trace_kprobe.c 45.6 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
#include <linux/stringify.h>
33
#include <linux/limits.h>
34
#include <asm/bitsperlong.h>
35 36 37 38

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

39
#define MAX_TRACE_ARGS 128
40
#define MAX_ARGSTR_LEN 63
41
#define MAX_EVENT_NAME_LEN 64
42
#define MAX_STRING_SIZE PATH_MAX
43
#define KPROBE_EVENT_SYSTEM "kprobes"
44

45
/* Reserved field names */
46 47 48
#define FIELD_STRING_IP "__probe_ip"
#define FIELD_STRING_RETIP "__probe_ret_ip"
#define FIELD_STRING_FUNC "__probe_func"
49 50 51 52 53 54 55 56 57 58 59 60 61

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,
};

62
/* Printing function type */
63 64
typedef int (*print_type_func_t)(struct trace_seq *, const char *, void *,
				 void *);
65 66 67 68 69 70
#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,	\
71 72
						const char *name,	\
						void *data, void *ent)\
73 74 75 76 77 78 79 80 81 82 83 84 85 86
{									\
	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)

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
/* data_rloc: data relative location, compatible with u32 */
#define make_data_rloc(len, roffs)	\
	(((u32)(len) << 16) | ((u32)(roffs) & 0xffff))
#define get_rloc_len(dl)	((u32)(dl) >> 16)
#define get_rloc_offs(dl)	((u32)(dl) & 0xffff)

static inline void *get_rloc_data(u32 *dl)
{
	return (u8 *)dl + get_rloc_offs(*dl);
}

/* For data_loc conversion */
static inline void *get_loc_data(u32 *dl, void *ent)
{
	return (u8 *)ent + get_rloc_offs(*dl);
}

/*
 * Convert data_rloc to data_loc:
 *  data_rloc stores the offset from data_rloc itself, but data_loc
 *  stores the offset from event entry.
 */
#define convert_rloc_to_loc(dl, offs)	((u32)(dl) + (offs))

/* For defining macros, define string/string_size types */
typedef u32 string;
typedef u32 string_size;

/* Print type function for string type */
static __kprobes int PRINT_TYPE_FUNC_NAME(string)(struct trace_seq *s,
						  const char *name,
						  void *data, void *ent)
{
	int len = *(u32 *)data >> 16;

	if (!len)
		return trace_seq_printf(s, " %s=(fault)", name);
	else
		return trace_seq_printf(s, " %s=\"%s\"", name,
					(const char *)get_loc_data(data, ent));
}
static const char PRINT_TYPE_FMT_NAME(string)[] = "\\\"%s\\\"";

130 131 132 133 134
/* Data fetch function type */
typedef	void (*fetch_func_t)(struct pt_regs *, void *, void *);

struct fetch_param {
	fetch_func_t	fn;
135 136 137
	void *data;
};

138 139
static __kprobes void call_fetch(struct fetch_param *fprm,
				 struct pt_regs *regs, void *dest)
140
{
141
	return fprm->fn(regs, fprm->data, dest);
142 143
}

144
#define FETCH_FUNC_NAME(method, type)	fetch_##method##_##type
145 146 147 148
/*
 * Define macro for basic types - we don't need to define s* types, because
 * we have to care only about bitwidth at recording time.
 */
149 150 151 152 153 154 155 156 157 158 159 160 161 162
#define DEFINE_BASIC_FETCH_FUNCS(method) \
DEFINE_FETCH_##method(u8)		\
DEFINE_FETCH_##method(u16)		\
DEFINE_FETCH_##method(u32)		\
DEFINE_FETCH_##method(u64)

#define CHECK_FETCH_FUNCS(method, fn)			\
	(((FETCH_FUNC_NAME(method, u8) == fn) ||	\
	  (FETCH_FUNC_NAME(method, u16) == fn) ||	\
	  (FETCH_FUNC_NAME(method, u32) == fn) ||	\
	  (FETCH_FUNC_NAME(method, u64) == fn) ||	\
	  (FETCH_FUNC_NAME(method, string) == fn) ||	\
	  (FETCH_FUNC_NAME(method, string_size) == fn)) \
	 && (fn != NULL))
163 164 165 166

/* Data fetch function templates */
#define DEFINE_FETCH_reg(type)						\
static __kprobes void FETCH_FUNC_NAME(reg, type)(struct pt_regs *regs,	\
167
					void *offset, void *dest)	\
168 169 170
{									\
	*(type *)dest = (type)regs_get_register(regs,			\
				(unsigned int)((unsigned long)offset));	\
171
}
172
DEFINE_BASIC_FETCH_FUNCS(reg)
173 174 175
/* No string on the register */
#define fetch_reg_string NULL
#define fetch_reg_string_size NULL
176 177 178 179 180 181 182

#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));	\
183
}
184
DEFINE_BASIC_FETCH_FUNCS(stack)
185 186 187
/* No string on the stack entry */
#define fetch_stack_string NULL
#define fetch_stack_string_size NULL
188

189 190 191 192 193
#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);			\
194
}
195
DEFINE_BASIC_FETCH_FUNCS(retval)
196 197 198
/* No string on the retval */
#define fetch_retval_string NULL
#define fetch_retval_string_size NULL
199 200 201 202 203 204 205 206 207 208

#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;					\
209
}
210
DEFINE_BASIC_FETCH_FUNCS(memory)
211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
/*
 * Fetch a null-terminated string. Caller MUST set *(u32 *)dest with max
 * length and relative data location.
 */
static __kprobes void FETCH_FUNC_NAME(memory, string)(struct pt_regs *regs,
						      void *addr, void *dest)
{
	long ret;
	int maxlen = get_rloc_len(*(u32 *)dest);
	u8 *dst = get_rloc_data(dest);
	u8 *src = addr;
	mm_segment_t old_fs = get_fs();
	if (!maxlen)
		return;
	/*
	 * Try to get string again, since the string can be changed while
	 * probing.
	 */
	set_fs(KERNEL_DS);
	pagefault_disable();
	do
		ret = __copy_from_user_inatomic(dst++, src++, 1);
	while (dst[-1] && ret == 0 && src - (u8 *)addr < maxlen);
	dst[-1] = '\0';
	pagefault_enable();
	set_fs(old_fs);

	if (ret < 0) {	/* Failed to fetch string */
		((u8 *)get_rloc_data(dest))[0] = '\0';
		*(u32 *)dest = make_data_rloc(0, get_rloc_offs(*(u32 *)dest));
	} else
		*(u32 *)dest = make_data_rloc(src - (u8 *)addr,
					      get_rloc_offs(*(u32 *)dest));
}
/* Return the length of string -- including null terminal byte */
static __kprobes void FETCH_FUNC_NAME(memory, string_size)(struct pt_regs *regs,
							void *addr, void *dest)
{
	int ret, len = 0;
	u8 c;
	mm_segment_t old_fs = get_fs();

	set_fs(KERNEL_DS);
	pagefault_disable();
	do {
		ret = __copy_from_user_inatomic(&c, (u8 *)addr + len, 1);
		len++;
	} while (c && ret == 0 && len < MAX_STRING_SIZE);
	pagefault_enable();
	set_fs(old_fs);

	if (ret < 0)	/* Failed to check the length */
		*(u32 *)dest = 0;
	else
		*(u32 *)dest = len;
}
267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309

/* 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;
}

310 311 312 313 314 315 316 317 318
#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;					\
319
}
320
DEFINE_BASIC_FETCH_FUNCS(symbol)
321 322
DEFINE_FETCH_symbol(string)
DEFINE_FETCH_symbol(string_size)
323

324 325 326
/* Dereference memory access function */
struct deref_fetch_param {
	struct fetch_param orig;
327 328 329
	long offset;
};

330 331 332 333 334 335 336 337 338 339 340 341
#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;					\
342
}
343
DEFINE_BASIC_FETCH_FUNCS(deref)
344 345
DEFINE_FETCH_deref(string)
DEFINE_FETCH_deref(string_size)
346

347
static __kprobes void free_deref_fetch_param(struct deref_fetch_param *data)
348
{
349
	if (CHECK_FETCH_FUNCS(deref, data->orig.fn))
350
		free_deref_fetch_param(data->orig.data);
351
	else if (CHECK_FETCH_FUNCS(symbol, data->orig.fn))
352 353 354 355
		free_symbol_cache(data->orig.data);
	kfree(data);
}

356 357 358 359 360 361
/* 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)

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
/* Fetch types */
enum {
	FETCH_MTD_reg = 0,
	FETCH_MTD_stack,
	FETCH_MTD_retval,
	FETCH_MTD_memory,
	FETCH_MTD_symbol,
	FETCH_MTD_deref,
	FETCH_MTD_END,
};

#define ASSIGN_FETCH_FUNC(method, type)	\
	[FETCH_MTD_##method] = FETCH_FUNC_NAME(method, type)

#define __ASSIGN_FETCH_TYPE(_name, ptype, ftype, _size, sign, _fmttype)	\
	{.name = _name,				\
	 .size = _size,					\
	 .is_signed = sign,				\
	 .print = PRINT_TYPE_FUNC_NAME(ptype),		\
	 .fmt = PRINT_TYPE_FMT_NAME(ptype),		\
	 .fmttype = _fmttype,				\
	 .fetch = {					\
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),			\
	  }						\
391 392
	}

393 394 395 396 397 398
#define ASSIGN_FETCH_TYPE(ptype, ftype, sign)			\
	__ASSIGN_FETCH_TYPE(#ptype, ptype, ftype, sizeof(ftype), sign, #ptype)

#define FETCH_TYPE_STRING 0
#define FETCH_TYPE_STRSIZE 1

399 400 401 402 403 404 405
/* 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 */
406
	const char	*fmttype;	/* Name in format file */
407
	/* Fetch functions */
408
	fetch_func_t	fetch[FETCH_MTD_END];
409
} fetch_type_table[] = {
410 411 412 413 414 415
	/* Special types */
	[FETCH_TYPE_STRING] = __ASSIGN_FETCH_TYPE("string", string, string,
					sizeof(u32), 1, "__data_loc char[]"),
	[FETCH_TYPE_STRSIZE] = __ASSIGN_FETCH_TYPE("string_size", u32,
					string_size, sizeof(u32), 0, "u32"),
	/* Basic types */
416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445
	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);
}

446 447 448 449 450 451 452 453 454 455 456 457 458 459 460
static fetch_func_t get_fetch_size_function(const struct fetch_type *type,
					    fetch_func_t orig_fn)
{
	int i;

	if (type != &fetch_type_table[FETCH_TYPE_STRING])
		return NULL;	/* Only string type needs size function */
	for (i = 0; i < FETCH_MTD_END; i++)
		if (type->fetch[i] == orig_fn)
			return fetch_type_table[FETCH_TYPE_STRSIZE].fetch[i];

	WARN_ON(1);	/* This should not happen */
	return NULL;
}

461
/**
462
 * Kprobe event core functions
463 464
 */

465
struct probe_arg {
466
	struct fetch_param	fetch;
467
	struct fetch_param	fetch_size;
468 469 470 471
	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 */
472 473
};

474 475 476 477
/* Flags for trace_probe */
#define TP_FLAG_TRACE	1
#define TP_FLAG_PROFILE	2

478 479
struct trace_probe {
	struct list_head	list;
480
	struct kretprobe	rp;	/* Use rp.kp for kprobe use */
481
	unsigned long 		nhit;
482
	unsigned int		flags;	/* For TP_FLAG_* */
483
	const char		*symbol;	/* symbol name */
484
	struct ftrace_event_class	class;
485
	struct ftrace_event_call	call;
486
	ssize_t			size;		/* trace entry size */
487
	unsigned int		nr_args;
488
	struct probe_arg	args[];
489 490
};

491 492
#define SIZEOF_TRACE_PROBE(n)			\
	(offsetof(struct trace_probe, args) +	\
493
	(sizeof(struct probe_arg) * (n)))
494

495

496 497
static __kprobes int probe_is_return(struct trace_probe *tp)
{
498
	return tp->rp.handler != NULL;
499 500 501 502 503 504 505 506 507 508 509 510 511
}

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);

512 513 514 515
static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs);
static int kretprobe_dispatcher(struct kretprobe_instance *ri,
				struct pt_regs *regs);

516 517
/* Check the name is good for event/group/fields */
static int is_good_name(const char *name)
518 519 520 521 522 523 524 525 526 527
{
	if (!isalpha(*name) && *name != '_')
		return 0;
	while (*++name != '\0') {
		if (!isalpha(*name) && !isdigit(*name) && *name != '_')
			return 0;
	}
	return 1;
}

528 529 530
/*
 * Allocate new trace_probe and initialize it (including kprobes).
 */
531 532
static struct trace_probe *alloc_trace_probe(const char *group,
					     const char *event,
533 534 535 536
					     void *addr,
					     const char *symbol,
					     unsigned long offs,
					     int nargs, int is_return)
537 538
{
	struct trace_probe *tp;
539
	int ret = -ENOMEM;
540

541
	tp = kzalloc(SIZEOF_TRACE_PROBE(nargs), GFP_KERNEL);
542
	if (!tp)
543
		return ERR_PTR(ret);
544 545 546 547 548

	if (symbol) {
		tp->symbol = kstrdup(symbol, GFP_KERNEL);
		if (!tp->symbol)
			goto error;
549 550 551 552 553 554
		tp->rp.kp.symbol_name = tp->symbol;
		tp->rp.kp.offset = offs;
	} else
		tp->rp.kp.addr = addr;

	if (is_return)
555
		tp->rp.handler = kretprobe_dispatcher;
556
	else
557
		tp->rp.kp.pre_handler = kprobe_dispatcher;
558

559
	if (!event || !is_good_name(event)) {
560
		ret = -EINVAL;
561
		goto error;
562 563
	}

564
	tp->call.class = &tp->class;
565 566 567
	tp->call.name = kstrdup(event, GFP_KERNEL);
	if (!tp->call.name)
		goto error;
568

569
	if (!group || !is_good_name(group)) {
570
		ret = -EINVAL;
571
		goto error;
572 573
	}

574 575
	tp->class.system = kstrdup(group, GFP_KERNEL);
	if (!tp->class.system)
576 577
		goto error;

578 579 580
	INIT_LIST_HEAD(&tp->list);
	return tp;
error:
581
	kfree(tp->call.name);
582 583
	kfree(tp->symbol);
	kfree(tp);
584
	return ERR_PTR(ret);
585 586
}

587 588
static void free_probe_arg(struct probe_arg *arg)
{
589
	if (CHECK_FETCH_FUNCS(deref, arg->fetch.fn))
590
		free_deref_fetch_param(arg->fetch.data);
591
	else if (CHECK_FETCH_FUNCS(symbol, arg->fetch.fn))
592 593
		free_symbol_cache(arg->fetch.data);
	kfree(arg->name);
594
	kfree(arg->comm);
595 596
}

597 598 599 600 601
static void free_trace_probe(struct trace_probe *tp)
{
	int i;

	for (i = 0; i < tp->nr_args; i++)
602
		free_probe_arg(&tp->args[i]);
603

604
	kfree(tp->call.class->system);
605 606 607 608 609
	kfree(tp->call.name);
	kfree(tp->symbol);
	kfree(tp);
}

610 611
static struct trace_probe *find_probe_event(const char *event,
					    const char *group)
612 613 614 615
{
	struct trace_probe *tp;

	list_for_each_entry(tp, &probe_list, list)
616
		if (strcmp(tp->call.name, event) == 0 &&
617
		    strcmp(tp->call.class->system, group) == 0)
618 619 620 621
			return tp;
	return NULL;
}

622 623
/* Unregister a trace_probe and probe_event: call with locking probe_lock */
static void unregister_trace_probe(struct trace_probe *tp)
624 625 626 627
{
	if (probe_is_return(tp))
		unregister_kretprobe(&tp->rp);
	else
628
		unregister_kprobe(&tp->rp.kp);
629
	list_del(&tp->list);
630
	unregister_probe_event(tp);
631 632 633 634 635 636 637 638 639 640
}

/* 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);

641
	/* register as an event */
642
	old_tp = find_probe_event(tp->call.name, tp->call.class->system);
643 644 645 646 647 648 649
	if (old_tp) {
		/* delete old event */
		unregister_trace_probe(old_tp);
		free_trace_probe(old_tp);
	}
	ret = register_probe_event(tp);
	if (ret) {
P
Paul Bolle 已提交
650
		pr_warning("Failed to register probe event(%d)\n", ret);
651 652 653
		goto end;
	}

654
	tp->rp.kp.flags |= KPROBE_FLAG_DISABLED;
655 656 657
	if (probe_is_return(tp))
		ret = register_kretprobe(&tp->rp);
	else
658
		ret = register_kprobe(&tp->rp.kp);
659 660 661 662 663 664

	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",
665
				   tp->rp.kp.addr);
666 667
			ret = -EINVAL;
		}
668 669 670
		unregister_probe_event(tp);
	} else
		list_add_tail(&tp->list, &probe_list);
671 672 673 674 675 676
end:
	mutex_unlock(&probe_lock);
	return ret;
}

/* Split symbol and offset. */
677
static int split_symbol_offset(char *symbol, unsigned long *offset)
678 679 680 681 682 683 684 685 686 687
{
	char *tmp;
	int ret;

	if (!offset)
		return -EINVAL;

	tmp = strchr(symbol, '+');
	if (tmp) {
		/* skip sign because strict_strtol doesn't accept '+' */
688
		ret = strict_strtoul(tmp + 1, 0, offset);
689 690 691 692 693 694 695 696 697 698 699
		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))

700 701
static int parse_probe_vars(char *arg, const struct fetch_type *t,
			    struct fetch_param *f, int is_return)
702 703 704 705
{
	int ret = 0;
	unsigned long param;

706
	if (strcmp(arg, "retval") == 0) {
707
		if (is_return)
708
			f->fn = t->fetch[FETCH_MTD_retval];
709
		else
710
			ret = -EINVAL;
711 712
	} else if (strncmp(arg, "stack", 5) == 0) {
		if (arg[5] == '\0') {
713 714 715 716
			if (strcmp(t->name, DEFAULT_FETCH_TYPE_STR) == 0)
				f->fn = fetch_stack_address;
			else
				ret = -EINVAL;
717 718
		} else if (isdigit(arg[5])) {
			ret = strict_strtoul(arg + 5, 10, &param);
719 720 721
			if (ret || param > PARAM_MAX_STACK)
				ret = -EINVAL;
			else {
722
				f->fn = t->fetch[FETCH_MTD_stack];
723
				f->data = (void *)param;
724
			}
725 726 727
		} else
			ret = -EINVAL;
	} else
728 729 730 731
		ret = -EINVAL;
	return ret;
}

732
/* Recursive argument parser */
733 734
static int __parse_probe_arg(char *arg, const struct fetch_type *t,
			     struct fetch_param *f, int is_return)
735 736 737 738 739 740 741 742
{
	int ret = 0;
	unsigned long param;
	long offset;
	char *tmp;

	switch (arg[0]) {
	case '$':
743
		ret = parse_probe_vars(arg + 1, t, f, is_return);
744 745 746 747
		break;
	case '%':	/* named register */
		ret = regs_query_register_offset(arg + 1);
		if (ret >= 0) {
748
			f->fn = t->fetch[FETCH_MTD_reg];
749
			f->data = (void *)(unsigned long)ret;
750 751 752
			ret = 0;
		}
		break;
753 754 755 756 757
	case '@':	/* memory or symbol */
		if (isdigit(arg[1])) {
			ret = strict_strtoul(arg + 1, 0, &param);
			if (ret)
				break;
758
			f->fn = t->fetch[FETCH_MTD_memory];
759
			f->data = (void *)param;
760 761 762 763
		} else {
			ret = split_symbol_offset(arg + 1, &offset);
			if (ret)
				break;
764 765
			f->data = alloc_symbol_cache(arg + 1, offset);
			if (f->data)
766
				f->fn = t->fetch[FETCH_MTD_symbol];
767 768
		}
		break;
769
	case '+':	/* deref memory */
770 771
	case '-':
		tmp = strchr(arg, '(');
772
		if (!tmp)
773 774 775 776 777 778 779 780 781 782
			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) {
783 784
			struct deref_fetch_param *dprm;
			const struct fetch_type *t2 = find_fetch_type(NULL);
785
			*tmp = '\0';
786 787 788
			dprm = kzalloc(sizeof(struct deref_fetch_param),
				       GFP_KERNEL);
			if (!dprm)
789
				return -ENOMEM;
790 791 792
			dprm->offset = offset;
			ret = __parse_probe_arg(arg, t2, &dprm->orig,
						is_return);
793
			if (ret)
794
				kfree(dprm);
795
			else {
796
				f->fn = t->fetch[FETCH_MTD_deref];
797
				f->data = (void *)dprm;
798
			}
799
		}
800 801
		break;
	}
802 803 804
	if (!ret && !f->fn) {	/* Parsed, but do not find fetch method */
		pr_info("%s type has no corresponding fetch method.\n",
			t->name);
805
		ret = -EINVAL;
806
	}
807 808 809
	return ret;
}

810
/* String length checking wrapper */
811 812
static int parse_probe_arg(char *arg, struct trace_probe *tp,
			   struct probe_arg *parg, int is_return)
813
{
814
	const char *t;
815
	int ret;
816

817 818 819 820
	if (strlen(arg) > MAX_ARGSTR_LEN) {
		pr_info("Argument is too long.: %s\n",  arg);
		return -ENOSPC;
	}
821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837
	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;
838 839 840 841 842 843 844
	ret = __parse_probe_arg(arg, parg->type, &parg->fetch, is_return);
	if (ret >= 0) {
		parg->fetch_size.fn = get_fetch_size_function(parg->type,
							      parg->fetch.fn);
		parg->fetch_size.data = parg->fetch.data;
	}
	return ret;
845 846
}

847 848 849 850 851 852 853 854 855 856 857 858 859 860
/* 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;
}

861 862 863 864
static int create_trace_probe(int argc, char **argv)
{
	/*
	 * Argument syntax:
865 866
	 *  - Add kprobe: p[:[GRP/]EVENT] KSYM[+OFFS]|KADDR [FETCHARGS]
	 *  - Add kretprobe: r[:[GRP/]EVENT] KSYM[+0] [FETCHARGS]
867
	 * Fetch args:
868 869 870
	 *  $retval	: fetch return value
	 *  $stack	: fetch stack address
	 *  $stackN	: fetch Nth of stack (N:0-)
871 872 873
	 *  @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
874
	 * Dereferencing memory fetch:
875
	 *  +|-offs(ARG) : fetch memory at ARG +|- offs address.
876 877
	 * Alias name of args:
	 *  NAME=FETCHARG : set NAME as alias of FETCHARG.
878 879
	 * Type of args:
	 *  FETCHARG:TYPE : use TYPE instead of unsigned long.
880 881 882
	 */
	struct trace_probe *tp;
	int i, ret = 0;
883
	int is_return = 0, is_delete = 0;
884
	char *symbol = NULL, *event = NULL, *group = NULL;
885
	char *arg;
886
	unsigned long offset = 0;
887
	void *addr = NULL;
888
	char buf[MAX_EVENT_NAME_LEN];
889

890
	/* argc must be >= 1 */
891 892 893 894
	if (argv[0][0] == 'p')
		is_return = 0;
	else if (argv[0][0] == 'r')
		is_return = 1;
895 896
	else if (argv[0][0] == '-')
		is_delete = 1;
897
	else {
898 899
		pr_info("Probe definition must be started with 'p', 'r' or"
			" '-'.\n");
900
		return -EINVAL;
901
	}
902 903 904

	if (argv[0][1] == ':') {
		event = &argv[0][2];
905 906 907 908 909
		if (strchr(event, '/')) {
			group = event;
			event = strchr(group, '/') + 1;
			event[-1] = '\0';
			if (strlen(group) == 0) {
910
				pr_info("Group name is not specified\n");
911 912 913
				return -EINVAL;
			}
		}
914
		if (strlen(event) == 0) {
915
			pr_info("Event name is not specified\n");
916 917 918
			return -EINVAL;
		}
	}
919 920
	if (!group)
		group = KPROBE_EVENT_SYSTEM;
921

922 923 924 925 926
	if (is_delete) {
		if (!event) {
			pr_info("Delete command needs an event name.\n");
			return -EINVAL;
		}
927
		mutex_lock(&probe_lock);
928 929
		tp = find_probe_event(event, group);
		if (!tp) {
930
			mutex_unlock(&probe_lock);
931 932 933 934 935 936
			pr_info("Event %s/%s doesn't exist.\n", group, event);
			return -ENOENT;
		}
		/* delete an event */
		unregister_trace_probe(tp);
		free_trace_probe(tp);
937
		mutex_unlock(&probe_lock);
938 939 940 941 942 943 944
		return 0;
	}

	if (argc < 2) {
		pr_info("Probe point is not specified.\n");
		return -EINVAL;
	}
945
	if (isdigit(argv[1][0])) {
946 947
		if (is_return) {
			pr_info("Return probe point must be a symbol.\n");
948
			return -EINVAL;
949
		}
950
		/* an address specified */
951
		ret = strict_strtoul(&argv[1][0], 0, (unsigned long *)&addr);
952 953
		if (ret) {
			pr_info("Failed to parse address.\n");
954
			return ret;
955
		}
956 957 958 959 960
	} else {
		/* a symbol specified */
		symbol = argv[1];
		/* TODO: support .init module functions */
		ret = split_symbol_offset(symbol, &offset);
961 962
		if (ret) {
			pr_info("Failed to parse symbol.\n");
963
			return ret;
964 965 966
		}
		if (offset && is_return) {
			pr_info("Return probe must be used without offset.\n");
967
			return -EINVAL;
968
		}
969
	}
970
	argc -= 2; argv += 2;
971 972

	/* setup a probe */
973 974 975
	if (!event) {
		/* Make a new event name */
		if (symbol)
976
			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_%s_%ld",
977 978
				 is_return ? 'r' : 'p', symbol, offset);
		else
979
			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_0x%p",
980
				 is_return ? 'r' : 'p', addr);
981 982
		event = buf;
	}
983 984
	tp = alloc_trace_probe(group, event, addr, symbol, offset, argc,
			       is_return);
985 986 987
	if (IS_ERR(tp)) {
		pr_info("Failed to allocate trace_probe.(%d)\n",
			(int)PTR_ERR(tp));
988
		return PTR_ERR(tp);
989
	}
990 991

	/* parse arguments */
992 993
	ret = 0;
	for (i = 0; i < argc && i < MAX_TRACE_ARGS; i++) {
994 995 996
		/* Increment count for freeing args in error case */
		tp->nr_args++;

997 998
		/* Parse argument name */
		arg = strchr(argv[i], '=');
999
		if (arg) {
1000
			*arg++ = '\0';
1001 1002
			tp->args[i].name = kstrdup(argv[i], GFP_KERNEL);
		} else {
1003
			arg = argv[i];
1004 1005 1006 1007
			/* If argument name is omitted, set "argN" */
			snprintf(buf, MAX_EVENT_NAME_LEN, "arg%d", i + 1);
			tp->args[i].name = kstrdup(buf, GFP_KERNEL);
		}
1008

1009
		if (!tp->args[i].name) {
1010
			pr_info("Failed to allocate argument[%d] name.\n", i);
1011
			ret = -ENOMEM;
1012 1013
			goto error;
		}
1014 1015 1016 1017 1018 1019 1020

		if (!is_good_name(tp->args[i].name)) {
			pr_info("Invalid argument[%d] name: %s\n",
				i, tp->args[i].name);
			ret = -EINVAL;
			goto error;
		}
1021 1022

		if (conflict_field_name(tp->args[i].name, tp->args, i)) {
1023
			pr_info("Argument[%d] name '%s' conflicts with "
1024 1025 1026 1027
				"another field.\n", i, argv[i]);
			ret = -EINVAL;
			goto error;
		}
1028 1029

		/* Parse fetch argument */
1030
		ret = parse_probe_arg(arg, tp, &tp->args[i], is_return);
1031
		if (ret) {
1032
			pr_info("Parse error at argument[%d]. (%d)\n", i, ret);
1033
			goto error;
1034
		}
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081
	}

	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;
1082
	int i;
1083 1084

	seq_printf(m, "%c", probe_is_return(tp) ? 'r' : 'p');
1085
	seq_printf(m, ":%s/%s", tp->call.class->system, tp->call.name);
1086

1087 1088 1089
	if (!tp->symbol)
		seq_printf(m, " 0x%p", tp->rp.kp.addr);
	else if (tp->rp.kp.offset)
1090
		seq_printf(m, " %s+%u", probe_symbol(tp), tp->rp.kp.offset);
1091
	else
1092
		seq_printf(m, " %s", probe_symbol(tp));
1093

1094 1095
	for (i = 0; i < tp->nr_args; i++)
		seq_printf(m, " %s=%s", tp->args[i].name, tp->args[i].comm);
1096
	seq_printf(m, "\n");
1097

1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
	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,
};

1192 1193 1194 1195 1196 1197
/* 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,
1198
		   tp->rp.kp.nmissed);
1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222

	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,
};

1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270
/* Sum up total data length for dynamic arraies (strings) */
static __kprobes int __get_data_size(struct trace_probe *tp,
				     struct pt_regs *regs)
{
	int i, ret = 0;
	u32 len;

	for (i = 0; i < tp->nr_args; i++)
		if (unlikely(tp->args[i].fetch_size.fn)) {
			call_fetch(&tp->args[i].fetch_size, regs, &len);
			ret += len;
		}

	return ret;
}

/* Store the value of each argument */
static __kprobes void store_trace_args(int ent_size, struct trace_probe *tp,
				       struct pt_regs *regs,
				       u8 *data, int maxlen)
{
	int i;
	u32 end = tp->size;
	u32 *dl;	/* Data (relative) location */

	for (i = 0; i < tp->nr_args; i++) {
		if (unlikely(tp->args[i].fetch_size.fn)) {
			/*
			 * First, we set the relative location and
			 * maximum data length to *dl
			 */
			dl = (u32 *)(data + tp->args[i].offset);
			*dl = make_data_rloc(maxlen, end - tp->args[i].offset);
			/* Then try to fetch string or dynamic array data */
			call_fetch(&tp->args[i].fetch, regs, dl);
			/* Reduce maximum length */
			end += get_rloc_len(*dl);
			maxlen -= get_rloc_len(*dl);
			/* Trick here, convert data_rloc to data_loc */
			*dl = convert_rloc_to_loc(*dl,
				 ent_size + tp->args[i].offset);
		} else
			/* Just fetching data normally */
			call_fetch(&tp->args[i].fetch, regs,
				   data + tp->args[i].offset);
	}
}

1271
/* Kprobe handler */
1272
static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs)
1273
{
1274
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
1275
	struct kprobe_trace_entry_head *entry;
1276
	struct ring_buffer_event *event;
1277
	struct ring_buffer *buffer;
1278
	int size, dsize, pc;
1279
	unsigned long irq_flags;
1280
	struct ftrace_event_call *call = &tp->call;
1281

1282 1283
	tp->nhit++;

1284 1285 1286
	local_save_flags(irq_flags);
	pc = preempt_count();

1287 1288
	dsize = __get_data_size(tp, regs);
	size = sizeof(*entry) + tp->size + dsize;
1289

1290 1291
	event = trace_current_buffer_lock_reserve(&buffer, call->event.type,
						  size, irq_flags, pc);
1292
	if (!event)
1293
		return;
1294 1295 1296

	entry = ring_buffer_event_data(event);
	entry->ip = (unsigned long)kp->addr;
1297
	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
1298

1299 1300
	if (!filter_current_check_discard(buffer, call, entry, event))
		trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc);
1301 1302 1303
}

/* Kretprobe handler */
1304
static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri,
1305 1306 1307
					  struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp);
1308
	struct kretprobe_trace_entry_head *entry;
1309
	struct ring_buffer_event *event;
1310
	struct ring_buffer *buffer;
1311
	int size, pc, dsize;
1312
	unsigned long irq_flags;
1313
	struct ftrace_event_call *call = &tp->call;
1314 1315 1316 1317

	local_save_flags(irq_flags);
	pc = preempt_count();

1318 1319
	dsize = __get_data_size(tp, regs);
	size = sizeof(*entry) + tp->size + dsize;
1320

1321 1322
	event = trace_current_buffer_lock_reserve(&buffer, call->event.type,
						  size, irq_flags, pc);
1323
	if (!event)
1324
		return;
1325 1326

	entry = ring_buffer_event_data(event);
1327
	entry->func = (unsigned long)tp->rp.kp.addr;
1328
	entry->ret_ip = (unsigned long)ri->ret_addr;
1329
	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
1330

1331 1332
	if (!filter_current_check_discard(buffer, call, entry, event))
		trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc);
1333 1334 1335 1336
}

/* Event entry printers */
enum print_line_t
1337 1338
print_kprobe_event(struct trace_iterator *iter, int flags,
		   struct trace_event *event)
1339
{
1340
	struct kprobe_trace_entry_head *field;
1341
	struct trace_seq *s = &iter->seq;
1342
	struct trace_probe *tp;
1343
	u8 *data;
1344 1345
	int i;

1346
	field = (struct kprobe_trace_entry_head *)iter->ent;
1347
	tp = container_of(event, struct trace_probe, call.event);
1348

1349 1350 1351
	if (!trace_seq_printf(s, "%s: (", tp->call.name))
		goto partial;

1352 1353 1354
	if (!seq_print_ip_sym(s, field->ip, flags | TRACE_ITER_SYM_OFFSET))
		goto partial;

1355
	if (!trace_seq_puts(s, ")"))
1356 1357
		goto partial;

1358 1359 1360
	data = (u8 *)&field[1];
	for (i = 0; i < tp->nr_args; i++)
		if (!tp->args[i].type->print(s, tp->args[i].name,
1361
					     data + tp->args[i].offset, field))
1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
			goto partial;

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

	return TRACE_TYPE_HANDLED;
partial:
	return TRACE_TYPE_PARTIAL_LINE;
}

enum print_line_t
1373 1374
print_kretprobe_event(struct trace_iterator *iter, int flags,
		      struct trace_event *event)
1375
{
1376
	struct kretprobe_trace_entry_head *field;
1377
	struct trace_seq *s = &iter->seq;
1378
	struct trace_probe *tp;
1379
	u8 *data;
1380 1381
	int i;

1382
	field = (struct kretprobe_trace_entry_head *)iter->ent;
1383
	tp = container_of(event, struct trace_probe, call.event);
1384

1385 1386 1387
	if (!trace_seq_printf(s, "%s: (", tp->call.name))
		goto partial;

1388 1389 1390 1391 1392 1393 1394 1395 1396
	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;

1397
	if (!trace_seq_puts(s, ")"))
1398 1399
		goto partial;

1400 1401 1402
	data = (u8 *)&field[1];
	for (i = 0; i < tp->nr_args; i++)
		if (!tp->args[i].type->print(s, tp->args[i].name,
1403
					     data + tp->args[i].offset, field))
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417
			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;

1418 1419
	tp->flags |= TP_FLAG_TRACE;
	if (probe_is_return(tp))
1420
		return enable_kretprobe(&tp->rp);
1421
	else
1422
		return enable_kprobe(&tp->rp.kp);
1423 1424 1425 1426 1427 1428
}

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

1429 1430 1431 1432 1433 1434 1435
	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);
	}
1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451
}

#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;
1452
	struct kprobe_trace_entry_head field;
1453 1454
	struct trace_probe *tp = (struct trace_probe *)event_call->data;

1455
	DEFINE_FIELD(unsigned long, ip, FIELD_STRING_IP, 0);
1456
	/* Set argument names as fields */
1457
	for (i = 0; i < tp->nr_args; i++) {
1458
		ret = trace_define_field(event_call, tp->args[i].type->fmttype,
1459 1460 1461 1462 1463 1464 1465 1466
					 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;
	}
1467 1468 1469 1470 1471 1472
	return 0;
}

static int kretprobe_event_define_fields(struct ftrace_event_call *event_call)
{
	int ret, i;
1473
	struct kretprobe_trace_entry_head field;
1474 1475
	struct trace_probe *tp = (struct trace_probe *)event_call->data;

1476 1477
	DEFINE_FIELD(unsigned long, func, FIELD_STRING_FUNC, 0);
	DEFINE_FIELD(unsigned long, ret_ip, FIELD_STRING_RETIP, 0);
1478
	/* Set argument names as fields */
1479
	for (i = 0; i < tp->nr_args; i++) {
1480
		ret = trace_define_field(event_call, tp->args[i].type->fmttype,
1481 1482 1483 1484 1485 1486 1487 1488
					 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;
	}
1489 1490 1491
	return 0;
}

1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512
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++) {
1513 1514
		pos += snprintf(buf + pos, LEN_OR_ZERO, " %s=%s",
				tp->args[i].name, tp->args[i].type->fmt);
1515 1516 1517 1518 1519
	}

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

	for (i = 0; i < tp->nr_args; i++) {
1520 1521 1522 1523 1524 1525 1526
		if (strcmp(tp->args[i].type->name, "string") == 0)
			pos += snprintf(buf + pos, LEN_OR_ZERO,
					", __get_str(%s)",
					tp->args[i].name);
		else
			pos += snprintf(buf + pos, LEN_OR_ZERO, ", REC->%s",
					tp->args[i].name);
1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
	}

#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;
}

1553
#ifdef CONFIG_PERF_EVENTS
1554 1555

/* Kprobe profile handler */
1556
static __kprobes void kprobe_perf_func(struct kprobe *kp,
1557 1558 1559 1560
					 struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
	struct ftrace_event_call *call = &tp->call;
1561
	struct kprobe_trace_entry_head *entry;
1562
	struct hlist_head *head;
1563
	int size, __size, dsize;
1564
	int rctx;
1565

1566 1567
	dsize = __get_data_size(tp, regs);
	__size = sizeof(*entry) + tp->size + dsize;
1568 1569
	size = ALIGN(__size + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);
1570
	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
1571
		     "profile buffer not large enough"))
1572
		return;
1573

S
Steven Rostedt 已提交
1574
	entry = perf_trace_buf_prepare(size, call->event.type, regs, &rctx);
1575
	if (!entry)
1576
		return;
1577 1578

	entry->ip = (unsigned long)kp->addr;
1579 1580
	memset(&entry[1], 0, dsize);
	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
1581

1582
	head = this_cpu_ptr(call->perf_events);
1583
	perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head);
1584 1585 1586
}

/* Kretprobe profile handler */
1587
static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri,
1588 1589 1590 1591
					    struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp);
	struct ftrace_event_call *call = &tp->call;
1592
	struct kretprobe_trace_entry_head *entry;
1593
	struct hlist_head *head;
1594
	int size, __size, dsize;
1595
	int rctx;
1596

1597 1598
	dsize = __get_data_size(tp, regs);
	__size = sizeof(*entry) + tp->size + dsize;
1599 1600
	size = ALIGN(__size + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);
1601
	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
1602
		     "profile buffer not large enough"))
1603
		return;
1604

S
Steven Rostedt 已提交
1605
	entry = perf_trace_buf_prepare(size, call->event.type, regs, &rctx);
1606
	if (!entry)
1607
		return;
1608

1609 1610
	entry->func = (unsigned long)tp->rp.kp.addr;
	entry->ret_ip = (unsigned long)ri->ret_addr;
1611
	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
1612

1613
	head = this_cpu_ptr(call->perf_events);
1614
	perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head);
1615 1616
}

1617
static int probe_perf_enable(struct ftrace_event_call *call)
1618 1619 1620
{
	struct trace_probe *tp = (struct trace_probe *)call->data;

1621
	tp->flags |= TP_FLAG_PROFILE;
1622

1623
	if (probe_is_return(tp))
1624
		return enable_kretprobe(&tp->rp);
1625
	else
1626 1627 1628
		return enable_kprobe(&tp->rp.kp);
}

1629
static void probe_perf_disable(struct ftrace_event_call *call)
1630
{
1631 1632
	struct trace_probe *tp = (struct trace_probe *)call->data;

1633
	tp->flags &= ~TP_FLAG_PROFILE;
1634

1635
	if (!(tp->flags & TP_FLAG_TRACE)) {
1636 1637 1638 1639 1640
		if (probe_is_return(tp))
			disable_kretprobe(&tp->rp);
		else
			disable_kprobe(&tp->rp.kp);
	}
1641
}
1642
#endif	/* CONFIG_PERF_EVENTS */
1643

1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663
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;
}
1664 1665 1666 1667 1668

static __kprobes
int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs)
{
	struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp);
1669

1670 1671
	if (tp->flags & TP_FLAG_TRACE)
		kprobe_trace_func(kp, regs);
1672
#ifdef CONFIG_PERF_EVENTS
1673
	if (tp->flags & TP_FLAG_PROFILE)
1674
		kprobe_perf_func(kp, regs);
1675
#endif
1676 1677 1678 1679 1680 1681 1682 1683 1684 1685
	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);
1686
#ifdef CONFIG_PERF_EVENTS
1687
	if (tp->flags & TP_FLAG_PROFILE)
1688
		kretprobe_perf_func(ri, regs);
1689
#endif
1690 1691
	return 0;	/* We don't tweek kernel, so just return 0 */
}
1692

1693 1694 1695 1696 1697 1698 1699 1700
static struct trace_event_functions kretprobe_funcs = {
	.trace		= print_kretprobe_event
};

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

1701 1702 1703 1704 1705 1706
static int register_probe_event(struct trace_probe *tp)
{
	struct ftrace_event_call *call = &tp->call;
	int ret;

	/* Initialize ftrace_event_call */
1707
	INIT_LIST_HEAD(&call->class->fields);
1708
	if (probe_is_return(tp)) {
1709
		call->event.funcs = &kretprobe_funcs;
1710
		call->class->define_fields = kretprobe_event_define_fields;
1711
	} else {
1712
		call->event.funcs = &kprobe_funcs;
1713
		call->class->define_fields = kprobe_event_define_fields;
1714
	}
1715 1716
	if (set_print_fmt(tp) < 0)
		return -ENOMEM;
1717 1718
	ret = register_ftrace_event(&call->event);
	if (!ret) {
1719
		kfree(call->print_fmt);
1720
		return -ENODEV;
1721
	}
1722
	call->flags = 0;
1723
	call->class->reg = kprobe_register;
1724 1725
	call->data = tp;
	ret = trace_add_event_call(call);
1726
	if (ret) {
1727
		pr_info("Failed to register kprobe event: %s\n", call->name);
1728
		kfree(call->print_fmt);
1729
		unregister_ftrace_event(&call->event);
1730
	}
1731 1732 1733 1734 1735
	return ret;
}

static void unregister_probe_event(struct trace_probe *tp)
{
1736
	/* tp->event is unregistered in trace_remove_event_call() */
1737
	trace_remove_event_call(&tp->call);
1738
	kfree(tp->call.print_fmt);
1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753
}

/* 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);

1754
	/* Event list interface */
1755 1756 1757
	if (!entry)
		pr_warning("Could not create debugfs "
			   "'kprobe_events' entry\n");
1758 1759 1760 1761 1762 1763 1764 1765

	/* 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");
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780
	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)
{
1781
	int ret, warn = 0;
1782
	int (*target)(int, int, int, int, int, int);
1783
	struct trace_probe *tp;
1784 1785 1786 1787 1788 1789

	target = kprobe_trace_selftest_target;

	pr_info("Testing kprobe tracing: ");

	ret = command_trace_probe("p:testprobe kprobe_trace_selftest_target "
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802
				  "$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);
	}
1803 1804

	ret = command_trace_probe("r:testprobe2 kprobe_trace_selftest_target "
1805
				  "$retval");
1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820
	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;
1821 1822 1823

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

1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834
	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++;
	}
1835

1836 1837 1838 1839 1840 1841
end:
	cleanup_all_probes();
	if (warn)
		pr_cont("NG: Some tests are failed. Please check them.\n");
	else
		pr_cont("OK\n");
1842 1843 1844 1845 1846 1847
	return 0;
}

late_initcall(kprobe_trace_self_tests_init);

#endif