftrace_page	*ftrace_pages_start;
static struct ftrace_page	*ftrace_pages;

1139 1140 1141 1142 1143 1144 1145 1146 1147
static __always_inline unsigned long
ftrace_hash_key(struct ftrace_hash *hash, unsigned long ip)
{
	if (hash->size_bits > 0)
		return hash_long(ip, hash->size_bits);

	return 0;
}

1148 1149 1150
/* Only use this function if ftrace_hash_empty() has already been tested */
static __always_inline struct ftrace_func_entry *
__ftrace_lookup_ip(struct ftrace_hash *hash, unsigned long ip)
1151 1152 1153 1154 1155
{
	unsigned long key;
	struct ftrace_func_entry *entry;
	struct hlist_head *hhd;

1156
	key = ftrace_hash_key(hash, ip);
1157 1158
	hhd = &hash->buckets[key];

1159
	hlist_for_each_entry_rcu_notrace(entry, hhd, hlist) {
1160 1161 1162 1163 1164 1165
		if (entry->ip == ip)
			return entry;
	}
	return NULL;
}

1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184
/**
 * ftrace_lookup_ip - Test to see if an ip exists in an ftrace_hash
 * @hash: The hash to look at
 * @ip: The instruction pointer to test
 *
 * Search a given @hash to see if a given instruction pointer (@ip)
 * exists in it.
 *
 * Returns the entry that holds the @ip if found. NULL otherwise.
 */
struct ftrace_func_entry *
ftrace_lookup_ip(struct ftrace_hash *hash, unsigned long ip)
{
	if (ftrace_hash_empty(hash))
		return NULL;

	return __ftrace_lookup_ip(hash, ip);
}

1185 1186
static void __add_hash_entry(struct ftrace_hash *hash,
			     struct ftrace_func_entry *entry)
1187 1188 1189 1190
{
	struct hlist_head *hhd;
	unsigned long key;

1191
	key = ftrace_hash_key(hash, entry->ip);
1192 1193 1194
	hhd = &hash->buckets[key];
	hlist_add_head(&entry->hlist, hhd);
	hash->count++;
1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
}

static int add_hash_entry(struct ftrace_hash *hash, unsigned long ip)
{
	struct ftrace_func_entry *entry;

	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
	if (!entry)
		return -ENOMEM;

	entry->ip = ip;
	__add_hash_entry(hash, entry);
1207 1208 1209 1210 1211

	return 0;
}

static void
1212
free_hash_entry(struct ftrace_hash *hash,
1213 1214 1215 1216 1217 1218 1219
		  struct ftrace_func_entry *entry)
{
	hlist_del(&entry->hlist);
	kfree(entry);
	hash->count--;
}

1220 1221 1222 1223
static void
remove_hash_entry(struct ftrace_hash *hash,
		  struct ftrace_func_entry *entry)
{
1224
	hlist_del_rcu(&entry->hlist);
1225 1226 1227
	hash->count--;
}

1228 1229 1230
static void ftrace_hash_clear(struct ftrace_hash *hash)
{
	struct hlist_head *hhd;
1231
	struct hlist_node *tn;
1232 1233 1234 1235
	struct ftrace_func_entry *entry;
	int size = 1 << hash->size_bits;
	int i;

1236 1237 1238
	if (!hash->count)
		return;

1239 1240
	for (i = 0; i < size; i++) {
		hhd = &hash->buckets[i];
1241
		hlist_for_each_entry_safe(entry, tn, hhd, hlist)
1242
			free_hash_entry(hash, entry);
1243 1244 1245 1246
	}
	FTRACE_WARN_ON(hash->count);
}

1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268
static void free_ftrace_mod(struct ftrace_mod_load *ftrace_mod)
{
	list_del(&ftrace_mod->list);
	kfree(ftrace_mod->module);
	kfree(ftrace_mod->func);
	kfree(ftrace_mod);
}

static void clear_ftrace_mod_list(struct list_head *head)
{
	struct ftrace_mod_load *p, *n;

	/* stack tracer isn't supported yet */
	if (!head)
		return;

	mutex_lock(&ftrace_lock);
	list_for_each_entry_safe(p, n, head, list)
		free_ftrace_mod(p);
	mutex_unlock(&ftrace_lock);
}

1269 1270 1271 1272 1273 1274 1275 1276 1277
static void free_ftrace_hash(struct ftrace_hash *hash)
{
	if (!hash || hash == EMPTY_HASH)
		return;
	ftrace_hash_clear(hash);
	kfree(hash->buckets);
	kfree(hash);
}

1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
static void __free_ftrace_hash_rcu(struct rcu_head *rcu)
{
	struct ftrace_hash *hash;

	hash = container_of(rcu, struct ftrace_hash, rcu);
	free_ftrace_hash(hash);
}

static void free_ftrace_hash_rcu(struct ftrace_hash *hash)
{
	if (!hash || hash == EMPTY_HASH)
		return;
	call_rcu_sched(&hash->rcu, __free_ftrace_hash_rcu);
}

1293 1294
void ftrace_free_filter(struct ftrace_ops *ops)
{
1295
	ftrace_ops_init(ops);
1296 1297
	free_ftrace_hash(ops->func_hash->filter_hash);
	free_ftrace_hash(ops->func_hash->notrace_hash);
1298 1299
}

1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
static struct ftrace_hash *alloc_ftrace_hash(int size_bits)
{
	struct ftrace_hash *hash;
	int size;

	hash = kzalloc(sizeof(*hash), GFP_KERNEL);
	if (!hash)
		return NULL;

	size = 1 << size_bits;
1310
	hash->buckets = kcalloc(size, sizeof(*hash->buckets), GFP_KERNEL);
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321

	if (!hash->buckets) {
		kfree(hash);
		return NULL;
	}

	hash->size_bits = size_bits;

	return hash;
}

1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350

static int ftrace_add_mod(struct trace_array *tr,
			  const char *func, const char *module,
			  int enable)
{
	struct ftrace_mod_load *ftrace_mod;
	struct list_head *mod_head = enable ? &tr->mod_trace : &tr->mod_notrace;

	ftrace_mod = kzalloc(sizeof(*ftrace_mod), GFP_KERNEL);
	if (!ftrace_mod)
		return -ENOMEM;

	ftrace_mod->func = kstrdup(func, GFP_KERNEL);
	ftrace_mod->module = kstrdup(module, GFP_KERNEL);
	ftrace_mod->enable = enable;

	if (!ftrace_mod->func || !ftrace_mod->module)
		goto out_free;

	list_add(&ftrace_mod->list, mod_head);

	return 0;

 out_free:
	free_ftrace_mod(ftrace_mod);

	return -ENOMEM;
}

1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363
static struct ftrace_hash *
alloc_and_copy_ftrace_hash(int size_bits, struct ftrace_hash *hash)
{
	struct ftrace_func_entry *entry;
	struct ftrace_hash *new_hash;
	int size;
	int ret;
	int i;

	new_hash = alloc_ftrace_hash(size_bits);
	if (!new_hash)
		return NULL;

1364 1365 1366
	if (hash)
		new_hash->flags = hash->flags;

1367
	/* Empty hash? */
1368
	if (ftrace_hash_empty(hash))
1369 1370 1371 1372
		return new_hash;

	size = 1 << hash->size_bits;
	for (i = 0; i < size; i++) {
1373
		hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388
			ret = add_hash_entry(new_hash, entry->ip);
			if (ret < 0)
				goto free_hash;
		}
	}

	FTRACE_WARN_ON(new_hash->count != hash->count);

	return new_hash;

 free_hash:
	free_ftrace_hash(new_hash);
	return NULL;
}

1389
static void
1390
ftrace_hash_rec_disable_modify(struct ftrace_ops *ops, int filter_hash);
1391
static void
1392
ftrace_hash_rec_enable_modify(struct ftrace_ops *ops, int filter_hash);
1393

1394 1395 1396
static int ftrace_hash_ipmodify_update(struct ftrace_ops *ops,
				       struct ftrace_hash *new_hash);

1397 1398
static struct ftrace_hash *
__ftrace_hash_move(struct ftrace_hash *src)
1399 1400
{
	struct ftrace_func_entry *entry;
1401
	struct hlist_node *tn;
1402
	struct hlist_head *hhd;
1403
	struct ftrace_hash *new_hash;
1404 1405 1406 1407 1408
	int size = src->count;
	int bits = 0;
	int i;

	/*
1409
	 * If the new source is empty, just return the empty_hash.
1410
	 */
1411
	if (ftrace_hash_empty(src))
1412
		return EMPTY_HASH;
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423

	/*
	 * Make the hash size about 1/2 the # found
	 */
	for (size /= 2; size; size >>= 1)
		bits++;

	/* Don't allocate too much */
	if (bits > FTRACE_HASH_MAX_BITS)
		bits = FTRACE_HASH_MAX_BITS;

1424 1425
	new_hash = alloc_ftrace_hash(bits);
	if (!new_hash)
1426
		return NULL;
1427

1428 1429
	new_hash->flags = src->flags;

1430 1431 1432
	size = 1 << src->size_bits;
	for (i = 0; i < size; i++) {
		hhd = &src->buckets[i];
1433
		hlist_for_each_entry_safe(entry, tn, hhd, hlist) {
1434
			remove_hash_entry(src, entry);
1435
			__add_hash_entry(new_hash, entry);
1436 1437 1438
		}
	}

1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
	return new_hash;
}

static int
ftrace_hash_move(struct ftrace_ops *ops, int enable,
		 struct ftrace_hash **dst, struct ftrace_hash *src)
{
	struct ftrace_hash *new_hash;
	int ret;

	/* Reject setting notrace hash on IPMODIFY ftrace_ops */
	if (ops->flags & FTRACE_OPS_FL_IPMODIFY && !enable)
		return -EINVAL;

	new_hash = __ftrace_hash_move(src);
	if (!new_hash)
		return -ENOMEM;

1457 1458 1459 1460 1461 1462 1463 1464 1465 1466
	/* Make sure this can be applied if it is IPMODIFY ftrace_ops */
	if (enable) {
		/* IPMODIFY should be updated only when filter_hash updating */
		ret = ftrace_hash_ipmodify_update(ops, new_hash);
		if (ret < 0) {
			free_ftrace_hash(new_hash);
			return ret;
		}
	}

1467 1468 1469 1470
	/*
	 * Remove the current set, update the hash and add
	 * them back.
	 */
1471
	ftrace_hash_rec_disable_modify(ops, enable);
1472

1473 1474
	rcu_assign_pointer(*dst, new_hash);

1475
	ftrace_hash_rec_enable_modify(ops, enable);
1476

1477
	return 0;
1478 1479
}

1480 1481 1482 1483 1484 1485 1486 1487 1488 1489
static bool hash_contains_ip(unsigned long ip,
			     struct ftrace_ops_hash *hash)
{
	/*
	 * The function record is a match if it exists in the filter
	 * hash and not in the notrace hash. Note, an emty hash is
	 * considered a match for the filter hash, but an empty
	 * notrace hash is considered not in the notrace hash.
	 */
	return (ftrace_hash_empty(hash->filter_hash) ||
1490
		__ftrace_lookup_ip(hash->filter_hash, ip)) &&
1491
		(ftrace_hash_empty(hash->notrace_hash) ||
1492
		 !__ftrace_lookup_ip(hash->notrace_hash, ip));
1493 1494
}

1495 1496 1497 1498 1499 1500 1501 1502
/*
 * Test the hashes for this ops to see if we want to call
 * the ops->func or not.
 *
 * It's a match if the ip is in the ops->filter_hash or
 * the filter_hash does not exist or is empty,
 *  AND
 * the ip is not in the ops->notrace_hash.
1503 1504 1505
 *
 * This needs to be called with preemption disabled as
 * the hashes are freed with call_rcu_sched().
1506 1507
 */
static int
1508
ftrace_ops_test(struct ftrace_ops *ops, unsigned long ip, void *regs)
1509
{
1510
	struct ftrace_ops_hash hash;
1511 1512
	int ret;

1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
	/*
	 * There's a small race when adding ops that the ftrace handler
	 * that wants regs, may be called without them. We can not
	 * allow that handler to be called if regs is NULL.
	 */
	if (regs == NULL && (ops->flags & FTRACE_OPS_FL_SAVE_REGS))
		return 0;
#endif

1523 1524
	rcu_assign_pointer(hash.filter_hash, ops->func_hash->filter_hash);
	rcu_assign_pointer(hash.notrace_hash, ops->func_hash->notrace_hash);
1525

1526
	if (hash_contains_ip(ip, &hash))
1527 1528 1529 1530 1531 1532 1533
		ret = 1;
	else
		ret = 0;

	return ret;
}

1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547
/*
 * This is a double for. Do not use 'break' to break out of the loop,
 * you must use a goto.
 */
#define do_for_each_ftrace_rec(pg, rec)					\
	for (pg = ftrace_pages_start; pg; pg = pg->next) {		\
		int _____i;						\
		for (_____i = 0; _____i < pg->index; _____i++) {	\
			rec = &pg->records[_____i];

#define while_for_each_ftrace_rec()		\
		}				\
	}

1548 1549 1550

static int ftrace_cmp_recs(const void *a, const void *b)
{
1551 1552
	const struct dyn_ftrace *key = a;
	const struct dyn_ftrace *rec = b;
1553

1554
	if (key->flags < rec->ip)
1555
		return -1;
1556 1557
	if (key->ip >= rec->ip + MCOUNT_INSN_SIZE)
		return 1;
1558 1559 1560
	return 0;
}

1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573
/**
 * ftrace_location_range - return the first address of a traced location
 *	if it touches the given ip range
 * @start: start of range to search.
 * @end: end of range to search (inclusive). @end points to the last byte
 *	to check.
 *
 * Returns rec->ip if the related ftrace location is a least partly within
 * the given address range. That is, the first address of the instruction
 * that is either a NOP or call to the function tracer. It checks the ftrace
 * internal tables to determine if the address belongs or not.
 */
unsigned long ftrace_location_range(unsigned long start, unsigned long end)
1574 1575 1576
{
	struct ftrace_page *pg;
	struct dyn_ftrace *rec;
1577
	struct dyn_ftrace key;
1578

1579 1580
	key.ip = start;
	key.flags = end;	/* overload flags, as it is unsigned long */
1581 1582

	for (pg = ftrace_pages_start; pg; pg = pg->next) {
1583 1584
		if (end < pg->records[0].ip ||
		    start >= (pg->records[pg->index - 1].ip + MCOUNT_INSN_SIZE))
1585
			continue;
1586 1587 1588 1589
		rec = bsearch(&key, pg->records, pg->index,
			      sizeof(struct dyn_ftrace),
			      ftrace_cmp_recs);
		if (rec)
1590
			return rec->ip;
1591
	}
1592 1593 1594 1595

	return 0;
}

1596 1597 1598 1599
/**
 * ftrace_location - return true if the ip giving is a traced location
 * @ip: the instruction pointer to check
 *
1600
 * Returns rec->ip if @ip given is a pointer to a ftrace location.
1601 1602 1603 1604
 * That is, the instruction that is either a NOP or call to
 * the function tracer. It checks the ftrace internal tables to
 * determine if the address belongs or not.
 */
1605
unsigned long ftrace_location(unsigned long ip)
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619
{
	return ftrace_location_range(ip, ip);
}

/**
 * ftrace_text_reserved - return true if range contains an ftrace location
 * @start: start of range to search
 * @end: end of range to search (inclusive). @end points to the last byte to check.
 *
 * Returns 1 if @start and @end contains a ftrace location.
 * That is, the instruction that is either a NOP or call to
 * the function tracer. It checks the ftrace internal tables to
 * determine if the address belongs or not.
 */
1620
int ftrace_text_reserved(const void *start, const void *end)
1621
{
1622 1623 1624 1625 1626 1627
	unsigned long ret;

	ret = ftrace_location_range((unsigned long)start,
				    (unsigned long)end);

	return (int)!!ret;
1628 1629
}

1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
/* Test if ops registered to this rec needs regs */
static bool test_rec_ops_needs_regs(struct dyn_ftrace *rec)
{
	struct ftrace_ops *ops;
	bool keep_regs = false;

	for (ops = ftrace_ops_list;
	     ops != &ftrace_list_end; ops = ops->next) {
		/* pass rec in as regs to have non-NULL val */
		if (ftrace_ops_test(ops, rec->ip, rec)) {
			if (ops->flags & FTRACE_OPS_FL_SAVE_REGS) {
				keep_regs = true;
				break;
			}
		}
	}

	return  keep_regs;
}

1650
static bool __ftrace_hash_rec_update(struct ftrace_ops *ops,
1651 1652 1653 1654 1655 1656 1657
				     int filter_hash,
				     bool inc)
{
	struct ftrace_hash *hash;
	struct ftrace_hash *other_hash;
	struct ftrace_page *pg;
	struct dyn_ftrace *rec;
1658
	bool update = false;
1659
	int count = 0;
1660
	int all = false;
1661 1662 1663

	/* Only update if the ops has been registered */
	if (!(ops->flags & FTRACE_OPS_FL_ENABLED))
1664
		return false;
1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677

	/*
	 * In the filter_hash case:
	 *   If the count is zero, we update all records.
	 *   Otherwise we just update the items in the hash.
	 *
	 * In the notrace_hash case:
	 *   We enable the update in the hash.
	 *   As disabling notrace means enabling the tracing,
	 *   and enabling notrace means disabling, the inc variable
	 *   gets inversed.
	 */
	if (filter_hash) {
1678 1679
		hash = ops->func_hash->filter_hash;
		other_hash = ops->func_hash->notrace_hash;
1680
		if (ftrace_hash_empty(hash))
1681
			all = true;
1682 1683
	} else {
		inc = !inc;
1684 1685
		hash = ops->func_hash->notrace_hash;
		other_hash = ops->func_hash->filter_hash;
1686 1687 1688 1689
		/*
		 * If the notrace hash has no items,
		 * then there's nothing to do.
		 */
1690
		if (ftrace_hash_empty(hash))
1691
			return false;
1692 1693 1694 1695 1696 1697 1698
	}

	do_for_each_ftrace_rec(pg, rec) {
		int in_other_hash = 0;
		int in_hash = 0;
		int match = 0;

1699 1700 1701
		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

1702 1703 1704 1705 1706
		if (all) {
			/*
			 * Only the filter_hash affects all records.
			 * Update if the record is not in the notrace hash.
			 */
1707
			if (!other_hash || !ftrace_lookup_ip(other_hash, rec->ip))
1708 1709
				match = 1;
		} else {
1710 1711
			in_hash = !!ftrace_lookup_ip(hash, rec->ip);
			in_other_hash = !!ftrace_lookup_ip(other_hash, rec->ip);
1712 1713

			/*
1714 1715
			 * If filter_hash is set, we want to match all functions
			 * that are in the hash but not in the other hash.
1716
			 *
1717 1718 1719 1720 1721
			 * If filter_hash is not set, then we are decrementing.
			 * That means we match anything that is in the hash
			 * and also in the other_hash. That is, we need to turn
			 * off functions in the other hash because they are disabled
			 * by this hash.
1722 1723 1724 1725
			 */
			if (filter_hash && in_hash && !in_other_hash)
				match = 1;
			else if (!filter_hash && in_hash &&
1726
				 (in_other_hash || ftrace_hash_empty(other_hash)))
1727 1728 1729 1730 1731 1732 1733
				match = 1;
		}
		if (!match)
			continue;

		if (inc) {
			rec->flags++;
1734
			if (FTRACE_WARN_ON(ftrace_rec_count(rec) == FTRACE_REF_MAX))
1735
				return false;
1736 1737 1738 1739 1740 1741

			/*
			 * If there's only a single callback registered to a
			 * function, and the ops has a trampoline registered
			 * for it, then we can call it directly.
			 */
1742
			if (ftrace_rec_count(rec) == 1 && ops->trampoline)
1743
				rec->flags |= FTRACE_FL_TRAMP;
1744
			else
1745 1746 1747
				/*
				 * If we are adding another function callback
				 * to this function, and the previous had a
1748 1749
				 * custom trampoline in use, then we need to go
				 * back to the default trampoline.
1750
				 */
1751
				rec->flags &= ~FTRACE_FL_TRAMP;
1752

1753 1754 1755 1756 1757 1758
			/*
			 * If any ops wants regs saved for this function
			 * then all ops will get saved regs.
			 */
			if (ops->flags & FTRACE_OPS_FL_SAVE_REGS)
				rec->flags |= FTRACE_FL_REGS;
1759
		} else {
1760
			if (FTRACE_WARN_ON(ftrace_rec_count(rec) == 0))
1761
				return false;
1762
			rec->flags--;
1763

1764 1765 1766 1767 1768 1769
			/*
			 * If the rec had REGS enabled and the ops that is
			 * being removed had REGS set, then see if there is
			 * still any ops for this record that wants regs.
			 * If not, we can stop recording them.
			 */
1770
			if (ftrace_rec_count(rec) > 0 &&
1771 1772 1773 1774 1775
			    rec->flags & FTRACE_FL_REGS &&
			    ops->flags & FTRACE_OPS_FL_SAVE_REGS) {
				if (!test_rec_ops_needs_regs(rec))
					rec->flags &= ~FTRACE_FL_REGS;
			}
1776

1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787
			/*
			 * If the rec had TRAMP enabled, then it needs to
			 * be cleared. As TRAMP can only be enabled iff
			 * there is only a single ops attached to it.
			 * In otherwords, always disable it on decrementing.
			 * In the future, we may set it if rec count is
			 * decremented to one, and the ops that is left
			 * has a trampoline.
			 */
			rec->flags &= ~FTRACE_FL_TRAMP;

1788 1789 1790 1791
			/*
			 * flags will be cleared in ftrace_check_record()
			 * if rec count is zero.
			 */
1792 1793
		}
		count++;
1794 1795 1796 1797

		/* Must match FTRACE_UPDATE_CALLS in ftrace_modify_all_code() */
		update |= ftrace_test_record(rec, 1) != FTRACE_UPDATE_IGNORE;

1798 1799
		/* Shortcut, if we handled all records, we are done. */
		if (!all && count == hash->count)
1800
			return update;
1801
	} while_for_each_ftrace_rec();
1802 1803

	return update;
1804 1805
}

1806
static bool ftrace_hash_rec_disable(struct ftrace_ops *ops,
1807 1808
				    int filter_hash)
{
1809
	return __ftrace_hash_rec_update(ops, filter_hash, 0);
1810 1811
}

1812
static bool ftrace_hash_rec_enable(struct ftrace_ops *ops,
1813 1814
				   int filter_hash)
{
1815
	return __ftrace_hash_rec_update(ops, filter_hash, 1);
1816 1817
}

1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852
static void ftrace_hash_rec_update_modify(struct ftrace_ops *ops,
					  int filter_hash, int inc)
{
	struct ftrace_ops *op;

	__ftrace_hash_rec_update(ops, filter_hash, inc);

	if (ops->func_hash != &global_ops.local_hash)
		return;

	/*
	 * If the ops shares the global_ops hash, then we need to update
	 * all ops that are enabled and use this hash.
	 */
	do_for_each_ftrace_op(op, ftrace_ops_list) {
		/* Already done */
		if (op == ops)
			continue;
		if (op->func_hash == &global_ops.local_hash)
			__ftrace_hash_rec_update(op, filter_hash, inc);
	} while_for_each_ftrace_op(op);
}

static void ftrace_hash_rec_disable_modify(struct ftrace_ops *ops,
					   int filter_hash)
{
	ftrace_hash_rec_update_modify(ops, filter_hash, 0);
}

static void ftrace_hash_rec_enable_modify(struct ftrace_ops *ops,
					  int filter_hash)
{
	ftrace_hash_rec_update_modify(ops, filter_hash, 1);
}

1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885
/*
 * Try to update IPMODIFY flag on each ftrace_rec. Return 0 if it is OK
 * or no-needed to update, -EBUSY if it detects a conflict of the flag
 * on a ftrace_rec, and -EINVAL if the new_hash tries to trace all recs.
 * Note that old_hash and new_hash has below meanings
 *  - If the hash is NULL, it hits all recs (if IPMODIFY is set, this is rejected)
 *  - If the hash is EMPTY_HASH, it hits nothing
 *  - Anything else hits the recs which match the hash entries.
 */
static int __ftrace_hash_update_ipmodify(struct ftrace_ops *ops,
					 struct ftrace_hash *old_hash,
					 struct ftrace_hash *new_hash)
{
	struct ftrace_page *pg;
	struct dyn_ftrace *rec, *end = NULL;
	int in_old, in_new;

	/* Only update if the ops has been registered */
	if (!(ops->flags & FTRACE_OPS_FL_ENABLED))
		return 0;

	if (!(ops->flags & FTRACE_OPS_FL_IPMODIFY))
		return 0;

	/*
	 * Since the IPMODIFY is a very address sensitive action, we do not
	 * allow ftrace_ops to set all functions to new hash.
	 */
	if (!new_hash || !old_hash)
		return -EINVAL;

	/* Update rec->flags */
	do_for_each_ftrace_rec(pg, rec) {
1886 1887 1888 1889

		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911
		/* We need to update only differences of filter_hash */
		in_old = !!ftrace_lookup_ip(old_hash, rec->ip);
		in_new = !!ftrace_lookup_ip(new_hash, rec->ip);
		if (in_old == in_new)
			continue;

		if (in_new) {
			/* New entries must ensure no others are using it */
			if (rec->flags & FTRACE_FL_IPMODIFY)
				goto rollback;
			rec->flags |= FTRACE_FL_IPMODIFY;
		} else /* Removed entry */
			rec->flags &= ~FTRACE_FL_IPMODIFY;
	} while_for_each_ftrace_rec();

	return 0;

rollback:
	end = rec;

	/* Roll back what we did above */
	do_for_each_ftrace_rec(pg, rec) {
1912 1913 1914 1915

		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
		if (rec == end)
			goto err_out;

		in_old = !!ftrace_lookup_ip(old_hash, rec->ip);
		in_new = !!ftrace_lookup_ip(new_hash, rec->ip);
		if (in_old == in_new)
			continue;

		if (in_new)
			rec->flags &= ~FTRACE_FL_IPMODIFY;
		else
			rec->flags |= FTRACE_FL_IPMODIFY;
	} while_for_each_ftrace_rec();

err_out:
	return -EBUSY;
}

static int ftrace_hash_ipmodify_enable(struct ftrace_ops *ops)
{
	struct ftrace_hash *hash = ops->func_hash->filter_hash;

	if (ftrace_hash_empty(hash))
		hash = NULL;

	return __ftrace_hash_update_ipmodify(ops, EMPTY_HASH, hash);
}

/* Disabling always succeeds */
static void ftrace_hash_ipmodify_disable(struct ftrace_ops *ops)
{
	struct ftrace_hash *hash = ops->func_hash->filter_hash;

	if (ftrace_hash_empty(hash))
		hash = NULL;

	__ftrace_hash_update_ipmodify(ops, hash, EMPTY_HASH);
}

static int ftrace_hash_ipmodify_update(struct ftrace_ops *ops,
				       struct ftrace_hash *new_hash)
{
	struct ftrace_hash *old_hash = ops->func_hash->filter_hash;

	if (ftrace_hash_empty(old_hash))
		old_hash = NULL;

	if (ftrace_hash_empty(new_hash))
		new_hash = NULL;

	return __ftrace_hash_update_ipmodify(ops, old_hash, new_hash);
}

1969
static void print_ip_ins(const char *fmt, const unsigned char *p)
1970 1971 1972 1973 1974 1975 1976 1977 1978
{
	int i;

	printk(KERN_CONT "%s", fmt);

	for (i = 0; i < MCOUNT_INSN_SIZE; i++)
		printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]);
}

1979 1980
static struct ftrace_ops *
ftrace_find_tramp_ops_any(struct dyn_ftrace *rec);
1981 1982
static struct ftrace_ops *
ftrace_find_tramp_ops_next(struct dyn_ftrace *rec, struct ftrace_ops *ops);
1983

1984
enum ftrace_bug_type ftrace_bug_type;
1985
const void *ftrace_expected;
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

static void print_bug_type(void)
{
	switch (ftrace_bug_type) {
	case FTRACE_BUG_UNKNOWN:
		break;
	case FTRACE_BUG_INIT:
		pr_info("Initializing ftrace call sites\n");
		break;
	case FTRACE_BUG_NOP:
		pr_info("Setting ftrace call site to NOP\n");
		break;
	case FTRACE_BUG_CALL:
		pr_info("Setting ftrace call site to call ftrace function\n");
		break;
	case FTRACE_BUG_UPDATE:
		pr_info("Updating ftrace call site to call a different ftrace function\n");
		break;
	}
}

2007 2008 2009
/**
 * ftrace_bug - report and shutdown function tracer
 * @failed: The failed type (EFAULT, EINVAL, EPERM)
2010
 * @rec: The record that failed
2011 2012 2013 2014 2015 2016 2017 2018
 *
 * The arch code that enables or disables the function tracing
 * can call ftrace_bug() when it has detected a problem in
 * modifying the code. @failed should be one of either:
 * EFAULT - if the problem happens on reading the @ip address
 * EINVAL - if what is read at @ip is not what was expected
 * EPERM - if the problem happens on writting to the @ip address
 */
2019
void ftrace_bug(int failed, struct dyn_ftrace *rec)
2020
{
2021 2022
	unsigned long ip = rec ? rec->ip : 0;

2023 2024 2025 2026 2027 2028 2029 2030 2031 2032
	switch (failed) {
	case -EFAULT:
		FTRACE_WARN_ON_ONCE(1);
		pr_info("ftrace faulted on modifying ");
		print_ip_sym(ip);
		break;
	case -EINVAL:
		FTRACE_WARN_ON_ONCE(1);
		pr_info("ftrace failed to modify ");
		print_ip_sym(ip);
2033
		print_ip_ins(" actual:   ", (unsigned char *)ip);
2034
		pr_cont("\n");
2035 2036 2037 2038
		if (ftrace_expected) {
			print_ip_ins(" expected: ", ftrace_expected);
			pr_cont("\n");
		}
2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049
		break;
	case -EPERM:
		FTRACE_WARN_ON_ONCE(1);
		pr_info("ftrace faulted on writing ");
		print_ip_sym(ip);
		break;
	default:
		FTRACE_WARN_ON_ONCE(1);
		pr_info("ftrace faulted on unknown error ");
		print_ip_sym(ip);
	}
2050
	print_bug_type();
2051 2052 2053 2054 2055 2056 2057 2058
	if (rec) {
		struct ftrace_ops *ops = NULL;

		pr_info("ftrace record flags: %lx\n", rec->flags);
		pr_cont(" (%ld)%s", ftrace_rec_count(rec),
			rec->flags & FTRACE_FL_REGS ? " R" : "  ");
		if (rec->flags & FTRACE_FL_TRAMP_EN) {
			ops = ftrace_find_tramp_ops_any(rec);
2059 2060 2061 2062 2063 2064 2065 2066
			if (ops) {
				do {
					pr_cont("\ttramp: %pS (%pS)",
						(void *)ops->trampoline,
						(void *)ops->func);
					ops = ftrace_find_tramp_ops_next(rec, ops);
				} while (ops);
			} else
2067 2068 2069 2070
				pr_cont("\ttramp: ERROR!");

		}
		ip = ftrace_get_addr_curr(rec);
2071
		pr_cont("\n expected tramp: %lx\n", ip);
2072
	}
2073 2074
}

2075
static int ftrace_check_record(struct dyn_ftrace *rec, int enable, int update)
2076
{
2077
	unsigned long flag = 0UL;
2078

2079 2080
	ftrace_bug_type = FTRACE_BUG_UNKNOWN;

2081 2082 2083
	if (rec->flags & FTRACE_FL_DISABLED)
		return FTRACE_UPDATE_IGNORE;

S
Steven Rostedt 已提交
2084
	/*
2085
	 * If we are updating calls:
S
Steven Rostedt 已提交
2086
	 *
2087 2088
	 *   If the record has a ref count, then we need to enable it
	 *   because someone is using it.
S
Steven Rostedt 已提交
2089
	 *
2090 2091
	 *   Otherwise we make sure its disabled.
	 *
2092
	 * If we are disabling calls, then disable all records that
2093
	 * are enabled.
S
Steven Rostedt 已提交
2094
	 */
2095
	if (enable && ftrace_rec_count(rec))
2096
		flag = FTRACE_FL_ENABLED;
S
Steven Rostedt 已提交
2097

2098
	/*
2099 2100 2101
	 * If enabling and the REGS flag does not match the REGS_EN, or
	 * the TRAMP flag doesn't match the TRAMP_EN, then do not ignore
	 * this record. Set flags to fail the compare against ENABLED.
2102
	 */
2103 2104 2105 2106 2107 2108 2109 2110 2111
	if (flag) {
		if (!(rec->flags & FTRACE_FL_REGS) != 
		    !(rec->flags & FTRACE_FL_REGS_EN))
			flag |= FTRACE_FL_REGS;

		if (!(rec->flags & FTRACE_FL_TRAMP) != 
		    !(rec->flags & FTRACE_FL_TRAMP_EN))
			flag |= FTRACE_FL_TRAMP;
	}
2112

2113 2114
	/* If the state of this record hasn't changed, then do nothing */
	if ((rec->flags & FTRACE_FL_ENABLED) == flag)
2115
		return FTRACE_UPDATE_IGNORE;
S
Steven Rostedt 已提交
2116

2117
	if (flag) {
2118 2119 2120 2121
		/* Save off if rec is being enabled (for return value) */
		flag ^= rec->flags & FTRACE_FL_ENABLED;

		if (update) {
2122
			rec->flags |= FTRACE_FL_ENABLED;
2123 2124 2125 2126 2127 2128
			if (flag & FTRACE_FL_REGS) {
				if (rec->flags & FTRACE_FL_REGS)
					rec->flags |= FTRACE_FL_REGS_EN;
				else
					rec->flags &= ~FTRACE_FL_REGS_EN;
			}
2129 2130 2131 2132 2133 2134
			if (flag & FTRACE_FL_TRAMP) {
				if (rec->flags & FTRACE_FL_TRAMP)
					rec->flags |= FTRACE_FL_TRAMP_EN;
				else
					rec->flags &= ~FTRACE_FL_TRAMP_EN;
			}
2135 2136 2137 2138 2139 2140 2141
		}

		/*
		 * If this record is being updated from a nop, then
		 *   return UPDATE_MAKE_CALL.
		 * Otherwise,
		 *   return UPDATE_MODIFY_CALL to tell the caller to convert
2142
		 *   from the save regs, to a non-save regs function or
2143
		 *   vice versa, or from a trampoline call.
2144
		 */
2145 2146
		if (flag & FTRACE_FL_ENABLED) {
			ftrace_bug_type = FTRACE_BUG_CALL;
2147
			return FTRACE_UPDATE_MAKE_CALL;
2148
		}
2149

2150
		ftrace_bug_type = FTRACE_BUG_UPDATE;
2151
		return FTRACE_UPDATE_MODIFY_CALL;
2152 2153
	}

2154 2155
	if (update) {
		/* If there's no more users, clear all flags */
2156
		if (!ftrace_rec_count(rec))
2157 2158
			rec->flags = 0;
		else
2159 2160 2161 2162 2163 2164
			/*
			 * Just disable the record, but keep the ops TRAMP
			 * and REGS states. The _EN flags must be disabled though.
			 */
			rec->flags &= ~(FTRACE_FL_ENABLED | FTRACE_FL_TRAMP_EN |
					FTRACE_FL_REGS_EN);
2165
	}
2166

2167
	ftrace_bug_type = FTRACE_BUG_NOP;
2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197
	return FTRACE_UPDATE_MAKE_NOP;
}

/**
 * ftrace_update_record, set a record that now is tracing or not
 * @rec: the record to update
 * @enable: set to 1 if the record is tracing, zero to force disable
 *
 * The records that represent all functions that can be traced need
 * to be updated when tracing has been enabled.
 */
int ftrace_update_record(struct dyn_ftrace *rec, int enable)
{
	return ftrace_check_record(rec, enable, 1);
}

/**
 * ftrace_test_record, check if the record has been enabled or not
 * @rec: the record to test
 * @enable: set to 1 to check if enabled, 0 if it is disabled
 *
 * The arch code may need to test if a record is already set to
 * tracing to determine how to modify the function code that it
 * represents.
 */
int ftrace_test_record(struct dyn_ftrace *rec, int enable)
{
	return ftrace_check_record(rec, enable, 0);
}

2198 2199 2200 2201
static struct ftrace_ops *
ftrace_find_tramp_ops_any(struct dyn_ftrace *rec)
{
	struct ftrace_ops *op;
2202
	unsigned long ip = rec->ip;
2203 2204 2205 2206 2207 2208

	do_for_each_ftrace_op(op, ftrace_ops_list) {

		if (!op->trampoline)
			continue;

2209
		if (hash_contains_ip(ip, op->func_hash))
2210 2211 2212 2213 2214 2215
			return op;
	} while_for_each_ftrace_op(op);

	return NULL;
}

2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233
static struct ftrace_ops *
ftrace_find_tramp_ops_next(struct dyn_ftrace *rec,
			   struct ftrace_ops *op)
{
	unsigned long ip = rec->ip;

	while_for_each_ftrace_op(op) {

		if (!op->trampoline)
			continue;

		if (hash_contains_ip(ip, op->func_hash))
			return op;
	} 

	return NULL;
}

2234 2235 2236 2237
static struct ftrace_ops *
ftrace_find_tramp_ops_curr(struct dyn_ftrace *rec)
{
	struct ftrace_ops *op;
2238
	unsigned long ip = rec->ip;
2239

2240 2241 2242 2243 2244 2245 2246 2247
	/*
	 * Need to check removed ops first.
	 * If they are being removed, and this rec has a tramp,
	 * and this rec is in the ops list, then it would be the
	 * one with the tramp.
	 */
	if (removed_ops) {
		if (hash_contains_ip(ip, &removed_ops->old_hash))
2248 2249 2250
			return removed_ops;
	}

2251 2252 2253 2254 2255 2256 2257
	/*
	 * Need to find the current trampoline for a rec.
	 * Now, a trampoline is only attached to a rec if there
	 * was a single 'ops' attached to it. But this can be called
	 * when we are adding another op to the rec or removing the
	 * current one. Thus, if the op is being added, we can
	 * ignore it because it hasn't attached itself to the rec
2258 2259 2260 2261 2262 2263 2264 2265 2266 2267
	 * yet.
	 *
	 * If an ops is being modified (hooking to different functions)
	 * then we don't care about the new functions that are being
	 * added, just the old ones (that are probably being removed).
	 *
	 * If we are adding an ops to a function that already is using
	 * a trampoline, it needs to be removed (trampolines are only
	 * for single ops connected), then an ops that is not being
	 * modified also needs to be checked.
2268
	 */
2269
	do_for_each_ftrace_op(op, ftrace_ops_list) {
2270 2271 2272 2273 2274 2275 2276 2277 2278

		if (!op->trampoline)
			continue;

		/*
		 * If the ops is being added, it hasn't gotten to
		 * the point to be removed from this tree yet.
		 */
		if (op->flags & FTRACE_OPS_FL_ADDING)
2279 2280
			continue;

2281

2282
		/*
2283 2284 2285
		 * If the ops is being modified and is in the old
		 * hash, then it is probably being removed from this
		 * function.
2286 2287 2288
		 */
		if ((op->flags & FTRACE_OPS_FL_MODIFYING) &&
		    hash_contains_ip(ip, &op->old_hash))
2289
			return op;
2290 2291 2292 2293 2294 2295 2296 2297
		/*
		 * If the ops is not being added or modified, and it's
		 * in its normal filter hash, then this must be the one
		 * we want!
		 */
		if (!(op->flags & FTRACE_OPS_FL_MODIFYING) &&
		    hash_contains_ip(ip, op->func_hash))
			return op;
2298 2299 2300 2301 2302 2303 2304 2305 2306 2307

	} while_for_each_ftrace_op(op);

	return NULL;
}

static struct ftrace_ops *
ftrace_find_tramp_ops_new(struct dyn_ftrace *rec)
{
	struct ftrace_ops *op;
2308
	unsigned long ip = rec->ip;
2309 2310 2311

	do_for_each_ftrace_op(op, ftrace_ops_list) {
		/* pass rec in as regs to have non-NULL val */
2312
		if (hash_contains_ip(ip, op->func_hash))
2313 2314 2315 2316 2317 2318
			return op;
	} while_for_each_ftrace_op(op);

	return NULL;
}

2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330
/**
 * ftrace_get_addr_new - Get the call address to set to
 * @rec:  The ftrace record descriptor
 *
 * If the record has the FTRACE_FL_REGS set, that means that it
 * wants to convert to a callback that saves all regs. If FTRACE_FL_REGS
 * is not not set, then it wants to convert to the normal callback.
 *
 * Returns the address of the trampoline to set to
 */
unsigned long ftrace_get_addr_new(struct dyn_ftrace *rec)
{
2331 2332 2333 2334 2335 2336
	struct ftrace_ops *ops;

	/* Trampolines take precedence over regs */
	if (rec->flags & FTRACE_FL_TRAMP) {
		ops = ftrace_find_tramp_ops_new(rec);
		if (FTRACE_WARN_ON(!ops || !ops->trampoline)) {
2337 2338
			pr_warn("Bad trampoline accounting at: %p (%pS) (%lx)\n",
				(void *)rec->ip, (void *)rec->ip, rec->flags);
2339 2340 2341 2342 2343 2344
			/* Ftrace is shutting down, return anything */
			return (unsigned long)FTRACE_ADDR;
		}
		return ops->trampoline;
	}

2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362
	if (rec->flags & FTRACE_FL_REGS)
		return (unsigned long)FTRACE_REGS_ADDR;
	else
		return (unsigned long)FTRACE_ADDR;
}

/**
 * ftrace_get_addr_curr - Get the call address that is already there
 * @rec:  The ftrace record descriptor
 *
 * The FTRACE_FL_REGS_EN is set when the record already points to
 * a function that saves all the regs. Basically the '_EN' version
 * represents the current state of the function.
 *
 * Returns the address of the trampoline that is currently being called
 */
unsigned long ftrace_get_addr_curr(struct dyn_ftrace *rec)
{
2363 2364 2365 2366 2367 2368
	struct ftrace_ops *ops;

	/* Trampolines take precedence over regs */
	if (rec->flags & FTRACE_FL_TRAMP_EN) {
		ops = ftrace_find_tramp_ops_curr(rec);
		if (FTRACE_WARN_ON(!ops)) {
2369 2370
			pr_warn("Bad trampoline accounting at: %p (%pS)\n",
				(void *)rec->ip, (void *)rec->ip);
2371 2372 2373 2374 2375 2376
			/* Ftrace is shutting down, return anything */
			return (unsigned long)FTRACE_ADDR;
		}
		return ops->trampoline;
	}

2377 2378 2379 2380 2381 2382
	if (rec->flags & FTRACE_FL_REGS_EN)
		return (unsigned long)FTRACE_REGS_ADDR;
	else
		return (unsigned long)FTRACE_ADDR;
}

2383 2384 2385
static int
__ftrace_replace_code(struct dyn_ftrace *rec, int enable)
{
2386
	unsigned long ftrace_old_addr;
2387 2388 2389
	unsigned long ftrace_addr;
	int ret;

2390
	ftrace_addr = ftrace_get_addr_new(rec);
2391

2392 2393 2394 2395
	/* This needs to be done before we call ftrace_update_record */
	ftrace_old_addr = ftrace_get_addr_curr(rec);

	ret = ftrace_update_record(rec, enable);
2396

2397 2398
	ftrace_bug_type = FTRACE_BUG_UNKNOWN;

2399 2400 2401 2402 2403
	switch (ret) {
	case FTRACE_UPDATE_IGNORE:
		return 0;

	case FTRACE_UPDATE_MAKE_CALL:
2404
		ftrace_bug_type = FTRACE_BUG_CALL;
2405
		return ftrace_make_call(rec, ftrace_addr);
2406 2407

	case FTRACE_UPDATE_MAKE_NOP:
2408
		ftrace_bug_type = FTRACE_BUG_NOP;
2409
		return ftrace_make_nop(NULL, rec, ftrace_old_addr);
2410 2411

	case FTRACE_UPDATE_MODIFY_CALL:
2412
		ftrace_bug_type = FTRACE_BUG_UPDATE;
2413
		return ftrace_modify_call(rec, ftrace_old_addr, ftrace_addr);
2414 2415
	}

2416
	return -1; /* unknow ftrace bug */
2417 2418
}

2419
void __weak ftrace_replace_code(int enable)
2420 2421 2422
{
	struct dyn_ftrace *rec;
	struct ftrace_page *pg;
S
Steven Rostedt 已提交
2423
	int failed;
2424

2425 2426 2427
	if (unlikely(ftrace_disabled))
		return;

2428
	do_for_each_ftrace_rec(pg, rec) {
2429 2430 2431 2432

		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

2433
		failed = __ftrace_replace_code(rec, enable);
2434
		if (failed) {
2435
			ftrace_bug(failed, rec);
2436 2437
			/* Stop processing */
			return;
2438
		}
2439
	} while_for_each_ftrace_rec();
2440 2441
}

2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513
struct ftrace_rec_iter {
	struct ftrace_page	*pg;
	int			index;
};

/**
 * ftrace_rec_iter_start, start up iterating over traced functions
 *
 * Returns an iterator handle that is used to iterate over all
 * the records that represent address locations where functions
 * are traced.
 *
 * May return NULL if no records are available.
 */
struct ftrace_rec_iter *ftrace_rec_iter_start(void)
{
	/*
	 * We only use a single iterator.
	 * Protected by the ftrace_lock mutex.
	 */
	static struct ftrace_rec_iter ftrace_rec_iter;
	struct ftrace_rec_iter *iter = &ftrace_rec_iter;

	iter->pg = ftrace_pages_start;
	iter->index = 0;

	/* Could have empty pages */
	while (iter->pg && !iter->pg->index)
		iter->pg = iter->pg->next;

	if (!iter->pg)
		return NULL;

	return iter;
}

/**
 * ftrace_rec_iter_next, get the next record to process.
 * @iter: The handle to the iterator.
 *
 * Returns the next iterator after the given iterator @iter.
 */
struct ftrace_rec_iter *ftrace_rec_iter_next(struct ftrace_rec_iter *iter)
{
	iter->index++;

	if (iter->index >= iter->pg->index) {
		iter->pg = iter->pg->next;
		iter->index = 0;

		/* Could have empty pages */
		while (iter->pg && !iter->pg->index)
			iter->pg = iter->pg->next;
	}

	if (!iter->pg)
		return NULL;

	return iter;
}

/**
 * ftrace_rec_iter_record, get the record at the iterator location
 * @iter: The current iterator location
 *
 * Returns the record that the current @iter is at.
 */
struct dyn_ftrace *ftrace_rec_iter_record(struct ftrace_rec_iter *iter)
{
	return &iter->pg->records[iter->index];
}

2514
static int
2515
ftrace_code_disable(struct module *mod, struct dyn_ftrace *rec)
2516
{
2517
	int ret;
2518

2519 2520 2521
	if (unlikely(ftrace_disabled))
		return 0;

2522
	ret = ftrace_make_nop(mod, rec, MCOUNT_ADDR);
2523
	if (ret) {
2524
		ftrace_bug_type = FTRACE_BUG_INIT;
2525
		ftrace_bug(ret, rec);
2526
		return 0;
2527
	}
2528
	return 1;
2529 2530
}

2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548
/*
 * archs can override this function if they must do something
 * before the modifying code is performed.
 */
int __weak ftrace_arch_code_modify_prepare(void)
{
	return 0;
}

/*
 * archs can override this function if they must do something
 * after the modifying code is performed.
 */
int __weak ftrace_arch_code_modify_post_process(void)
{
	return 0;
}

2549
void ftrace_modify_all_code(int command)
2550
{
2551
	int update = command & FTRACE_UPDATE_TRACE_FUNC;
2552
	int err = 0;
2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563

	/*
	 * If the ftrace_caller calls a ftrace_ops func directly,
	 * we need to make sure that it only traces functions it
	 * expects to trace. When doing the switch of functions,
	 * we need to update to the ftrace_ops_list_func first
	 * before the transition between old and new calls are set,
	 * as the ftrace_ops_list_func will check the ops hashes
	 * to make sure the ops are having the right functions
	 * traced.
	 */
2564 2565 2566 2567 2568
	if (update) {
		err = ftrace_update_ftrace_func(ftrace_ops_list_func);
		if (FTRACE_WARN_ON(err))
			return;
	}
2569

2570
	if (command & FTRACE_UPDATE_CALLS)
2571
		ftrace_replace_code(1);
2572
	else if (command & FTRACE_DISABLE_CALLS)
2573 2574
		ftrace_replace_code(0);

2575 2576 2577 2578 2579 2580
	if (update && ftrace_trace_function != ftrace_ops_list_func) {
		function_trace_op = set_function_trace_op;
		smp_wmb();
		/* If irqs are disabled, we are in stop machine */
		if (!irqs_disabled())
			smp_call_function(ftrace_sync_ipi, NULL, 1);
2581 2582 2583
		err = ftrace_update_ftrace_func(ftrace_trace_function);
		if (FTRACE_WARN_ON(err))
			return;
2584
	}
2585

2586
	if (command & FTRACE_START_FUNC_RET)
2587
		err = ftrace_enable_ftrace_graph_caller();
2588
	else if (command & FTRACE_STOP_FUNC_RET)
2589 2590
		err = ftrace_disable_ftrace_graph_caller();
	FTRACE_WARN_ON(err);
2591 2592 2593 2594 2595 2596 2597
}

static int __ftrace_modify_code(void *data)
{
	int *command = data;

	ftrace_modify_all_code(*command);
2598

2599
	return 0;
2600 2601
}

2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
/**
 * ftrace_run_stop_machine, go back to the stop machine method
 * @command: The command to tell ftrace what to do
 *
 * If an arch needs to fall back to the stop machine method, the
 * it can call this function.
 */
void ftrace_run_stop_machine(int command)
{
	stop_machine(__ftrace_modify_code, &command, NULL);
}

/**
 * arch_ftrace_update_code, modify the code to trace or not trace
 * @command: The command that needs to be done
 *
 * Archs can override this function if it does not need to
 * run stop_machine() to modify code.
 */
void __weak arch_ftrace_update_code(int command)
{
	ftrace_run_stop_machine(command);
}

I
Ingo Molnar 已提交
2626
static void ftrace_run_update_code(int command)
2627
{
2628 2629 2630 2631 2632 2633 2634
	int ret;

	ret = ftrace_arch_code_modify_prepare();
	FTRACE_WARN_ON(ret);
	if (ret)
		return;

2635 2636 2637 2638 2639 2640 2641 2642
	/*
	 * By default we use stop_machine() to modify the code.
	 * But archs can do what ever they want as long as it
	 * is safe. The stop_machine() is the safest, but also
	 * produces the most overhead.
	 */
	arch_ftrace_update_code(command);

2643 2644
	ret = ftrace_arch_code_modify_post_process();
	FTRACE_WARN_ON(ret);
2645 2646
}

2647
static void ftrace_run_modify_code(struct ftrace_ops *ops, int command,
2648
				   struct ftrace_ops_hash *old_hash)
2649 2650
{
	ops->flags |= FTRACE_OPS_FL_MODIFYING;
2651 2652
	ops->old_hash.filter_hash = old_hash->filter_hash;
	ops->old_hash.notrace_hash = old_hash->notrace_hash;
2653
	ftrace_run_update_code(command);
2654
	ops->old_hash.filter_hash = NULL;
2655
	ops->old_hash.notrace_hash = NULL;
2656 2657 2658
	ops->flags &= ~FTRACE_OPS_FL_MODIFYING;
}

2659
static ftrace_func_t saved_ftrace_func;
2660
static int ftrace_start_up;
2661

2662 2663 2664 2665
void __weak arch_ftrace_trampoline_free(struct ftrace_ops *ops)
{
}

2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677
static void ftrace_startup_enable(int command)
{
	if (saved_ftrace_func != ftrace_trace_function) {
		saved_ftrace_func = ftrace_trace_function;
		command |= FTRACE_UPDATE_TRACE_FUNC;
	}

	if (!command || !ftrace_enabled)
		return;

	ftrace_run_update_code(command);
}
2678

2679 2680 2681 2682 2683 2684 2685
static void ftrace_startup_all(int command)
{
	update_all_ops = true;
	ftrace_startup_enable(command);
	update_all_ops = false;
}

2686
static int ftrace_startup(struct ftrace_ops *ops, int command)
2687
{
2688
	int ret;
2689

2690
	if (unlikely(ftrace_disabled))
2691
		return -ENODEV;
2692

2693 2694 2695 2696
	ret = __register_ftrace_function(ops);
	if (ret)
		return ret;

2697
	ftrace_start_up++;
2698

2699 2700 2701 2702 2703 2704 2705 2706 2707
	/*
	 * Note that ftrace probes uses this to start up
	 * and modify functions it will probe. But we still
	 * set the ADDING flag for modification, as probes
	 * do not have trampolines. If they add them in the
	 * future, then the probes will need to distinguish
	 * between adding and updating probes.
	 */
	ops->flags |= FTRACE_OPS_FL_ENABLED | FTRACE_OPS_FL_ADDING;
2708

2709 2710 2711 2712 2713 2714 2715 2716 2717
	ret = ftrace_hash_ipmodify_enable(ops);
	if (ret < 0) {
		/* Rollback registration process */
		__unregister_ftrace_function(ops);
		ftrace_start_up--;
		ops->flags &= ~FTRACE_OPS_FL_ENABLED;
		return ret;
	}

2718 2719
	if (ftrace_hash_rec_enable(ops, 1))
		command |= FTRACE_UPDATE_CALLS;
2720

2721
	ftrace_startup_enable(command);
2722

2723 2724
	ops->flags &= ~FTRACE_OPS_FL_ADDING;

2725
	return 0;
2726 2727
}

2728
static int ftrace_shutdown(struct ftrace_ops *ops, int command)
2729
{
2730
	int ret;
2731

2732
	if (unlikely(ftrace_disabled))
2733 2734 2735 2736 2737
		return -ENODEV;

	ret = __unregister_ftrace_function(ops);
	if (ret)
		return ret;
2738

2739
	ftrace_start_up--;
2740 2741 2742 2743 2744 2745 2746
	/*
	 * Just warn in case of unbalance, no need to kill ftrace, it's not
	 * critical but the ftrace_call callers may be never nopped again after
	 * further ftrace uses.
	 */
	WARN_ON_ONCE(ftrace_start_up < 0);

2747 2748
	/* Disabling ipmodify never fails */
	ftrace_hash_ipmodify_disable(ops);
2749

2750 2751
	if (ftrace_hash_rec_disable(ops, 1))
		command |= FTRACE_UPDATE_CALLS;
2752

2753
	ops->flags &= ~FTRACE_OPS_FL_ENABLED;
2754

2755 2756 2757 2758
	if (saved_ftrace_func != ftrace_trace_function) {
		saved_ftrace_func = ftrace_trace_function;
		command |= FTRACE_UPDATE_TRACE_FUNC;
	}
2759

2760 2761
	if (!command || !ftrace_enabled) {
		/*
2762 2763
		 * If these are dynamic or per_cpu ops, they still
		 * need their data freed. Since, function tracing is
2764 2765 2766
		 * not currently active, we can just free them
		 * without synchronizing all CPUs.
		 */
2767
		if (ops->flags & FTRACE_OPS_FL_DYNAMIC)
2768 2769
			goto free_ops;

2770
		return 0;
2771
	}
2772

2773 2774 2775 2776
	/*
	 * If the ops uses a trampoline, then it needs to be
	 * tested first on update.
	 */
2777
	ops->flags |= FTRACE_OPS_FL_REMOVING;
2778 2779
	removed_ops = ops;

2780 2781 2782 2783
	/* The trampoline logic checks the old hashes */
	ops->old_hash.filter_hash = ops->func_hash->filter_hash;
	ops->old_hash.notrace_hash = ops->func_hash->notrace_hash;

2784
	ftrace_run_update_code(command);
2785

2786 2787 2788 2789
	/*
	 * If there's no more ops registered with ftrace, run a
	 * sanity check to make sure all rec flags are cleared.
	 */
2790 2791
	if (rcu_dereference_protected(ftrace_ops_list,
			lockdep_is_held(&ftrace_lock)) == &ftrace_list_end) {
2792 2793 2794 2795
		struct ftrace_page *pg;
		struct dyn_ftrace *rec;

		do_for_each_ftrace_rec(pg, rec) {
2796
			if (FTRACE_WARN_ON_ONCE(rec->flags & ~FTRACE_FL_DISABLED))
2797 2798 2799 2800 2801
				pr_warn("  %pS flags:%lx\n",
					(void *)rec->ip, rec->flags);
		} while_for_each_ftrace_rec();
	}

2802 2803 2804 2805
	ops->old_hash.filter_hash = NULL;
	ops->old_hash.notrace_hash = NULL;

	removed_ops = NULL;
2806
	ops->flags &= ~FTRACE_OPS_FL_REMOVING;
2807

2808 2809 2810
	/*
	 * Dynamic ops may be freed, we must make sure that all
	 * callers are done before leaving this function.
2811
	 * The same goes for freeing the per_cpu data of the per_cpu
2812 2813
	 * ops.
	 */
2814
	if (ops->flags & FTRACE_OPS_FL_DYNAMIC) {
2815 2816 2817 2818 2819 2820 2821 2822
		/*
		 * We need to do a hard force of sched synchronization.
		 * This is because we use preempt_disable() to do RCU, but
		 * the function tracers can be called where RCU is not watching
		 * (like before user_exit()). We can not rely on the RCU
		 * infrastructure to do the synchronization, thus we must do it
		 * ourselves.
		 */
2823 2824
		schedule_on_each_cpu(ftrace_sync);

2825 2826 2827 2828 2829 2830 2831 2832 2833 2834
		/*
		 * When the kernel is preeptive, tasks can be preempted
		 * while on a ftrace trampoline. Just scheduling a task on
		 * a CPU is not good enough to flush them. Calling
		 * synchornize_rcu_tasks() will wait for those tasks to
		 * execute and either schedule voluntarily or enter user space.
		 */
		if (IS_ENABLED(CONFIG_PREEMPT))
			synchronize_rcu_tasks();

2835
 free_ops:
2836
		arch_ftrace_trampoline_free(ops);
2837 2838
	}

2839
	return 0;
2840 2841
}

I
Ingo Molnar 已提交
2842
static void ftrace_startup_sysctl(void)
2843
{
2844 2845
	int command;

2846 2847 2848
	if (unlikely(ftrace_disabled))
		return;

2849 2850
	/* Force update next time */
	saved_ftrace_func = NULL;
2851
	/* ftrace_start_up is true if we want ftrace running */
2852 2853 2854 2855
	if (ftrace_start_up) {
		command = FTRACE_UPDATE_CALLS;
		if (ftrace_graph_active)
			command |= FTRACE_START_FUNC_RET;
2856
		ftrace_startup_enable(command);
2857
	}
2858 2859
}

I
Ingo Molnar 已提交
2860
static void ftrace_shutdown_sysctl(void)
2861
{
2862 2863
	int command;

2864 2865 2866
	if (unlikely(ftrace_disabled))
		return;

2867
	/* ftrace_start_up is true if ftrace is running */
2868 2869 2870 2871 2872 2873
	if (ftrace_start_up) {
		command = FTRACE_DISABLE_CALLS;
		if (ftrace_graph_active)
			command |= FTRACE_STOP_FUNC_RET;
		ftrace_run_update_code(command);
	}
2874 2875
}

2876
static u64		ftrace_update_time;
2877 2878
unsigned long		ftrace_update_tot_cnt;

2879
static inline int ops_traces_mod(struct ftrace_ops *ops)
2880
{
2881 2882 2883 2884
	/*
	 * Filter_hash being empty will default to trace module.
	 * But notrace hash requires a test of individual module functions.
	 */
2885 2886
	return ftrace_hash_empty(ops->func_hash->filter_hash) &&
		ftrace_hash_empty(ops->func_hash->notrace_hash);
2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900
}

/*
 * Check if the current ops references the record.
 *
 * If the ops traces all functions, then it was already accounted for.
 * If the ops does not trace the current record function, skip it.
 * If the ops ignores the function via notrace filter, skip it.
 */
static inline bool
ops_references_rec(struct ftrace_ops *ops, struct dyn_ftrace *rec)
{
	/* If ops isn't enabled, ignore it */
	if (!(ops->flags & FTRACE_OPS_FL_ENABLED))
2901
		return false;
2902

2903
	/* If ops traces all then it includes this function */
2904
	if (ops_traces_mod(ops))
2905
		return true;
2906 2907

	/* The function must be in the filter */
2908
	if (!ftrace_hash_empty(ops->func_hash->filter_hash) &&
2909
	    !__ftrace_lookup_ip(ops->func_hash->filter_hash, rec->ip))
2910
		return false;
2911

2912
	/* If in notrace hash, we ignore it too */
2913
	if (ftrace_lookup_ip(ops->func_hash->notrace_hash, rec->ip))
2914
		return false;
2915

2916
	return true;
2917 2918
}

2919
static int ftrace_update_code(struct module *mod, struct ftrace_page *new_pgs)
2920
{
2921
	struct ftrace_page *pg;
2922
	struct dyn_ftrace *p;
2923
	u64 start, stop;
2924
	unsigned long update_cnt = 0;
2925
	unsigned long rec_flags = 0;
2926
	int i;
2927

2928 2929
	start = ftrace_now(raw_smp_processor_id());

2930
	/*
2931 2932 2933 2934 2935 2936 2937 2938 2939
	 * When a module is loaded, this function is called to convert
	 * the calls to mcount in its text to nops, and also to create
	 * an entry in the ftrace data. Now, if ftrace is activated
	 * after this call, but before the module sets its text to
	 * read-only, the modification of enabling ftrace can fail if
	 * the read-only is done while ftrace is converting the calls.
	 * To prevent this, the module's records are set as disabled
	 * and will be enabled after the call to set the module's text
	 * to read-only.
2940
	 */
2941 2942
	if (mod)
		rec_flags |= FTRACE_FL_DISABLED;
2943

2944
	for (pg = new_pgs; pg; pg = pg->next) {
2945

2946
		for (i = 0; i < pg->index; i++) {
2947

2948 2949 2950
			/* If something went wrong, bail without enabling anything */
			if (unlikely(ftrace_disabled))
				return -1;
2951

2952
			p = &pg->records[i];
2953
			p->flags = rec_flags;
2954

2955 2956 2957 2958 2959 2960
			/*
			 * Do the initial record conversion from mcount jump
			 * to the NOP instructions.
			 */
			if (!ftrace_code_disable(mod, p))
				break;
2961

2962
			update_cnt++;
2963
		}
2964 2965
	}

I
Ingo Molnar 已提交
2966
	stop = ftrace_now(raw_smp_processor_id());
2967
	ftrace_update_time = stop - start;
2968
	ftrace_update_tot_cnt += update_cnt;
2969

2970 2971 2972
	return 0;
}

2973
static int ftrace_allocate_records(struct ftrace_page *pg, int count)
2974
{
2975
	int order;
2976 2977
	int cnt;

2978 2979 2980 2981
	if (WARN_ON(!count))
		return -EINVAL;

	order = get_count_order(DIV_ROUND_UP(count, ENTRIES_PER_PAGE));
2982 2983

	/*
2984 2985
	 * We want to fill as much as possible. No more than a page
	 * may be empty.
2986
	 */
2987 2988
	while ((PAGE_SIZE << order) / ENTRY_SIZE >= count + ENTRIES_PER_PAGE)
		order--;
2989

2990 2991
 again:
	pg->records = (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, order);
2992

2993 2994 2995 2996 2997 2998 2999
	if (!pg->records) {
		/* if we can't allocate this size, try something smaller */
		if (!order)
			return -ENOMEM;
		order >>= 1;
		goto again;
	}
3000

3001 3002
	cnt = (PAGE_SIZE << order) / ENTRY_SIZE;
	pg->size = cnt;
3003

3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036
	if (cnt > count)
		cnt = count;

	return cnt;
}

static struct ftrace_page *
ftrace_allocate_pages(unsigned long num_to_init)
{
	struct ftrace_page *start_pg;
	struct ftrace_page *pg;
	int order;
	int cnt;

	if (!num_to_init)
		return 0;

	start_pg = pg = kzalloc(sizeof(*pg), GFP_KERNEL);
	if (!pg)
		return NULL;

	/*
	 * Try to allocate as much as possible in one continues
	 * location that fills in all of the space. We want to
	 * waste as little space as possible.
	 */
	for (;;) {
		cnt = ftrace_allocate_records(pg, num_to_init);
		if (cnt < 0)
			goto free_pages;

		num_to_init -= cnt;
		if (!num_to_init)
3037 3038
			break;

3039 3040 3041 3042
		pg->next = kzalloc(sizeof(*pg), GFP_KERNEL);
		if (!pg->next)
			goto free_pages;

3043 3044 3045
		pg = pg->next;
	}

3046 3047 3048
	return start_pg;

 free_pages:
3049 3050
	pg = start_pg;
	while (pg) {
3051 3052 3053 3054 3055 3056 3057 3058 3059 3060
		order = get_count_order(pg->size / ENTRIES_PER_PAGE);
		free_pages((unsigned long)pg->records, order);
		start_pg = pg->next;
		kfree(pg);
		pg = start_pg;
	}
	pr_info("ftrace: FAILED to allocate memory for functions\n");
	return NULL;
}

3061 3062 3063
#define FTRACE_BUFF_MAX (KSYM_SYMBOL_LEN+4) /* room for wildcards */

struct ftrace_iterator {
3064
	loff_t				pos;
3065
	loff_t				func_pos;
3066
	loff_t				mod_pos;
3067 3068 3069
	struct ftrace_page		*pg;
	struct dyn_ftrace		*func;
	struct ftrace_func_probe	*probe;
3070
	struct ftrace_func_entry	*probe_entry;
3071
	struct trace_parser		parser;
3072
	struct ftrace_hash		*hash;
3073
	struct ftrace_ops		*ops;
3074 3075
	struct trace_array		*tr;
	struct list_head		*mod_list;
3076
	int				pidx;
3077 3078
	int				idx;
	unsigned			flags;
3079 3080
};

3081
static void *
3082
t_probe_next(struct seq_file *m, loff_t *pos)
3083 3084
{
	struct ftrace_iterator *iter = m->private;
3085
	struct trace_array *tr = iter->ops->private;
3086
	struct list_head *func_probes;
3087 3088
	struct ftrace_hash *hash;
	struct list_head *next;
3089
	struct hlist_node *hnd = NULL;
3090
	struct hlist_head *hhd;
3091
	int size;
3092 3093

	(*pos)++;
3094
	iter->pos = *pos;
3095

3096
	if (!tr)
3097 3098
		return NULL;

3099 3100
	func_probes = &tr->func_probes;
	if (list_empty(func_probes))
3101 3102
		return NULL;

3103
	if (!iter->probe) {
3104
		next = func_probes->next;
3105
		iter->probe = list_entry(next, struct ftrace_func_probe, list);
3106 3107 3108 3109 3110 3111 3112 3113 3114 3115
	}

	if (iter->probe_entry)
		hnd = &iter->probe_entry->hlist;

	hash = iter->probe->ops.func_hash->filter_hash;
	size = 1 << hash->size_bits;

 retry:
	if (iter->pidx >= size) {
3116
		if (iter->probe->list.next == func_probes)
3117 3118
			return NULL;
		next = iter->probe->list.next;
3119
		iter->probe = list_entry(next, struct ftrace_func_probe, list);
3120 3121 3122 3123 3124 3125
		hash = iter->probe->ops.func_hash->filter_hash;
		size = 1 << hash->size_bits;
		iter->pidx = 0;
	}

	hhd = &hash->buckets[iter->pidx];
3126 3127

	if (hlist_empty(hhd)) {
3128
		iter->pidx++;
3129 3130 3131 3132 3133 3134 3135 3136 3137
		hnd = NULL;
		goto retry;
	}

	if (!hnd)
		hnd = hhd->first;
	else {
		hnd = hnd->next;
		if (!hnd) {
3138
			iter->pidx++;
3139 3140 3141 3142
			goto retry;
		}
	}

3143 3144 3145
	if (WARN_ON_ONCE(!hnd))
		return NULL;

3146
	iter->probe_entry = hlist_entry(hnd, struct ftrace_func_entry, hlist);
3147 3148

	return iter;
3149 3150
}

3151
static void *t_probe_start(struct seq_file *m, loff_t *pos)
3152 3153 3154
{
	struct ftrace_iterator *iter = m->private;
	void *p = NULL;
L
Li Zefan 已提交
3155 3156
	loff_t l;

3157
	if (!(iter->flags & FTRACE_ITER_DO_PROBES))
3158 3159
		return NULL;

3160
	if (iter->mod_pos > *pos)
3161
		return NULL;
3162

3163 3164 3165
	iter->probe = NULL;
	iter->probe_entry = NULL;
	iter->pidx = 0;
3166
	for (l = 0; l <= (*pos - iter->mod_pos); ) {
3167
		p = t_probe_next(m, &l);
L
Li Zefan 已提交
3168 3169 3170
		if (!p)
			break;
	}
3171 3172 3173
	if (!p)
		return NULL;

3174
	/* Only set this if we have an item */
3175
	iter->flags |= FTRACE_ITER_PROBE;
3176

3177
	return iter;
3178 3179
}

3180
static int
3181
t_probe_show(struct seq_file *m, struct ftrace_iterator *iter)
3182
{
3183
	struct ftrace_func_entry *probe_entry;
3184 3185
	struct ftrace_probe_ops *probe_ops;
	struct ftrace_func_probe *probe;
3186

3187 3188
	probe = iter->probe;
	probe_entry = iter->probe_entry;
3189

3190
	if (WARN_ON_ONCE(!probe || !probe_entry))
3191
		return -EIO;
3192

3193
	probe_ops = probe->probe_ops;
3194

3195
	if (probe_ops->print)
3196
		return probe_ops->print(m, probe_entry->ip, probe_ops, probe->data);
3197

3198 3199
	seq_printf(m, "%ps:%ps\n", (void *)probe_entry->ip,
		   (void *)probe_ops->func);
3200 3201 3202 3203

	return 0;
}

3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279
static void *
t_mod_next(struct seq_file *m, loff_t *pos)
{
	struct ftrace_iterator *iter = m->private;
	struct trace_array *tr = iter->tr;

	(*pos)++;
	iter->pos = *pos;

	iter->mod_list = iter->mod_list->next;

	if (iter->mod_list == &tr->mod_trace ||
	    iter->mod_list == &tr->mod_notrace) {
		iter->flags &= ~FTRACE_ITER_MOD;
		return NULL;
	}

	iter->mod_pos = *pos;

	return iter;
}

static void *t_mod_start(struct seq_file *m, loff_t *pos)
{
	struct ftrace_iterator *iter = m->private;
	void *p = NULL;
	loff_t l;

	if (iter->func_pos > *pos)
		return NULL;

	iter->mod_pos = iter->func_pos;

	/* probes are only available if tr is set */
	if (!iter->tr)
		return NULL;

	for (l = 0; l <= (*pos - iter->func_pos); ) {
		p = t_mod_next(m, &l);
		if (!p)
			break;
	}
	if (!p) {
		iter->flags &= ~FTRACE_ITER_MOD;
		return t_probe_start(m, pos);
	}

	/* Only set this if we have an item */
	iter->flags |= FTRACE_ITER_MOD;

	return iter;
}

static int
t_mod_show(struct seq_file *m, struct ftrace_iterator *iter)
{
	struct ftrace_mod_load *ftrace_mod;
	struct trace_array *tr = iter->tr;

	if (WARN_ON_ONCE(!iter->mod_list) ||
			 iter->mod_list == &tr->mod_trace ||
			 iter->mod_list == &tr->mod_notrace)
		return -EIO;

	ftrace_mod = list_entry(iter->mod_list, struct ftrace_mod_load, list);

	if (ftrace_mod->func)
		seq_printf(m, "%s", ftrace_mod->func);
	else
		seq_putc(m, '*');

	seq_printf(m, ":mod:%s\n", ftrace_mod->module);

	return 0;
}

I
Ingo Molnar 已提交
3280
static void *
3281
t_func_next(struct seq_file *m, loff_t *pos)
3282 3283 3284 3285 3286
{
	struct ftrace_iterator *iter = m->private;
	struct dyn_ftrace *rec = NULL;

	(*pos)++;
3287

3288 3289 3290 3291 3292 3293 3294 3295 3296
 retry:
	if (iter->idx >= iter->pg->index) {
		if (iter->pg->next) {
			iter->pg = iter->pg->next;
			iter->idx = 0;
			goto retry;
		}
	} else {
		rec = &iter->pg->records[iter->idx++];
3297 3298
		if (((iter->flags & (FTRACE_ITER_FILTER | FTRACE_ITER_NOTRACE)) &&
		     !ftrace_lookup_ip(iter->hash, rec->ip)) ||
3299 3300

		    ((iter->flags & FTRACE_ITER_ENABLED) &&
3301
		     !(rec->flags & FTRACE_FL_ENABLED))) {
3302

3303 3304 3305 3306 3307
			rec = NULL;
			goto retry;
		}
	}

3308
	if (!rec)
3309
		return NULL;
3310

3311
	iter->pos = iter->func_pos = *pos;
3312 3313 3314
	iter->func = rec;

	return iter;
3315 3316
}

3317 3318 3319 3320
static void *
t_next(struct seq_file *m, void *v, loff_t *pos)
{
	struct ftrace_iterator *iter = m->private;
3321
	loff_t l = *pos; /* t_probe_start() must use original pos */
3322 3323 3324 3325 3326
	void *ret;

	if (unlikely(ftrace_disabled))
		return NULL;

3327 3328
	if (iter->flags & FTRACE_ITER_PROBE)
		return t_probe_next(m, pos);
3329

3330 3331 3332
	if (iter->flags & FTRACE_ITER_MOD)
		return t_mod_next(m, pos);

3333
	if (iter->flags & FTRACE_ITER_PRINTALL) {
3334
		/* next must increment pos, and t_probe_start does not */
3335
		(*pos)++;
3336
		return t_mod_start(m, &l);
3337 3338 3339 3340 3341
	}

	ret = t_func_next(m, pos);

	if (!ret)
3342
		return t_mod_start(m, &l);
3343 3344 3345 3346

	return ret;
}

3347 3348 3349 3350
static void reset_iter_read(struct ftrace_iterator *iter)
{
	iter->pos = 0;
	iter->func_pos = 0;
3351
	iter->flags &= ~(FTRACE_ITER_PRINTALL | FTRACE_ITER_PROBE | FTRACE_ITER_MOD);
3352 3353 3354 3355 3356 3357
}

static void *t_start(struct seq_file *m, loff_t *pos)
{
	struct ftrace_iterator *iter = m->private;
	void *p = NULL;
3358
	loff_t l;
3359

3360
	mutex_lock(&ftrace_lock);
3361 3362 3363 3364

	if (unlikely(ftrace_disabled))
		return NULL;

3365 3366 3367 3368 3369 3370
	/*
	 * If an lseek was done, then reset and start from beginning.
	 */
	if (*pos < iter->pos)
		reset_iter_read(iter);

3371 3372 3373 3374 3375
	/*
	 * For set_ftrace_filter reading, if we have the filter
	 * off, we can short cut and just print out that all
	 * functions are enabled.
	 */
3376 3377
	if ((iter->flags & (FTRACE_ITER_FILTER | FTRACE_ITER_NOTRACE)) &&
	    ftrace_hash_empty(iter->hash)) {
3378
		iter->func_pos = 1; /* Account for the message */
3379
		if (*pos > 0)
3380
			return t_mod_start(m, pos);
3381
		iter->flags |= FTRACE_ITER_PRINTALL;
3382
		/* reset in case of seek/pread */
3383
		iter->flags &= ~FTRACE_ITER_PROBE;
3384 3385 3386
		return iter;
	}

3387 3388
	if (iter->flags & FTRACE_ITER_MOD)
		return t_mod_start(m, pos);
3389

3390 3391 3392 3393 3394
	/*
	 * Unfortunately, we need to restart at ftrace_pages_start
	 * every time we let go of the ftrace_mutex. This is because
	 * those pointers can change without the lock.
	 */
3395 3396 3397
	iter->pg = ftrace_pages_start;
	iter->idx = 0;
	for (l = 0; l <= *pos; ) {
3398
		p = t_func_next(m, &l);
3399 3400
		if (!p)
			break;
3401
	}
3402

3403
	if (!p)
3404
		return t_mod_start(m, pos);
3405 3406

	return iter;
3407 3408 3409 3410
}

static void t_stop(struct seq_file *m, void *p)
{
3411
	mutex_unlock(&ftrace_lock);
3412 3413
}

3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429
void * __weak
arch_ftrace_trampoline_func(struct ftrace_ops *ops, struct dyn_ftrace *rec)
{
	return NULL;
}

static void add_trampoline_func(struct seq_file *m, struct ftrace_ops *ops,
				struct dyn_ftrace *rec)
{
	void *ptr;

	ptr = arch_ftrace_trampoline_func(ops, rec);
	if (ptr)
		seq_printf(m, " ->%pS", ptr);
}

3430 3431
static int t_show(struct seq_file *m, void *v)
{
3432
	struct ftrace_iterator *iter = m->private;
3433
	struct dyn_ftrace *rec;
3434

3435 3436
	if (iter->flags & FTRACE_ITER_PROBE)
		return t_probe_show(m, iter);
3437

3438 3439 3440
	if (iter->flags & FTRACE_ITER_MOD)
		return t_mod_show(m, iter);

3441
	if (iter->flags & FTRACE_ITER_PRINTALL) {
3442
		if (iter->flags & FTRACE_ITER_NOTRACE)
3443
			seq_puts(m, "#### no functions disabled ####\n");
3444
		else
3445
			seq_puts(m, "#### all functions enabled ####\n");
3446 3447 3448
		return 0;
	}

3449 3450
	rec = iter->func;

3451 3452 3453
	if (!rec)
		return 0;

3454
	seq_printf(m, "%ps", (void *)rec->ip);
3455
	if (iter->flags & FTRACE_ITER_ENABLED) {
3456
		struct ftrace_ops *ops;
3457

3458
		seq_printf(m, " (%ld)%s%s",
3459
			   ftrace_rec_count(rec),
3460 3461
			   rec->flags & FTRACE_FL_REGS ? " R" : "  ",
			   rec->flags & FTRACE_FL_IPMODIFY ? " I" : "  ");
3462
		if (rec->flags & FTRACE_FL_TRAMP_EN) {
3463
			ops = ftrace_find_tramp_ops_any(rec);
3464 3465 3466 3467 3468
			if (ops) {
				do {
					seq_printf(m, "\ttramp: %pS (%pS)",
						   (void *)ops->trampoline,
						   (void *)ops->func);
3469
					add_trampoline_func(m, ops, rec);
3470 3471 3472
					ops = ftrace_find_tramp_ops_next(rec, ops);
				} while (ops);
			} else
3473
				seq_puts(m, "\ttramp: ERROR!");
3474 3475
		} else {
			add_trampoline_func(m, NULL, rec);
3476 3477 3478
		}
	}	

3479
	seq_putc(m, '\n');
3480 3481 3482 3483

	return 0;
}

J
James Morris 已提交
3484
static const struct seq_operations show_ftrace_seq_ops = {
3485 3486 3487 3488 3489 3490
	.start = t_start,
	.next = t_next,
	.stop = t_stop,
	.show = t_show,
};

I
Ingo Molnar 已提交
3491
static int
3492 3493 3494 3495
ftrace_avail_open(struct inode *inode, struct file *file)
{
	struct ftrace_iterator *iter;

3496 3497 3498
	if (unlikely(ftrace_disabled))
		return -ENODEV;

3499
	iter = __seq_open_private(file, &show_ftrace_seq_ops, sizeof(*iter));
3500 3501
	if (!iter)
		return -ENOMEM;
3502

3503 3504 3505 3506
	iter->pg = ftrace_pages_start;
	iter->ops = &global_ops;

	return 0;
3507 3508
}

3509 3510 3511 3512 3513
static int
ftrace_enabled_open(struct inode *inode, struct file *file)
{
	struct ftrace_iterator *iter;

3514
	iter = __seq_open_private(file, &show_ftrace_seq_ops, sizeof(*iter));
3515 3516
	if (!iter)
		return -ENOMEM;
3517

3518 3519 3520 3521 3522
	iter->pg = ftrace_pages_start;
	iter->flags = FTRACE_ITER_ENABLED;
	iter->ops = &global_ops;

	return 0;
3523 3524
}

3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537
/**
 * ftrace_regex_open - initialize function tracer filter files
 * @ops: The ftrace_ops that hold the hash filters
 * @flag: The type of filter to process
 * @inode: The inode, usually passed in to your open routine
 * @file: The file, usually passed in to your open routine
 *
 * ftrace_regex_open() initializes the filter files for the
 * @ops. Depending on @flag it may process the filter hash or
 * the notrace hash of @ops. With this called from the open
 * routine, you can use ftrace_filter_write() for the write
 * routine if @flag has FTRACE_ITER_FILTER set, or
 * ftrace_notrace_write() if @flag has FTRACE_ITER_NOTRACE set.
3538
 * tracing_lseek() should be used as the lseek routine, and
3539 3540 3541
 * release must call ftrace_regex_release().
 */
int
3542
ftrace_regex_open(struct ftrace_ops *ops, int flag,
3543
		  struct inode *inode, struct file *file)
3544 3545
{
	struct ftrace_iterator *iter;
3546
	struct ftrace_hash *hash;
3547 3548
	struct list_head *mod_head;
	struct trace_array *tr = ops->private;
3549 3550
	int ret = 0;

3551 3552
	ftrace_ops_init(ops);

3553 3554 3555
	if (unlikely(ftrace_disabled))
		return -ENODEV;

3556 3557 3558 3559
	iter = kzalloc(sizeof(*iter), GFP_KERNEL);
	if (!iter)
		return -ENOMEM;

3560 3561 3562 3563 3564
	if (trace_parser_get_init(&iter->parser, FTRACE_BUFF_MAX)) {
		kfree(iter);
		return -ENOMEM;
	}

3565 3566
	iter->ops = ops;
	iter->flags = flag;
3567
	iter->tr = tr;
3568

3569
	mutex_lock(&ops->func_hash->regex_lock);
3570

3571
	if (flag & FTRACE_ITER_NOTRACE) {
3572
		hash = ops->func_hash->notrace_hash;
3573
		mod_head = tr ? &tr->mod_notrace : NULL;
3574
	} else {
3575
		hash = ops->func_hash->filter_hash;
3576
		mod_head = tr ? &tr->mod_trace : NULL;
3577
	}
3578

3579 3580
	iter->mod_list = mod_head;

3581
	if (file->f_mode & FMODE_WRITE) {
3582 3583
		const int size_bits = FTRACE_HASH_DEFAULT_BITS;

3584
		if (file->f_flags & O_TRUNC) {
3585
			iter->hash = alloc_ftrace_hash(size_bits);
3586 3587
			clear_ftrace_mod_list(mod_head);
	        } else {
3588
			iter->hash = alloc_and_copy_ftrace_hash(size_bits, hash);
3589
		}
3590

3591 3592 3593
		if (!iter->hash) {
			trace_parser_put(&iter->parser);
			kfree(iter);
3594 3595
			ret = -ENOMEM;
			goto out_unlock;
3596
		}
3597 3598
	} else
		iter->hash = hash;
3599

3600 3601 3602 3603 3604 3605 3606
	if (file->f_mode & FMODE_READ) {
		iter->pg = ftrace_pages_start;

		ret = seq_open(file, &show_ftrace_seq_ops);
		if (!ret) {
			struct seq_file *m = file->private_data;
			m->private = iter;
3607
		} else {
3608 3609
			/* Failed */
			free_ftrace_hash(iter->hash);
3610
			trace_parser_put(&iter->parser);
3611
			kfree(iter);
3612
		}
3613 3614
	} else
		file->private_data = iter;
3615 3616

 out_unlock:
3617
	mutex_unlock(&ops->func_hash->regex_lock);
3618 3619 3620 3621

	return ret;
}

3622 3623 3624
static int
ftrace_filter_open(struct inode *inode, struct file *file)
{
3625 3626 3627
	struct ftrace_ops *ops = inode->i_private;

	return ftrace_regex_open(ops,
3628
			FTRACE_ITER_FILTER | FTRACE_ITER_DO_PROBES,
3629
			inode, file);
3630 3631 3632 3633 3634
}

static int
ftrace_notrace_open(struct inode *inode, struct file *file)
{
3635 3636 3637
	struct ftrace_ops *ops = inode->i_private;

	return ftrace_regex_open(ops, FTRACE_ITER_NOTRACE,
3638
				 inode, file);
3639 3640
}

3641 3642 3643 3644 3645 3646 3647
/* Type for quick search ftrace basic regexes (globs) from filter_parse_regex */
struct ftrace_glob {
	char *search;
	unsigned len;
	int type;
};

3648 3649 3650 3651 3652 3653 3654 3655 3656 3657
/*
 * If symbols in an architecture don't correspond exactly to the user-visible
 * name of what they represent, it is possible to define this function to
 * perform the necessary adjustments.
*/
char * __weak arch_ftrace_match_adjust(char *str, const char *search)
{
	return str;
}

3658
static int ftrace_match(char *str, struct ftrace_glob *g)
3659 3660
{
	int matched = 0;
3661
	int slen;
3662

3663 3664
	str = arch_ftrace_match_adjust(str, g->search);

3665
	switch (g->type) {
3666
	case MATCH_FULL:
3667
		if (strcmp(str, g->search) == 0)
3668 3669 3670
			matched = 1;
		break;
	case MATCH_FRONT_ONLY:
3671
		if (strncmp(str, g->search, g->len) == 0)
3672 3673 3674
			matched = 1;
		break;
	case MATCH_MIDDLE_ONLY:
3675
		if (strstr(str, g->search))
3676 3677 3678
			matched = 1;
		break;
	case MATCH_END_ONLY:
3679
		slen = strlen(str);
3680 3681
		if (slen >= g->len &&
		    memcmp(str + slen - g->len, g->search, g->len) == 0)
3682 3683
			matched = 1;
		break;
3684 3685 3686 3687
	case MATCH_GLOB:
		if (glob_match(g->search, str))
			matched = 1;
		break;
3688 3689 3690 3691 3692
	}

	return matched;
}

3693
static int
3694
enter_record(struct ftrace_hash *hash, struct dyn_ftrace *rec, int clear_filter)
3695
{
3696 3697 3698
	struct ftrace_func_entry *entry;
	int ret = 0;

3699
	entry = ftrace_lookup_ip(hash, rec->ip);
3700
	if (clear_filter) {
3701 3702 3703
		/* Do nothing if it doesn't exist */
		if (!entry)
			return 0;
3704

3705
		free_hash_entry(hash, entry);
3706 3707 3708 3709
	} else {
		/* Do nothing if it exists */
		if (entry)
			return 0;
3710

3711
		ret = add_hash_entry(hash, rec->ip);
3712 3713
	}
	return ret;
3714 3715
}

3716
static int
D
Dmitry Safonov 已提交
3717 3718
ftrace_match_record(struct dyn_ftrace *rec, struct ftrace_glob *func_g,
		struct ftrace_glob *mod_g, int exclude_mod)
3719 3720
{
	char str[KSYM_SYMBOL_LEN];
3721 3722 3723 3724
	char *modname;

	kallsyms_lookup(rec->ip, NULL, NULL, &modname, str);

D
Dmitry Safonov 已提交
3725 3726 3727 3728 3729 3730
	if (mod_g) {
		int mod_matches = (modname) ? ftrace_match(modname, mod_g) : 0;

		/* blank module name to match all modules */
		if (!mod_g->len) {
			/* blank module globbing: modname xor exclude_mod */
3731
			if (!exclude_mod != !modname)
D
Dmitry Safonov 已提交
3732 3733 3734 3735
				goto func_match;
			return 0;
		}

3736 3737 3738 3739 3740 3741 3742 3743
		/*
		 * exclude_mod is set to trace everything but the given
		 * module. If it is set and the module matches, then
		 * return 0. If it is not set, and the module doesn't match
		 * also return 0. Otherwise, check the function to see if
		 * that matches.
		 */
		if (!mod_matches == !exclude_mod)
3744
			return 0;
D
Dmitry Safonov 已提交
3745
func_match:
3746
		/* blank search means to match all funcs in the mod */
3747
		if (!func_g->len)
3748 3749
			return 1;
	}
3750

3751
	return ftrace_match(str, func_g);
3752 3753
}

3754
static int
3755
match_records(struct ftrace_hash *hash, char *func, int len, char *mod)
3756 3757 3758
{
	struct ftrace_page *pg;
	struct dyn_ftrace *rec;
3759
	struct ftrace_glob func_g = { .type = MATCH_FULL };
D
Dmitry Safonov 已提交
3760 3761 3762
	struct ftrace_glob mod_g = { .type = MATCH_FULL };
	struct ftrace_glob *mod_match = (mod) ? &mod_g : NULL;
	int exclude_mod = 0;
3763
	int found = 0;
3764
	int ret;
3765
	int clear_filter = 0;
3766

D
Dmitry Safonov 已提交
3767
	if (func) {
3768 3769 3770
		func_g.type = filter_parse_regex(func, len, &func_g.search,
						 &clear_filter);
		func_g.len = strlen(func_g.search);
3771
	}
3772

D
Dmitry Safonov 已提交
3773 3774 3775 3776
	if (mod) {
		mod_g.type = filter_parse_regex(mod, strlen(mod),
				&mod_g.search, &exclude_mod);
		mod_g.len = strlen(mod_g.search);
3777
	}
3778

3779
	mutex_lock(&ftrace_lock);
3780

3781 3782
	if (unlikely(ftrace_disabled))
		goto out_unlock;
3783

3784
	do_for_each_ftrace_rec(pg, rec) {
3785 3786 3787 3788

		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

D
Dmitry Safonov 已提交
3789
		if (ftrace_match_record(rec, &func_g, mod_match, exclude_mod)) {
3790
			ret = enter_record(hash, rec, clear_filter);
3791 3792 3793 3794
			if (ret < 0) {
				found = ret;
				goto out_unlock;
			}
3795
			found = 1;
3796 3797
		}
	} while_for_each_ftrace_rec();
3798
 out_unlock:
3799
	mutex_unlock(&ftrace_lock);
3800 3801

	return found;
3802 3803
}

3804
static int
3805
ftrace_match_records(struct ftrace_hash *hash, char *buff, int len)
3806
{
3807
	return match_records(hash, buff, len, NULL);
3808 3809
}

3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859
static void ftrace_ops_update_code(struct ftrace_ops *ops,
				   struct ftrace_ops_hash *old_hash)
{
	struct ftrace_ops *op;

	if (!ftrace_enabled)
		return;

	if (ops->flags & FTRACE_OPS_FL_ENABLED) {
		ftrace_run_modify_code(ops, FTRACE_UPDATE_CALLS, old_hash);
		return;
	}

	/*
	 * If this is the shared global_ops filter, then we need to
	 * check if there is another ops that shares it, is enabled.
	 * If so, we still need to run the modify code.
	 */
	if (ops->func_hash != &global_ops.local_hash)
		return;

	do_for_each_ftrace_op(op, ftrace_ops_list) {
		if (op->func_hash == &global_ops.local_hash &&
		    op->flags & FTRACE_OPS_FL_ENABLED) {
			ftrace_run_modify_code(op, FTRACE_UPDATE_CALLS, old_hash);
			/* Only need to do this once */
			return;
		}
	} while_for_each_ftrace_op(op);
}

static int ftrace_hash_move_and_update_ops(struct ftrace_ops *ops,
					   struct ftrace_hash **orig_hash,
					   struct ftrace_hash *hash,
					   int enable)
{
	struct ftrace_ops_hash old_hash_ops;
	struct ftrace_hash *old_hash;
	int ret;

	old_hash = *orig_hash;
	old_hash_ops.filter_hash = ops->func_hash->filter_hash;
	old_hash_ops.notrace_hash = ops->func_hash->notrace_hash;
	ret = ftrace_hash_move(ops, enable, orig_hash, hash);
	if (!ret) {
		ftrace_ops_update_code(ops, &old_hash_ops);
		free_ftrace_hash_rcu(old_hash);
	}
	return ret;
}
3860

3861 3862 3863 3864
static bool module_exists(const char *module)
{
	/* All modules have the symbol __this_module */
	const char this_mod[] = "__this_module";
3865
	char modname[MAX_PARAM_PREFIX_LEN + sizeof(this_mod) + 2];
3866 3867 3868
	unsigned long val;
	int n;

3869
	n = snprintf(modname, sizeof(modname), "%s:%s", module, this_mod);
3870

3871
	if (n > sizeof(modname) - 1)
3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897
		return false;

	val = module_kallsyms_lookup_name(modname);
	return val != 0;
}

static int cache_mod(struct trace_array *tr,
		     const char *func, char *module, int enable)
{
	struct ftrace_mod_load *ftrace_mod, *n;
	struct list_head *head = enable ? &tr->mod_trace : &tr->mod_notrace;
	int ret;

	mutex_lock(&ftrace_lock);

	/* We do not cache inverse filters */
	if (func[0] == '!') {
		func++;
		ret = -EINVAL;

		/* Look to remove this hash */
		list_for_each_entry_safe(ftrace_mod, n, head, list) {
			if (strcmp(ftrace_mod->module, module) != 0)
				continue;

			/* no func matches all */
3898
			if (strcmp(func, "*") == 0 ||
3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921
			    (ftrace_mod->func &&
			     strcmp(ftrace_mod->func, func) == 0)) {
				ret = 0;
				free_ftrace_mod(ftrace_mod);
				continue;
			}
		}
		goto out;
	}

	ret = -EINVAL;
	/* We only care about modules that have not been loaded yet */
	if (module_exists(module))
		goto out;

	/* Save this string off, and execute it when the module is loaded */
	ret = ftrace_add_mod(tr, func, module, enable);
 out:
	mutex_unlock(&ftrace_lock);

	return ret;
}

3922 3923 3924 3925
static int
ftrace_set_regex(struct ftrace_ops *ops, unsigned char *buf, int len,
		 int reset, int enable);

3926
#ifdef CONFIG_MODULES
3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945
static void process_mod_list(struct list_head *head, struct ftrace_ops *ops,
			     char *mod, bool enable)
{
	struct ftrace_mod_load *ftrace_mod, *n;
	struct ftrace_hash **orig_hash, *new_hash;
	LIST_HEAD(process_mods);
	char *func;
	int ret;

	mutex_lock(&ops->func_hash->regex_lock);

	if (enable)
		orig_hash = &ops->func_hash->filter_hash;
	else
		orig_hash = &ops->func_hash->notrace_hash;

	new_hash = alloc_and_copy_ftrace_hash(FTRACE_HASH_DEFAULT_BITS,
					      *orig_hash);
	if (!new_hash)
3946
		goto out; /* warn? */
3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981

	mutex_lock(&ftrace_lock);

	list_for_each_entry_safe(ftrace_mod, n, head, list) {

		if (strcmp(ftrace_mod->module, mod) != 0)
			continue;

		if (ftrace_mod->func)
			func = kstrdup(ftrace_mod->func, GFP_KERNEL);
		else
			func = kstrdup("*", GFP_KERNEL);

		if (!func) /* warn? */
			continue;

		list_del(&ftrace_mod->list);
		list_add(&ftrace_mod->list, &process_mods);

		/* Use the newly allocated func, as it may be "*" */
		kfree(ftrace_mod->func);
		ftrace_mod->func = func;
	}

	mutex_unlock(&ftrace_lock);

	list_for_each_entry_safe(ftrace_mod, n, &process_mods, list) {

		func = ftrace_mod->func;

		/* Grabs ftrace_lock, which is why we have this extra step */
		match_records(new_hash, func, strlen(func), mod);
		free_ftrace_mod(ftrace_mod);
	}

3982 3983 3984
	if (enable && list_empty(head))
		new_hash->flags &= ~FTRACE_HASH_FL_MOD;

3985 3986 3987 3988 3989 3990
	mutex_lock(&ftrace_lock);

	ret = ftrace_hash_move_and_update_ops(ops, orig_hash,
					      new_hash, enable);
	mutex_unlock(&ftrace_lock);

3991
 out:
3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016
	mutex_unlock(&ops->func_hash->regex_lock);

	free_ftrace_hash(new_hash);
}

static void process_cached_mods(const char *mod_name)
{
	struct trace_array *tr;
	char *mod;

	mod = kstrdup(mod_name, GFP_KERNEL);
	if (!mod)
		return;

	mutex_lock(&trace_types_lock);
	list_for_each_entry(tr, &ftrace_trace_arrays, list) {
		if (!list_empty(&tr->mod_trace))
			process_mod_list(&tr->mod_trace, tr->ops, mod, true);
		if (!list_empty(&tr->mod_notrace))
			process_mod_list(&tr->mod_notrace, tr->ops, mod, false);
	}
	mutex_unlock(&trace_types_lock);

	kfree(mod);
}
4017
#endif
4018

4019 4020 4021 4022 4023 4024
/*
 * We register the module command as a template to show others how
 * to register the a command as well.
 */

static int
4025
ftrace_mod_callback(struct trace_array *tr, struct ftrace_hash *hash,
4026
		    char *func_orig, char *cmd, char *module, int enable)
4027
{
4028
	char *func;
4029
	int ret;
4030

4031 4032 4033 4034 4035
	/* match_records() modifies func, and we need the original */
	func = kstrdup(func_orig, GFP_KERNEL);
	if (!func)
		return -ENOMEM;

4036 4037 4038 4039 4040 4041 4042
	/*
	 * cmd == 'mod' because we only registered this func
	 * for the 'mod' ftrace_func_command.
	 * But if you register one func with multiple commands,
	 * you can tell which command was used by the cmd
	 * parameter.
	 */
4043
	ret = match_records(hash, func, strlen(func), module);
4044 4045
	kfree(func);

4046
	if (!ret)
4047
		return cache_mod(tr, func_orig, module, enable);
4048 4049 4050
	if (ret < 0)
		return ret;
	return 0;
4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061
}

static struct ftrace_func_command ftrace_mod_cmd = {
	.name			= "mod",
	.func			= ftrace_mod_callback,
};

static int __init ftrace_mod_cmd_init(void)
{
	return register_ftrace_command(&ftrace_mod_cmd);
}
4062
core_initcall(ftrace_mod_cmd_init);
4063

4064
static void function_trace_probe_call(unsigned long ip, unsigned long parent_ip,
4065
				      struct ftrace_ops *op, struct pt_regs *pt_regs)
4066
{
4067
	struct ftrace_probe_ops *probe_ops;
4068
	struct ftrace_func_probe *probe;
4069

4070 4071
	probe = container_of(op, struct ftrace_func_probe, ops);
	probe_ops = probe->probe_ops;
4072 4073 4074 4075 4076 4077

	/*
	 * Disable preemption for these calls to prevent a RCU grace
	 * period. This syncs the hash iteration and freeing of items
	 * on the hash. rcu_read_lock is too dangerous here.
	 */
4078
	preempt_disable_notrace();
4079
	probe_ops->func(ip, parent_ip, probe->tr, probe_ops, probe->data);
4080
	preempt_enable_notrace();
4081 4082
}

4083 4084 4085
struct ftrace_func_map {
	struct ftrace_func_entry	entry;
	void				*data;
4086 4087
};

4088 4089 4090
struct ftrace_func_mapper {
	struct ftrace_hash		hash;
};
4091

4092 4093 4094 4095 4096 4097
/**
 * allocate_ftrace_func_mapper - allocate a new ftrace_func_mapper
 *
 * Returns a ftrace_func_mapper descriptor that can be used to map ips to data.
 */
struct ftrace_func_mapper *allocate_ftrace_func_mapper(void)
4098
{
4099
	struct ftrace_hash *hash;
4100

4101 4102 4103 4104 4105 4106 4107 4108
	/*
	 * The mapper is simply a ftrace_hash, but since the entries
	 * in the hash are not ftrace_func_entry type, we define it
	 * as a separate structure.
	 */
	hash = alloc_ftrace_hash(FTRACE_HASH_DEFAULT_BITS);
	return (struct ftrace_func_mapper *)hash;
}
4109

4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125
/**
 * ftrace_func_mapper_find_ip - Find some data mapped to an ip
 * @mapper: The mapper that has the ip maps
 * @ip: the instruction pointer to find the data for
 *
 * Returns the data mapped to @ip if found otherwise NULL. The return
 * is actually the address of the mapper data pointer. The address is
 * returned for use cases where the data is no bigger than a long, and
 * the user can use the data pointer as its data instead of having to
 * allocate more memory for the reference.
 */
void **ftrace_func_mapper_find_ip(struct ftrace_func_mapper *mapper,
				  unsigned long ip)
{
	struct ftrace_func_entry *entry;
	struct ftrace_func_map *map;
4126

4127 4128 4129
	entry = ftrace_lookup_ip(&mapper->hash, ip);
	if (!entry)
		return NULL;
4130

4131 4132
	map = (struct ftrace_func_map *)entry;
	return &map->data;
4133 4134
}

4135 4136 4137 4138 4139 4140 4141 4142 4143 4144
/**
 * ftrace_func_mapper_add_ip - Map some data to an ip
 * @mapper: The mapper that has the ip maps
 * @ip: The instruction pointer address to map @data to
 * @data: The data to map to @ip
 *
 * Returns 0 on succes otherwise an error.
 */
int ftrace_func_mapper_add_ip(struct ftrace_func_mapper *mapper,
			      unsigned long ip, void *data)
4145
{
4146 4147
	struct ftrace_func_entry *entry;
	struct ftrace_func_map *map;
4148

4149 4150 4151
	entry = ftrace_lookup_ip(&mapper->hash, ip);
	if (entry)
		return -EBUSY;
4152

4153 4154 4155
	map = kmalloc(sizeof(*map), GFP_KERNEL);
	if (!map)
		return -ENOMEM;
4156

4157 4158
	map->entry.ip = ip;
	map->data = data;
4159

4160
	__add_hash_entry(&mapper->hash, &map->entry);
4161

4162 4163
	return 0;
}
4164

4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176
/**
 * ftrace_func_mapper_remove_ip - Remove an ip from the mapping
 * @mapper: The mapper that has the ip maps
 * @ip: The instruction pointer address to remove the data from
 *
 * Returns the data if it is found, otherwise NULL.
 * Note, if the data pointer is used as the data itself, (see 
 * ftrace_func_mapper_find_ip(), then the return value may be meaningless,
 * if the data pointer was set to zero.
 */
void *ftrace_func_mapper_remove_ip(struct ftrace_func_mapper *mapper,
				   unsigned long ip)
4177
{
4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189
	struct ftrace_func_entry *entry;
	struct ftrace_func_map *map;
	void *data;

	entry = ftrace_lookup_ip(&mapper->hash, ip);
	if (!entry)
		return NULL;

	map = (struct ftrace_func_map *)entry;
	data = map->data;

	remove_hash_entry(&mapper->hash, entry);
4190
	kfree(entry);
4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223

	return data;
}

/**
 * free_ftrace_func_mapper - free a mapping of ips and data
 * @mapper: The mapper that has the ip maps
 * @free_func: A function to be called on each data item.
 *
 * This is used to free the function mapper. The @free_func is optional
 * and can be used if the data needs to be freed as well.
 */
void free_ftrace_func_mapper(struct ftrace_func_mapper *mapper,
			     ftrace_mapper_func free_func)
{
	struct ftrace_func_entry *entry;
	struct ftrace_func_map *map;
	struct hlist_head *hhd;
	int size = 1 << mapper->hash.size_bits;
	int i;

	if (free_func && mapper->hash.count) {
		for (i = 0; i < size; i++) {
			hhd = &mapper->hash.buckets[i];
			hlist_for_each_entry(entry, hhd, hlist) {
				map = (struct ftrace_func_map *)entry;
				free_func(map);
			}
		}
	}
	free_ftrace_hash(&mapper->hash);
}

4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236
static void release_probe(struct ftrace_func_probe *probe)
{
	struct ftrace_probe_ops *probe_ops;

	mutex_lock(&ftrace_lock);

	WARN_ON(probe->ref <= 0);

	/* Subtract the ref that was used to protect this instance */
	probe->ref--;

	if (!probe->ref) {
		probe_ops = probe->probe_ops;
4237 4238 4239 4240 4241 4242
		/*
		 * Sending zero as ip tells probe_ops to free
		 * the probe->data itself
		 */
		if (probe_ops->free)
			probe_ops->free(probe_ops, probe->tr, 0, probe->data);
4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255
		list_del(&probe->list);
		kfree(probe);
	}
	mutex_unlock(&ftrace_lock);
}

static void acquire_probe_locked(struct ftrace_func_probe *probe)
{
	/*
	 * Add one ref to keep it from being freed when releasing the
	 * ftrace_lock mutex.
	 */
	probe->ref++;
4256 4257 4258
}

int
4259
register_ftrace_function_probe(char *glob, struct trace_array *tr,
4260 4261
			       struct ftrace_probe_ops *probe_ops,
			       void *data)
4262
{
4263
	struct ftrace_func_entry *entry;
4264
	struct ftrace_func_probe *probe;
4265 4266
	struct ftrace_hash **orig_hash;
	struct ftrace_hash *old_hash;
4267
	struct ftrace_hash *hash;
4268
	int count = 0;
4269
	int size;
4270
	int ret;
4271
	int i;
4272

4273
	if (WARN_ON(!tr))
4274 4275
		return -EINVAL;

4276 4277
	/* We do not support '!' for function probes */
	if (WARN_ON(glob[0] == '!'))
4278 4279
		return -EINVAL;

4280

4281 4282 4283 4284 4285
	mutex_lock(&ftrace_lock);
	/* Check if the probe_ops is already registered */
	list_for_each_entry(probe, &tr->func_probes, list) {
		if (probe->probe_ops == probe_ops)
			break;
4286
	}
4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297
	if (&probe->list == &tr->func_probes) {
		probe = kzalloc(sizeof(*probe), GFP_KERNEL);
		if (!probe) {
			mutex_unlock(&ftrace_lock);
			return -ENOMEM;
		}
		probe->probe_ops = probe_ops;
		probe->ops.func = function_trace_probe_call;
		probe->tr = tr;
		ftrace_ops_init(&probe->ops);
		list_add(&probe->list, &tr->func_probes);
4298
	}
4299

4300
	acquire_probe_locked(probe);
4301

4302
	mutex_unlock(&ftrace_lock);
4303

4304
	mutex_lock(&probe->ops.func_hash->regex_lock);
4305

4306
	orig_hash = &probe->ops.func_hash->filter_hash;
4307 4308
	old_hash = *orig_hash;
	hash = alloc_and_copy_ftrace_hash(FTRACE_HASH_DEFAULT_BITS, old_hash);
4309

4310
	ret = ftrace_match_records(hash, glob, strlen(glob));
4311

4312 4313 4314
	/* Nothing found? */
	if (!ret)
		ret = -EINVAL;
4315

4316 4317
	if (ret < 0)
		goto out;
4318

4319 4320 4321 4322
	size = 1 << hash->size_bits;
	for (i = 0; i < size; i++) {
		hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
			if (ftrace_lookup_ip(old_hash, entry->ip))
4323
				continue;
4324 4325 4326 4327 4328
			/*
			 * The caller might want to do something special
			 * for each function we find. We call the callback
			 * to give the caller an opportunity to do so.
			 */
4329 4330
			if (probe_ops->init) {
				ret = probe_ops->init(probe_ops, tr,
4331 4332 4333 4334 4335 4336 4337
						      entry->ip, data,
						      &probe->data);
				if (ret < 0) {
					if (probe_ops->free && count)
						probe_ops->free(probe_ops, tr,
								0, probe->data);
					probe->data = NULL;
4338
					goto out;
4339
				}
4340
			}
4341
			count++;
4342
		}
4343
	}
4344

4345
	mutex_lock(&ftrace_lock);
4346

4347 4348 4349 4350 4351
	if (!count) {
		/* Nothing was added? */
		ret = -EINVAL;
		goto out_unlock;
	}
4352

4353 4354
	ret = ftrace_hash_move_and_update_ops(&probe->ops, orig_hash,
					      hash, 1);
4355
	if (ret < 0)
4356
		goto err_unlock;
4357

4358 4359
	/* One ref for each new function traced */
	probe->ref += count;
4360

4361 4362
	if (!(probe->ops.flags & FTRACE_OPS_FL_ENABLED))
		ret = ftrace_startup(&probe->ops, 0);
4363

4364
 out_unlock:
4365
	mutex_unlock(&ftrace_lock);
4366

4367
	if (!ret)
4368
		ret = count;
4369
 out:
4370
	mutex_unlock(&probe->ops.func_hash->regex_lock);
4371
	free_ftrace_hash(hash);
4372

4373
	release_probe(probe);
4374

4375
	return ret;
4376

4377
 err_unlock:
4378
	if (!probe_ops->free || !count)
4379 4380 4381 4382 4383 4384 4385
		goto out_unlock;

	/* Failed to do the move, need to call the free functions */
	for (i = 0; i < size; i++) {
		hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
			if (ftrace_lookup_ip(old_hash, entry->ip))
				continue;
4386
			probe_ops->free(probe_ops, tr, entry->ip, probe->data);
4387 4388 4389
		}
	}
	goto out_unlock;
4390 4391
}

4392
int
4393 4394
unregister_ftrace_function_probe_func(char *glob, struct trace_array *tr,
				      struct ftrace_probe_ops *probe_ops)
4395
{
4396
	struct ftrace_ops_hash old_hash_ops;
4397
	struct ftrace_func_entry *entry;
4398
	struct ftrace_func_probe *probe;
4399
	struct ftrace_glob func_g;
4400 4401 4402
	struct ftrace_hash **orig_hash;
	struct ftrace_hash *old_hash;
	struct ftrace_hash *hash = NULL;
4403
	struct hlist_node *tmp;
4404
	struct hlist_head hhd;
4405
	char str[KSYM_SYMBOL_LEN];
4406 4407
	int count = 0;
	int i, ret = -ENODEV;
4408
	int size;
4409

4410
	if (!glob || !strlen(glob) || !strcmp(glob, "*"))
4411
		func_g.search = NULL;
4412
	else {
4413 4414
		int not;

4415 4416 4417
		func_g.type = filter_parse_regex(glob, strlen(glob),
						 &func_g.search, &not);
		func_g.len = strlen(func_g.search);
4418

S
Steven Rostedt 已提交
4419
		/* we do not support '!' for function probes */
4420
		if (WARN_ON(not))
4421
			return -EINVAL;
4422 4423
	}

4424 4425 4426 4427 4428
	mutex_lock(&ftrace_lock);
	/* Check if the probe_ops is already registered */
	list_for_each_entry(probe, &tr->func_probes, list) {
		if (probe->probe_ops == probe_ops)
			break;
4429
	}
4430 4431 4432 4433 4434 4435 4436 4437 4438 4439
	if (&probe->list == &tr->func_probes)
		goto err_unlock_ftrace;

	ret = -EINVAL;
	if (!(probe->ops.flags & FTRACE_OPS_FL_INITIALIZED))
		goto err_unlock_ftrace;

	acquire_probe_locked(probe);

	mutex_unlock(&ftrace_lock);
4440

4441
	mutex_lock(&probe->ops.func_hash->regex_lock);
4442

4443
	orig_hash = &probe->ops.func_hash->filter_hash;
4444 4445 4446 4447
	old_hash = *orig_hash;

	if (ftrace_hash_empty(old_hash))
		goto out_unlock;
4448

4449 4450 4451 4452
	old_hash_ops.filter_hash = old_hash;
	/* Probes only have filters */
	old_hash_ops.notrace_hash = NULL;

4453
	ret = -ENOMEM;
4454
	hash = alloc_and_copy_ftrace_hash(FTRACE_HASH_DEFAULT_BITS, old_hash);
4455 4456 4457
	if (!hash)
		goto out_unlock;

4458
	INIT_HLIST_HEAD(&hhd);
4459

4460 4461 4462
	size = 1 << hash->size_bits;
	for (i = 0; i < size; i++) {
		hlist_for_each_entry_safe(entry, tmp, &hash->buckets[i], hlist) {
4463

4464
			if (func_g.search) {
4465 4466
				kallsyms_lookup(entry->ip, NULL, NULL,
						NULL, str);
4467
				if (!ftrace_match(str, &func_g))
4468 4469
					continue;
			}
4470
			count++;
4471 4472
			remove_hash_entry(hash, entry);
			hlist_add_head(&entry->hlist, &hhd);
4473 4474
		}
	}
4475 4476

	/* Nothing found? */
4477
	if (!count) {
4478 4479 4480 4481
		ret = -EINVAL;
		goto out_unlock;
	}

4482
	mutex_lock(&ftrace_lock);
4483

4484
	WARN_ON(probe->ref < count);
4485

4486
	probe->ref -= count;
4487

4488 4489 4490 4491
	if (ftrace_hash_empty(hash))
		ftrace_shutdown(&probe->ops, 0);

	ret = ftrace_hash_move_and_update_ops(&probe->ops, orig_hash,
4492
					      hash, 1);
4493 4494

	/* still need to update the function call sites */
4495
	if (ftrace_enabled && !ftrace_hash_empty(hash))
4496
		ftrace_run_modify_code(&probe->ops, FTRACE_UPDATE_CALLS,
4497
				       &old_hash_ops);
4498
	synchronize_sched();
4499

4500 4501
	hlist_for_each_entry_safe(entry, tmp, &hhd, hlist) {
		hlist_del(&entry->hlist);
4502
		if (probe_ops->free)
4503
			probe_ops->free(probe_ops, tr, entry->ip, probe->data);
4504
		kfree(entry);
4505
	}
4506
	mutex_unlock(&ftrace_lock);
4507

4508
 out_unlock:
4509
	mutex_unlock(&probe->ops.func_hash->regex_lock);
4510
	free_ftrace_hash(hash);
4511

4512
	release_probe(probe);
4513

4514
	return ret;
4515

4516 4517
 err_unlock_ftrace:
	mutex_unlock(&ftrace_lock);
4518
	return ret;
4519 4520
}

4521 4522 4523 4524 4525 4526 4527 4528
void clear_ftrace_function_probes(struct trace_array *tr)
{
	struct ftrace_func_probe *probe, *n;

	list_for_each_entry_safe(probe, n, &tr->func_probes, list)
		unregister_ftrace_function_probe_func(NULL, tr, probe->probe_ops);
}

4529 4530 4531
static LIST_HEAD(ftrace_commands);
static DEFINE_MUTEX(ftrace_cmd_mutex);

4532 4533 4534 4535 4536
/*
 * Currently we only register ftrace commands from __init, so mark this
 * __init too.
 */
__init int register_ftrace_command(struct ftrace_func_command *cmd)
4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554
{
	struct ftrace_func_command *p;
	int ret = 0;

	mutex_lock(&ftrace_cmd_mutex);
	list_for_each_entry(p, &ftrace_commands, list) {
		if (strcmp(cmd->name, p->name) == 0) {
			ret = -EBUSY;
			goto out_unlock;
		}
	}
	list_add(&cmd->list, &ftrace_commands);
 out_unlock:
	mutex_unlock(&ftrace_cmd_mutex);

	return ret;
}

4555 4556 4557 4558 4559
/*
 * Currently we only unregister ftrace commands from __init, so mark
 * this __init too.
 */
__init int unregister_ftrace_command(struct ftrace_func_command *cmd)
4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577
{
	struct ftrace_func_command *p, *n;
	int ret = -ENODEV;

	mutex_lock(&ftrace_cmd_mutex);
	list_for_each_entry_safe(p, n, &ftrace_commands, list) {
		if (strcmp(cmd->name, p->name) == 0) {
			ret = 0;
			list_del_init(&p->list);
			goto out_unlock;
		}
	}
 out_unlock:
	mutex_unlock(&ftrace_cmd_mutex);

	return ret;
}

4578
static int ftrace_process_regex(struct ftrace_iterator *iter,
4579
				char *buff, int len, int enable)
4580
{
4581
	struct ftrace_hash *hash = iter->hash;
4582
	struct trace_array *tr = iter->ops->private;
4583
	char *func, *command, *next = buff;
S
Steven Rostedt 已提交
4584
	struct ftrace_func_command *p;
4585
	int ret = -EINVAL;
4586 4587 4588 4589

	func = strsep(&next, ":");

	if (!next) {
4590
		ret = ftrace_match_records(hash, func, len);
4591 4592 4593 4594 4595
		if (!ret)
			ret = -EINVAL;
		if (ret < 0)
			return ret;
		return 0;
4596 4597
	}

4598
	/* command found */
4599 4600 4601

	command = strsep(&next, ":");

4602 4603 4604
	mutex_lock(&ftrace_cmd_mutex);
	list_for_each_entry(p, &ftrace_commands, list) {
		if (strcmp(p->name, command) == 0) {
4605
			ret = p->func(tr, hash, func, command, next, enable);
4606 4607
			goto out_unlock;
		}
4608
	}
4609 4610
 out_unlock:
	mutex_unlock(&ftrace_cmd_mutex);
4611

4612
	return ret;
4613 4614
}

I
Ingo Molnar 已提交
4615
static ssize_t
4616 4617
ftrace_regex_write(struct file *file, const char __user *ubuf,
		   size_t cnt, loff_t *ppos, int enable)
4618 4619
{
	struct ftrace_iterator *iter;
4620 4621
	struct trace_parser *parser;
	ssize_t ret, read;
4622

4623
	if (!cnt)
4624 4625 4626 4627 4628 4629 4630 4631
		return 0;

	if (file->f_mode & FMODE_READ) {
		struct seq_file *m = file->private_data;
		iter = m->private;
	} else
		iter = file->private_data;

4632
	if (unlikely(ftrace_disabled))
4633 4634 4635
		return -ENODEV;

	/* iter->hash is a local copy, so we don't need regex_lock */
4636

4637 4638
	parser = &iter->parser;
	read = trace_get_user(parser, ubuf, cnt, ppos);
4639

4640
	if (read >= 0 && trace_parser_loaded(parser) &&
4641
	    !trace_parser_cont(parser)) {
4642
		ret = ftrace_process_regex(iter, parser->buffer,
4643
					   parser->idx, enable);
4644
		trace_parser_clear(parser);
4645
		if (ret < 0)
4646
			goto out;
4647
	}
4648 4649

	ret = read;
4650
 out:
4651 4652 4653
	return ret;
}

4654
ssize_t
4655 4656 4657 4658 4659 4660
ftrace_filter_write(struct file *file, const char __user *ubuf,
		    size_t cnt, loff_t *ppos)
{
	return ftrace_regex_write(file, ubuf, cnt, ppos, 1);
}

4661
ssize_t
4662 4663 4664 4665 4666 4667
ftrace_notrace_write(struct file *file, const char __user *ubuf,
		     size_t cnt, loff_t *ppos)
{
	return ftrace_regex_write(file, ubuf, cnt, ppos, 0);
}

4668
static int
4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689
ftrace_match_addr(struct ftrace_hash *hash, unsigned long ip, int remove)
{
	struct ftrace_func_entry *entry;

	if (!ftrace_location(ip))
		return -EINVAL;

	if (remove) {
		entry = ftrace_lookup_ip(hash, ip);
		if (!entry)
			return -ENOENT;
		free_hash_entry(hash, entry);
		return 0;
	}

	return add_hash_entry(hash, ip);
}

static int
ftrace_set_hash(struct ftrace_ops *ops, unsigned char *buf, int len,
		unsigned long ip, int remove, int reset, int enable)
4690
{
4691
	struct ftrace_hash **orig_hash;
4692
	struct ftrace_hash *hash;
4693
	int ret;
4694

4695
	if (unlikely(ftrace_disabled))
4696
		return -ENODEV;
4697

4698
	mutex_lock(&ops->func_hash->regex_lock);
4699

4700
	if (enable)
4701
		orig_hash = &ops->func_hash->filter_hash;
4702
	else
4703
		orig_hash = &ops->func_hash->notrace_hash;
4704

4705 4706 4707 4708 4709
	if (reset)
		hash = alloc_ftrace_hash(FTRACE_HASH_DEFAULT_BITS);
	else
		hash = alloc_and_copy_ftrace_hash(FTRACE_HASH_DEFAULT_BITS, *orig_hash);

4710 4711 4712 4713
	if (!hash) {
		ret = -ENOMEM;
		goto out_regex_unlock;
	}
4714

4715 4716 4717 4718
	if (buf && !ftrace_match_records(hash, buf, len)) {
		ret = -EINVAL;
		goto out_regex_unlock;
	}
4719 4720 4721 4722 4723
	if (ip) {
		ret = ftrace_match_addr(hash, ip, remove);
		if (ret < 0)
			goto out_regex_unlock;
	}
4724 4725

	mutex_lock(&ftrace_lock);
4726
	ret = ftrace_hash_move_and_update_ops(ops, orig_hash, hash, enable);
4727 4728
	mutex_unlock(&ftrace_lock);

4729
 out_regex_unlock:
4730
	mutex_unlock(&ops->func_hash->regex_lock);
4731 4732 4733

	free_ftrace_hash(hash);
	return ret;
4734 4735
}

4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755
static int
ftrace_set_addr(struct ftrace_ops *ops, unsigned long ip, int remove,
		int reset, int enable)
{
	return ftrace_set_hash(ops, 0, 0, ip, remove, reset, enable);
}

/**
 * ftrace_set_filter_ip - set a function to filter on in ftrace by address
 * @ops - the ops to set the filter with
 * @ip - the address to add to or remove from the filter.
 * @remove - non zero to remove the ip from the filter
 * @reset - non zero to reset all filters before applying this filter.
 *
 * Filters denote which functions should be enabled when tracing is enabled
 * If @ip is NULL, it failes to update filter.
 */
int ftrace_set_filter_ip(struct ftrace_ops *ops, unsigned long ip,
			 int remove, int reset)
{
4756
	ftrace_ops_init(ops);
4757 4758 4759 4760
	return ftrace_set_addr(ops, ip, remove, reset, 1);
}
EXPORT_SYMBOL_GPL(ftrace_set_filter_ip);

4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777
/**
 * ftrace_ops_set_global_filter - setup ops to use global filters
 * @ops - the ops which will use the global filters
 *
 * ftrace users who need global function trace filtering should call this.
 * It can set the global filter only if ops were not initialized before.
 */
void ftrace_ops_set_global_filter(struct ftrace_ops *ops)
{
	if (ops->flags & FTRACE_OPS_FL_INITIALIZED)
		return;

	ftrace_ops_init(ops);
	ops->func_hash = &global_ops.local_hash;
}
EXPORT_SYMBOL_GPL(ftrace_ops_set_global_filter);

4778 4779 4780 4781 4782 4783 4784
static int
ftrace_set_regex(struct ftrace_ops *ops, unsigned char *buf, int len,
		 int reset, int enable)
{
	return ftrace_set_hash(ops, buf, len, 0, 0, reset, enable);
}

4785 4786
/**
 * ftrace_set_filter - set a function to filter on in ftrace
4787 4788 4789 4790 4791 4792 4793 4794
 * @ops - the ops to set the filter with
 * @buf - the string that holds the function filter text.
 * @len - the length of the string.
 * @reset - non zero to reset all filters before applying this filter.
 *
 * Filters denote which functions should be enabled when tracing is enabled.
 * If @buf is NULL and reset is set, all functions will be enabled for tracing.
 */
4795
int ftrace_set_filter(struct ftrace_ops *ops, unsigned char *buf,
4796 4797
		       int len, int reset)
{
4798
	ftrace_ops_init(ops);
4799
	return ftrace_set_regex(ops, buf, len, reset, 1);
4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813
}
EXPORT_SYMBOL_GPL(ftrace_set_filter);

/**
 * ftrace_set_notrace - set a function to not trace in ftrace
 * @ops - the ops to set the notrace filter with
 * @buf - the string that holds the function notrace text.
 * @len - the length of the string.
 * @reset - non zero to reset all filters before applying this filter.
 *
 * Notrace Filters denote which functions should not be enabled when tracing
 * is enabled. If @buf is NULL and reset is set, all functions will be enabled
 * for tracing.
 */
4814
int ftrace_set_notrace(struct ftrace_ops *ops, unsigned char *buf,
4815 4816
			int len, int reset)
{
4817
	ftrace_ops_init(ops);
4818
	return ftrace_set_regex(ops, buf, len, reset, 0);
4819 4820 4821
}
EXPORT_SYMBOL_GPL(ftrace_set_notrace);
/**
4822
 * ftrace_set_global_filter - set a function to filter on with global tracers
4823 4824 4825 4826 4827 4828 4829
 * @buf - the string that holds the function filter text.
 * @len - the length of the string.
 * @reset - non zero to reset all filters before applying this filter.
 *
 * Filters denote which functions should be enabled when tracing is enabled.
 * If @buf is NULL and reset is set, all functions will be enabled for tracing.
 */
4830
void ftrace_set_global_filter(unsigned char *buf, int len, int reset)
4831
{
4832
	ftrace_set_regex(&global_ops, buf, len, reset, 1);
4833
}
4834
EXPORT_SYMBOL_GPL(ftrace_set_global_filter);
4835

4836
/**
4837
 * ftrace_set_global_notrace - set a function to not trace with global tracers
4838 4839 4840 4841 4842 4843 4844 4845
 * @buf - the string that holds the function notrace text.
 * @len - the length of the string.
 * @reset - non zero to reset all filters before applying this filter.
 *
 * Notrace Filters denote which functions should not be enabled when tracing
 * is enabled. If @buf is NULL and reset is set, all functions will be enabled
 * for tracing.
 */
4846
void ftrace_set_global_notrace(unsigned char *buf, int len, int reset)
4847
{
4848
	ftrace_set_regex(&global_ops, buf, len, reset, 0);
4849
}
4850
EXPORT_SYMBOL_GPL(ftrace_set_global_notrace);
4851

4852 4853 4854 4855 4856 4857 4858
/*
 * command line interface to allow users to set filters on boot up.
 */
#define FTRACE_FILTER_SIZE		COMMAND_LINE_SIZE
static char ftrace_notrace_buf[FTRACE_FILTER_SIZE] __initdata;
static char ftrace_filter_buf[FTRACE_FILTER_SIZE] __initdata;

4859 4860 4861
/* Used by function selftest to not test if filter is set */
bool ftrace_filter_param __initdata;

4862 4863
static int __init set_ftrace_notrace(char *str)
{
4864
	ftrace_filter_param = true;
4865
	strlcpy(ftrace_notrace_buf, str, FTRACE_FILTER_SIZE);
4866 4867 4868 4869 4870 4871
	return 1;
}
__setup("ftrace_notrace=", set_ftrace_notrace);

static int __init set_ftrace_filter(char *str)
{
4872
	ftrace_filter_param = true;
4873
	strlcpy(ftrace_filter_buf, str, FTRACE_FILTER_SIZE);
4874 4875 4876 4877
	return 1;
}
__setup("ftrace_filter=", set_ftrace_filter);

4878
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
4879
static char ftrace_graph_buf[FTRACE_FILTER_SIZE] __initdata;
4880
static char ftrace_graph_notrace_buf[FTRACE_FILTER_SIZE] __initdata;
4881
static int ftrace_graph_set_hash(struct ftrace_hash *hash, char *buffer);
4882

4883 4884
static int __init set_graph_function(char *str)
{
4885
	strlcpy(ftrace_graph_buf, str, FTRACE_FILTER_SIZE);
4886 4887 4888 4889
	return 1;
}
__setup("ftrace_graph_filter=", set_graph_function);

4890 4891 4892 4893 4894 4895 4896
static int __init set_graph_notrace_function(char *str)
{
	strlcpy(ftrace_graph_notrace_buf, str, FTRACE_FILTER_SIZE);
	return 1;
}
__setup("ftrace_graph_notrace=", set_graph_notrace_function);

4897 4898 4899 4900 4901 4902 4903 4904
static int __init set_graph_max_depth_function(char *str)
{
	if (!str)
		return 0;
	fgraph_max_depth = simple_strtoul(str, NULL, 0);
	return 1;
}
__setup("ftrace_graph_max_depth=", set_graph_max_depth_function);
4905 4906

static void __init set_ftrace_early_graph(char *buf, int enable)
4907 4908 4909
{
	int ret;
	char *func;
4910
	struct ftrace_hash *hash;
4911

4912 4913 4914
	hash = alloc_ftrace_hash(FTRACE_HASH_DEFAULT_BITS);
	if (WARN_ON(!hash))
		return;
4915 4916 4917 4918

	while (buf) {
		func = strsep(&buf, ",");
		/* we allow only one expression at a time */
4919
		ret = ftrace_graph_set_hash(hash, func);
4920 4921 4922 4923
		if (ret)
			printk(KERN_DEBUG "ftrace: function %s not "
					  "traceable\n", func);
	}
4924 4925 4926 4927 4928

	if (enable)
		ftrace_graph_hash = hash;
	else
		ftrace_graph_notrace_hash = hash;
4929 4930 4931
}
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */

4932 4933
void __init
ftrace_set_early_filter(struct ftrace_ops *ops, char *buf, int enable)
4934 4935 4936
{
	char *func;

4937 4938
	ftrace_ops_init(ops);

4939 4940
	while (buf) {
		func = strsep(&buf, ",");
4941
		ftrace_set_regex(ops, func, strlen(func), 0, enable);
4942 4943 4944 4945 4946 4947
	}
}

static void __init set_ftrace_early_filters(void)
{
	if (ftrace_filter_buf[0])
4948
		ftrace_set_early_filter(&global_ops, ftrace_filter_buf, 1);
4949
	if (ftrace_notrace_buf[0])
4950
		ftrace_set_early_filter(&global_ops, ftrace_notrace_buf, 0);
4951 4952
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
	if (ftrace_graph_buf[0])
4953 4954 4955
		set_ftrace_early_graph(ftrace_graph_buf, 1);
	if (ftrace_graph_notrace_buf[0])
		set_ftrace_early_graph(ftrace_graph_notrace_buf, 0);
4956
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
4957 4958
}

4959
int ftrace_regex_release(struct inode *inode, struct file *file)
4960 4961 4962
{
	struct seq_file *m = (struct seq_file *)file->private_data;
	struct ftrace_iterator *iter;
4963
	struct ftrace_hash **orig_hash;
4964
	struct trace_parser *parser;
4965
	int filter_hash;
4966
	int ret;
4967 4968 4969 4970 4971 4972 4973

	if (file->f_mode & FMODE_READ) {
		iter = m->private;
		seq_release(inode, file);
	} else
		iter = file->private_data;

4974 4975
	parser = &iter->parser;
	if (trace_parser_loaded(parser)) {
4976
		ftrace_match_records(iter->hash, parser->buffer, parser->idx);
4977 4978
	}

4979 4980
	trace_parser_put(parser);

4981
	mutex_lock(&iter->ops->func_hash->regex_lock);
4982

4983
	if (file->f_mode & FMODE_WRITE) {
4984 4985
		filter_hash = !!(iter->flags & FTRACE_ITER_FILTER);

4986
		if (filter_hash) {
4987
			orig_hash = &iter->ops->func_hash->filter_hash;
4988
			if (iter->tr && !list_empty(&iter->tr->mod_trace))
4989 4990
				iter->hash->flags |= FTRACE_HASH_FL_MOD;
		} else
4991
			orig_hash = &iter->ops->func_hash->notrace_hash;
4992

4993
		mutex_lock(&ftrace_lock);
4994 4995
		ret = ftrace_hash_move_and_update_ops(iter->ops, orig_hash,
						      iter->hash, filter_hash);
4996
		mutex_unlock(&ftrace_lock);
4997 4998 4999
	} else {
		/* For read only, the hash is the ops hash */
		iter->hash = NULL;
5000
	}
5001

5002
	mutex_unlock(&iter->ops->func_hash->regex_lock);
5003 5004
	free_ftrace_hash(iter->hash);
	kfree(iter);
5005

5006 5007 5008
	return 0;
}

5009
static const struct file_operations ftrace_avail_fops = {
5010 5011 5012
	.open = ftrace_avail_open,
	.read = seq_read,
	.llseek = seq_lseek,
L
Li Zefan 已提交
5013
	.release = seq_release_private,
5014 5015
};

5016 5017 5018 5019 5020 5021 5022
static const struct file_operations ftrace_enabled_fops = {
	.open = ftrace_enabled_open,
	.read = seq_read,
	.llseek = seq_lseek,
	.release = seq_release_private,
};

5023
static const struct file_operations ftrace_filter_fops = {
5024
	.open = ftrace_filter_open,
L
Lai Jiangshan 已提交
5025
	.read = seq_read,
5026
	.write = ftrace_filter_write,
5027
	.llseek = tracing_lseek,
5028
	.release = ftrace_regex_release,
5029 5030
};

5031
static const struct file_operations ftrace_notrace_fops = {
5032
	.open = ftrace_notrace_open,
L
Lai Jiangshan 已提交
5033
	.read = seq_read,
5034
	.write = ftrace_notrace_write,
5035
	.llseek = tracing_lseek,
5036
	.release = ftrace_regex_release,
5037 5038
};

5039 5040 5041 5042
#ifdef CONFIG_FUNCTION_GRAPH_TRACER

static DEFINE_MUTEX(graph_lock);

5043 5044 5045 5046 5047 5048 5049
struct ftrace_hash *ftrace_graph_hash = EMPTY_HASH;
struct ftrace_hash *ftrace_graph_notrace_hash = EMPTY_HASH;

enum graph_filter_type {
	GRAPH_FILTER_NOTRACE	= 0,
	GRAPH_FILTER_FUNCTION,
};
5050

5051 5052
#define FTRACE_GRAPH_EMPTY	((void *)1)

5053
struct ftrace_graph_data {
5054 5055 5056 5057 5058 5059 5060
	struct ftrace_hash		*hash;
	struct ftrace_func_entry	*entry;
	int				idx;   /* for hash table iteration */
	enum graph_filter_type		type;
	struct ftrace_hash		*new_hash;
	const struct seq_operations	*seq_ops;
	struct trace_parser		parser;
5061 5062
};

5063
static void *
5064
__g_next(struct seq_file *m, loff_t *pos)
5065
{
5066
	struct ftrace_graph_data *fgd = m->private;
5067 5068 5069
	struct ftrace_func_entry *entry = fgd->entry;
	struct hlist_head *head;
	int i, idx = fgd->idx;
5070

5071
	if (*pos >= fgd->hash->count)
5072
		return NULL;
5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091

	if (entry) {
		hlist_for_each_entry_continue(entry, hlist) {
			fgd->entry = entry;
			return entry;
		}

		idx++;
	}

	for (i = idx; i < 1 << fgd->hash->size_bits; i++) {
		head = &fgd->hash->buckets[i];
		hlist_for_each_entry(entry, head, hlist) {
			fgd->entry = entry;
			fgd->idx = i;
			return entry;
		}
	}
	return NULL;
5092
}
5093

5094 5095 5096 5097 5098
static void *
g_next(struct seq_file *m, void *v, loff_t *pos)
{
	(*pos)++;
	return __g_next(m, pos);
5099 5100 5101 5102
}

static void *g_start(struct seq_file *m, loff_t *pos)
{
5103 5104
	struct ftrace_graph_data *fgd = m->private;

5105 5106
	mutex_lock(&graph_lock);

5107 5108 5109 5110 5111 5112 5113
	if (fgd->type == GRAPH_FILTER_FUNCTION)
		fgd->hash = rcu_dereference_protected(ftrace_graph_hash,
					lockdep_is_held(&graph_lock));
	else
		fgd->hash = rcu_dereference_protected(ftrace_graph_notrace_hash,
					lockdep_is_held(&graph_lock));

5114
	/* Nothing, tell g_show to print all functions are enabled */
5115
	if (ftrace_hash_empty(fgd->hash) && !*pos)
5116
		return FTRACE_GRAPH_EMPTY;
5117

5118 5119
	fgd->idx = 0;
	fgd->entry = NULL;
5120
	return __g_next(m, pos);
5121 5122 5123 5124 5125 5126 5127 5128 5129
}

static void g_stop(struct seq_file *m, void *p)
{
	mutex_unlock(&graph_lock);
}

static int g_show(struct seq_file *m, void *v)
{
5130
	struct ftrace_func_entry *entry = v;
5131

5132
	if (!entry)
5133 5134
		return 0;

5135
	if (entry == FTRACE_GRAPH_EMPTY) {
5136 5137
		struct ftrace_graph_data *fgd = m->private;

5138
		if (fgd->type == GRAPH_FILTER_FUNCTION)
5139
			seq_puts(m, "#### all functions enabled ####\n");
5140
		else
5141
			seq_puts(m, "#### no functions disabled ####\n");
5142 5143 5144
		return 0;
	}

5145
	seq_printf(m, "%ps\n", (void *)entry->ip);
5146 5147 5148 5149

	return 0;
}

J
James Morris 已提交
5150
static const struct seq_operations ftrace_graph_seq_ops = {
5151 5152 5153 5154 5155 5156 5157
	.start = g_start,
	.next = g_next,
	.stop = g_stop,
	.show = g_show,
};

static int
5158 5159
__ftrace_graph_open(struct inode *inode, struct file *file,
		    struct ftrace_graph_data *fgd)
5160 5161
{
	int ret = 0;
5162
	struct ftrace_hash *new_hash = NULL;
5163

5164 5165 5166
	if (file->f_mode & FMODE_WRITE) {
		const int size_bits = FTRACE_HASH_DEFAULT_BITS;

5167 5168 5169
		if (trace_parser_get_init(&fgd->parser, FTRACE_BUFF_MAX))
			return -ENOMEM;

5170 5171 5172 5173 5174 5175 5176 5177 5178
		if (file->f_flags & O_TRUNC)
			new_hash = alloc_ftrace_hash(size_bits);
		else
			new_hash = alloc_and_copy_ftrace_hash(size_bits,
							      fgd->hash);
		if (!new_hash) {
			ret = -ENOMEM;
			goto out;
		}
5179 5180
	}

5181
	if (file->f_mode & FMODE_READ) {
5182
		ret = seq_open(file, &ftrace_graph_seq_ops);
5183 5184 5185
		if (!ret) {
			struct seq_file *m = file->private_data;
			m->private = fgd;
5186 5187 5188 5189
		} else {
			/* Failed */
			free_ftrace_hash(new_hash);
			new_hash = NULL;
5190 5191 5192
		}
	} else
		file->private_data = fgd;
5193

5194
out:
5195 5196 5197
	if (ret < 0 && file->f_mode & FMODE_WRITE)
		trace_parser_put(&fgd->parser);

5198
	fgd->new_hash = new_hash;
5199 5200 5201 5202 5203 5204 5205 5206

	/*
	 * All uses of fgd->hash must be taken with the graph_lock
	 * held. The graph_lock is going to be released, so force
	 * fgd->hash to be reinitialized when it is taken again.
	 */
	fgd->hash = NULL;

5207 5208 5209
	return ret;
}

5210 5211 5212 5213
static int
ftrace_graph_open(struct inode *inode, struct file *file)
{
	struct ftrace_graph_data *fgd;
5214
	int ret;
5215 5216 5217 5218 5219 5220 5221 5222

	if (unlikely(ftrace_disabled))
		return -ENODEV;

	fgd = kmalloc(sizeof(*fgd), GFP_KERNEL);
	if (fgd == NULL)
		return -ENOMEM;

5223 5224
	mutex_lock(&graph_lock);

5225 5226
	fgd->hash = rcu_dereference_protected(ftrace_graph_hash,
					lockdep_is_held(&graph_lock));
5227
	fgd->type = GRAPH_FILTER_FUNCTION;
5228 5229
	fgd->seq_ops = &ftrace_graph_seq_ops;

5230 5231 5232 5233 5234 5235
	ret = __ftrace_graph_open(inode, file, fgd);
	if (ret < 0)
		kfree(fgd);

	mutex_unlock(&graph_lock);
	return ret;
5236 5237
}

5238 5239 5240 5241
static int
ftrace_graph_notrace_open(struct inode *inode, struct file *file)
{
	struct ftrace_graph_data *fgd;
5242
	int ret;
5243 5244 5245 5246 5247 5248 5249 5250

	if (unlikely(ftrace_disabled))
		return -ENODEV;

	fgd = kmalloc(sizeof(*fgd), GFP_KERNEL);
	if (fgd == NULL)
		return -ENOMEM;

5251 5252
	mutex_lock(&graph_lock);

5253 5254
	fgd->hash = rcu_dereference_protected(ftrace_graph_notrace_hash,
					lockdep_is_held(&graph_lock));
5255
	fgd->type = GRAPH_FILTER_NOTRACE;
5256 5257
	fgd->seq_ops = &ftrace_graph_seq_ops;

5258 5259 5260 5261 5262 5263
	ret = __ftrace_graph_open(inode, file, fgd);
	if (ret < 0)
		kfree(fgd);

	mutex_unlock(&graph_lock);
	return ret;
5264 5265
}

5266 5267 5268
static int
ftrace_graph_release(struct inode *inode, struct file *file)
{
5269
	struct ftrace_graph_data *fgd;
5270 5271 5272
	struct ftrace_hash *old_hash, *new_hash;
	struct trace_parser *parser;
	int ret = 0;
5273

5274 5275 5276
	if (file->f_mode & FMODE_READ) {
		struct seq_file *m = file->private_data;

5277
		fgd = m->private;
5278
		seq_release(inode, file);
5279
	} else {
5280
		fgd = file->private_data;
5281 5282
	}

5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321

	if (file->f_mode & FMODE_WRITE) {

		parser = &fgd->parser;

		if (trace_parser_loaded((parser))) {
			ret = ftrace_graph_set_hash(fgd->new_hash,
						    parser->buffer);
		}

		trace_parser_put(parser);

		new_hash = __ftrace_hash_move(fgd->new_hash);
		if (!new_hash) {
			ret = -ENOMEM;
			goto out;
		}

		mutex_lock(&graph_lock);

		if (fgd->type == GRAPH_FILTER_FUNCTION) {
			old_hash = rcu_dereference_protected(ftrace_graph_hash,
					lockdep_is_held(&graph_lock));
			rcu_assign_pointer(ftrace_graph_hash, new_hash);
		} else {
			old_hash = rcu_dereference_protected(ftrace_graph_notrace_hash,
					lockdep_is_held(&graph_lock));
			rcu_assign_pointer(ftrace_graph_notrace_hash, new_hash);
		}

		mutex_unlock(&graph_lock);

		/* Wait till all users are no longer using the old hash */
		synchronize_sched();

		free_ftrace_hash(old_hash);
	}

 out:
5322
	free_ftrace_hash(fgd->new_hash);
5323 5324
	kfree(fgd);

5325
	return ret;
5326 5327
}

5328
static int
5329
ftrace_graph_set_hash(struct ftrace_hash *hash, char *buffer)
5330
{
5331
	struct ftrace_glob func_g;
5332 5333
	struct dyn_ftrace *rec;
	struct ftrace_page *pg;
5334
	struct ftrace_func_entry *entry;
5335
	int fail = 1;
5336
	int not;
5337

5338
	/* decode regex */
5339 5340
	func_g.type = filter_parse_regex(buffer, strlen(buffer),
					 &func_g.search, &not);
5341

5342
	func_g.len = strlen(func_g.search);
5343

5344
	mutex_lock(&ftrace_lock);
5345 5346 5347 5348 5349 5350

	if (unlikely(ftrace_disabled)) {
		mutex_unlock(&ftrace_lock);
		return -ENODEV;
	}

5351 5352
	do_for_each_ftrace_rec(pg, rec) {

5353 5354 5355
		if (rec->flags & FTRACE_FL_DISABLED)
			continue;

D
Dmitry Safonov 已提交
5356
		if (ftrace_match_record(rec, &func_g, NULL, 0)) {
5357
			entry = ftrace_lookup_ip(hash, rec->ip);
5358 5359 5360

			if (!not) {
				fail = 0;
5361 5362 5363 5364 5365

				if (entry)
					continue;
				if (add_hash_entry(hash, rec->ip) < 0)
					goto out;
5366
			} else {
5367 5368
				if (entry) {
					free_hash_entry(hash, entry);
5369 5370 5371
					fail = 0;
				}
			}
5372
		}
5373
	} while_for_each_ftrace_rec();
5374
out:
5375
	mutex_unlock(&ftrace_lock);
5376

5377 5378 5379 5380
	if (fail)
		return -EINVAL;

	return 0;
5381 5382 5383 5384 5385 5386
}

static ssize_t
ftrace_graph_write(struct file *file, const char __user *ubuf,
		   size_t cnt, loff_t *ppos)
{
5387
	ssize_t read, ret = 0;
5388
	struct ftrace_graph_data *fgd = file->private_data;
5389
	struct trace_parser *parser;
5390

5391
	if (!cnt)
5392 5393
		return 0;

5394 5395 5396 5397 5398 5399
	/* Read mode uses seq functions */
	if (file->f_mode & FMODE_READ) {
		struct seq_file *m = file->private_data;
		fgd = m->private;
	}

5400
	parser = &fgd->parser;
5401

5402
	read = trace_get_user(parser, ubuf, cnt, ppos);
5403

5404 5405
	if (read >= 0 && trace_parser_loaded(parser) &&
	    !trace_parser_cont(parser)) {
5406

5407
		ret = ftrace_graph_set_hash(fgd->new_hash,
5408 5409
					    parser->buffer);
		trace_parser_clear(parser);
5410 5411
	}

5412 5413
	if (!ret)
		ret = read;
5414

5415 5416 5417 5418
	return ret;
}

static const struct file_operations ftrace_graph_fops = {
5419 5420 5421
	.open		= ftrace_graph_open,
	.read		= seq_read,
	.write		= ftrace_graph_write,
5422
	.llseek		= tracing_lseek,
5423
	.release	= ftrace_graph_release,
5424
};
5425 5426 5427 5428 5429

static const struct file_operations ftrace_graph_notrace_fops = {
	.open		= ftrace_graph_notrace_open,
	.read		= seq_read,
	.write		= ftrace_graph_write,
5430
	.llseek		= tracing_lseek,
5431 5432
	.release	= ftrace_graph_release,
};
5433 5434
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */

5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464
void ftrace_create_filter_files(struct ftrace_ops *ops,
				struct dentry *parent)
{

	trace_create_file("set_ftrace_filter", 0644, parent,
			  ops, &ftrace_filter_fops);

	trace_create_file("set_ftrace_notrace", 0644, parent,
			  ops, &ftrace_notrace_fops);
}

/*
 * The name "destroy_filter_files" is really a misnomer. Although
 * in the future, it may actualy delete the files, but this is
 * really intended to make sure the ops passed in are disabled
 * and that when this function returns, the caller is free to
 * free the ops.
 *
 * The "destroy" name is only to match the "create" name that this
 * should be paired with.
 */
void ftrace_destroy_filter_files(struct ftrace_ops *ops)
{
	mutex_lock(&ftrace_lock);
	if (ops->flags & FTRACE_OPS_FL_ENABLED)
		ftrace_shutdown(ops, 0);
	ops->flags |= FTRACE_OPS_FL_DELETED;
	mutex_unlock(&ftrace_lock);
}

5465
static __init int ftrace_init_dyn_tracefs(struct dentry *d_tracer)
5466 5467
{

5468 5469
	trace_create_file("available_filter_functions", 0444,
			d_tracer, NULL, &ftrace_avail_fops);
5470

5471 5472 5473
	trace_create_file("enabled_functions", 0444,
			d_tracer, NULL, &ftrace_enabled_fops);

5474
	ftrace_create_filter_files(&global_ops, d_tracer);
5475

5476
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
5477
	trace_create_file("set_graph_function", 0644, d_tracer,
5478 5479
				    NULL,
				    &ftrace_graph_fops);
5480
	trace_create_file("set_graph_notrace", 0644, d_tracer,
5481 5482
				    NULL,
				    &ftrace_graph_notrace_fops);
5483 5484
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */

5485 5486 5487
	return 0;
}

5488
static int ftrace_cmp_ips(const void *a, const void *b)
5489
{
5490 5491
	const unsigned long *ipa = a;
	const unsigned long *ipb = b;
5492

5493 5494 5495 5496 5497 5498 5499
	if (*ipa > *ipb)
		return 1;
	if (*ipa < *ipb)
		return -1;
	return 0;
}

5500
static int ftrace_process_locs(struct module *mod,
5501
			       unsigned long *start,
5502 5503
			       unsigned long *end)
{
5504
	struct ftrace_page *start_pg;
5505
	struct ftrace_page *pg;
5506
	struct dyn_ftrace *rec;
5507
	unsigned long count;
5508 5509
	unsigned long *p;
	unsigned long addr;
5510
	unsigned long flags = 0; /* Shut up gcc */
5511 5512 5513 5514 5515 5516 5517
	int ret = -ENOMEM;

	count = end - start;

	if (!count)
		return 0;

5518
	sort(start, count, sizeof(*start),
5519
	     ftrace_cmp_ips, NULL);
5520

5521 5522
	start_pg = ftrace_allocate_pages(count);
	if (!start_pg)
5523
		return -ENOMEM;
5524

S
Steven Rostedt 已提交
5525
	mutex_lock(&ftrace_lock);
5526

5527 5528 5529 5530 5531
	/*
	 * Core and each module needs their own pages, as
	 * modules will free them when they are removed.
	 * Force a new page to be allocated for modules.
	 */
5532 5533 5534
	if (!mod) {
		WARN_ON(ftrace_pages || ftrace_pages_start);
		/* First initialization */
5535
		ftrace_pages = ftrace_pages_start = start_pg;
5536
	} else {
5537
		if (!ftrace_pages)
5538
			goto out;
5539

5540 5541 5542 5543
		if (WARN_ON(ftrace_pages->next)) {
			/* Hmm, we have free pages? */
			while (ftrace_pages->next)
				ftrace_pages = ftrace_pages->next;
5544
		}
5545

5546
		ftrace_pages->next = start_pg;
5547 5548
	}

5549
	p = start;
5550
	pg = start_pg;
5551 5552
	while (p < end) {
		addr = ftrace_call_adjust(*p++);
5553 5554 5555 5556 5557 5558 5559 5560
		/*
		 * Some architecture linkers will pad between
		 * the different mcount_loc sections of different
		 * object files to satisfy alignments.
		 * Skip any NULL pointers.
		 */
		if (!addr)
			continue;
5561 5562 5563 5564 5565 5566 5567 5568 5569 5570

		if (pg->index == pg->size) {
			/* We should have allocated enough */
			if (WARN_ON(!pg->next))
				break;
			pg = pg->next;
		}

		rec = &pg->records[pg->index++];
		rec->ip = addr;
5571 5572
	}

5573 5574 5575 5576 5577 5578
	/* We should have used all pages */
	WARN_ON(pg->next);

	/* Assign the last page to ftrace_pages */
	ftrace_pages = pg;

5579
	/*
5580 5581 5582 5583 5584 5585
	 * We only need to disable interrupts on start up
	 * because we are modifying code that an interrupt
	 * may execute, and the modification is not atomic.
	 * But for modules, nothing runs the code we modify
	 * until we are finished with it, and there's no
	 * reason to cause large interrupt latencies while we do it.
5586
	 */
5587 5588
	if (!mod)
		local_irq_save(flags);
5589
	ftrace_update_code(mod, start_pg);
5590 5591
	if (!mod)
		local_irq_restore(flags);
5592 5593
	ret = 0;
 out:
S
Steven Rostedt 已提交
5594
	mutex_unlock(&ftrace_lock);
5595

5596
	return ret;
5597 5598
}

5599 5600 5601 5602 5603 5604 5605 5606
struct ftrace_mod_func {
	struct list_head	list;
	char			*name;
	unsigned long		ip;
	unsigned int		size;
};

struct ftrace_mod_map {
5607
	struct rcu_head		rcu;
5608 5609 5610 5611 5612
	struct list_head	list;
	struct module		*mod;
	unsigned long		start_addr;
	unsigned long		end_addr;
	struct list_head	funcs;
5613
	unsigned int		num_funcs;
5614 5615
};

5616
#ifdef CONFIG_MODULES
5617 5618 5619

#define next_to_ftrace_page(p) container_of(p, struct ftrace_page, next)

5620 5621
static LIST_HEAD(ftrace_mod_maps);

5622 5623 5624 5625 5626 5627 5628 5629 5630 5631 5632 5633 5634
static int referenced_filters(struct dyn_ftrace *rec)
{
	struct ftrace_ops *ops;
	int cnt = 0;

	for (ops = ftrace_ops_list; ops != &ftrace_list_end; ops = ops->next) {
		if (ops_references_rec(ops, rec))
		    cnt++;
	}

	return cnt;
}

5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665 5666 5667 5668 5669 5670 5671 5672 5673 5674
static void
clear_mod_from_hash(struct ftrace_page *pg, struct ftrace_hash *hash)
{
	struct ftrace_func_entry *entry;
	struct dyn_ftrace *rec;
	int i;

	if (ftrace_hash_empty(hash))
		return;

	for (i = 0; i < pg->index; i++) {
		rec = &pg->records[i];
		entry = __ftrace_lookup_ip(hash, rec->ip);
		/*
		 * Do not allow this rec to match again.
		 * Yeah, it may waste some memory, but will be removed
		 * if/when the hash is modified again.
		 */
		if (entry)
			entry->ip = 0;
	}
}

/* Clear any records from hashs */
static void clear_mod_from_hashes(struct ftrace_page *pg)
{
	struct trace_array *tr;

	mutex_lock(&trace_types_lock);
	list_for_each_entry(tr, &ftrace_trace_arrays, list) {
		if (!tr->ops || !tr->ops->func_hash)
			continue;
		mutex_lock(&tr->ops->func_hash->regex_lock);
		clear_mod_from_hash(pg, tr->ops->func_hash->filter_hash);
		clear_mod_from_hash(pg, tr->ops->func_hash->notrace_hash);
		mutex_unlock(&tr->ops->func_hash->regex_lock);
	}
	mutex_unlock(&trace_types_lock);
}

5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690
static void ftrace_free_mod_map(struct rcu_head *rcu)
{
	struct ftrace_mod_map *mod_map = container_of(rcu, struct ftrace_mod_map, rcu);
	struct ftrace_mod_func *mod_func;
	struct ftrace_mod_func *n;

	/* All the contents of mod_map are now not visible to readers */
	list_for_each_entry_safe(mod_func, n, &mod_map->funcs, list) {
		kfree(mod_func->name);
		list_del(&mod_func->list);
		kfree(mod_func);
	}

	kfree(mod_map);
}

5691
void ftrace_release_mod(struct module *mod)
5692
{
5693 5694
	struct ftrace_mod_map *mod_map;
	struct ftrace_mod_map *n;
5695
	struct dyn_ftrace *rec;
5696
	struct ftrace_page **last_pg;
5697
	struct ftrace_page *tmp_page = NULL;
5698
	struct ftrace_page *pg;
5699
	int order;
5700

5701 5702
	mutex_lock(&ftrace_lock);

5703
	if (ftrace_disabled)
5704
		goto out_unlock;
5705

5706 5707 5708 5709 5710 5711 5712 5713
	list_for_each_entry_safe(mod_map, n, &ftrace_mod_maps, list) {
		if (mod_map->mod == mod) {
			list_del_rcu(&mod_map->list);
			call_rcu_sched(&mod_map->rcu, ftrace_free_mod_map);
			break;
		}
	}

5714 5715 5716 5717 5718 5719 5720
	/*
	 * Each module has its own ftrace_pages, remove
	 * them from the list.
	 */
	last_pg = &ftrace_pages_start;
	for (pg = ftrace_pages_start; pg; pg = *last_pg) {
		rec = &pg->records[0];
5721 5722
		if (within_module_core(rec->ip, mod) ||
		    within_module_init(rec->ip, mod)) {
5723
			/*
5724 5725
			 * As core pages are first, the first
			 * page should never be a module page.
5726
			 */
5727 5728 5729 5730 5731 5732 5733
			if (WARN_ON(pg == ftrace_pages_start))
				goto out_unlock;

			/* Check if we are deleting the last page */
			if (pg == ftrace_pages)
				ftrace_pages = next_to_ftrace_page(last_pg);

5734
			ftrace_update_tot_cnt -= pg->index;
5735
			*last_pg = pg->next;
5736 5737 5738

			pg->next = tmp_page;
			tmp_page = pg;
5739 5740 5741
		} else
			last_pg = &pg->next;
	}
5742
 out_unlock:
5743
	mutex_unlock(&ftrace_lock);
5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754

	for (pg = tmp_page; pg; pg = tmp_page) {

		/* Needs to be called outside of ftrace_lock */
		clear_mod_from_hashes(pg);

		order = get_count_order(pg->size / ENTRIES_PER_PAGE);
		free_pages((unsigned long)pg->records, order);
		tmp_page = pg->next;
		kfree(pg);
	}
5755 5756
}

5757
void ftrace_module_enable(struct module *mod)
5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790
{
	struct dyn_ftrace *rec;
	struct ftrace_page *pg;

	mutex_lock(&ftrace_lock);

	if (ftrace_disabled)
		goto out_unlock;

	/*
	 * If the tracing is enabled, go ahead and enable the record.
	 *
	 * The reason not to enable the record immediatelly is the
	 * inherent check of ftrace_make_nop/ftrace_make_call for
	 * correct previous instructions.  Making first the NOP
	 * conversion puts the module to the correct state, thus
	 * passing the ftrace_make_call check.
	 *
	 * We also delay this to after the module code already set the
	 * text to read-only, as we now need to set it back to read-write
	 * so that we can modify the text.
	 */
	if (ftrace_start_up)
		ftrace_arch_code_modify_prepare();

	do_for_each_ftrace_rec(pg, rec) {
		int cnt;
		/*
		 * do_for_each_ftrace_rec() is a double loop.
		 * module text shares the pg. If a record is
		 * not part of this module, then skip this pg,
		 * which the "break" will do.
		 */
5791 5792
		if (!within_module_core(rec->ip, mod) &&
		    !within_module_init(rec->ip, mod))
5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819 5820 5821 5822 5823 5824
			break;

		cnt = 0;

		/*
		 * When adding a module, we need to check if tracers are
		 * currently enabled and if they are, and can trace this record,
		 * we need to enable the module functions as well as update the
		 * reference counts for those function records.
		 */
		if (ftrace_start_up)
			cnt += referenced_filters(rec);

		/* This clears FTRACE_FL_DISABLED */
		rec->flags = cnt;

		if (ftrace_start_up && cnt) {
			int failed = __ftrace_replace_code(rec, 1);
			if (failed) {
				ftrace_bug(failed, rec);
				goto out_loop;
			}
		}

	} while_for_each_ftrace_rec();

 out_loop:
	if (ftrace_start_up)
		ftrace_arch_code_modify_post_process();

 out_unlock:
	mutex_unlock(&ftrace_lock);
5825 5826

	process_cached_mods(mod->name);
5827 5828
}

5829
void ftrace_module_init(struct module *mod)
5830
{
5831
	if (ftrace_disabled || !mod->num_ftrace_callsites)
5832
		return;
5833

5834 5835
	ftrace_process_locs(mod, mod->ftrace_callsites,
			    mod->ftrace_callsites + mod->num_ftrace_callsites);
5836
}
5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864

static void save_ftrace_mod_rec(struct ftrace_mod_map *mod_map,
				struct dyn_ftrace *rec)
{
	struct ftrace_mod_func *mod_func;
	unsigned long symsize;
	unsigned long offset;
	char str[KSYM_SYMBOL_LEN];
	char *modname;
	const char *ret;

	ret = kallsyms_lookup(rec->ip, &symsize, &offset, &modname, str);
	if (!ret)
		return;

	mod_func = kmalloc(sizeof(*mod_func), GFP_KERNEL);
	if (!mod_func)
		return;

	mod_func->name = kstrdup(str, GFP_KERNEL);
	if (!mod_func->name) {
		kfree(mod_func);
		return;
	}

	mod_func->ip = rec->ip - offset;
	mod_func->size = symsize;

5865 5866
	mod_map->num_funcs++;

5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882
	list_add_rcu(&mod_func->list, &mod_map->funcs);
}

static struct ftrace_mod_map *
allocate_ftrace_mod_map(struct module *mod,
			unsigned long start, unsigned long end)
{
	struct ftrace_mod_map *mod_map;

	mod_map = kmalloc(sizeof(*mod_map), GFP_KERNEL);
	if (!mod_map)
		return NULL;

	mod_map->mod = mod;
	mod_map->start_addr = start;
	mod_map->end_addr = end;
5883
	mod_map->num_funcs = 0;
5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928

	INIT_LIST_HEAD_RCU(&mod_map->funcs);

	list_add_rcu(&mod_map->list, &ftrace_mod_maps);

	return mod_map;
}

static const char *
ftrace_func_address_lookup(struct ftrace_mod_map *mod_map,
			   unsigned long addr, unsigned long *size,
			   unsigned long *off, char *sym)
{
	struct ftrace_mod_func *found_func =  NULL;
	struct ftrace_mod_func *mod_func;

	list_for_each_entry_rcu(mod_func, &mod_map->funcs, list) {
		if (addr >= mod_func->ip &&
		    addr < mod_func->ip + mod_func->size) {
			found_func = mod_func;
			break;
		}
	}

	if (found_func) {
		if (size)
			*size = found_func->size;
		if (off)
			*off = addr - found_func->ip;
		if (sym)
			strlcpy(sym, found_func->name, KSYM_NAME_LEN);

		return found_func->name;
	}

	return NULL;
}

const char *
ftrace_mod_address_lookup(unsigned long addr, unsigned long *size,
		   unsigned long *off, char **modname, char *sym)
{
	struct ftrace_mod_map *mod_map;
	const char *ret = NULL;

5929
	/* mod_map is freed via call_rcu_sched() */
5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943
	preempt_disable();
	list_for_each_entry_rcu(mod_map, &ftrace_mod_maps, list) {
		ret = ftrace_func_address_lookup(mod_map, addr, size, off, sym);
		if (ret) {
			if (modname)
				*modname = mod_map->mod->name;
			break;
		}
	}
	preempt_enable();

	return ret;
}

5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979
int ftrace_mod_get_kallsym(unsigned int symnum, unsigned long *value,
			   char *type, char *name,
			   char *module_name, int *exported)
{
	struct ftrace_mod_map *mod_map;
	struct ftrace_mod_func *mod_func;

	preempt_disable();
	list_for_each_entry_rcu(mod_map, &ftrace_mod_maps, list) {

		if (symnum >= mod_map->num_funcs) {
			symnum -= mod_map->num_funcs;
			continue;
		}

		list_for_each_entry_rcu(mod_func, &mod_map->funcs, list) {
			if (symnum > 1) {
				symnum--;
				continue;
			}

			*value = mod_func->ip;
			*type = 'T';
			strlcpy(name, mod_func->name, KSYM_NAME_LEN);
			strlcpy(module_name, mod_map->mod->name, MODULE_NAME_LEN);
			*exported = 1;
			preempt_enable();
			return 0;
		}
		WARN_ON(1);
		break;
	}
	preempt_enable();
	return -ERANGE;
}

5980 5981 5982 5983 5984 5985 5986 5987 5988
#else
static void save_ftrace_mod_rec(struct ftrace_mod_map *mod_map,
				struct dyn_ftrace *rec) { }
static inline struct ftrace_mod_map *
allocate_ftrace_mod_map(struct module *mod,
			unsigned long start, unsigned long end)
{
	return NULL;
}
5989 5990
#endif /* CONFIG_MODULES */

5991 5992 5993 5994 5995 5996 5997 5998
struct ftrace_init_func {
	struct list_head list;
	unsigned long ip;
};

/* Clear any init ips from hashes */
static void
clear_func_from_hash(struct ftrace_init_func *func, struct ftrace_hash *hash)
5999
{
6000 6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047
	struct ftrace_func_entry *entry;

	if (ftrace_hash_empty(hash))
		return;

	entry = __ftrace_lookup_ip(hash, func->ip);

	/*
	 * Do not allow this rec to match again.
	 * Yeah, it may waste some memory, but will be removed
	 * if/when the hash is modified again.
	 */
	if (entry)
		entry->ip = 0;
}

static void
clear_func_from_hashes(struct ftrace_init_func *func)
{
	struct trace_array *tr;

	mutex_lock(&trace_types_lock);
	list_for_each_entry(tr, &ftrace_trace_arrays, list) {
		if (!tr->ops || !tr->ops->func_hash)
			continue;
		mutex_lock(&tr->ops->func_hash->regex_lock);
		clear_func_from_hash(func, tr->ops->func_hash->filter_hash);
		clear_func_from_hash(func, tr->ops->func_hash->notrace_hash);
		mutex_unlock(&tr->ops->func_hash->regex_lock);
	}
	mutex_unlock(&trace_types_lock);
}

static void add_to_clear_hash_list(struct list_head *clear_list,
				   struct dyn_ftrace *rec)
{
	struct ftrace_init_func *func;

	func = kmalloc(sizeof(*func), GFP_KERNEL);
	if (!func) {
		WARN_ONCE(1, "alloc failure, ftrace filter could be stale\n");
		return;
	}

	func->ip = rec->ip;
	list_add(&func->list, clear_list);
}

6048
void ftrace_free_mem(struct module *mod, void *start_ptr, void *end_ptr)
6049
{
6050 6051
	unsigned long start = (unsigned long)(start_ptr);
	unsigned long end = (unsigned long)(end_ptr);
6052 6053 6054 6055
	struct ftrace_page **last_pg = &ftrace_pages_start;
	struct ftrace_page *pg;
	struct dyn_ftrace *rec;
	struct dyn_ftrace key;
6056
	struct ftrace_mod_map *mod_map = NULL;
6057 6058
	struct ftrace_init_func *func, *func_next;
	struct list_head clear_hash;
6059 6060
	int order;

6061 6062
	INIT_LIST_HEAD(&clear_hash);

6063 6064 6065 6066 6067
	key.ip = start;
	key.flags = end;	/* overload flags, as it is unsigned long */

	mutex_lock(&ftrace_lock);

6068 6069 6070 6071 6072 6073 6074 6075
	/*
	 * If we are freeing module init memory, then check if
	 * any tracer is active. If so, we need to save a mapping of
	 * the module functions being freed with the address.
	 */
	if (mod && ftrace_ops_list != &ftrace_list_end)
		mod_map = allocate_ftrace_mod_map(mod, start, end);

6076 6077 6078 6079 6080 6081 6082 6083 6084 6085
	for (pg = ftrace_pages_start; pg; last_pg = &pg->next, pg = *last_pg) {
		if (end < pg->records[0].ip ||
		    start >= (pg->records[pg->index - 1].ip + MCOUNT_INSN_SIZE))
			continue;
 again:
		rec = bsearch(&key, pg->records, pg->index,
			      sizeof(struct dyn_ftrace),
			      ftrace_cmp_recs);
		if (!rec)
			continue;
6086

6087 6088 6089
		/* rec will be cleared from hashes after ftrace_lock unlock */
		add_to_clear_hash_list(&clear_hash, rec);

6090 6091 6092
		if (mod_map)
			save_ftrace_mod_rec(mod_map, rec);

6093
		pg->index--;
6094
		ftrace_update_tot_cnt--;
6095 6096 6097 6098 6099 6100 6101 6102 6103 6104 6105 6106 6107 6108 6109 6110
		if (!pg->index) {
			*last_pg = pg->next;
			order = get_count_order(pg->size / ENTRIES_PER_PAGE);
			free_pages((unsigned long)pg->records, order);
			kfree(pg);
			pg = container_of(last_pg, struct ftrace_page, next);
			if (!(*last_pg))
				ftrace_pages = pg;
			continue;
		}
		memmove(rec, rec + 1,
			(pg->index - (rec - pg->records)) * sizeof(*rec));
		/* More than one function may be in this block */
		goto again;
	}
	mutex_unlock(&ftrace_lock);
6111 6112 6113 6114 6115

	list_for_each_entry_safe(func, func_next, &clear_hash, list) {
		clear_func_from_hashes(func);
		kfree(func);
	}
6116 6117
}

6118 6119 6120 6121 6122
void __init ftrace_free_init_mem(void)
{
	void *start = (void *)(&__init_begin);
	void *end = (void *)(&__init_end);

6123
	ftrace_free_mem(NULL, start, end);
6124 6125
}

6126 6127
void __init ftrace_init(void)
{
6128 6129
	extern unsigned long __start_mcount_loc[];
	extern unsigned long __stop_mcount_loc[];
6130
	unsigned long count, flags;
6131 6132 6133
	int ret;

	local_irq_save(flags);
6134
	ret = ftrace_dyn_arch_init();
6135
	local_irq_restore(flags);
6136
	if (ret)
6137 6138 6139
		goto failed;

	count = __stop_mcount_loc - __start_mcount_loc;
6140 6141
	if (!count) {
		pr_info("ftrace: No functions to be traced?\n");
6142
		goto failed;
6143 6144 6145 6146
	}

	pr_info("ftrace: allocating %ld entries in %ld pages\n",
		count, count / ENTRIES_PER_PAGE + 1);
6147 6148 6149

	last_ftrace_enabled = ftrace_enabled = 1;

6150
	ret = ftrace_process_locs(NULL,
6151
				  __start_mcount_loc,
6152 6153
				  __stop_mcount_loc);

6154 6155
	set_ftrace_early_filters();

6156 6157 6158 6159 6160
	return;
 failed:
	ftrace_disabled = 1;
}

6161 6162 6163 6164 6165 6166 6167 6168 6169 6170
/* Do nothing if arch does not support this */
void __weak arch_ftrace_update_trampoline(struct ftrace_ops *ops)
{
}

static void ftrace_update_trampoline(struct ftrace_ops *ops)
{
	arch_ftrace_update_trampoline(ops);
}

6171 6172 6173
void ftrace_init_trace_array(struct trace_array *tr)
{
	INIT_LIST_HEAD(&tr->func_probes);
6174 6175
	INIT_LIST_HEAD(&tr->mod_trace);
	INIT_LIST_HEAD(&tr->mod_notrace);
6176
}
6177
#else
6178

6179
static struct ftrace_ops global_ops = {
6180
	.func			= ftrace_stub,
6181 6182 6183
	.flags			= FTRACE_OPS_FL_RECURSION_SAFE |
				  FTRACE_OPS_FL_INITIALIZED |
				  FTRACE_OPS_FL_PID,
6184 6185
};

6186 6187 6188 6189 6190
static int __init ftrace_nodyn_init(void)
{
	ftrace_enabled = 1;
	return 0;
}
6191
core_initcall(ftrace_nodyn_init);
6192

6193
static inline int ftrace_init_dyn_tracefs(struct dentry *d_tracer) { return 0; }
6194
static inline void ftrace_startup_enable(int command) { }
6195
static inline void ftrace_startup_all(int command) { }
6196
/* Keep as macros so we do not need to define the commands */
6197 6198 6199 6200 6201 6202
# define ftrace_startup(ops, command)					\
	({								\
		int ___ret = __register_ftrace_function(ops);		\
		if (!___ret)						\
			(ops)->flags |= FTRACE_OPS_FL_ENABLED;		\
		___ret;							\
6203
	})
6204 6205 6206 6207 6208 6209 6210
# define ftrace_shutdown(ops, command)					\
	({								\
		int ___ret = __unregister_ftrace_function(ops);		\
		if (!___ret)						\
			(ops)->flags &= ~FTRACE_OPS_FL_ENABLED;		\
		___ret;							\
	})
6211

I
Ingo Molnar 已提交
6212 6213
# define ftrace_startup_sysctl()	do { } while (0)
# define ftrace_shutdown_sysctl()	do { } while (0)
6214 6215

static inline int
6216
ftrace_ops_test(struct ftrace_ops *ops, unsigned long ip, void *regs)
6217 6218 6219 6220
{
	return 1;
}

6221 6222 6223 6224
static void ftrace_update_trampoline(struct ftrace_ops *ops)
{
}

6225 6226
#endif /* CONFIG_DYNAMIC_FTRACE */

6227 6228 6229 6230
__init void ftrace_init_global_array_ops(struct trace_array *tr)
{
	tr->ops = &global_ops;
	tr->ops->private = tr;
6231
	ftrace_init_trace_array(tr);
6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250
}

void ftrace_init_array_ops(struct trace_array *tr, ftrace_func_t func)
{
	/* If we filter on pids, update to use the pid function */
	if (tr->flags & TRACE_ARRAY_FL_GLOBAL) {
		if (WARN_ON(tr->ops->func != ftrace_stub))
			printk("ftrace ops had %pS for function\n",
			       tr->ops->func);
	}
	tr->ops->func = func;
	tr->ops->private = tr;
}

void ftrace_reset_array_ops(struct trace_array *tr)
{
	tr->ops->func = ftrace_stub;
}

6251 6252
static inline void
__ftrace_ops_list_func(unsigned long ip, unsigned long parent_ip,
6253
		       struct ftrace_ops *ignored, struct pt_regs *regs)
6254
{
6255
	struct ftrace_ops *op;
6256
	int bit;
6257

6258 6259 6260
	bit = trace_test_and_set_recursion(TRACE_LIST_START, TRACE_LIST_MAX);
	if (bit < 0)
		return;
6261

6262 6263 6264 6265 6266
	/*
	 * Some of the ops may be dynamically allocated,
	 * they must be freed after a synchronize_sched().
	 */
	preempt_disable_notrace();
6267

6268
	do_for_each_ftrace_op(op, ftrace_ops_list) {
6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279
		/*
		 * Check the following for each ops before calling their func:
		 *  if RCU flag is set, then rcu_is_watching() must be true
		 *  if PER_CPU is set, then ftrace_function_local_disable()
		 *                          must be false
		 *  Otherwise test if the ip matches the ops filter
		 *
		 * If any of the above fails then the op->func() is not executed.
		 */
		if ((!(op->flags & FTRACE_OPS_FL_RCU) || rcu_is_watching()) &&
		    ftrace_ops_test(op, ip, regs)) {
6280 6281
			if (FTRACE_WARN_ON(!op->func)) {
				pr_warn("op=%p %pS\n", op, op);
6282 6283
				goto out;
			}
6284
			op->func(ip, parent_ip, op, regs);
6285
		}
6286
	} while_for_each_ftrace_op(op);
6287
out:
6288
	preempt_enable_notrace();
6289
	trace_clear_recursion(bit);
6290 6291
}

6292 6293 6294 6295 6296
/*
 * Some archs only support passing ip and parent_ip. Even though
 * the list function ignores the op parameter, we do not want any
 * C side effects, where a function is called without the caller
 * sending a third parameter.
6297 6298 6299
 * Archs are to support both the regs and ftrace_ops at the same time.
 * If they support ftrace_ops, it is assumed they support regs.
 * If call backs want to use regs, they must either check for regs
6300 6301
 * being NULL, or CONFIG_DYNAMIC_FTRACE_WITH_REGS.
 * Note, CONFIG_DYNAMIC_FTRACE_WITH_REGS expects a full regs to be saved.
6302
 * An architecture can pass partial regs with ftrace_ops and still
L
Li Bin 已提交
6303
 * set the ARCH_SUPPORTS_FTRACE_OPS.
6304 6305 6306
 */
#if ARCH_SUPPORTS_FTRACE_OPS
static void ftrace_ops_list_func(unsigned long ip, unsigned long parent_ip,
6307
				 struct ftrace_ops *op, struct pt_regs *regs)
6308
{
6309
	__ftrace_ops_list_func(ip, parent_ip, NULL, regs);
6310 6311 6312 6313
}
#else
static void ftrace_ops_no_ops(unsigned long ip, unsigned long parent_ip)
{
6314
	__ftrace_ops_list_func(ip, parent_ip, NULL, NULL);
6315 6316 6317
}
#endif

6318 6319
/*
 * If there's only one function registered but it does not support
6320 6321
 * recursion, needs RCU protection and/or requires per cpu handling, then
 * this function will be called by the mcount trampoline.
6322
 */
6323
static void ftrace_ops_assist_func(unsigned long ip, unsigned long parent_ip,
6324 6325 6326 6327
				   struct ftrace_ops *op, struct pt_regs *regs)
{
	int bit;

6328 6329 6330
	if ((op->flags & FTRACE_OPS_FL_RCU) && !rcu_is_watching())
		return;

6331 6332 6333 6334
	bit = trace_test_and_set_recursion(TRACE_LIST_START, TRACE_LIST_MAX);
	if (bit < 0)
		return;

6335
	preempt_disable_notrace();
6336

6337
	op->func(ip, parent_ip, op, regs);
6338 6339

	preempt_enable_notrace();
6340 6341 6342
	trace_clear_recursion(bit);
}

6343 6344 6345 6346 6347 6348 6349
/**
 * ftrace_ops_get_func - get the function a trampoline should call
 * @ops: the ops to get the function for
 *
 * Normally the mcount trampoline will call the ops->func, but there
 * are times that it should not. For example, if the ops does not
 * have its own recursion protection, then it should call the
6350
 * ftrace_ops_assist_func() instead.
6351 6352 6353 6354 6355 6356
 *
 * Returns the function that the trampoline should call for @ops.
 */
ftrace_func_t ftrace_ops_get_func(struct ftrace_ops *ops)
{
	/*
6357 6358
	 * If the function does not handle recursion, needs to be RCU safe,
	 * or does per cpu logic, then we need to call the assist handler.
6359
	 */
6360
	if (!(ops->flags & FTRACE_OPS_FL_RECURSION_SAFE) ||
6361
	    ops->flags & FTRACE_OPS_FL_RCU)
6362
		return ftrace_ops_assist_func;
6363 6364 6365 6366

	return ops->func;
}

6367 6368 6369
static void
ftrace_filter_pid_sched_switch_probe(void *data, bool preempt,
		    struct task_struct *prev, struct task_struct *next)
S
Steven Rostedt 已提交
6370
{
6371 6372
	struct trace_array *tr = data;
	struct trace_pid_list *pid_list;
S
Steven Rostedt 已提交
6373

6374
	pid_list = rcu_dereference_sched(tr->function_pids);
6375

6376 6377
	this_cpu_write(tr->trace_buffer.data->ftrace_ignore_pid,
		       trace_ignore_this_task(pid_list, next));
S
Steven Rostedt 已提交
6378 6379
}

6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416
static void
ftrace_pid_follow_sched_process_fork(void *data,
				     struct task_struct *self,
				     struct task_struct *task)
{
	struct trace_pid_list *pid_list;
	struct trace_array *tr = data;

	pid_list = rcu_dereference_sched(tr->function_pids);
	trace_filter_add_remove_task(pid_list, self, task);
}

static void
ftrace_pid_follow_sched_process_exit(void *data, struct task_struct *task)
{
	struct trace_pid_list *pid_list;
	struct trace_array *tr = data;

	pid_list = rcu_dereference_sched(tr->function_pids);
	trace_filter_add_remove_task(pid_list, NULL, task);
}

void ftrace_pid_follow_fork(struct trace_array *tr, bool enable)
{
	if (enable) {
		register_trace_sched_process_fork(ftrace_pid_follow_sched_process_fork,
						  tr);
		register_trace_sched_process_exit(ftrace_pid_follow_sched_process_exit,
						  tr);
	} else {
		unregister_trace_sched_process_fork(ftrace_pid_follow_sched_process_fork,
						    tr);
		unregister_trace_sched_process_exit(ftrace_pid_follow_sched_process_exit,
						    tr);
	}
}

6417
static void clear_ftrace_pids(struct trace_array *tr)
6418
{
6419 6420
	struct trace_pid_list *pid_list;
	int cpu;
6421

6422 6423 6424 6425
	pid_list = rcu_dereference_protected(tr->function_pids,
					     lockdep_is_held(&ftrace_lock));
	if (!pid_list)
		return;
6426

6427
	unregister_trace_sched_switch(ftrace_filter_pid_sched_switch_probe, tr);
6428

6429 6430
	for_each_possible_cpu(cpu)
		per_cpu_ptr(tr->trace_buffer.data, cpu)->ftrace_ignore_pid = false;
S
Steven Rostedt 已提交
6431

6432
	rcu_assign_pointer(tr->function_pids, NULL);
S
Steven Rostedt 已提交
6433

6434 6435
	/* Wait till all users are no longer using pid filtering */
	synchronize_sched();
6436

6437
	trace_free_pid_list(pid_list);
6438 6439
}

6440 6441 6442 6443 6444 6445 6446 6447 6448
void ftrace_clear_pids(struct trace_array *tr)
{
	mutex_lock(&ftrace_lock);

	clear_ftrace_pids(tr);

	mutex_unlock(&ftrace_lock);
}

6449
static void ftrace_pid_reset(struct trace_array *tr)
6450
{
6451
	mutex_lock(&ftrace_lock);
6452
	clear_ftrace_pids(tr);
S
Steven Rostedt 已提交
6453

6454
	ftrace_update_pid_func();
6455
	ftrace_startup_all(0);
6456 6457 6458 6459

	mutex_unlock(&ftrace_lock);
}

6460 6461
/* Greater than any max PID */
#define FTRACE_NO_PIDS		(void *)(PID_MAX_LIMIT + 1)
6462

6463
static void *fpid_start(struct seq_file *m, loff_t *pos)
6464
	__acquires(RCU)
6465
{
6466 6467 6468
	struct trace_pid_list *pid_list;
	struct trace_array *tr = m->private;

6469
	mutex_lock(&ftrace_lock);
6470 6471 6472
	rcu_read_lock_sched();

	pid_list = rcu_dereference_sched(tr->function_pids);
6473

6474 6475
	if (!pid_list)
		return !(*pos) ? FTRACE_NO_PIDS : NULL;
6476

6477
	return trace_pid_start(pid_list, pos);
6478 6479 6480 6481
}

static void *fpid_next(struct seq_file *m, void *v, loff_t *pos)
{
6482 6483 6484 6485
	struct trace_array *tr = m->private;
	struct trace_pid_list *pid_list = rcu_dereference_sched(tr->function_pids);

	if (v == FTRACE_NO_PIDS)
6486 6487
		return NULL;

6488
	return trace_pid_next(pid_list, v, pos);
6489 6490 6491
}

static void fpid_stop(struct seq_file *m, void *p)
6492
	__releases(RCU)
6493
{
6494
	rcu_read_unlock_sched();
6495 6496 6497 6498 6499
	mutex_unlock(&ftrace_lock);
}

static int fpid_show(struct seq_file *m, void *v)
{
6500
	if (v == FTRACE_NO_PIDS) {
6501
		seq_puts(m, "no pid\n");
6502 6503 6504
		return 0;
	}

6505
	return trace_pid_show(m, v);
6506 6507 6508 6509 6510 6511 6512 6513 6514 6515 6516 6517
}

static const struct seq_operations ftrace_pid_sops = {
	.start = fpid_start,
	.next = fpid_next,
	.stop = fpid_stop,
	.show = fpid_show,
};

static int
ftrace_pid_open(struct inode *inode, struct file *file)
{
6518 6519
	struct trace_array *tr = inode->i_private;
	struct seq_file *m;
6520 6521
	int ret = 0;

6522 6523 6524
	if (trace_array_get(tr) < 0)
		return -ENODEV;

6525 6526
	if ((file->f_mode & FMODE_WRITE) &&
	    (file->f_flags & O_TRUNC))
6527
		ftrace_pid_reset(tr);
6528

6529 6530 6531 6532 6533 6534 6535 6536
	ret = seq_open(file, &ftrace_pid_sops);
	if (ret < 0) {
		trace_array_put(tr);
	} else {
		m = file->private_data;
		/* copy tr over to seq ops */
		m->private = tr;
	}
6537 6538 6539 6540

	return ret;
}

6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556
static void ignore_task_cpu(void *data)
{
	struct trace_array *tr = data;
	struct trace_pid_list *pid_list;

	/*
	 * This function is called by on_each_cpu() while the
	 * event_mutex is held.
	 */
	pid_list = rcu_dereference_protected(tr->function_pids,
					     mutex_is_locked(&ftrace_lock));

	this_cpu_write(tr->trace_buffer.data->ftrace_ignore_pid,
		       trace_ignore_this_task(pid_list, current));
}

6557 6558 6559 6560
static ssize_t
ftrace_pid_write(struct file *filp, const char __user *ubuf,
		   size_t cnt, loff_t *ppos)
{
6561 6562 6563 6564 6565
	struct seq_file *m = filp->private_data;
	struct trace_array *tr = m->private;
	struct trace_pid_list *filtered_pids = NULL;
	struct trace_pid_list *pid_list;
	ssize_t ret;
6566

6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577
	if (!cnt)
		return 0;

	mutex_lock(&ftrace_lock);

	filtered_pids = rcu_dereference_protected(tr->function_pids,
					     lockdep_is_held(&ftrace_lock));

	ret = trace_pid_write(filtered_pids, &pid_list, ubuf, cnt);
	if (ret < 0)
		goto out;
6578

6579
	rcu_assign_pointer(tr->function_pids, pid_list);
6580

6581 6582 6583 6584 6585 6586 6587
	if (filtered_pids) {
		synchronize_sched();
		trace_free_pid_list(filtered_pids);
	} else if (pid_list) {
		/* Register a probe to set whether to ignore the tracing of a task */
		register_trace_sched_switch(ftrace_filter_pid_sched_switch_probe, tr);
	}
6588

6589
	/*
6590 6591 6592
	 * Ignoring of pids is done at task switch. But we have to
	 * check for those tasks that are currently running.
	 * Always do this in case a pid was appended or removed.
6593
	 */
6594
	on_each_cpu(ignore_task_cpu, tr, 1);
6595

6596 6597 6598 6599
	ftrace_update_pid_func();
	ftrace_startup_all(0);
 out:
	mutex_unlock(&ftrace_lock);
6600

6601 6602
	if (ret > 0)
		*ppos += ret;
6603

6604
	return ret;
6605
}
6606

6607 6608 6609
static int
ftrace_pid_release(struct inode *inode, struct file *file)
{
6610
	struct trace_array *tr = inode->i_private;
6611

6612 6613 6614
	trace_array_put(tr);

	return seq_release(inode, file);
6615 6616
}

6617
static const struct file_operations ftrace_pid_fops = {
6618 6619 6620
	.open		= ftrace_pid_open,
	.write		= ftrace_pid_write,
	.read		= seq_read,
6621
	.llseek		= tracing_lseek,
6622
	.release	= ftrace_pid_release,
6623 6624
};

6625
void ftrace_init_tracefs(struct trace_array *tr, struct dentry *d_tracer)
6626
{
6627
	trace_create_file("set_ftrace_pid", 0644, d_tracer,
6628
			    tr, &ftrace_pid_fops);
6629 6630
}

6631 6632 6633 6634 6635 6636 6637 6638 6639 6640
void __init ftrace_init_tracefs_toplevel(struct trace_array *tr,
					 struct dentry *d_tracer)
{
	/* Only the top level directory has the dyn_tracefs and profile */
	WARN_ON(!(tr->flags & TRACE_ARRAY_FL_GLOBAL));

	ftrace_init_dyn_tracefs(d_tracer);
	ftrace_profile_tracefs(d_tracer);
}

S
Steven Rostedt 已提交
6641
/**
6642
 * ftrace_kill - kill ftrace
S
Steven Rostedt 已提交
6643 6644 6645 6646 6647
 *
 * This function should be used by panic code. It stops ftrace
 * but in a not so nice way. If you need to simply kill ftrace
 * from a non-atomic section, use ftrace_kill.
 */
6648
void ftrace_kill(void)
S
Steven Rostedt 已提交
6649 6650 6651
{
	ftrace_disabled = 1;
	ftrace_enabled = 0;
Y
Yisheng Xie 已提交
6652
	ftrace_trace_function = ftrace_stub;
S
Steven Rostedt 已提交
6653 6654
}

6655 6656 6657 6658 6659 6660 6661 6662
/**
 * Test if ftrace is dead or not.
 */
int ftrace_is_dead(void)
{
	return ftrace_disabled;
}

6663
/**
6664 6665
 * register_ftrace_function - register a function for profiling
 * @ops - ops structure that holds the function for profiling.
6666
 *
6667 6668 6669 6670 6671 6672
 * Register a function to be called by all functions in the
 * kernel.
 *
 * Note: @ops->func and all the functions it calls must be labeled
 *       with "notrace", otherwise it will go into a
 *       recursive loop.
6673
 */
6674
int register_ftrace_function(struct ftrace_ops *ops)
6675
{
6676
	int ret = -1;
6677

6678 6679
	ftrace_ops_init(ops);

S
Steven Rostedt 已提交
6680
	mutex_lock(&ftrace_lock);
6681

6682
	ret = ftrace_startup(ops, 0);
6683

S
Steven Rostedt 已提交
6684
	mutex_unlock(&ftrace_lock);
6685

6686
	return ret;
6687
}
6688
EXPORT_SYMBOL_GPL(register_ftrace_function);
6689 6690

/**
6691
 * unregister_ftrace_function - unregister a function for profiling.
6692 6693 6694 6695 6696 6697 6698 6699
 * @ops - ops structure that holds the function to unregister
 *
 * Unregister a function that was added to be called by ftrace profiling.
 */
int unregister_ftrace_function(struct ftrace_ops *ops)
{
	int ret;

S
Steven Rostedt 已提交
6700
	mutex_lock(&ftrace_lock);
6701
	ret = ftrace_shutdown(ops, 0);
S
Steven Rostedt 已提交
6702
	mutex_unlock(&ftrace_lock);
6703 6704 6705

	return ret;
}
6706
EXPORT_SYMBOL_GPL(unregister_ftrace_function);
6707

I
Ingo Molnar 已提交
6708
int
6709
ftrace_enable_sysctl(struct ctl_table *table, int write,
6710
		     void __user *buffer, size_t *lenp,
6711 6712
		     loff_t *ppos)
{
6713
	int ret = -ENODEV;
6714

S
Steven Rostedt 已提交
6715
	mutex_lock(&ftrace_lock);
6716

6717 6718 6719 6720
	if (unlikely(ftrace_disabled))
		goto out;

	ret = proc_dointvec(table, write, buffer, lenp, ppos);
6721

6722
	if (ret || !write || (last_ftrace_enabled == !!ftrace_enabled))
6723 6724
		goto out;

6725
	last_ftrace_enabled = !!ftrace_enabled;
6726 6727 6728 6729

	if (ftrace_enabled) {

		/* we are starting ftrace again */
6730 6731
		if (rcu_dereference_protected(ftrace_ops_list,
			lockdep_is_held(&ftrace_lock)) != &ftrace_list_end)
6732
			update_ftrace_function();
6733

6734 6735
		ftrace_startup_sysctl();

6736 6737 6738 6739 6740 6741 6742 6743
	} else {
		/* stopping ftrace calls (just send to ftrace_stub) */
		ftrace_trace_function = ftrace_stub;

		ftrace_shutdown_sysctl();
	}

 out:
S
Steven Rostedt 已提交
6744
	mutex_unlock(&ftrace_lock);
6745
	return ret;
6746
}
6747

6748
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
6749

6750 6751 6752 6753
static struct ftrace_ops graph_ops = {
	.func			= ftrace_stub,
	.flags			= FTRACE_OPS_FL_RECURSION_SAFE |
				   FTRACE_OPS_FL_INITIALIZED |
6754
				   FTRACE_OPS_FL_PID |
6755 6756 6757
				   FTRACE_OPS_FL_STUB,
#ifdef FTRACE_GRAPH_TRAMP_ADDR
	.trampoline		= FTRACE_GRAPH_TRAMP_ADDR,
6758
	/* trampoline_size is only needed for dynamically allocated tramps */
6759 6760 6761 6762
#endif
	ASSIGN_OPS_HASH(graph_ops, &global_ops.local_hash)
};

6763 6764 6765 6766 6767 6768 6769 6770 6771 6772
void ftrace_graph_sleep_time_control(bool enable)
{
	fgraph_sleep_time = enable;
}

void ftrace_graph_graph_time_control(bool enable)
{
	fgraph_graph_time = enable;
}

6773 6774 6775 6776 6777
int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace)
{
	return 0;
}

6778 6779 6780
/* The callbacks that hook a function */
trace_func_graph_ret_t ftrace_graph_return =
			(trace_func_graph_ret_t)ftrace_stub;
6781
trace_func_graph_ent_t ftrace_graph_entry = ftrace_graph_entry_stub;
6782
static trace_func_graph_ent_t __ftrace_graph_entry = ftrace_graph_entry_stub;
6783 6784 6785 6786 6787 6788 6789 6790 6791 6792

/* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */
static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list)
{
	int i;
	int ret = 0;
	int start = 0, end = FTRACE_RETSTACK_ALLOC_SIZE;
	struct task_struct *g, *t;

	for (i = 0; i < FTRACE_RETSTACK_ALLOC_SIZE; i++) {
6793 6794 6795 6796
		ret_stack_list[i] =
			kmalloc_array(FTRACE_RETFUNC_DEPTH,
				      sizeof(struct ftrace_ret_stack),
				      GFP_KERNEL);
6797 6798 6799 6800 6801 6802 6803 6804
		if (!ret_stack_list[i]) {
			start = 0;
			end = i;
			ret = -ENOMEM;
			goto free;
		}
	}

6805
	read_lock(&tasklist_lock);
6806 6807 6808 6809 6810 6811 6812
	do_each_thread(g, t) {
		if (start == end) {
			ret = -EAGAIN;
			goto unlock;
		}

		if (t->ret_stack == NULL) {
6813
			atomic_set(&t->tracing_graph_pause, 0);
6814
			atomic_set(&t->trace_overrun, 0);
6815 6816 6817 6818
			t->curr_ret_stack = -1;
			/* Make sure the tasks see the -1 first: */
			smp_wmb();
			t->ret_stack = ret_stack_list[start++];
6819 6820 6821 6822
		}
	} while_each_thread(g, t);

unlock:
6823
	read_unlock(&tasklist_lock);
6824 6825 6826 6827 6828 6829
free:
	for (i = start; i < end; i++)
		kfree(ret_stack_list[i]);
	return ret;
}

6830
static void
6831
ftrace_graph_probe_sched_switch(void *ignore, bool preempt,
6832
			struct task_struct *prev, struct task_struct *next)
6833 6834 6835 6836
{
	unsigned long long timestamp;
	int index;

6837 6838 6839 6840
	/*
	 * Does the user want to count the time a function was asleep.
	 * If so, do not update the time stamps.
	 */
6841
	if (fgraph_sleep_time)
6842 6843
		return;

6844 6845 6846 6847 6848 6849 6850 6851 6852 6853 6854 6855 6856 6857 6858 6859 6860 6861
	timestamp = trace_clock_local();

	prev->ftrace_timestamp = timestamp;

	/* only process tasks that we timestamped */
	if (!next->ftrace_timestamp)
		return;

	/*
	 * Update all the counters in next to make up for the
	 * time next was sleeping.
	 */
	timestamp -= next->ftrace_timestamp;

	for (index = next->curr_ret_stack; index >= 0; index--)
		next->ret_stack[index].calltime += timestamp;
}

6862
/* Allocate a return stack for each task */
6863
static int start_graph_tracing(void)
6864 6865
{
	struct ftrace_ret_stack **ret_stack_list;
6866
	int ret, cpu;
6867

6868 6869 6870
	ret_stack_list = kmalloc_array(FTRACE_RETSTACK_ALLOC_SIZE,
				       sizeof(struct ftrace_ret_stack *),
				       GFP_KERNEL);
6871 6872 6873 6874

	if (!ret_stack_list)
		return -ENOMEM;

6875
	/* The cpu_boot init_task->ret_stack will never be freed */
6876 6877
	for_each_online_cpu(cpu) {
		if (!idle_task(cpu)->ret_stack)
6878
			ftrace_graph_init_idle_task(idle_task(cpu), cpu);
6879
	}
6880

6881 6882 6883 6884
	do {
		ret = alloc_retstack_tasklist(ret_stack_list);
	} while (ret == -EAGAIN);

6885
	if (!ret) {
6886
		ret = register_trace_sched_switch(ftrace_graph_probe_sched_switch, NULL);
6887 6888 6889 6890 6891
		if (ret)
			pr_info("ftrace_graph: Couldn't activate tracepoint"
				" probe to kernel_sched_switch\n");
	}

6892 6893 6894 6895
	kfree(ret_stack_list);
	return ret;
}

6896 6897 6898 6899 6900 6901 6902 6903 6904 6905 6906 6907 6908 6909 6910 6911 6912 6913 6914 6915 6916
/*
 * Hibernation protection.
 * The state of the current task is too much unstable during
 * suspend/restore to disk. We want to protect against that.
 */
static int
ftrace_suspend_notifier_call(struct notifier_block *bl, unsigned long state,
							void *unused)
{
	switch (state) {
	case PM_HIBERNATION_PREPARE:
		pause_graph_tracing();
		break;

	case PM_POST_HIBERNATION:
		unpause_graph_tracing();
		break;
	}
	return NOTIFY_DONE;
}

6917 6918 6919 6920 6921 6922 6923 6924 6925 6926 6927 6928 6929 6930 6931 6932
static int ftrace_graph_entry_test(struct ftrace_graph_ent *trace)
{
	if (!ftrace_ops_test(&global_ops, trace->func, NULL))
		return 0;
	return __ftrace_graph_entry(trace);
}

/*
 * The function graph tracer should only trace the functions defined
 * by set_ftrace_filter and set_ftrace_notrace. If another function
 * tracer ops is registered, the graph tracer requires testing the
 * function against the global ops, and not just trace any function
 * that any ftrace_ops registered.
 */
static void update_function_graph_func(void)
{
6933 6934 6935 6936 6937 6938 6939 6940 6941 6942 6943 6944 6945 6946 6947 6948 6949 6950 6951
	struct ftrace_ops *op;
	bool do_test = false;

	/*
	 * The graph and global ops share the same set of functions
	 * to test. If any other ops is on the list, then
	 * the graph tracing needs to test if its the function
	 * it should call.
	 */
	do_for_each_ftrace_op(op, ftrace_ops_list) {
		if (op != &global_ops && op != &graph_ops &&
		    op != &ftrace_list_end) {
			do_test = true;
			/* in double loop, break out with goto */
			goto out;
		}
	} while_for_each_ftrace_op(op);
 out:
	if (do_test)
6952
		ftrace_graph_entry = ftrace_graph_entry_test;
6953 6954
	else
		ftrace_graph_entry = __ftrace_graph_entry;
6955 6956
}

6957 6958 6959 6960
static struct notifier_block ftrace_suspend_notifier = {
	.notifier_call = ftrace_suspend_notifier_call,
};

6961 6962
int register_ftrace_graph(trace_func_graph_ret_t retfunc,
			trace_func_graph_ent_t entryfunc)
6963
{
6964 6965
	int ret = 0;

S
Steven Rostedt 已提交
6966
	mutex_lock(&ftrace_lock);
6967

6968
	/* we currently allow only one tracer registered at a time */
6969
	if (ftrace_graph_active) {
6970 6971 6972 6973
		ret = -EBUSY;
		goto out;
	}

6974 6975
	register_pm_notifier(&ftrace_suspend_notifier);

6976
	ftrace_graph_active++;
6977
	ret = start_graph_tracing();
6978
	if (ret) {
6979
		ftrace_graph_active--;
6980 6981
		goto out;
	}
6982

6983
	ftrace_graph_return = retfunc;
6984 6985 6986 6987 6988 6989 6990 6991 6992 6993

	/*
	 * Update the indirect function to the entryfunc, and the
	 * function that gets called to the entry_test first. Then
	 * call the update fgraph entry function to determine if
	 * the entryfunc should be called directly or not.
	 */
	__ftrace_graph_entry = entryfunc;
	ftrace_graph_entry = ftrace_graph_entry_test;
	update_function_graph_func();
6994

6995
	ret = ftrace_startup(&graph_ops, FTRACE_START_FUNC_RET);
6996
out:
S
Steven Rostedt 已提交
6997
	mutex_unlock(&ftrace_lock);
6998
	return ret;
6999 7000
}

7001
void unregister_ftrace_graph(void)
7002
{
S
Steven Rostedt 已提交
7003
	mutex_lock(&ftrace_lock);
7004

7005
	if (unlikely(!ftrace_graph_active))
7006 7007
		goto out;

7008
	ftrace_graph_active--;
7009
	ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub;
7010
	ftrace_graph_entry = ftrace_graph_entry_stub;
7011
	__ftrace_graph_entry = ftrace_graph_entry_stub;
7012
	ftrace_shutdown(&graph_ops, FTRACE_STOP_FUNC_RET);
7013
	unregister_pm_notifier(&ftrace_suspend_notifier);
7014
	unregister_trace_sched_switch(ftrace_graph_probe_sched_switch, NULL);
7015

7016
 out:
S
Steven Rostedt 已提交
7017
	mutex_unlock(&ftrace_lock);
7018
}
7019

7020 7021 7022 7023 7024 7025 7026 7027
static DEFINE_PER_CPU(struct ftrace_ret_stack *, idle_ret_stack);

static void
graph_init_task(struct task_struct *t, struct ftrace_ret_stack *ret_stack)
{
	atomic_set(&t->tracing_graph_pause, 0);
	atomic_set(&t->trace_overrun, 0);
	t->ftrace_timestamp = 0;
L
Lucas De Marchi 已提交
7028
	/* make curr_ret_stack visible before we add the ret_stack */
7029 7030 7031 7032 7033 7034 7035 7036 7037 7038 7039 7040 7041 7042 7043 7044 7045 7046 7047 7048 7049 7050 7051
	smp_wmb();
	t->ret_stack = ret_stack;
}

/*
 * Allocate a return stack for the idle task. May be the first
 * time through, or it may be done by CPU hotplug online.
 */
void ftrace_graph_init_idle_task(struct task_struct *t, int cpu)
{
	t->curr_ret_stack = -1;
	/*
	 * The idle task has no parent, it either has its own
	 * stack or no stack at all.
	 */
	if (t->ret_stack)
		WARN_ON(t->ret_stack != per_cpu(idle_ret_stack, cpu));

	if (ftrace_graph_active) {
		struct ftrace_ret_stack *ret_stack;

		ret_stack = per_cpu(idle_ret_stack, cpu);
		if (!ret_stack) {
7052 7053 7054 7055
			ret_stack =
				kmalloc_array(FTRACE_RETFUNC_DEPTH,
					      sizeof(struct ftrace_ret_stack),
					      GFP_KERNEL);
7056 7057 7058 7059 7060 7061 7062 7063
			if (!ret_stack)
				return;
			per_cpu(idle_ret_stack, cpu) = ret_stack;
		}
		graph_init_task(t, ret_stack);
	}
}

7064
/* Allocate a return stack for newly created task */
7065
void ftrace_graph_init_task(struct task_struct *t)
7066
{
7067 7068
	/* Make sure we do not use the parent ret_stack */
	t->ret_stack = NULL;
7069
	t->curr_ret_stack = -1;
7070

7071
	if (ftrace_graph_active) {
7072 7073
		struct ftrace_ret_stack *ret_stack;

7074 7075 7076
		ret_stack = kmalloc_array(FTRACE_RETFUNC_DEPTH,
					  sizeof(struct ftrace_ret_stack),
					  GFP_KERNEL);
7077
		if (!ret_stack)
7078
			return;
7079
		graph_init_task(t, ret_stack);
7080
	}
7081 7082
}

7083
void ftrace_graph_exit_task(struct task_struct *t)
7084
{
7085 7086
	struct ftrace_ret_stack	*ret_stack = t->ret_stack;

7087
	t->ret_stack = NULL;
7088 7089 7090 7091
	/* NULL must become visible to IRQs before we free it: */
	barrier();

	kfree(ret_stack);
7092
}
7093
#endif