[deliverable/linux.git] / include / linux / trace_events.h


#ifndef _LINUX_TRACE_EVENT_H
#define _LINUX_TRACE_EVENT_H

#include <linux/ring_buffer.h>
#include <linux/trace_seq.h>
#include <linux/percpu.h>
#include <linux/hardirq.h>
#include <linux/perf_event.h>
#include <linux/tracepoint.h>

struct trace_array;
struct trace_buffer;
struct tracer;
struct dentry;
struct bpf_prog;

const char *trace_print_flags_seq(struct trace_seq *p, const char *delim,
				  unsigned long flags,
				  const struct trace_print_flags *flag_array);

const char *trace_print_symbols_seq(struct trace_seq *p, unsigned long val,
				    const struct trace_print_flags *symbol_array);

#if BITS_PER_LONG == 32
const char *trace_print_symbols_seq_u64(struct trace_seq *p,
					unsigned long long val,
					const struct trace_print_flags_u64
								 *symbol_array);
#endif

const char *trace_print_bitmask_seq(struct trace_seq *p, void *bitmask_ptr,
				    unsigned int bitmask_size);

const char *trace_print_hex_seq(struct trace_seq *p,
				const unsigned char *buf, int len);

const char *trace_print_array_seq(struct trace_seq *p,
				   const void *buf, int count,
				   size_t el_size);

struct trace_iterator;
struct trace_event;

int trace_raw_output_prep(struct trace_iterator *iter,
			  struct trace_event *event);

/*
 * The trace entry - the most basic unit of tracing. This is what
 * is printed in the end as a single line in the trace output, such as:
 *
 *     bash-15816 [01]   235.197585: idle_cpu <- irq_enter
 */
struct trace_entry {
	unsigned short		type;
	unsigned char		flags;
	unsigned char		preempt_count;
	int			pid;
};

#define TRACE_EVENT_TYPE_MAX						\
	((1 << (sizeof(((struct trace_entry *)0)->type) * 8)) - 1)

/*
 * Trace iterator - used by printout routines who present trace
 * results to users and which routines might sleep, etc:
 */
struct trace_iterator {
	struct trace_array	*tr;
	struct tracer		*trace;
	struct trace_buffer	*trace_buffer;
	void			*private;
	int			cpu_file;
	struct mutex		mutex;
	struct ring_buffer_iter	**buffer_iter;
	unsigned long		iter_flags;

	/* trace_seq for __print_flags() and __print_symbolic() etc. */
	struct trace_seq	tmp_seq;

	cpumask_var_t		started;

	/* it's true when current open file is snapshot */
	bool			snapshot;

	/* The below is zeroed out in pipe_read */
	struct trace_seq	seq;
	struct trace_entry	*ent;
	unsigned long		lost_events;
	int			leftover;
	int			ent_size;
	int			cpu;
	u64			ts;

	loff_t			pos;
	long			idx;

	/* All new field here will be zeroed out in pipe_read */
};

enum trace_iter_flags {
	TRACE_FILE_LAT_FMT	= 1,
	TRACE_FILE_ANNOTATE	= 2,
	TRACE_FILE_TIME_IN_NS	= 4,
};


typedef enum print_line_t (*trace_print_func)(struct trace_iterator *iter,
				      int flags, struct trace_event *event);

struct trace_event_functions {
	trace_print_func	trace;
	trace_print_func	raw;
	trace_print_func	hex;
	trace_print_func	binary;
};

struct trace_event {
	struct hlist_node		node;
	struct list_head		list;
	int				type;
	struct trace_event_functions	*funcs;
};

extern int register_trace_event(struct trace_event *event);
extern int unregister_trace_event(struct trace_event *event);

/* Return values for print_line callback */
enum print_line_t {
	TRACE_TYPE_PARTIAL_LINE	= 0,	/* Retry after flushing the seq */
	TRACE_TYPE_HANDLED	= 1,
	TRACE_TYPE_UNHANDLED	= 2,	/* Relay to other output functions */
	TRACE_TYPE_NO_CONSUME	= 3	/* Handled but ask to not consume */
};

/*
 * Several functions return TRACE_TYPE_PARTIAL_LINE if the trace_seq
 * overflowed, and TRACE_TYPE_HANDLED otherwise. This helper function
 * simplifies those functions and keeps them in sync.
 */
static inline enum print_line_t trace_handle_return(struct trace_seq *s)
{
	return trace_seq_has_overflowed(s) ?
		TRACE_TYPE_PARTIAL_LINE : TRACE_TYPE_HANDLED;
}

void tracing_generic_entry_update(struct trace_entry *entry,
				  unsigned long flags,
				  int pc);
struct trace_event_file;

struct ring_buffer_event *
trace_event_buffer_lock_reserve(struct ring_buffer **current_buffer,
				struct trace_event_file *trace_file,
				int type, unsigned long len,
				unsigned long flags, int pc);

void tracing_record_cmdline(struct task_struct *tsk);

int trace_output_call(struct trace_iterator *iter, char *name, char *fmt, ...);

struct event_filter;

enum trace_reg {
	TRACE_REG_REGISTER,
	TRACE_REG_UNREGISTER,
#ifdef CONFIG_PERF_EVENTS
	TRACE_REG_PERF_REGISTER,
	TRACE_REG_PERF_UNREGISTER,
	TRACE_REG_PERF_OPEN,
	TRACE_REG_PERF_CLOSE,
	TRACE_REG_PERF_ADD,
	TRACE_REG_PERF_DEL,
#endif
};

struct trace_event_call;

struct trace_event_class {
	const char		*system;
	void			*probe;
#ifdef CONFIG_PERF_EVENTS
	void			*perf_probe;
#endif
	int			(*reg)(struct trace_event_call *event,
				       enum trace_reg type, void *data);
	int			(*define_fields)(struct trace_event_call *);
	struct list_head	*(*get_fields)(struct trace_event_call *);
	struct list_head	fields;
	int			(*raw_init)(struct trace_event_call *);
};

extern int trace_event_reg(struct trace_event_call *event,
			    enum trace_reg type, void *data);

struct trace_event_buffer {
	struct ring_buffer		*buffer;
	struct ring_buffer_event	*event;
	struct trace_event_file		*trace_file;
	void				*entry;
	unsigned long			flags;
	int				pc;
};

void *trace_event_buffer_reserve(struct trace_event_buffer *fbuffer,
				  struct trace_event_file *trace_file,
				  unsigned long len);

void trace_event_buffer_commit(struct trace_event_buffer *fbuffer);

enum {
	TRACE_EVENT_FL_FILTERED_BIT,
	TRACE_EVENT_FL_CAP_ANY_BIT,
	TRACE_EVENT_FL_NO_SET_FILTER_BIT,
	TRACE_EVENT_FL_IGNORE_ENABLE_BIT,
	TRACE_EVENT_FL_WAS_ENABLED_BIT,
	TRACE_EVENT_FL_TRACEPOINT_BIT,
	TRACE_EVENT_FL_KPROBE_BIT,
	TRACE_EVENT_FL_UPROBE_BIT,
	TRACE_EVENT_FL_MAP_BIT,
};

/*
 * Event flags:
 *  FILTERED	  - The event has a filter attached
 *  CAP_ANY	  - Any user can enable for perf
 *  NO_SET_FILTER - Set when filter has error and is to be ignored
 *  IGNORE_ENABLE - For trace internal events, do not enable with debugfs file
 *  WAS_ENABLED   - Set and stays set when an event was ever enabled
 *                    (used for module unloading, if a module event is enabled,
 *                     it is best to clear the buffers that used it).
 *  TRACEPOINT    - Event is a tracepoint
 *  KPROBE        - Event is a kprobe
 *  UPROBE        - Event is a uprobe
 *  MAP           - Event maps to a tracepoint as an alias
 */
enum {
	TRACE_EVENT_FL_FILTERED		= (1 << TRACE_EVENT_FL_FILTERED_BIT),
	TRACE_EVENT_FL_CAP_ANY		= (1 << TRACE_EVENT_FL_CAP_ANY_BIT),
	TRACE_EVENT_FL_NO_SET_FILTER	= (1 << TRACE_EVENT_FL_NO_SET_FILTER_BIT),
	TRACE_EVENT_FL_IGNORE_ENABLE	= (1 << TRACE_EVENT_FL_IGNORE_ENABLE_BIT),
	TRACE_EVENT_FL_WAS_ENABLED	= (1 << TRACE_EVENT_FL_WAS_ENABLED_BIT),
	TRACE_EVENT_FL_TRACEPOINT	= (1 << TRACE_EVENT_FL_TRACEPOINT_BIT),
	TRACE_EVENT_FL_KPROBE		= (1 << TRACE_EVENT_FL_KPROBE_BIT),
	TRACE_EVENT_FL_UPROBE		= (1 << TRACE_EVENT_FL_UPROBE_BIT),
	TRACE_EVENT_FL_MAP		= (1 << TRACE_EVENT_FL_MAP_BIT),
};

#define TRACE_EVENT_FL_UKPROBE (TRACE_EVENT_FL_KPROBE | TRACE_EVENT_FL_UPROBE)

struct trace_event_map {
	struct tracepoint	*tp;
	char			*name;
};

struct trace_event_call {
	struct list_head	list;
	struct trace_event_class *class;
	union {
		char			*name;
		/* Set TRACE_EVENT_FL_TRACEPOINT flag when using "tp" */
		struct tracepoint	*tp;
		/* Set TRACE_EVENT_FL_MAP flag when using "map" instead */
		struct trace_event_map	*map;
	};
	struct trace_event	event;
	char			*print_fmt;
	struct event_filter	*filter;
	void			*mod;
	void			*data;
	/*
	 *   bit 0:		filter_active
	 *   bit 1:		allow trace by non root (cap any)
	 *   bit 2:		failed to apply filter
	 *   bit 3:		trace internal event (do not enable)
	 *   bit 4:		Event was enabled by module
	 *   bit 5:		use call filter rather than file filter
	 *   bit 6:		Event is a tracepoint
	 */
	int			flags; /* static flags of different events */

#ifdef CONFIG_PERF_EVENTS
	int				perf_refcount;
	struct hlist_head __percpu	*perf_events;
	struct bpf_prog			*prog;

	int	(*perf_perm)(struct trace_event_call *,
			     struct perf_event *);
#endif
};

static inline const char *
trace_event_name(struct trace_event_call *call)
{
	if (call->flags & TRACE_EVENT_FL_MAP)
		return call->map->name;
	else if (call->flags & TRACE_EVENT_FL_TRACEPOINT)
		return call->tp ? call->tp->name : NULL;
	else
		return call->name;
}

struct trace_array;
struct trace_subsystem_dir;

enum {
	EVENT_FILE_FL_ENABLED_BIT,
	EVENT_FILE_FL_RECORDED_CMD_BIT,
	EVENT_FILE_FL_FILTERED_BIT,
	EVENT_FILE_FL_NO_SET_FILTER_BIT,
	EVENT_FILE_FL_SOFT_MODE_BIT,
	EVENT_FILE_FL_SOFT_DISABLED_BIT,
	EVENT_FILE_FL_TRIGGER_MODE_BIT,
	EVENT_FILE_FL_TRIGGER_COND_BIT,
	EVENT_FILE_FL_PID_FILTER_BIT,
};

/*
 * Event file flags:
 *  ENABLED	  - The event is enabled
 *  RECORDED_CMD  - The comms should be recorded at sched_switch
 *  FILTERED	  - The event has a filter attached
 *  NO_SET_FILTER - Set when filter has error and is to be ignored
 *  SOFT_MODE     - The event is enabled/disabled by SOFT_DISABLED
 *  SOFT_DISABLED - When set, do not trace the event (even though its
 *                   tracepoint may be enabled)
 *  TRIGGER_MODE  - When set, invoke the triggers associated with the event
 *  TRIGGER_COND  - When set, one or more triggers has an associated filter
 *  PID_FILTER    - When set, the event is filtered based on pid
 */
enum {
	EVENT_FILE_FL_ENABLED		= (1 << EVENT_FILE_FL_ENABLED_BIT),
	EVENT_FILE_FL_RECORDED_CMD	= (1 << EVENT_FILE_FL_RECORDED_CMD_BIT),
	EVENT_FILE_FL_FILTERED		= (1 << EVENT_FILE_FL_FILTERED_BIT),
	EVENT_FILE_FL_NO_SET_FILTER	= (1 << EVENT_FILE_FL_NO_SET_FILTER_BIT),
	EVENT_FILE_FL_SOFT_MODE		= (1 << EVENT_FILE_FL_SOFT_MODE_BIT),
	EVENT_FILE_FL_SOFT_DISABLED	= (1 << EVENT_FILE_FL_SOFT_DISABLED_BIT),
	EVENT_FILE_FL_TRIGGER_MODE	= (1 << EVENT_FILE_FL_TRIGGER_MODE_BIT),
	EVENT_FILE_FL_TRIGGER_COND	= (1 << EVENT_FILE_FL_TRIGGER_COND_BIT),
	EVENT_FILE_FL_PID_FILTER	= (1 << EVENT_FILE_FL_PID_FILTER_BIT),
};

struct trace_event_file {
	struct list_head		list;
	struct trace_event_call		*event_call;
	struct event_filter		*filter;
	struct dentry			*dir;
	struct trace_array		*tr;
	struct trace_subsystem_dir	*system;
	struct list_head		triggers;

	/*
	 * 32 bit flags:
	 *   bit 0:		enabled
	 *   bit 1:		enabled cmd record
	 *   bit 2:		enable/disable with the soft disable bit
	 *   bit 3:		soft disabled
	 *   bit 4:		trigger enabled
	 *
	 * Note: The bits must be set atomically to prevent races
	 * from other writers. Reads of flags do not need to be in
	 * sync as they occur in critical sections. But the way flags
	 * is currently used, these changes do not affect the code
	 * except that when a change is made, it may have a slight
	 * delay in propagating the changes to other CPUs due to
	 * caching and such. Which is mostly OK ;-)
	 */
	unsigned long		flags;
	atomic_t		sm_ref;	/* soft-mode reference counter */
	atomic_t		tm_ref;	/* trigger-mode reference counter */
};

#define __TRACE_EVENT_FLAGS(name, value)				\
	static int __init trace_init_flags_##name(void)			\
	{								\
		event_##name.flags |= value;				\
		return 0;						\
	}								\
	early_initcall(trace_init_flags_##name);

#define __TRACE_EVENT_PERF_PERM(name, expr...)				\
	static int perf_perm_##name(struct trace_event_call *tp_event, \
				    struct perf_event *p_event)		\
	{								\
		return ({ expr; });					\
	}								\
	static int __init trace_init_perf_perm_##name(void)		\
	{								\
		event_##name.perf_perm = &perf_perm_##name;		\
		return 0;						\
	}								\
	early_initcall(trace_init_perf_perm_##name);

#define PERF_MAX_TRACE_SIZE	2048

#define MAX_FILTER_STR_VAL	256	/* Should handle KSYM_SYMBOL_LEN */

enum event_trigger_type {
	ETT_NONE		= (0),
	ETT_TRACE_ONOFF		= (1 << 0),
	ETT_SNAPSHOT		= (1 << 1),
	ETT_STACKTRACE		= (1 << 2),
	ETT_EVENT_ENABLE	= (1 << 3),
	ETT_EVENT_HIST		= (1 << 4),
	ETT_HIST_ENABLE		= (1 << 5),
};

extern int filter_match_preds(struct event_filter *filter, void *rec);

extern enum event_trigger_type event_triggers_call(struct trace_event_file *file,
						   void *rec);
extern void event_triggers_post_call(struct trace_event_file *file,
				     enum event_trigger_type tt,
				     void *rec);

bool trace_event_ignore_this_pid(struct trace_event_file *trace_file);

/**
 * trace_trigger_soft_disabled - do triggers and test if soft disabled
 * @file: The file pointer of the event to test
 *
 * If any triggers without filters are attached to this event, they
 * will be called here. If the event is soft disabled and has no
 * triggers that require testing the fields, it will return true,
 * otherwise false.
 */
static inline bool
trace_trigger_soft_disabled(struct trace_event_file *file)
{
	unsigned long eflags = file->flags;

	if (!(eflags & EVENT_FILE_FL_TRIGGER_COND)) {
		if (eflags & EVENT_FILE_FL_TRIGGER_MODE)
			event_triggers_call(file, NULL);
		if (eflags & EVENT_FILE_FL_SOFT_DISABLED)
			return true;
		if (eflags & EVENT_FILE_FL_PID_FILTER)
			return trace_event_ignore_this_pid(file);
	}
	return false;
}

#ifdef CONFIG_BPF_EVENTS
unsigned int trace_call_bpf(struct bpf_prog *prog, void *ctx);
#else
static inline unsigned int trace_call_bpf(struct bpf_prog *prog, void *ctx)
{
	return 1;
}
#endif

enum {
	FILTER_OTHER = 0,
	FILTER_STATIC_STRING,
	FILTER_DYN_STRING,
	FILTER_PTR_STRING,
	FILTER_TRACE_FN,
	FILTER_COMM,
	FILTER_CPU,
};

extern int trace_event_raw_init(struct trace_event_call *call);
extern int trace_define_field(struct trace_event_call *call, const char *type,
			      const char *name, int offset, int size,
			      int is_signed, int filter_type);
extern int trace_add_event_call(struct trace_event_call *call);
extern int trace_remove_event_call(struct trace_event_call *call);
extern int trace_event_get_offsets(struct trace_event_call *call);

#define is_signed_type(type)	(((type)(-1)) < (type)1)

int trace_set_clr_event(const char *system, const char *event, int set);

/*
 * The double __builtin_constant_p is because gcc will give us an error
 * if we try to allocate the static variable to fmt if it is not a
 * constant. Even with the outer if statement optimizing out.
 */
#define event_trace_printk(ip, fmt, args...)				\
do {									\
	__trace_printk_check_format(fmt, ##args);			\
	tracing_record_cmdline(current);				\
	if (__builtin_constant_p(fmt)) {				\
		static const char *trace_printk_fmt			\
		  __attribute__((section("__trace_printk_fmt"))) =	\
			__builtin_constant_p(fmt) ? fmt : NULL;		\
									\
		__trace_bprintk(ip, trace_printk_fmt, ##args);		\
	} else								\
		__trace_printk(ip, fmt, ##args);			\
} while (0)

#ifdef CONFIG_PERF_EVENTS
struct perf_event;

DECLARE_PER_CPU(struct pt_regs, perf_trace_regs);

extern int  perf_trace_init(struct perf_event *event);
extern void perf_trace_destroy(struct perf_event *event);
extern int  perf_trace_add(struct perf_event *event, int flags);
extern void perf_trace_del(struct perf_event *event, int flags);
extern int  ftrace_profile_set_filter(struct perf_event *event, int event_id,
				     char *filter_str);
extern void ftrace_profile_free_filter(struct perf_event *event);
void perf_trace_buf_update(void *record, u16 type);
void *perf_trace_buf_alloc(int size, struct pt_regs **regs, int *rctxp);

void perf_trace_run_bpf_submit(void *raw_data, int size, int rctx,
			       struct trace_event_call *call, u64 count,
			       struct pt_regs *regs, struct hlist_head *head,
			       struct task_struct *task);

static inline void
perf_trace_buf_submit(void *raw_data, int size, int rctx, u16 type,
		       u64 count, struct pt_regs *regs, void *head,
		       struct task_struct *task)
{
	perf_tp_event(type, count, raw_data, size, regs, head, rctx, task);
}
#endif

#endif /* _LINUX_TRACE_EVENT_H */
Commit	Line	Data
bac5fb97	1
645df987 SRRH	2	#ifndef _LINUX_TRACE_EVENT_H
645df987 SRRH	3	#define _LINUX_TRACE_EVENT_H
97f20251	4
97f20251	5	#include <linux/ring_buffer.h>
16bb8eb1	6	#include <linux/trace_seq.h>
be74b73a	7	#include <linux/percpu.h>
20ab4425	8	#include <linux/hardirq.h>
430ad5a6	9	#include <linux/perf_event.h>
de7b2973	10	#include <linux/tracepoint.h>
97f20251 SR	11
97f20251 SR	12	struct trace_array;
12883efb	13	struct trace_buffer;
97f20251	14	struct tracer;
6d723736	15	struct dentry;
2541517c	16	struct bpf_prog;
97f20251	17
645df987 SRRH	18	const char trace_print_flags_seq(struct trace_seq p, const char *delim,
	19	unsigned long flags,
	20	const struct trace_print_flags *flag_array);
be74b73a	21
645df987 SRRH	22	const char trace_print_symbols_seq(struct trace_seq p, unsigned long val,
645df987 SRRH	23	const struct trace_print_flags *symbol_array);
0f4fc29d	24
2fc1b6f0	25	#if BITS_PER_LONG == 32
645df987 SRRH	26	const char trace_print_symbols_seq_u64(struct trace_seq p,
	27	unsigned long long val,
	28	const struct trace_print_flags_u64
2fc1b6f0	29	*symbol_array);
	30	#endif
	31
645df987 SRRH	32	const char trace_print_bitmask_seq(struct trace_seq p, void *bitmask_ptr,
645df987 SRRH	33	unsigned int bitmask_size);
4449bf92	34
645df987 SRRH	35	const char trace_print_hex_seq(struct trace_seq p,
645df987 SRRH	36	const unsigned char *buf, int len);
5a2e3995	37
645df987	38	const char trace_print_array_seq(struct trace_seq p,
ac01ce14	39	const void *buf, int count,
6ea22486 DM	40	size_t el_size);
6ea22486 DM	41
f71130de LZ	42	struct trace_iterator;
	43	struct trace_event;
	44
892c505a SRRH	45	int trace_raw_output_prep(struct trace_iterator *iter,
892c505a SRRH	46	struct trace_event *event);
f71130de	47
97f20251 SR	48	/*
	49	* The trace entry - the most basic unit of tracing. This is what
	50	* is printed in the end as a single line in the trace output, such as:
	51	*
	52	* bash-15816 [01] 235.197585: idle_cpu <- irq_enter
	53	*/
	54	struct trace_entry {
89ec0dee	55	unsigned short type;
97f20251 SR	56	unsigned char flags;
	57	unsigned char preempt_count;
	58	int pid;
97f20251 SR	59	};
97f20251 SR	60
609a7404	61	#define TRACE_EVENT_TYPE_MAX \
89ec0dee SR	62	((1 << (sizeof(((struct trace_entry )0)->type) 8)) - 1)
89ec0dee SR	63
97f20251 SR	64	/*
	65	* Trace iterator - used by printout routines who present trace
	66	* results to users and which routines might sleep, etc:
	67	*/
	68	struct trace_iterator {
	69	struct trace_array *tr;
	70	struct tracer *trace;
12883efb	71	struct trace_buffer *trace_buffer;
97f20251 SR	72	void *private;
	73	int cpu_file;
	74	struct mutex mutex;
6d158a81	75	struct ring_buffer_iter **buffer_iter;
112f38a7	76	unsigned long iter_flags;
97f20251	77
bc289ae9 LJ	78	/* trace_seq for __print_flags() and __print_symbolic() etc. */
	79	struct trace_seq tmp_seq;
	80
ed5467da AV	81	cpumask_var_t started;
	82
	83	/* it's true when current open file is snapshot */
	84	bool snapshot;
	85
97f20251 SR	86	/* The below is zeroed out in pipe_read */
	87	struct trace_seq seq;
	88	struct trace_entry *ent;
bc21b478	89	unsigned long lost_events;
a63ce5b3	90	int leftover;
4a9bd3f1	91	int ent_size;
97f20251 SR	92	int cpu;
	93	u64 ts;
	94
97f20251 SR	95	loff_t pos;
	96	long idx;
	97
ed5467da	98	/* All new field here will be zeroed out in pipe_read */
97f20251 SR	99	};
97f20251 SR	100
8be0709f DS	101	enum trace_iter_flags {
	102	TRACE_FILE_LAT_FMT = 1,
	103	TRACE_FILE_ANNOTATE = 2,
	104	TRACE_FILE_TIME_IN_NS = 4,
	105	};
	106
97f20251 SR	107
97f20251 SR	108	typedef enum print_line_t (trace_print_func)(struct trace_iterator iter,
a9a57763 SR	109	int flags, struct trace_event *event);
	110
	111	struct trace_event_functions {
97f20251 SR	112	trace_print_func trace;
	113	trace_print_func raw;
	114	trace_print_func hex;
	115	trace_print_func binary;
	116	};
	117
a9a57763 SR	118	struct trace_event {
	119	struct hlist_node node;
	120	struct list_head list;
	121	int type;
	122	struct trace_event_functions *funcs;
	123	};
	124
9023c930 SRRH	125	extern int register_trace_event(struct trace_event *event);
9023c930 SRRH	126	extern int unregister_trace_event(struct trace_event *event);
97f20251 SR	127
	128	/* Return values for print_line callback */
	129	enum print_line_t {
	130	TRACE_TYPE_PARTIAL_LINE = 0, /* Retry after flushing the seq */
	131	TRACE_TYPE_HANDLED = 1,
	132	TRACE_TYPE_UNHANDLED = 2, /* Relay to other output functions */
	133	TRACE_TYPE_NO_CONSUME = 3 /* Handled but ask to not consume */
	134	};
	135
19a7fe20 SRRH	136	/*
	137	* Several functions return TRACE_TYPE_PARTIAL_LINE if the trace_seq
	138	* overflowed, and TRACE_TYPE_HANDLED otherwise. This helper function
	139	* simplifies those functions and keeps them in sync.
	140	*/
	141	static inline enum print_line_t trace_handle_return(struct trace_seq *s)
	142	{
	143	return trace_seq_has_overflowed(s) ?
	144	TRACE_TYPE_PARTIAL_LINE : TRACE_TYPE_HANDLED;
	145	}
	146
f413cdb8 FW	147	void tracing_generic_entry_update(struct trace_entry *entry,
	148	unsigned long flags,
	149	int pc);
7f1d2f82	150	struct trace_event_file;
ccb469a1 SR	151
	152	struct ring_buffer_event *
	153	trace_event_buffer_lock_reserve(struct ring_buffer **current_buffer,
7f1d2f82	154	struct trace_event_file *trace_file,
ccb469a1 SR	155	int type, unsigned long len,
ccb469a1 SR	156	unsigned long flags, int pc);
97f20251 SR	157
	158	void tracing_record_cmdline(struct task_struct *tsk);
	159
892c505a	160	int trace_output_call(struct trace_iterator iter, char name, char *fmt, ...);
1d6bae96	161
1f9963cb LZ	162	struct event_filter;
1f9963cb LZ	163
2239291a SR	164	enum trace_reg {
	165	TRACE_REG_REGISTER,
	166	TRACE_REG_UNREGISTER,
37d73998	167	#ifdef CONFIG_PERF_EVENTS
2239291a SR	168	TRACE_REG_PERF_REGISTER,
2239291a SR	169	TRACE_REG_PERF_UNREGISTER,
ceec0b6f JO	170	TRACE_REG_PERF_OPEN,
ceec0b6f JO	171	TRACE_REG_PERF_CLOSE,
489c75c3 JO	172	TRACE_REG_PERF_ADD,
489c75c3 JO	173	TRACE_REG_PERF_DEL,
37d73998	174	#endif
2239291a SR	175	};
2239291a SR	176
2425bcb9	177	struct trace_event_call;
2239291a	178
2425bcb9	179	struct trace_event_class {
acd388fd	180	const char *system;
2239291a SR	181	void *probe;
	182	#ifdef CONFIG_PERF_EVENTS
	183	void *perf_probe;
	184	#endif
2425bcb9	185	int (reg)(struct trace_event_call event,
ceec0b6f	186	enum trace_reg type, void *data);
2425bcb9 SRRH	187	int (define_fields)(struct trace_event_call );
2425bcb9 SRRH	188	struct list_head (get_fields)(struct trace_event_call *);
2e33af02	189	struct list_head fields;
2425bcb9	190	int (raw_init)(struct trace_event_call );
8f082018 SR	191	};
8f082018 SR	192
2425bcb9	193	extern int trace_event_reg(struct trace_event_call *event,
ceec0b6f	194	enum trace_reg type, void *data);
a1d0ce82	195
3f795dcf	196	struct trace_event_buffer {
3fd40d1e SR	197	struct ring_buffer *buffer;
3fd40d1e SR	198	struct ring_buffer_event *event;
7f1d2f82	199	struct trace_event_file *trace_file;
3fd40d1e SR	200	void *entry;
	201	unsigned long flags;
	202	int pc;
	203	};
	204
3f795dcf	205	void trace_event_buffer_reserve(struct trace_event_buffer fbuffer,
7f1d2f82	206	struct trace_event_file *trace_file,
3fd40d1e SR	207	unsigned long len);
3fd40d1e SR	208
3f795dcf	209	void trace_event_buffer_commit(struct trace_event_buffer *fbuffer);
3fd40d1e	210
553552ce	211	enum {
553552ce	212	TRACE_EVENT_FL_FILTERED_BIT,
61c32659	213	TRACE_EVENT_FL_CAP_ANY_BIT,
27b14b56	214	TRACE_EVENT_FL_NO_SET_FILTER_BIT,
9b63776f	215	TRACE_EVENT_FL_IGNORE_ENABLE_BIT,
575380da	216	TRACE_EVENT_FL_WAS_ENABLED_BIT,
de7b2973	217	TRACE_EVENT_FL_TRACEPOINT_BIT,
72cbbc89	218	TRACE_EVENT_FL_KPROBE_BIT,
04a22fae	219	TRACE_EVENT_FL_UPROBE_BIT,
b1dace68	220	TRACE_EVENT_FL_MAP_BIT,
553552ce SR	221	};
553552ce SR	222
ae63b31e SR	223	/*
	224	* Event flags:
	225	* FILTERED - The event has a filter attached
	226	* CAP_ANY - Any user can enable for perf
	227	* NO_SET_FILTER - Set when filter has error and is to be ignored
5d6ad960	228	* IGNORE_ENABLE - For trace internal events, do not enable with debugfs file
575380da SRRH	229	* WAS_ENABLED - Set and stays set when an event was ever enabled
	230	* (used for module unloading, if a module event is enabled,
	231	* it is best to clear the buffers that used it).
de7b2973	232	* TRACEPOINT - Event is a tracepoint
72cbbc89	233	* KPROBE - Event is a kprobe
04a22fae	234	* UPROBE - Event is a uprobe
b1dace68	235	* MAP - Event maps to a tracepoint as an alias
ae63b31e	236	*/
553552ce	237	enum {
e870e9a1	238	TRACE_EVENT_FL_FILTERED = (1 << TRACE_EVENT_FL_FILTERED_BIT),
61c32659	239	TRACE_EVENT_FL_CAP_ANY = (1 << TRACE_EVENT_FL_CAP_ANY_BIT),
27b14b56	240	TRACE_EVENT_FL_NO_SET_FILTER = (1 << TRACE_EVENT_FL_NO_SET_FILTER_BIT),
9b63776f	241	TRACE_EVENT_FL_IGNORE_ENABLE = (1 << TRACE_EVENT_FL_IGNORE_ENABLE_BIT),
575380da	242	TRACE_EVENT_FL_WAS_ENABLED = (1 << TRACE_EVENT_FL_WAS_ENABLED_BIT),
de7b2973	243	TRACE_EVENT_FL_TRACEPOINT = (1 << TRACE_EVENT_FL_TRACEPOINT_BIT),
72cbbc89	244	TRACE_EVENT_FL_KPROBE = (1 << TRACE_EVENT_FL_KPROBE_BIT),
04a22fae	245	TRACE_EVENT_FL_UPROBE = (1 << TRACE_EVENT_FL_UPROBE_BIT),
b1dace68	246	TRACE_EVENT_FL_MAP = (1 << TRACE_EVENT_FL_MAP_BIT),
553552ce SR	247	};
553552ce SR	248
04a22fae WN	249	#define TRACE_EVENT_FL_UKPROBE (TRACE_EVENT_FL_KPROBE \| TRACE_EVENT_FL_UPROBE)
04a22fae WN	250
b1dace68 JD	251	struct trace_event_map {
	252	struct tracepoint *tp;
	253	char *name;
	254	};
	255
2425bcb9	256	struct trace_event_call {
a59fd602	257	struct list_head list;
2425bcb9	258	struct trace_event_class *class;
de7b2973 MD	259	union {
	260	char *name;
	261	/* Set TRACE_EVENT_FL_TRACEPOINT flag when using "tp" */
	262	struct tracepoint *tp;
b1dace68 JD	263	/* Set TRACE_EVENT_FL_MAP flag when using "map" instead */
b1dace68 JD	264	struct trace_event_map *map;
de7b2973	265	};
80decc70	266	struct trace_event event;
0c564a53	267	char *print_fmt;
1f9963cb	268	struct event_filter *filter;
6d723736	269	void *mod;
69fd4f0e	270	void *data;
57d01ad0 SRRH	271	/*
	272	* bit 0: filter_active
	273	* bit 1: allow trace by non root (cap any)
	274	* bit 2: failed to apply filter
5d6ad960	275	* bit 3: trace internal event (do not enable)
57d01ad0	276	* bit 4: Event was enabled by module
f306cc82	277	* bit 5: use call filter rather than file filter
de7b2973	278	* bit 6: Event is a tracepoint
57d01ad0	279	*/
ae63b31e SR	280	int flags; /* static flags of different events */
	281
	282	#ifdef CONFIG_PERF_EVENTS
	283	int perf_refcount;
	284	struct hlist_head __percpu *perf_events;
2541517c	285	struct bpf_prog *prog;
d5b5f391	286
2425bcb9	287	int (perf_perm)(struct trace_event_call ,
d5b5f391	288	struct perf_event *);
ae63b31e SR	289	#endif
	290	};
	291
de7b2973	292	static inline const char *
687fcc4a	293	trace_event_name(struct trace_event_call *call)
de7b2973	294	{
b1dace68 JD	295	if (call->flags & TRACE_EVENT_FL_MAP)
	296	return call->map->name;
	297	else if (call->flags & TRACE_EVENT_FL_TRACEPOINT)
de7b2973 MD	298	return call->tp ? call->tp->name : NULL;
	299	else
	300	return call->name;
	301	}
	302
ae63b31e	303	struct trace_array;
7967b3e0	304	struct trace_subsystem_dir;
ae63b31e SR	305
ae63b31e SR	306	enum {
5d6ad960 SRRH	307	EVENT_FILE_FL_ENABLED_BIT,
	308	EVENT_FILE_FL_RECORDED_CMD_BIT,
	309	EVENT_FILE_FL_FILTERED_BIT,
	310	EVENT_FILE_FL_NO_SET_FILTER_BIT,
	311	EVENT_FILE_FL_SOFT_MODE_BIT,
	312	EVENT_FILE_FL_SOFT_DISABLED_BIT,
	313	EVENT_FILE_FL_TRIGGER_MODE_BIT,
	314	EVENT_FILE_FL_TRIGGER_COND_BIT,
3fdaf80f	315	EVENT_FILE_FL_PID_FILTER_BIT,
ae63b31e SR	316	};
	317
	318	/*
5d6ad960	319	* Event file flags:
57d01ad0	320	* ENABLED - The event is enabled
ae63b31e	321	* RECORDED_CMD - The comms should be recorded at sched_switch
f306cc82 TZ	322	* FILTERED - The event has a filter attached
f306cc82 TZ	323	* NO_SET_FILTER - Set when filter has error and is to be ignored
417944c4 SRRH	324	* SOFT_MODE - The event is enabled/disabled by SOFT_DISABLED
	325	* SOFT_DISABLED - When set, do not trace the event (even though its
	326	* tracepoint may be enabled)
85f2b082	327	* TRIGGER_MODE - When set, invoke the triggers associated with the event
bac5fb97	328	* TRIGGER_COND - When set, one or more triggers has an associated filter
3fdaf80f	329	* PID_FILTER - When set, the event is filtered based on pid
ae63b31e SR	330	*/
ae63b31e SR	331	enum {
5d6ad960 SRRH	332	EVENT_FILE_FL_ENABLED = (1 << EVENT_FILE_FL_ENABLED_BIT),
	333	EVENT_FILE_FL_RECORDED_CMD = (1 << EVENT_FILE_FL_RECORDED_CMD_BIT),
	334	EVENT_FILE_FL_FILTERED = (1 << EVENT_FILE_FL_FILTERED_BIT),
	335	EVENT_FILE_FL_NO_SET_FILTER = (1 << EVENT_FILE_FL_NO_SET_FILTER_BIT),
	336	EVENT_FILE_FL_SOFT_MODE = (1 << EVENT_FILE_FL_SOFT_MODE_BIT),
	337	EVENT_FILE_FL_SOFT_DISABLED = (1 << EVENT_FILE_FL_SOFT_DISABLED_BIT),
	338	EVENT_FILE_FL_TRIGGER_MODE = (1 << EVENT_FILE_FL_TRIGGER_MODE_BIT),
	339	EVENT_FILE_FL_TRIGGER_COND = (1 << EVENT_FILE_FL_TRIGGER_COND_BIT),
3fdaf80f	340	EVENT_FILE_FL_PID_FILTER = (1 << EVENT_FILE_FL_PID_FILTER_BIT),
ae63b31e SR	341	};
ae63b31e SR	342
7f1d2f82	343	struct trace_event_file {
ae63b31e	344	struct list_head list;
2425bcb9	345	struct trace_event_call *event_call;
f306cc82	346	struct event_filter *filter;
ae63b31e SR	347	struct dentry *dir;
ae63b31e SR	348	struct trace_array *tr;
7967b3e0	349	struct trace_subsystem_dir *system;
85f2b082	350	struct list_head triggers;
97f20251	351
553552ce SR	352	/*
553552ce SR	353	* 32 bit flags:
57d01ad0 SRRH	354	* bit 0: enabled
57d01ad0 SRRH	355	* bit 1: enabled cmd record
417944c4 SRRH	356	* bit 2: enable/disable with the soft disable bit
417944c4 SRRH	357	* bit 3: soft disabled
85f2b082	358	* bit 4: trigger enabled
553552ce	359	*
417944c4 SRRH	360	* Note: The bits must be set atomically to prevent races
	361	* from other writers. Reads of flags do not need to be in
	362	* sync as they occur in critical sections. But the way flags
ae63b31e	363	* is currently used, these changes do not affect the code
1eaa4787 SR	364	* except that when a change is made, it may have a slight
1eaa4787 SR	365	* delay in propagating the changes to other CPUs due to
417944c4	366	* caching and such. Which is mostly OK ;-)
553552ce	367	*/
417944c4	368	unsigned long flags;
1cf4c073	369	atomic_t sm_ref; /* soft-mode reference counter */
85f2b082	370	atomic_t tm_ref; /* trigger-mode reference counter */
97f20251 SR	371	};
97f20251 SR	372
53cf810b FW	373	#define __TRACE_EVENT_FLAGS(name, value) \
	374	static int __init trace_init_flags_##name(void) \
	375	{ \
de7b2973	376	event_##name.flags \|= value; \
53cf810b FW	377	return 0; \
	378	} \
	379	early_initcall(trace_init_flags_##name);
	380
d5b5f391	381	#define __TRACE_EVENT_PERF_PERM(name, expr...) \
2425bcb9	382	static int perf_perm_##name(struct trace_event_call *tp_event, \
d5b5f391 PZ	383	struct perf_event *p_event) \
	384	{ \
	385	return ({ expr; }); \
	386	} \
	387	static int __init trace_init_perf_perm_##name(void) \
	388	{ \
	389	event_##name.perf_perm = &perf_perm_##name; \
	390	return 0; \
	391	} \
	392	early_initcall(trace_init_perf_perm_##name);
	393
97d5a220	394	#define PERF_MAX_TRACE_SIZE 2048
20ab4425	395
16bb8eb1	396	#define MAX_FILTER_STR_VAL 256 /* Should handle KSYM_SYMBOL_LEN */
97f20251	397
85f2b082 TZ	398	enum event_trigger_type {
85f2b082 TZ	399	ETT_NONE = (0),
2a2df321	400	ETT_TRACE_ONOFF = (1 << 0),
93e31ffb	401	ETT_SNAPSHOT = (1 << 1),
f21ecbb3	402	ETT_STACKTRACE = (1 << 2),
7862ad18	403	ETT_EVENT_ENABLE = (1 << 3),
7ef224d1	404	ETT_EVENT_HIST = (1 << 4),
d0bad49b	405	ETT_HIST_ENABLE = (1 << 5),
85f2b082 TZ	406	};
85f2b082 TZ	407
6fb2915d	408	extern int filter_match_preds(struct event_filter filter, void rec);
f306cc82	409
7f1d2f82	410	extern enum event_trigger_type event_triggers_call(struct trace_event_file *file,
bac5fb97	411	void *rec);
7f1d2f82	412	extern void event_triggers_post_call(struct trace_event_file *file,
c4a59230 TZ	413	enum event_trigger_type tt,
c4a59230 TZ	414	void *rec);
97f20251	415
3fdaf80f SRRH	416	bool trace_event_ignore_this_pid(struct trace_event_file *trace_file);
3fdaf80f SRRH	417
13a1e4ae	418	/**
09a5059a	419	* trace_trigger_soft_disabled - do triggers and test if soft disabled
13a1e4ae SRRH	420	* @file: The file pointer of the event to test
	421	*
	422	* If any triggers without filters are attached to this event, they
	423	* will be called here. If the event is soft disabled and has no
	424	* triggers that require testing the fields, it will return true,
	425	* otherwise false.
	426	*/
	427	static inline bool
09a5059a	428	trace_trigger_soft_disabled(struct trace_event_file *file)
13a1e4ae SRRH	429	{
	430	unsigned long eflags = file->flags;
	431
5d6ad960 SRRH	432	if (!(eflags & EVENT_FILE_FL_TRIGGER_COND)) {
5d6ad960 SRRH	433	if (eflags & EVENT_FILE_FL_TRIGGER_MODE)
13a1e4ae	434	event_triggers_call(file, NULL);
5d6ad960	435	if (eflags & EVENT_FILE_FL_SOFT_DISABLED)
13a1e4ae	436	return true;
3fdaf80f SRRH	437	if (eflags & EVENT_FILE_FL_PID_FILTER)
3fdaf80f SRRH	438	return trace_event_ignore_this_pid(file);
13a1e4ae SRRH	439	}
	440	return false;
	441	}
	442
098d2164	443	#ifdef CONFIG_BPF_EVENTS
2541517c AS	444	unsigned int trace_call_bpf(struct bpf_prog prog, void ctx);
	445	#else
	446	static inline unsigned int trace_call_bpf(struct bpf_prog prog, void ctx)
	447	{
	448	return 1;
	449	}
	450	#endif
	451
43b51ead LZ	452	enum {
	453	FILTER_OTHER = 0,
	454	FILTER_STATIC_STRING,
	455	FILTER_DYN_STRING,
87a342f5	456	FILTER_PTR_STRING,
02aa3162	457	FILTER_TRACE_FN,
e57cbaf0 SRRH	458	FILTER_COMM,
e57cbaf0 SRRH	459	FILTER_CPU,
43b51ead LZ	460	};
43b51ead LZ	461
2425bcb9 SRRH	462	extern int trace_event_raw_init(struct trace_event_call *call);
2425bcb9 SRRH	463	extern int trace_define_field(struct trace_event_call call, const char type,
aeaeae11 FW	464	const char *name, int offset, int size,
aeaeae11 FW	465	int is_signed, int filter_type);
2425bcb9 SRRH	466	extern int trace_add_event_call(struct trace_event_call *call);
2425bcb9 SRRH	467	extern int trace_remove_event_call(struct trace_event_call *call);
32bbe007	468	extern int trace_event_get_offsets(struct trace_event_call *call);
97f20251	469
d2802d07	470	#define is_signed_type(type) (((type)(-1)) < (type)1)
97f20251	471
4671c794 SR	472	int trace_set_clr_event(const char system, const char event, int set);
4671c794 SR	473
97f20251 SR	474	/*
	475	* The double __builtin_constant_p is because gcc will give us an error
	476	* if we try to allocate the static variable to fmt if it is not a
	477	* constant. Even with the outer if statement optimizing out.
	478	*/
	479	#define event_trace_printk(ip, fmt, args...) \
	480	do { \
	481	__trace_printk_check_format(fmt, ##args); \
	482	tracing_record_cmdline(current); \
	483	if (__builtin_constant_p(fmt)) { \
	484	static const char *trace_printk_fmt \
	485	__attribute__((section("__trace_printk_fmt"))) = \
	486	__builtin_constant_p(fmt) ? fmt : NULL; \
	487	\
	488	__trace_bprintk(ip, trace_printk_fmt, ##args); \
	489	} else \
	490	__trace_printk(ip, fmt, ##args); \
	491	} while (0)
	492
07b139c8	493	#ifdef CONFIG_PERF_EVENTS
6fb2915d	494	struct perf_event;
c530665c FW	495
	496	DECLARE_PER_CPU(struct pt_regs, perf_trace_regs);
	497
1c024eca PZ	498	extern int perf_trace_init(struct perf_event *event);
1c024eca PZ	499	extern void perf_trace_destroy(struct perf_event *event);
a4eaf7f1 PZ	500	extern int perf_trace_add(struct perf_event *event, int flags);
a4eaf7f1 PZ	501	extern void perf_trace_del(struct perf_event *event, int flags);
1c024eca	502	extern int ftrace_profile_set_filter(struct perf_event *event, int event_id,
6fb2915d LZ	503	char *filter_str);
6fb2915d LZ	504	extern void ftrace_profile_free_filter(struct perf_event *event);
1e1dcd93 AS	505	void perf_trace_buf_update(void *record, u16 type);
1e1dcd93 AS	506	void perf_trace_buf_alloc(int size, struct pt_regs regs, int rctxp);
430ad5a6	507
85b67bcb AS	508	void perf_trace_run_bpf_submit(void *raw_data, int size, int rctx,
	509	struct trace_event_call *call, u64 count,
	510	struct pt_regs regs, struct hlist_head head,
	511	struct task_struct *task);
	512
430ad5a6	513	static inline void
1e1dcd93	514	perf_trace_buf_submit(void *raw_data, int size, int rctx, u16 type,
e6dab5ff AV	515	u64 count, struct pt_regs regs, void head,
e6dab5ff AV	516	struct task_struct *task)
430ad5a6	517	{
1e1dcd93	518	perf_tp_event(type, count, raw_data, size, regs, head, rctx, task);
430ad5a6	519	}
6fb2915d LZ	520	#endif
6fb2915d LZ	521
2425bcb9	522	#endif /* _LINUX_TRACE_EVENT_H */