trace_kprobe.c 39.1 KB
Newer Older
1
/*
2
 * Kprobes-based tracing events
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
 *
 * Created by Masami Hiramatsu <mhiramat@redhat.com>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

#include <linux/module.h>
#include <linux/uaccess.h>
#include <linux/kprobes.h>
#include <linux/seq_file.h>
#include <linux/slab.h>
#include <linux/smp.h>
#include <linux/debugfs.h>
#include <linux/types.h>
#include <linux/string.h>
#include <linux/ctype.h>
#include <linux/ptrace.h>
31
#include <linux/perf_event.h>
32 33
#include <linux/stringify.h>
#include <asm/bitsperlong.h>
34 35 36 37

#include "trace.h"
#include "trace_output.h"

38
#define MAX_TRACE_ARGS 128
39
#define MAX_ARGSTR_LEN 63
40
#define MAX_EVENT_NAME_LEN 64
41
#define KPROBE_EVENT_SYSTEM "kprobes"
42

43
/* Reserved field names */
44 45 46
#define FIELD_STRING_IP "__probe_ip"
#define FIELD_STRING_RETIP "__probe_ret_ip"
#define FIELD_STRING_FUNC "__probe_func"
47 48 49 50 51 52 53 54 55 56 57 58 59

const char *reserved_field_names[] = {
	"common_type",
	"common_flags",
	"common_preempt_count",
	"common_pid",
	"common_tgid",
	"common_lock_depth",
	FIELD_STRING_IP,
	FIELD_STRING_RETIP,
	FIELD_STRING_FUNC,
};

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
/* Printing function type */
typedef int (*print_type_func_t)(struct trace_seq *, const char *, void *);
#define PRINT_TYPE_FUNC_NAME(type)	print_type_##type
#define PRINT_TYPE_FMT_NAME(type)	print_type_format_##type

/* Printing  in basic type function template */
#define DEFINE_BASIC_PRINT_TYPE_FUNC(type, fmt, cast)			\
static __kprobes int PRINT_TYPE_FUNC_NAME(type)(struct trace_seq *s,	\
						const char *name, void *data)\
{									\
	return trace_seq_printf(s, " %s=" fmt, name, (cast)*(type *)data);\
}									\
static const char PRINT_TYPE_FMT_NAME(type)[] = fmt;

DEFINE_BASIC_PRINT_TYPE_FUNC(u8, "%x", unsigned int)
DEFINE_BASIC_PRINT_TYPE_FUNC(u16, "%x", unsigned int)
DEFINE_BASIC_PRINT_TYPE_FUNC(u32, "%lx", unsigned long)
DEFINE_BASIC_PRINT_TYPE_FUNC(u64, "%llx", unsigned long long)
DEFINE_BASIC_PRINT_TYPE_FUNC(s8, "%d", int)
DEFINE_BASIC_PRINT_TYPE_FUNC(s16, "%d", int)
DEFINE_BASIC_PRINT_TYPE_FUNC(s32, "%ld", long)
DEFINE_BASIC_PRINT_TYPE_FUNC(s64, "%lld", long long)

/* Data fetch function type */
typedef	void (*fetch_func_t)(struct pt_regs *, void *, void *);

struct fetch_param {
	fetch_func_t	fn;
88 89 90
	void *data;
};

91 92
static __kprobes void call_fetch(struct fetch_param *fprm,
				 struct pt_regs *regs, void *dest)
93
{
94
	return fprm->fn(regs, fprm->data, dest);
95 96
}

97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
#define FETCH_FUNC_NAME(kind, type)	fetch_##kind##_##type
/*
 * Define macro for basic types - we don't need to define s* types, because
 * we have to care only about bitwidth at recording time.
 */
#define DEFINE_BASIC_FETCH_FUNCS(kind)  \
DEFINE_FETCH_##kind(u8)			\
DEFINE_FETCH_##kind(u16)		\
DEFINE_FETCH_##kind(u32)		\
DEFINE_FETCH_##kind(u64)

#define CHECK_BASIC_FETCH_FUNCS(kind, fn)	\
	((FETCH_FUNC_NAME(kind, u8) == fn) ||	\
	 (FETCH_FUNC_NAME(kind, u16) == fn) ||	\
	 (FETCH_FUNC_NAME(kind, u32) == fn) ||	\
	 (FETCH_FUNC_NAME(kind, u64) == fn))

/* Data fetch function templates */
#define DEFINE_FETCH_reg(type)						\
static __kprobes void FETCH_FUNC_NAME(reg, type)(struct pt_regs *regs,	\
					  void *offset, void *dest)	\
{									\
	*(type *)dest = (type)regs_get_register(regs,			\
				(unsigned int)((unsigned long)offset));	\
121
}
122 123 124 125 126 127 128 129
DEFINE_BASIC_FETCH_FUNCS(reg)

#define DEFINE_FETCH_stack(type)					\
static __kprobes void FETCH_FUNC_NAME(stack, type)(struct pt_regs *regs,\
					  void *offset, void *dest)	\
{									\
	*(type *)dest = (type)regs_get_kernel_stack_nth(regs,		\
				(unsigned int)((unsigned long)offset));	\
130
}
131
DEFINE_BASIC_FETCH_FUNCS(stack)
132

133 134 135 136 137
#define DEFINE_FETCH_retval(type)					\
static __kprobes void FETCH_FUNC_NAME(retval, type)(struct pt_regs *regs,\
					  void *dummy, void *dest)	\
{									\
	*(type *)dest = (type)regs_return_value(regs);			\
138
}
139 140 141 142 143 144 145 146 147 148 149
DEFINE_BASIC_FETCH_FUNCS(retval)

#define DEFINE_FETCH_memory(type)					\
static __kprobes void FETCH_FUNC_NAME(memory, type)(struct pt_regs *regs,\
					  void *addr, void *dest)	\
{									\
	type retval;							\
	if (probe_kernel_address(addr, retval))				\
		*(type *)dest = 0;					\
	else								\
		*(type *)dest = retval;					\
150
}
151
DEFINE_BASIC_FETCH_FUNCS(memory)
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194

/* Memory fetching by symbol */
struct symbol_cache {
	char *symbol;
	long offset;
	unsigned long addr;
};

static unsigned long update_symbol_cache(struct symbol_cache *sc)
{
	sc->addr = (unsigned long)kallsyms_lookup_name(sc->symbol);
	if (sc->addr)
		sc->addr += sc->offset;
	return sc->addr;
}

static void free_symbol_cache(struct symbol_cache *sc)
{
	kfree(sc->symbol);
	kfree(sc);
}

static struct symbol_cache *alloc_symbol_cache(const char *sym, long offset)
{
	struct symbol_cache *sc;

	if (!sym || strlen(sym) == 0)
		return NULL;
	sc = kzalloc(sizeof(struct symbol_cache), GFP_KERNEL);
	if (!sc)
		return NULL;

	sc->symbol = kstrdup(sym, GFP_KERNEL);
	if (!sc->symbol) {
		kfree(sc);
		return NULL;
	}
	sc->offset = offset;

	update_symbol_cache(sc);
	return sc;
}

195 196 197 198 199 200 201 202 203
#define DEFINE_FETCH_symbol(type)					\
static __kprobes void FETCH_FUNC_NAME(symbol, type)(struct pt_regs *regs,\
					  void *data, void *dest)	\
{									\
	struct symbol_cache *sc = data;					\
	if (sc->addr)							\
		fetch_memory_##type(regs, (void *)sc->addr, dest);	\
	else								\
		*(type *)dest = 0;					\
204
}
205
DEFINE_BASIC_FETCH_FUNCS(symbol)
206

207 208 209
/* Dereference memory access function */
struct deref_fetch_param {
	struct fetch_param orig;
210 211 212
	long offset;
};

213 214 215 216 217 218 219 220 221 222 223 224
#define DEFINE_FETCH_deref(type)					\
static __kprobes void FETCH_FUNC_NAME(deref, type)(struct pt_regs *regs,\
					    void *data, void *dest)	\
{									\
	struct deref_fetch_param *dprm = data;				\
	unsigned long addr;						\
	call_fetch(&dprm->orig, regs, &addr);				\
	if (addr) {							\
		addr += dprm->offset;					\
		fetch_memory_##type(regs, (void *)addr, dest);		\
	} else								\
		*(type *)dest = 0;					\
225
}
226
DEFINE_BASIC_FETCH_FUNCS(deref)
227

228
static __kprobes void free_deref_fetch_param(struct deref_fetch_param *data)
229
{
230 231 232
	if (CHECK_BASIC_FETCH_FUNCS(deref, data->orig.fn))
		free_deref_fetch_param(data->orig.data);
	else if (CHECK_BASIC_FETCH_FUNCS(symbol, data->orig.fn))
233 234 235 236
		free_symbol_cache(data->orig.data);
	kfree(data);
}

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
/* Default (unsigned long) fetch type */
#define __DEFAULT_FETCH_TYPE(t) u##t
#define _DEFAULT_FETCH_TYPE(t) __DEFAULT_FETCH_TYPE(t)
#define DEFAULT_FETCH_TYPE _DEFAULT_FETCH_TYPE(BITS_PER_LONG)
#define DEFAULT_FETCH_TYPE_STR __stringify(DEFAULT_FETCH_TYPE)

#define ASSIGN_FETCH_FUNC(kind, type)	\
	.kind = FETCH_FUNC_NAME(kind, type)

#define ASSIGN_FETCH_TYPE(ptype, ftype, sign)	\
	{.name = #ptype,			\
	 .size = sizeof(ftype),			\
	 .is_signed = sign,			\
	 .print = PRINT_TYPE_FUNC_NAME(ptype),	\
	 .fmt = PRINT_TYPE_FMT_NAME(ptype),	\
ASSIGN_FETCH_FUNC(reg, ftype),			\
ASSIGN_FETCH_FUNC(stack, ftype),		\
ASSIGN_FETCH_FUNC(retval, ftype),		\
ASSIGN_FETCH_FUNC(memory, ftype),		\
ASSIGN_FETCH_FUNC(symbol, ftype),		\
ASSIGN_FETCH_FUNC(deref, ftype),		\
	}

/* Fetch type information table */
static const struct fetch_type {
	const char	*name;		/* Name of type */
	size_t		size;		/* Byte size of type */
	int		is_signed;	/* Signed flag */
	print_type_func_t	print;	/* Print functions */
	const char	*fmt;		/* Fromat string */
	/* Fetch functions */
	fetch_func_t	reg;
	fetch_func_t	stack;
	fetch_func_t	retval;
	fetch_func_t	memory;
	fetch_func_t	symbol;
	fetch_func_t	deref;
} fetch_type_table[] = {
	ASSIGN_FETCH_TYPE(u8,  u8,  0),
	ASSIGN_FETCH_TYPE(u16, u16, 0),
	ASSIGN_FETCH_TYPE(u32, u32, 0),
	ASSIGN_FETCH_TYPE(u64, u64, 0),
	ASSIGN_FETCH_TYPE(s8,  u8,  1),
	ASSIGN_FETCH_TYPE(s16, u16, 1),
	ASSIGN_FETCH_TYPE(s32, u32, 1),
	ASSIGN_FETCH_TYPE(s64, u64, 1),
};

static const struct fetch_type *find_fetch_type(const char *type)
{
	int i;

	if (!type)
		type = DEFAULT_FETCH_TYPE_STR;

	for (i = 0; i < ARRAY_SIZE(fetch_type_table); i++)
		if (strcmp(type, fetch_type_table[i].name) == 0)
			return &fetch_type_table[i];
	return NULL;
}

/* Special function : only accept unsigned long */
static __kprobes void fetch_stack_address(struct pt_regs *regs,
					  void *dummy, void *dest)
{
	*(unsigned long *)dest = kernel_stack_pointer(regs);
}

305
/**
306
 * Kprobe event core functions
307 308
 */

309
struct probe_arg {
310 311 312 313 314
	struct fetch_param	fetch;
	unsigned int		offset;	/* Offset from argument entry */
	const char		*name;	/* Name of this argument */
	const char		*comm;	/* Command of this argument */
	const struct fetch_type	*type;	/* Type of this argument */
315 316
};

317 318 319 320
/* Flags for trace_probe */
#define TP_FLAG_TRACE	1
#define TP_FLAG_PROFILE	2

321 322
struct trace_probe {
	struct list_head	list;
323
	struct kretprobe	rp;	/* Use rp.kp for kprobe use */
324
	unsigned long 		nhit;
325
	unsigned int		flags;	/* For TP_FLAG_* */
326
	const char		*symbol;	/* symbol name */
327
	struct ftrace_event_class	class;
328
	struct ftrace_event_call	call;
329
	ssize_t			size;		/* trace entry size */
330
	unsigned int		nr_args;
331
	struct probe_arg	args[];
332 333
};

334 335
#define SIZEOF_TRACE_PROBE(n)			\
	(offsetof(struct trace_probe, args) +	\
336
	(sizeof(struct probe_arg) * (n)))
337

338

339 340
static __kprobes int probe_is_return(struct trace_probe *tp)
{
341
	return tp->rp.handler != NULL;
342 343 344 345 346 347 348 349 350 351 352 353 354
}

static __kprobes const char *probe_symbol(struct trace_probe *tp)
{
	return tp->symbol ? tp->symbol : "unknown";
}

static int register_probe_event(struct trace_probe *tp);
static void unregister_probe_event(struct trace_probe *tp);

static DEFINE_MUTEX(probe_lock);
static LIST_HEAD(probe_list);

355 356 357 358
static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs);
static int kretprobe_dispatcher(struct kretprobe_instance *ri,
				struct pt_regs *regs);

359 360 361 362 363 364 365 366 367 368 369 370
/* Check the name is good for event/group */
static int check_event_name(const char *name)
{
	if (!isalpha(*name) && *name != '_')
		return 0;
	while (*++name != '\0') {
		if (!isalpha(*name) && !isdigit(*name) && *name != '_')
			return 0;
	}
	return 1;
}

371 372 373
/*
 * Allocate new trace_probe and initialize it (including kprobes).
 */
374 375
static struct trace_probe *alloc_trace_probe(const char *group,
					     const char *event,
376 377 378 379
					     void *addr,
					     const char *symbol,
					     unsigned long offs,
					     int nargs, int is_return)
380 381
{
	struct trace_probe *tp;
382
	int ret = -ENOMEM;
383

384
	tp = kzalloc(SIZEOF_TRACE_PROBE(nargs), GFP_KERNEL);
385
	if (!tp)
386
		return ERR_PTR(ret);
387 388 389 390 391

	if (symbol) {
		tp->symbol = kstrdup(symbol, GFP_KERNEL);
		if (!tp->symbol)
			goto error;
392 393 394 395 396 397
		tp->rp.kp.symbol_name = tp->symbol;
		tp->rp.kp.offset = offs;
	} else
		tp->rp.kp.addr = addr;

	if (is_return)
398
		tp->rp.handler = kretprobe_dispatcher;
399
	else
400
		tp->rp.kp.pre_handler = kprobe_dispatcher;
401

402 403
	if (!event || !check_event_name(event)) {
		ret = -EINVAL;
404
		goto error;
405 406
	}

407
	tp->call.class = &tp->class;
408 409 410
	tp->call.name = kstrdup(event, GFP_KERNEL);
	if (!tp->call.name)
		goto error;
411

412 413
	if (!group || !check_event_name(group)) {
		ret = -EINVAL;
414
		goto error;
415 416
	}

417 418
	tp->class.system = kstrdup(group, GFP_KERNEL);
	if (!tp->class.system)
419 420
		goto error;

421 422 423
	INIT_LIST_HEAD(&tp->list);
	return tp;
error:
424
	kfree(tp->call.name);
425 426
	kfree(tp->symbol);
	kfree(tp);
427
	return ERR_PTR(ret);
428 429
}

430 431
static void free_probe_arg(struct probe_arg *arg)
{
432 433 434
	if (CHECK_BASIC_FETCH_FUNCS(deref, arg->fetch.fn))
		free_deref_fetch_param(arg->fetch.data);
	else if (CHECK_BASIC_FETCH_FUNCS(symbol, arg->fetch.fn))
435 436
		free_symbol_cache(arg->fetch.data);
	kfree(arg->name);
437
	kfree(arg->comm);
438 439
}

440 441 442 443 444
static void free_trace_probe(struct trace_probe *tp)
{
	int i;

	for (i = 0; i < tp->nr_args; i++)
445
		free_probe_arg(&tp->args[i]);
446

447
	kfree(tp->call.class->system);
448 449 450 451 452
	kfree(tp->call.name);
	kfree(tp->symbol);
	kfree(tp);
}

453 454
static struct trace_probe *find_probe_event(const char *event,
					    const char *group)
455 456 457 458
{
	struct trace_probe *tp;

	list_for_each_entry(tp, &probe_list, list)
459
		if (strcmp(tp->call.name, event) == 0 &&
460
		    strcmp(tp->call.class->system, group) == 0)
461 462 463 464
			return tp;
	return NULL;
}

465 466
/* Unregister a trace_probe and probe_event: call with locking probe_lock */
static void unregister_trace_probe(struct trace_probe *tp)
467 468 469 470
{
	if (probe_is_return(tp))
		unregister_kretprobe(&tp->rp);
	else
471
		unregister_kprobe(&tp->rp.kp);
472
	list_del(&tp->list);
473
	unregister_probe_event(tp);
474 475 476 477 478 479 480 481 482 483
}

/* Register a trace_probe and probe_event */
static int register_trace_probe(struct trace_probe *tp)
{
	struct trace_probe *old_tp;
	int ret;

	mutex_lock(&probe_lock);

484
	/* register as an event */
485
	old_tp = find_probe_event(tp->call.name, tp->call.class->system);
486 487 488 489 490 491 492 493 494 495 496
	if (old_tp) {
		/* delete old event */
		unregister_trace_probe(old_tp);
		free_trace_probe(old_tp);
	}
	ret = register_probe_event(tp);
	if (ret) {
		pr_warning("Faild to register probe event(%d)\n", ret);
		goto end;
	}

497
	tp->rp.kp.flags |= KPROBE_FLAG_DISABLED;
498 499 500
	if (probe_is_return(tp))
		ret = register_kretprobe(&tp->rp);
	else
501
		ret = register_kprobe(&tp->rp.kp);
502 503 504 505 506 507

	if (ret) {
		pr_warning("Could not insert probe(%d)\n", ret);
		if (ret == -EILSEQ) {
			pr_warning("Probing address(0x%p) is not an "
				   "instruction boundary.\n",
508
				   tp->rp.kp.addr);
509 510
			ret = -EINVAL;
		}
511 512 513
		unregister_probe_event(tp);
	} else
		list_add_tail(&tp->list, &probe_list);
514 515 516 517 518 519
end:
	mutex_unlock(&probe_lock);
	return ret;
}

/* Split symbol and offset. */
520
static int split_symbol_offset(char *symbol, unsigned long *offset)
521 522 523 524 525 526 527 528 529 530
{
	char *tmp;
	int ret;

	if (!offset)
		return -EINVAL;

	tmp = strchr(symbol, '+');
	if (tmp) {
		/* skip sign because strict_strtol doesn't accept '+' */
531
		ret = strict_strtoul(tmp + 1, 0, offset);
532 533 534 535 536 537 538 539 540 541 542
		if (ret)
			return ret;
		*tmp = '\0';
	} else
		*offset = 0;
	return 0;
}

#define PARAM_MAX_ARGS 16
#define PARAM_MAX_STACK (THREAD_SIZE / sizeof(unsigned long))

543 544
static int parse_probe_vars(char *arg, const struct fetch_type *t,
			    struct fetch_param *f, int is_return)
545 546 547 548
{
	int ret = 0;
	unsigned long param;

549
	if (strcmp(arg, "retval") == 0) {
550 551 552
		if (is_return)
			f->fn = t->retval;
		else
553
			ret = -EINVAL;
554 555
	} else if (strncmp(arg, "stack", 5) == 0) {
		if (arg[5] == '\0') {
556 557 558 559
			if (strcmp(t->name, DEFAULT_FETCH_TYPE_STR) == 0)
				f->fn = fetch_stack_address;
			else
				ret = -EINVAL;
560 561
		} else if (isdigit(arg[5])) {
			ret = strict_strtoul(arg + 5, 10, &param);
562 563 564
			if (ret || param > PARAM_MAX_STACK)
				ret = -EINVAL;
			else {
565 566
				f->fn = t->stack;
				f->data = (void *)param;
567
			}
568 569 570
		} else
			ret = -EINVAL;
	} else
571 572 573 574
		ret = -EINVAL;
	return ret;
}

575
/* Recursive argument parser */
576 577
static int __parse_probe_arg(char *arg, const struct fetch_type *t,
			     struct fetch_param *f, int is_return)
578 579 580 581 582 583 584 585
{
	int ret = 0;
	unsigned long param;
	long offset;
	char *tmp;

	switch (arg[0]) {
	case '$':
586
		ret = parse_probe_vars(arg + 1, t, f, is_return);
587 588 589 590
		break;
	case '%':	/* named register */
		ret = regs_query_register_offset(arg + 1);
		if (ret >= 0) {
591 592
			f->fn = t->reg;
			f->data = (void *)(unsigned long)ret;
593 594 595
			ret = 0;
		}
		break;
596 597 598 599 600
	case '@':	/* memory or symbol */
		if (isdigit(arg[1])) {
			ret = strict_strtoul(arg + 1, 0, &param);
			if (ret)
				break;
601 602
			f->fn = t->memory;
			f->data = (void *)param;
603 604 605 606
		} else {
			ret = split_symbol_offset(arg + 1, &offset);
			if (ret)
				break;
607 608 609
			f->data = alloc_symbol_cache(arg + 1, offset);
			if (f->data)
				f->fn = t->symbol;
610 611
		}
		break;
612
	case '+':	/* deref memory */
613 614
	case '-':
		tmp = strchr(arg, '(');
615
		if (!tmp)
616 617 618 619 620 621 622 623 624 625
			break;
		*tmp = '\0';
		ret = strict_strtol(arg + 1, 0, &offset);
		if (ret)
			break;
		if (arg[0] == '-')
			offset = -offset;
		arg = tmp + 1;
		tmp = strrchr(arg, ')');
		if (tmp) {
626 627
			struct deref_fetch_param *dprm;
			const struct fetch_type *t2 = find_fetch_type(NULL);
628
			*tmp = '\0';
629 630 631
			dprm = kzalloc(sizeof(struct deref_fetch_param),
				       GFP_KERNEL);
			if (!dprm)
632
				return -ENOMEM;
633 634 635
			dprm->offset = offset;
			ret = __parse_probe_arg(arg, t2, &dprm->orig,
						is_return);
636
			if (ret)
637
				kfree(dprm);
638
			else {
639 640
				f->fn = t->deref;
				f->data = (void *)dprm;
641
			}
642
		}
643 644
		break;
	}
645 646
	if (!ret && !f->fn)
		ret = -EINVAL;
647 648 649
	return ret;
}

650
/* String length checking wrapper */
651 652
static int parse_probe_arg(char *arg, struct trace_probe *tp,
			   struct probe_arg *parg, int is_return)
653
{
654 655
	const char *t;

656 657 658 659
	if (strlen(arg) > MAX_ARGSTR_LEN) {
		pr_info("Argument is too long.: %s\n",  arg);
		return -ENOSPC;
	}
660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677
	parg->comm = kstrdup(arg, GFP_KERNEL);
	if (!parg->comm) {
		pr_info("Failed to allocate memory for command '%s'.\n", arg);
		return -ENOMEM;
	}
	t = strchr(parg->comm, ':');
	if (t) {
		arg[t - parg->comm] = '\0';
		t++;
	}
	parg->type = find_fetch_type(t);
	if (!parg->type) {
		pr_info("Unsupported type: %s\n", t);
		return -EINVAL;
	}
	parg->offset = tp->size;
	tp->size += parg->type->size;
	return __parse_probe_arg(arg, parg->type, &parg->fetch, is_return);
678 679
}

680 681 682 683 684 685 686 687 688 689 690 691 692 693
/* Return 1 if name is reserved or already used by another argument */
static int conflict_field_name(const char *name,
			       struct probe_arg *args, int narg)
{
	int i;
	for (i = 0; i < ARRAY_SIZE(reserved_field_names); i++)
		if (strcmp(reserved_field_names[i], name) == 0)
			return 1;
	for (i = 0; i < narg; i++)
		if (strcmp(args[i].name, name) == 0)
			return 1;
	return 0;
}

694 695 696 697
static int create_trace_probe(int argc, char **argv)
{
	/*
	 * Argument syntax:
698 699
	 *  - Add kprobe: p[:[GRP/]EVENT] KSYM[+OFFS]|KADDR [FETCHARGS]
	 *  - Add kretprobe: r[:[GRP/]EVENT] KSYM[+0] [FETCHARGS]
700
	 * Fetch args:
701 702 703
	 *  $retval	: fetch return value
	 *  $stack	: fetch stack address
	 *  $stackN	: fetch Nth of stack (N:0-)
704 705 706
	 *  @ADDR	: fetch memory at ADDR (ADDR should be in kernel)
	 *  @SYM[+|-offs] : fetch memory at SYM +|- offs (SYM is a data symbol)
	 *  %REG	: fetch register REG
707
	 * Dereferencing memory fetch:
708
	 *  +|-offs(ARG) : fetch memory at ARG +|- offs address.
709 710
	 * Alias name of args:
	 *  NAME=FETCHARG : set NAME as alias of FETCHARG.
711 712
	 * Type of args:
	 *  FETCHARG:TYPE : use TYPE instead of unsigned long.
713 714 715
	 */
	struct trace_probe *tp;
	int i, ret = 0;
716
	int is_return = 0, is_delete = 0;
717 718
	char *symbol = NULL, *event = NULL, *group = NULL;
	char *arg, *tmp;
719
	unsigned long offset = 0;
720
	void *addr = NULL;
721
	char buf[MAX_EVENT_NAME_LEN];
722

723
	/* argc must be >= 1 */
724 725 726 727
	if (argv[0][0] == 'p')
		is_return = 0;
	else if (argv[0][0] == 'r')
		is_return = 1;
728 729
	else if (argv[0][0] == '-')
		is_delete = 1;
730
	else {
731 732
		pr_info("Probe definition must be started with 'p', 'r' or"
			" '-'.\n");
733
		return -EINVAL;
734
	}
735 736 737

	if (argv[0][1] == ':') {
		event = &argv[0][2];
738 739 740 741 742
		if (strchr(event, '/')) {
			group = event;
			event = strchr(group, '/') + 1;
			event[-1] = '\0';
			if (strlen(group) == 0) {
743
				pr_info("Group name is not specified\n");
744 745 746
				return -EINVAL;
			}
		}
747
		if (strlen(event) == 0) {
748
			pr_info("Event name is not specified\n");
749 750 751
			return -EINVAL;
		}
	}
752 753
	if (!group)
		group = KPROBE_EVENT_SYSTEM;
754

755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774
	if (is_delete) {
		if (!event) {
			pr_info("Delete command needs an event name.\n");
			return -EINVAL;
		}
		tp = find_probe_event(event, group);
		if (!tp) {
			pr_info("Event %s/%s doesn't exist.\n", group, event);
			return -ENOENT;
		}
		/* delete an event */
		unregister_trace_probe(tp);
		free_trace_probe(tp);
		return 0;
	}

	if (argc < 2) {
		pr_info("Probe point is not specified.\n");
		return -EINVAL;
	}
775
	if (isdigit(argv[1][0])) {
776 777
		if (is_return) {
			pr_info("Return probe point must be a symbol.\n");
778
			return -EINVAL;
779
		}
780
		/* an address specified */
781
		ret = strict_strtoul(&argv[1][0], 0, (unsigned long *)&addr);
782 783
		if (ret) {
			pr_info("Failed to parse address.\n");
784
			return ret;
785
		}
786 787 788 789 790
	} else {
		/* a symbol specified */
		symbol = argv[1];
		/* TODO: support .init module functions */
		ret = split_symbol_offset(symbol, &offset);
791 792
		if (ret) {
			pr_info("Failed to parse symbol.\n");
793
			return ret;
794 795 796
		}
		if (offset && is_return) {
			pr_info("Return probe must be used without offset.\n");
797
			return -EINVAL;
798
		}
799
	}
800
	argc -= 2; argv += 2;
801 802

	/* setup a probe */
803 804 805
	if (!event) {
		/* Make a new event name */
		if (symbol)
806
			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_%s_%ld",
807 808
				 is_return ? 'r' : 'p', symbol, offset);
		else
809
			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_0x%p",
810
				 is_return ? 'r' : 'p', addr);
811 812
		event = buf;
	}
813 814
	tp = alloc_trace_probe(group, event, addr, symbol, offset, argc,
			       is_return);
815 816 817
	if (IS_ERR(tp)) {
		pr_info("Failed to allocate trace_probe.(%d)\n",
			(int)PTR_ERR(tp));
818
		return PTR_ERR(tp);
819
	}
820 821

	/* parse arguments */
822 823
	ret = 0;
	for (i = 0; i < argc && i < MAX_TRACE_ARGS; i++) {
824 825 826 827 828 829
		/* Parse argument name */
		arg = strchr(argv[i], '=');
		if (arg)
			*arg++ = '\0';
		else
			arg = argv[i];
830

831
		tp->args[i].name = kstrdup(argv[i], GFP_KERNEL);
832 833 834 835
		if (!tp->args[i].name) {
			pr_info("Failed to allocate argument%d name '%s'.\n",
				i, argv[i]);
			ret = -ENOMEM;
836 837
			goto error;
		}
838 839 840 841 842 843 844 845 846 847
		tmp = strchr(tp->args[i].name, ':');
		if (tmp)
			*tmp = '_';	/* convert : to _ */

		if (conflict_field_name(tp->args[i].name, tp->args, i)) {
			pr_info("Argument%d name '%s' conflicts with "
				"another field.\n", i, argv[i]);
			ret = -EINVAL;
			goto error;
		}
848 849

		/* Parse fetch argument */
850
		ret = parse_probe_arg(arg, tp, &tp->args[i], is_return);
851 852
		if (ret) {
			pr_info("Parse error at argument%d. (%d)\n", i, ret);
L
Lai Jiangshan 已提交
853
			kfree(tp->args[i].name);
854
			goto error;
855
		}
L
Lai Jiangshan 已提交
856 857

		tp->nr_args++;
858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904
	}

	ret = register_trace_probe(tp);
	if (ret)
		goto error;
	return 0;

error:
	free_trace_probe(tp);
	return ret;
}

static void cleanup_all_probes(void)
{
	struct trace_probe *tp;

	mutex_lock(&probe_lock);
	/* TODO: Use batch unregistration */
	while (!list_empty(&probe_list)) {
		tp = list_entry(probe_list.next, struct trace_probe, list);
		unregister_trace_probe(tp);
		free_trace_probe(tp);
	}
	mutex_unlock(&probe_lock);
}


/* Probes listing interfaces */
static void *probes_seq_start(struct seq_file *m, loff_t *pos)
{
	mutex_lock(&probe_lock);
	return seq_list_start(&probe_list, *pos);
}

static void *probes_seq_next(struct seq_file *m, void *v, loff_t *pos)
{
	return seq_list_next(v, &probe_list, pos);
}

static void probes_seq_stop(struct seq_file *m, void *v)
{
	mutex_unlock(&probe_lock);
}

static int probes_seq_show(struct seq_file *m, void *v)
{
	struct trace_probe *tp = v;
905
	int i;
906 907

	seq_printf(m, "%c", probe_is_return(tp) ? 'r' : 'p');
908
	seq_printf(m, ":%s/%s", tp->call.class->system, tp->call.name);
909

910 911 912
	if (!tp->symbol)
		seq_printf(m, " 0x%p", tp->rp.kp.addr);
	else if (tp->rp.kp.offset)
913
		seq_printf(m, " %s+%u", probe_symbol(tp), tp->rp.kp.offset);
914
	else
915
		seq_printf(m, " %s", probe_symbol(tp));
916

917 918
	for (i = 0; i < tp->nr_args; i++)
		seq_printf(m, " %s=%s", tp->args[i].name, tp->args[i].comm);
919
	seq_printf(m, "\n");
920

921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
	return 0;
}

static const struct seq_operations probes_seq_op = {
	.start  = probes_seq_start,
	.next   = probes_seq_next,
	.stop   = probes_seq_stop,
	.show   = probes_seq_show
};

static int probes_open(struct inode *inode, struct file *file)
{
	if ((file->f_mode & FMODE_WRITE) &&
	    (file->f_flags & O_TRUNC))
		cleanup_all_probes();

	return seq_open(file, &probes_seq_op);
}

static int command_trace_probe(const char *buf)
{
	char **argv;
	int argc = 0, ret = 0;

	argv = argv_split(GFP_KERNEL, buf, &argc);
	if (!argv)
		return -ENOMEM;

	if (argc)
		ret = create_trace_probe(argc, argv);

	argv_free(argv);
	return ret;
}

#define WRITE_BUFSIZE 128

static ssize_t probes_write(struct file *file, const char __user *buffer,
			    size_t count, loff_t *ppos)
{
	char *kbuf, *tmp;
	int ret;
	size_t done;
	size_t size;

	kbuf = kmalloc(WRITE_BUFSIZE, GFP_KERNEL);
	if (!kbuf)
		return -ENOMEM;

	ret = done = 0;
	while (done < count) {
		size = count - done;
		if (size >= WRITE_BUFSIZE)
			size = WRITE_BUFSIZE - 1;
		if (copy_from_user(kbuf, buffer + done, size)) {
			ret = -EFAULT;
			goto out;
		}
		kbuf[size] = '\0';
		tmp = strchr(kbuf, '\n');
		if (tmp) {
			*tmp = '\0';
			size = tmp - kbuf + 1;
		} else if (done + size < count) {
			pr_warning("Line length is too long: "
				   "Should be less than %d.", WRITE_BUFSIZE);
			ret = -EINVAL;
			goto out;
		}
		done += size;
		/* Remove comments */
		tmp = strchr(kbuf, '#');
		if (tmp)
			*tmp = '\0';

		ret = command_trace_probe(kbuf);
		if (ret)
			goto out;
	}
	ret = done;
out:
	kfree(kbuf);
	return ret;
}

static const struct file_operations kprobe_events_ops = {
	.owner          = THIS_MODULE,
	.open           = probes_open,
	.read           = seq_read,
	.llseek         = seq_lseek,
	.release        = seq_release,
	.write		= probes_write,
};

1015 1016 1017 1018 1019 1020
/* Probes profiling interfaces */
static int probes_profile_seq_show(struct seq_file *m, void *v)
{
	struct trace_probe *tp = v;

	seq_printf(m, "  %-44s %15lu %15lu\n", tp->call.name, tp->nhit,
1021
		   tp->rp.kp.nmissed);
1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045

	return 0;
}

static const struct seq_operations profile_seq_op = {
	.start  = probes_seq_start,
	.next   = probes_seq_next,
	.stop   = probes_seq_stop,
	.show   = probes_profile_seq_show
};

static int profile_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &profile_seq_op);
}

static const struct file_operations kprobe_profile_ops = {
	.owner          = THIS_MODULE,
	.open           = profile_open,
	.read           = seq_read,
	.llseek         = seq_lseek,
	.release        = seq_release,
};

1046
/* Kprobe handler */
1047
static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs)
1048
{
1049
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
1050
	struct kprobe_trace_entry_head *entry;
1051
	struct ring_buffer_event *event;
1052
	struct ring_buffer *buffer;
1053
	u8 *data;
1054 1055
	int size, i, pc;
	unsigned long irq_flags;
1056
	struct ftrace_event_call *call = &tp->call;
1057

1058 1059
	tp->nhit++;

1060 1061 1062
	local_save_flags(irq_flags);
	pc = preempt_count();

1063
	size = sizeof(*entry) + tp->size;
1064

1065 1066
	event = trace_current_buffer_lock_reserve(&buffer, call->event.type,
						  size, irq_flags, pc);
1067
	if (!event)
1068
		return;
1069 1070 1071

	entry = ring_buffer_event_data(event);
	entry->ip = (unsigned long)kp->addr;
1072
	data = (u8 *)&entry[1];
1073
	for (i = 0; i < tp->nr_args; i++)
1074
		call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
1075

1076 1077
	if (!filter_current_check_discard(buffer, call, entry, event))
		trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc);
1078 1079 1080
}

/* Kretprobe handler */
1081
static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri,
1082 1083 1084
					  struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp);
1085
	struct kretprobe_trace_entry_head *entry;
1086
	struct ring_buffer_event *event;
1087
	struct ring_buffer *buffer;
1088
	u8 *data;
1089 1090
	int size, i, pc;
	unsigned long irq_flags;
1091
	struct ftrace_event_call *call = &tp->call;
1092 1093 1094 1095

	local_save_flags(irq_flags);
	pc = preempt_count();

1096
	size = sizeof(*entry) + tp->size;
1097

1098 1099
	event = trace_current_buffer_lock_reserve(&buffer, call->event.type,
						  size, irq_flags, pc);
1100
	if (!event)
1101
		return;
1102 1103

	entry = ring_buffer_event_data(event);
1104
	entry->func = (unsigned long)tp->rp.kp.addr;
1105
	entry->ret_ip = (unsigned long)ri->ret_addr;
1106
	data = (u8 *)&entry[1];
1107
	for (i = 0; i < tp->nr_args; i++)
1108
		call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
1109

1110 1111
	if (!filter_current_check_discard(buffer, call, entry, event))
		trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc);
1112 1113 1114 1115
}

/* Event entry printers */
enum print_line_t
1116 1117
print_kprobe_event(struct trace_iterator *iter, int flags,
		   struct trace_event *event)
1118
{
1119
	struct kprobe_trace_entry_head *field;
1120
	struct trace_seq *s = &iter->seq;
1121
	struct trace_probe *tp;
1122
	u8 *data;
1123 1124
	int i;

1125
	field = (struct kprobe_trace_entry_head *)iter->ent;
1126
	tp = container_of(event, struct trace_probe, call.event);
1127

1128 1129 1130
	if (!trace_seq_printf(s, "%s: (", tp->call.name))
		goto partial;

1131 1132 1133
	if (!seq_print_ip_sym(s, field->ip, flags | TRACE_ITER_SYM_OFFSET))
		goto partial;

1134
	if (!trace_seq_puts(s, ")"))
1135 1136
		goto partial;

1137 1138 1139 1140
	data = (u8 *)&field[1];
	for (i = 0; i < tp->nr_args; i++)
		if (!tp->args[i].type->print(s, tp->args[i].name,
					     data + tp->args[i].offset))
1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
			goto partial;

	if (!trace_seq_puts(s, "\n"))
		goto partial;

	return TRACE_TYPE_HANDLED;
partial:
	return TRACE_TYPE_PARTIAL_LINE;
}

enum print_line_t
1152 1153
print_kretprobe_event(struct trace_iterator *iter, int flags,
		      struct trace_event *event)
1154
{
1155
	struct kretprobe_trace_entry_head *field;
1156
	struct trace_seq *s = &iter->seq;
1157
	struct trace_probe *tp;
1158
	u8 *data;
1159 1160
	int i;

1161
	field = (struct kretprobe_trace_entry_head *)iter->ent;
1162
	tp = container_of(event, struct trace_probe, call.event);
1163

1164 1165 1166
	if (!trace_seq_printf(s, "%s: (", tp->call.name))
		goto partial;

1167 1168 1169 1170 1171 1172 1173 1174 1175
	if (!seq_print_ip_sym(s, field->ret_ip, flags | TRACE_ITER_SYM_OFFSET))
		goto partial;

	if (!trace_seq_puts(s, " <- "))
		goto partial;

	if (!seq_print_ip_sym(s, field->func, flags & ~TRACE_ITER_SYM_OFFSET))
		goto partial;

1176
	if (!trace_seq_puts(s, ")"))
1177 1178
		goto partial;

1179 1180 1181 1182
	data = (u8 *)&field[1];
	for (i = 0; i < tp->nr_args; i++)
		if (!tp->args[i].type->print(s, tp->args[i].name,
					     data + tp->args[i].offset))
1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
			goto partial;

	if (!trace_seq_puts(s, "\n"))
		goto partial;

	return TRACE_TYPE_HANDLED;
partial:
	return TRACE_TYPE_PARTIAL_LINE;
}

static int probe_event_enable(struct ftrace_event_call *call)
{
	struct trace_probe *tp = (struct trace_probe *)call->data;

1197 1198
	tp->flags |= TP_FLAG_TRACE;
	if (probe_is_return(tp))
1199
		return enable_kretprobe(&tp->rp);
1200
	else
1201
		return enable_kprobe(&tp->rp.kp);
1202 1203 1204 1205 1206 1207
}

static void probe_event_disable(struct ftrace_event_call *call)
{
	struct trace_probe *tp = (struct trace_probe *)call->data;

1208 1209 1210 1211 1212 1213 1214
	tp->flags &= ~TP_FLAG_TRACE;
	if (!(tp->flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE))) {
		if (probe_is_return(tp))
			disable_kretprobe(&tp->rp);
		else
			disable_kprobe(&tp->rp.kp);
	}
1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
}

#undef DEFINE_FIELD
#define DEFINE_FIELD(type, item, name, is_signed)			\
	do {								\
		ret = trace_define_field(event_call, #type, name,	\
					 offsetof(typeof(field), item),	\
					 sizeof(field.item), is_signed, \
					 FILTER_OTHER);			\
		if (ret)						\
			return ret;					\
	} while (0)

static int kprobe_event_define_fields(struct ftrace_event_call *event_call)
{
	int ret, i;
1231
	struct kprobe_trace_entry_head field;
1232 1233
	struct trace_probe *tp = (struct trace_probe *)event_call->data;

1234
	DEFINE_FIELD(unsigned long, ip, FIELD_STRING_IP, 0);
1235
	/* Set argument names as fields */
1236 1237 1238 1239 1240 1241 1242 1243 1244 1245
	for (i = 0; i < tp->nr_args; i++) {
		ret = trace_define_field(event_call, tp->args[i].type->name,
					 tp->args[i].name,
					 sizeof(field) + tp->args[i].offset,
					 tp->args[i].type->size,
					 tp->args[i].type->is_signed,
					 FILTER_OTHER);
		if (ret)
			return ret;
	}
1246 1247 1248 1249 1250 1251
	return 0;
}

static int kretprobe_event_define_fields(struct ftrace_event_call *event_call)
{
	int ret, i;
1252
	struct kretprobe_trace_entry_head field;
1253 1254
	struct trace_probe *tp = (struct trace_probe *)event_call->data;

1255 1256
	DEFINE_FIELD(unsigned long, func, FIELD_STRING_FUNC, 0);
	DEFINE_FIELD(unsigned long, ret_ip, FIELD_STRING_RETIP, 0);
1257
	/* Set argument names as fields */
1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
	for (i = 0; i < tp->nr_args; i++) {
		ret = trace_define_field(event_call, tp->args[i].type->name,
					 tp->args[i].name,
					 sizeof(field) + tp->args[i].offset,
					 tp->args[i].type->size,
					 tp->args[i].type->is_signed,
					 FILTER_OTHER);
		if (ret)
			return ret;
	}
1268 1269 1270
	return 0;
}

1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
static int __set_print_fmt(struct trace_probe *tp, char *buf, int len)
{
	int i;
	int pos = 0;

	const char *fmt, *arg;

	if (!probe_is_return(tp)) {
		fmt = "(%lx)";
		arg = "REC->" FIELD_STRING_IP;
	} else {
		fmt = "(%lx <- %lx)";
		arg = "REC->" FIELD_STRING_FUNC ", REC->" FIELD_STRING_RETIP;
	}

	/* When len=0, we just calculate the needed length */
#define LEN_OR_ZERO (len ? len - pos : 0)

	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"%s", fmt);

	for (i = 0; i < tp->nr_args; i++) {
1292 1293
		pos += snprintf(buf + pos, LEN_OR_ZERO, " %s=%s",
				tp->args[i].name, tp->args[i].type->fmt);
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326
	}

	pos += snprintf(buf + pos, LEN_OR_ZERO, "\", %s", arg);

	for (i = 0; i < tp->nr_args; i++) {
		pos += snprintf(buf + pos, LEN_OR_ZERO, ", REC->%s",
				tp->args[i].name);
	}

#undef LEN_OR_ZERO

	/* return the length of print_fmt */
	return pos;
}

static int set_print_fmt(struct trace_probe *tp)
{
	int len;
	char *print_fmt;

	/* First: called with 0 length to calculate the needed length */
	len = __set_print_fmt(tp, NULL, 0);
	print_fmt = kmalloc(len + 1, GFP_KERNEL);
	if (!print_fmt)
		return -ENOMEM;

	/* Second: actually write the @print_fmt */
	__set_print_fmt(tp, print_fmt, len + 1);
	tp->call.print_fmt = print_fmt;

	return 0;
}

1327
#ifdef CONFIG_PERF_EVENTS
1328 1329

/* Kprobe profile handler */
1330
static __kprobes void kprobe_perf_func(struct kprobe *kp,
1331 1332 1333 1334
					 struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
	struct ftrace_event_call *call = &tp->call;
1335
	struct kprobe_trace_entry_head *entry;
1336
	struct hlist_head *head;
1337
	u8 *data;
1338
	int size, __size, i;
1339
	int rctx;
1340

1341
	__size = sizeof(*entry) + tp->size;
1342 1343
	size = ALIGN(__size + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);
1344
	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
1345
		     "profile buffer not large enough"))
1346
		return;
1347

S
Steven Rostedt 已提交
1348
	entry = perf_trace_buf_prepare(size, call->event.type, regs, &rctx);
1349
	if (!entry)
1350
		return;
1351 1352

	entry->ip = (unsigned long)kp->addr;
1353
	data = (u8 *)&entry[1];
1354
	for (i = 0; i < tp->nr_args; i++)
1355
		call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
1356

1357
	head = this_cpu_ptr(call->perf_events);
1358
	perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head);
1359 1360 1361
}

/* Kretprobe profile handler */
1362
static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri,
1363 1364 1365 1366
					    struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp);
	struct ftrace_event_call *call = &tp->call;
1367
	struct kretprobe_trace_entry_head *entry;
1368
	struct hlist_head *head;
1369
	u8 *data;
1370
	int size, __size, i;
1371
	int rctx;
1372

1373
	__size = sizeof(*entry) + tp->size;
1374 1375
	size = ALIGN(__size + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);
1376
	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
1377
		     "profile buffer not large enough"))
1378
		return;
1379

S
Steven Rostedt 已提交
1380
	entry = perf_trace_buf_prepare(size, call->event.type, regs, &rctx);
1381
	if (!entry)
1382
		return;
1383

1384 1385
	entry->func = (unsigned long)tp->rp.kp.addr;
	entry->ret_ip = (unsigned long)ri->ret_addr;
1386
	data = (u8 *)&entry[1];
1387
	for (i = 0; i < tp->nr_args; i++)
1388
		call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset);
1389

1390
	head = this_cpu_ptr(call->perf_events);
1391
	perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head);
1392 1393
}

1394
static int probe_perf_enable(struct ftrace_event_call *call)
1395 1396 1397
{
	struct trace_probe *tp = (struct trace_probe *)call->data;

1398
	tp->flags |= TP_FLAG_PROFILE;
1399

1400
	if (probe_is_return(tp))
1401
		return enable_kretprobe(&tp->rp);
1402
	else
1403 1404 1405
		return enable_kprobe(&tp->rp.kp);
}

1406
static void probe_perf_disable(struct ftrace_event_call *call)
1407
{
1408 1409
	struct trace_probe *tp = (struct trace_probe *)call->data;

1410
	tp->flags &= ~TP_FLAG_PROFILE;
1411

1412
	if (!(tp->flags & TP_FLAG_TRACE)) {
1413 1414 1415 1416 1417
		if (probe_is_return(tp))
			disable_kretprobe(&tp->rp);
		else
			disable_kprobe(&tp->rp.kp);
	}
1418
}
1419
#endif	/* CONFIG_PERF_EVENTS */
1420

1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440
static __kprobes
int kprobe_register(struct ftrace_event_call *event, enum trace_reg type)
{
	switch (type) {
	case TRACE_REG_REGISTER:
		return probe_event_enable(event);
	case TRACE_REG_UNREGISTER:
		probe_event_disable(event);
		return 0;

#ifdef CONFIG_PERF_EVENTS
	case TRACE_REG_PERF_REGISTER:
		return probe_perf_enable(event);
	case TRACE_REG_PERF_UNREGISTER:
		probe_perf_disable(event);
		return 0;
#endif
	}
	return 0;
}
1441 1442 1443 1444 1445

static __kprobes
int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
1446

1447 1448
	if (tp->flags & TP_FLAG_TRACE)
		kprobe_trace_func(kp, regs);
1449
#ifdef CONFIG_PERF_EVENTS
1450
	if (tp->flags & TP_FLAG_PROFILE)
1451
		kprobe_perf_func(kp, regs);
1452
#endif
1453 1454 1455 1456 1457 1458 1459 1460 1461 1462
	return 0;	/* We don't tweek kernel, so just return 0 */
}

static __kprobes
int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp);

	if (tp->flags & TP_FLAG_TRACE)
		kretprobe_trace_func(ri, regs);
1463
#ifdef CONFIG_PERF_EVENTS
1464
	if (tp->flags & TP_FLAG_PROFILE)
1465
		kretprobe_perf_func(ri, regs);
1466
#endif
1467 1468
	return 0;	/* We don't tweek kernel, so just return 0 */
}
1469

1470 1471 1472 1473 1474 1475 1476 1477
static struct trace_event_functions kretprobe_funcs = {
	.trace		= print_kretprobe_event
};

static struct trace_event_functions kprobe_funcs = {
	.trace		= print_kprobe_event
};

1478 1479 1480 1481 1482 1483
static int register_probe_event(struct trace_probe *tp)
{
	struct ftrace_event_call *call = &tp->call;
	int ret;

	/* Initialize ftrace_event_call */
1484
	INIT_LIST_HEAD(&call->class->fields);
1485
	if (probe_is_return(tp)) {
1486
		call->event.funcs = &kretprobe_funcs;
1487
		call->class->define_fields = kretprobe_event_define_fields;
1488
	} else {
1489
		call->event.funcs = &kprobe_funcs;
1490
		call->class->define_fields = kprobe_event_define_fields;
1491
	}
1492 1493
	if (set_print_fmt(tp) < 0)
		return -ENOMEM;
1494 1495
	ret = register_ftrace_event(&call->event);
	if (!ret) {
1496
		kfree(call->print_fmt);
1497
		return -ENODEV;
1498
	}
1499
	call->flags = 0;
1500
	call->class->reg = kprobe_register;
1501 1502
	call->data = tp;
	ret = trace_add_event_call(call);
1503
	if (ret) {
1504
		pr_info("Failed to register kprobe event: %s\n", call->name);
1505
		kfree(call->print_fmt);
1506
		unregister_ftrace_event(&call->event);
1507
	}
1508 1509 1510 1511 1512
	return ret;
}

static void unregister_probe_event(struct trace_probe *tp)
{
1513
	/* tp->event is unregistered in trace_remove_event_call() */
1514
	trace_remove_event_call(&tp->call);
1515
	kfree(tp->call.print_fmt);
1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530
}

/* Make a debugfs interface for controling probe points */
static __init int init_kprobe_trace(void)
{
	struct dentry *d_tracer;
	struct dentry *entry;

	d_tracer = tracing_init_dentry();
	if (!d_tracer)
		return 0;

	entry = debugfs_create_file("kprobe_events", 0644, d_tracer,
				    NULL, &kprobe_events_ops);

1531
	/* Event list interface */
1532 1533 1534
	if (!entry)
		pr_warning("Could not create debugfs "
			   "'kprobe_events' entry\n");
1535 1536 1537 1538 1539 1540 1541 1542

	/* Profile interface */
	entry = debugfs_create_file("kprobe_profile", 0444, d_tracer,
				    NULL, &kprobe_profile_ops);

	if (!entry)
		pr_warning("Could not create debugfs "
			   "'kprobe_profile' entry\n");
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557
	return 0;
}
fs_initcall(init_kprobe_trace);


#ifdef CONFIG_FTRACE_STARTUP_TEST

static int kprobe_trace_selftest_target(int a1, int a2, int a3,
					int a4, int a5, int a6)
{
	return a1 + a2 + a3 + a4 + a5 + a6;
}

static __init int kprobe_trace_self_tests_init(void)
{
1558
	int ret, warn = 0;
1559
	int (*target)(int, int, int, int, int, int);
1560
	struct trace_probe *tp;
1561 1562 1563 1564 1565 1566

	target = kprobe_trace_selftest_target;

	pr_info("Testing kprobe tracing: ");

	ret = command_trace_probe("p:testprobe kprobe_trace_selftest_target "
1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579
				  "$stack $stack0 +0($stack)");
	if (WARN_ON_ONCE(ret)) {
		pr_warning("error on probing function entry.\n");
		warn++;
	} else {
		/* Enable trace point */
		tp = find_probe_event("testprobe", KPROBE_EVENT_SYSTEM);
		if (WARN_ON_ONCE(tp == NULL)) {
			pr_warning("error on getting new probe.\n");
			warn++;
		} else
			probe_event_enable(&tp->call);
	}
1580 1581

	ret = command_trace_probe("r:testprobe2 kprobe_trace_selftest_target "
1582
				  "$retval");
1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597
	if (WARN_ON_ONCE(ret)) {
		pr_warning("error on probing function return.\n");
		warn++;
	} else {
		/* Enable trace point */
		tp = find_probe_event("testprobe2", KPROBE_EVENT_SYSTEM);
		if (WARN_ON_ONCE(tp == NULL)) {
			pr_warning("error on getting new probe.\n");
			warn++;
		} else
			probe_event_enable(&tp->call);
	}

	if (warn)
		goto end;
1598 1599 1600

	ret = target(1, 2, 3, 4, 5, 6);

1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611
	ret = command_trace_probe("-:testprobe");
	if (WARN_ON_ONCE(ret)) {
		pr_warning("error on deleting a probe.\n");
		warn++;
	}

	ret = command_trace_probe("-:testprobe2");
	if (WARN_ON_ONCE(ret)) {
		pr_warning("error on deleting a probe.\n");
		warn++;
	}
1612

1613 1614 1615 1616 1617 1618
end:
	cleanup_all_probes();
	if (warn)
		pr_cont("NG: Some tests are failed. Please check them.\n");
	else
		pr_cont("OK\n");
1619 1620 1621 1622 1623 1624
	return 0;
}

late_initcall(kprobe_trace_self_tests_init);

#endif