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

#ifndef _LINUX_TRACEPOINT_H
#define _LINUX_TRACEPOINT_H

/*
 * Kernel Tracepoint API.
 *
 * See Documentation/trace/tracepoints.txt.
 *
 * Copyright (C) 2008-2014 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
 *
 * Heavily inspired from the Linux Kernel Markers.
 *
 * This file is released under the GPLv2.
 * See the file COPYING for more details.
 */

#include <linux/errno.h>
#include <linux/types.h>
#include <linux/rcupdate.h>
#include <linux/tracepoint-defs.h>

struct module;
struct tracepoint;
struct notifier_block;

struct trace_enum_map {
	const char		*system;
	const char		*enum_string;
	unsigned long		enum_value;
};

#define TRACEPOINT_DEFAULT_PRIO	10

extern int
tracepoint_probe_register(struct tracepoint *tp, void *probe, void *data);
extern int
tracepoint_probe_register_prio(struct tracepoint *tp, void *probe, void *data,
			       int prio);
extern int
tracepoint_probe_unregister(struct tracepoint *tp, void *probe, void *data);
extern void
for_each_kernel_tracepoint(void (*fct)(struct tracepoint *tp, void *priv),
		void *priv);

#ifdef CONFIG_MODULES
struct tp_module {
	struct list_head list;
	struct module *mod;
};

bool trace_module_has_bad_taint(struct module *mod);
extern int register_tracepoint_module_notifier(struct notifier_block *nb);
extern int unregister_tracepoint_module_notifier(struct notifier_block *nb);
#else
static inline bool trace_module_has_bad_taint(struct module *mod)
{
	return false;
}
static inline
int register_tracepoint_module_notifier(struct notifier_block *nb)
{
	return 0;
}
static inline
int unregister_tracepoint_module_notifier(struct notifier_block *nb)
{
	return 0;
}
#endif /* CONFIG_MODULES */

/*
 * tracepoint_synchronize_unregister must be called between the last tracepoint
 * probe unregistration and the end of module exit to make sure there is no
 * caller executing a probe when it is freed.
 */
static inline void tracepoint_synchronize_unregister(void)
{
	synchronize_sched();
}

#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS
extern void syscall_regfunc(void);
extern void syscall_unregfunc(void);
#endif /* CONFIG_HAVE_SYSCALL_TRACEPOINTS */

#define PARAMS(args...) args

#define TRACE_DEFINE_ENUM(x)

#endif /* _LINUX_TRACEPOINT_H */

/*
 * Note: we keep the TRACE_EVENT and DECLARE_TRACE outside the include
 *  file ifdef protection.
 *  This is due to the way trace events work. If a file includes two
 *  trace event headers under one "CREATE_TRACE_POINTS" the first include
 *  will override the TRACE_EVENT and break the second include.
 */

#ifndef DECLARE_TRACE

#define TP_PROTO(args...)	args
#define TP_ARGS(args...)	args
#define TP_CONDITION(args...)	args

/*
 * Individual subsystem my have a separate configuration to
 * enable their tracepoints. By default, this file will create
 * the tracepoints if CONFIG_TRACEPOINT is defined. If a subsystem
 * wants to be able to disable its tracepoints from being created
 * it can define NOTRACE before including the tracepoint headers.
 */
#if defined(CONFIG_TRACEPOINTS) && !defined(NOTRACE)
#define TRACEPOINTS_ENABLED
#endif

#ifdef TRACEPOINTS_ENABLED

/*
 * it_func[0] is never NULL because there is at least one element in the array
 * when the array itself is non NULL.
 *
 * Note, the proto and args passed in includes "__data" as the first parameter.
 * The reason for this is to handle the "void" prototype. If a tracepoint
 * has a "void" prototype, then it is invalid to declare a function
 * as "(void *, void)". The DECLARE_TRACE_NOARGS() will pass in just
 * "void *data", where as the DECLARE_TRACE() will pass in "void *data, proto".
 */
#define __DO_TRACE(tp, proto, args, cond, prercu, postrcu)		\
	do {								\
		struct tracepoint_func *it_func_ptr;			\
		void *it_func;						\
		void *__data;						\
									\
		if (!(cond))						\
			return;						\
		prercu;							\
		rcu_read_lock_sched_notrace();				\
		it_func_ptr = rcu_dereference_sched((tp)->funcs);	\
		if (it_func_ptr) {					\
			do {						\
				it_func = (it_func_ptr)->func;		\
				__data = (it_func_ptr)->data;		\
				((void(*)(proto))(it_func))(args);	\
			} while ((++it_func_ptr)->func);		\
		}							\
		rcu_read_unlock_sched_notrace();			\
		postrcu;						\
	} while (0)

#ifndef MODULE
#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)	\
	static inline void trace_##name##_rcuidle(proto)		\
	{								\
		if (static_key_false(&__tracepoint_##name.key))		\
			__DO_TRACE(&__tracepoint_##name,		\
				TP_PROTO(data_proto),			\
				TP_ARGS(data_args),			\
				TP_CONDITION(cond),			\
				rcu_irq_enter_irqson(),			\
				rcu_irq_exit_irqson());			\
	}
#else
#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)
#endif

/*
 * Make sure the alignment of the structure in the __tracepoints section will
 * not add unwanted padding between the beginning of the section and the
 * structure. Force alignment to the same alignment as the section start.
 *
 * When lockdep is enabled, we make sure to always do the RCU portions of
 * the tracepoint code, regardless of whether tracing is on. However,
 * don't check if the condition is false, due to interaction with idle
 * instrumentation. This lets us find RCU issues triggered with tracepoints
 * even when this tracepoint is off. This code has no purpose other than
 * poking RCU a bit.
 */
#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
	extern struct tracepoint __tracepoint_##name;			\
	static inline void trace_##name(proto)				\
	{								\
		if (static_key_false(&__tracepoint_##name.key))		\
			__DO_TRACE(&__tracepoint_##name,		\
				TP_PROTO(data_proto),			\
				TP_ARGS(data_args),			\
				TP_CONDITION(cond),,);			\
		if (IS_ENABLED(CONFIG_LOCKDEP) && (cond)) {		\
			rcu_read_lock_sched_notrace();			\
			rcu_dereference_sched(__tracepoint_##name.funcs);\
			rcu_read_unlock_sched_notrace();		\
		}							\
	}								\
	__DECLARE_TRACE_RCU(name, PARAMS(proto), PARAMS(args),		\
		PARAMS(cond), PARAMS(data_proto), PARAMS(data_args))	\
	static inline int						\
	register_trace_##name(void (*probe)(data_proto), void *data)	\
	{								\
		return tracepoint_probe_register(&__tracepoint_##name,	\
						(void *)probe, data);	\
	}								\
	static inline int						\
	register_trace_prio_##name(void (*probe)(data_proto), void *data,\
				   int prio)				\
	{								\
		return tracepoint_probe_register_prio(&__tracepoint_##name, \
					      (void *)probe, data, prio); \
	}								\
	static inline int						\
	unregister_trace_##name(void (*probe)(data_proto), void *data)	\
	{								\
		return tracepoint_probe_unregister(&__tracepoint_##name,\
						(void *)probe, data);	\
	}								\
	static inline void						\
	check_trace_callback_type_##name(void (*cb)(data_proto))	\
	{								\
	}								\
	static inline bool						\
	trace_##name##_enabled(void)					\
	{								\
		return static_key_false(&__tracepoint_##name.key);	\
	}

/*
 * We have no guarantee that gcc and the linker won't up-align the tracepoint
 * structures, so we create an array of pointers that will be used for iteration
 * on the tracepoints.
 */
#define DEFINE_TRACE_FN(name, reg, unreg)				 \
	static const char __tpstrtab_##name[]				 \
	__attribute__((section("__tracepoints_strings"))) = #name;	 \
	struct tracepoint __tracepoint_##name				 \
	__attribute__((section("__tracepoints"))) =			 \
		{ __tpstrtab_##name, STATIC_KEY_INIT_FALSE, reg, unreg, NULL };\
	static struct tracepoint * const __tracepoint_ptr_##name __used	 \
	__attribute__((section("__tracepoints_ptrs"))) =		 \
		&__tracepoint_##name;

#define DEFINE_TRACE(name)						\
	DEFINE_TRACE_FN(name, NULL, NULL);

#define EXPORT_TRACEPOINT_SYMBOL_GPL(name)				\
	EXPORT_SYMBOL_GPL(__tracepoint_##name)
#define EXPORT_TRACEPOINT_SYMBOL(name)					\
	EXPORT_SYMBOL(__tracepoint_##name)

#else /* !TRACEPOINTS_ENABLED */
#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
	static inline void trace_##name(proto)				\
	{ }								\
	static inline void trace_##name##_rcuidle(proto)		\
	{ }								\
	static inline int						\
	register_trace_##name(void (*probe)(data_proto),		\
			      void *data)				\
	{								\
		return -ENOSYS;						\
	}								\
	static inline int						\
	unregister_trace_##name(void (*probe)(data_proto),		\
				void *data)				\
	{								\
		return -ENOSYS;						\
	}								\
	static inline void check_trace_callback_type_##name(void (*cb)(data_proto)) \
	{								\
	}								\
	static inline bool						\
	trace_##name##_enabled(void)					\
	{								\
		return false;						\
	}

#define DEFINE_TRACE_FN(name, reg, unreg)
#define DEFINE_TRACE(name)
#define EXPORT_TRACEPOINT_SYMBOL_GPL(name)
#define EXPORT_TRACEPOINT_SYMBOL(name)

#endif /* TRACEPOINTS_ENABLED */

#ifdef CONFIG_TRACING
/**
 * tracepoint_string - register constant persistent string to trace system
 * @str - a constant persistent string that will be referenced in tracepoints
 *
 * If constant strings are being used in tracepoints, it is faster and
 * more efficient to just save the pointer to the string and reference
 * that with a printf "%s" instead of saving the string in the ring buffer
 * and wasting space and time.
 *
 * The problem with the above approach is that userspace tools that read
 * the binary output of the trace buffers do not have access to the string.
 * Instead they just show the address of the string which is not very
 * useful to users.
 *
 * With tracepoint_string(), the string will be registered to the tracing
 * system and exported to userspace via the debugfs/tracing/printk_formats
 * file that maps the string address to the string text. This way userspace
 * tools that read the binary buffers have a way to map the pointers to
 * the ASCII strings they represent.
 *
 * The @str used must be a constant string and persistent as it would not
 * make sense to show a string that no longer exists. But it is still fine
 * to be used with modules, because when modules are unloaded, if they
 * had tracepoints, the ring buffers are cleared too. As long as the string
 * does not change during the life of the module, it is fine to use
 * tracepoint_string() within a module.
 */
#define tracepoint_string(str)						\
	({								\
		static const char *___tp_str __tracepoint_string = str; \
		___tp_str;						\
	})
#define __tracepoint_string	__attribute__((section("__tracepoint_str")))
#else
/*
 * tracepoint_string() is used to save the string address for userspace
 * tracing tools. When tracing isn't configured, there's no need to save
 * anything.
 */
# define tracepoint_string(str) str
# define __tracepoint_string
#endif

/*
 * The need for the DECLARE_TRACE_NOARGS() is to handle the prototype
 * (void). "void" is a special value in a function prototype and can
 * not be combined with other arguments. Since the DECLARE_TRACE()
 * macro adds a data element at the beginning of the prototype,
 * we need a way to differentiate "(void *data, proto)" from
 * "(void *data, void)". The second prototype is invalid.
 *
 * DECLARE_TRACE_NOARGS() passes "void" as the tracepoint prototype
 * and "void *__data" as the callback prototype.
 *
 * DECLARE_TRACE() passes "proto" as the tracepoint protoype and
 * "void *__data, proto" as the callback prototype.
 */
#define DECLARE_TRACE_NOARGS(name)					\
		__DECLARE_TRACE(name, void, , 1, void *__data, __data)

#define DECLARE_TRACE(name, proto, args)				\
		__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), 1,	\
				PARAMS(void *__data, proto),		\
				PARAMS(__data, args))

#define DECLARE_TRACE_CONDITION(name, proto, args, cond)		\
	__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), PARAMS(cond), \
			PARAMS(void *__data, proto),			\
			PARAMS(__data, args))

#define TRACE_EVENT_FLAGS(event, flag)

#define TRACE_EVENT_PERF_PERM(event, expr...)

#endif /* DECLARE_TRACE */

#ifndef TRACE_EVENT
/*
 * For use with the TRACE_EVENT macro:
 *
 * We define a tracepoint, its arguments, its printk format
 * and its 'fast binary record' layout.
 *
 * Firstly, name your tracepoint via TRACE_EVENT(name : the
 * 'subsystem_event' notation is fine.
 *
 * Think about this whole construct as the
 * 'trace_sched_switch() function' from now on.
 *
 *
 *  TRACE_EVENT(sched_switch,
 *
 *	*
 *	* A function has a regular function arguments
 *	* prototype, declare it via TP_PROTO():
 *	*
 *
 *	TP_PROTO(struct rq *rq, struct task_struct *prev,
 *		 struct task_struct *next),
 *
 *	*
 *	* Define the call signature of the 'function'.
 *	* (Design sidenote: we use this instead of a
 *	*  TP_PROTO1/TP_PROTO2/TP_PROTO3 ugliness.)
 *	*
 *
 *	TP_ARGS(rq, prev, next),
 *
 *	*
 *	* Fast binary tracing: define the trace record via
 *	* TP_STRUCT__entry(). You can think about it like a
 *	* regular C structure local variable definition.
 *	*
 *	* This is how the trace record is structured and will
 *	* be saved into the ring buffer. These are the fields
 *	* that will be exposed to user-space in
 *	* /sys/kernel/debug/tracing/events/<*>/format.
 *	*
 *	* The declared 'local variable' is called '__entry'
 *	*
 *	* __field(pid_t, prev_prid) is equivalent to a standard declariton:
 *	*
 *	*	pid_t	prev_pid;
 *	*
 *	* __array(char, prev_comm, TASK_COMM_LEN) is equivalent to:
 *	*
 *	*	char	prev_comm[TASK_COMM_LEN];
 *	*
 *
 *	TP_STRUCT__entry(
 *		__array(	char,	prev_comm,	TASK_COMM_LEN	)
 *		__field(	pid_t,	prev_pid			)
 *		__field(	int,	prev_prio			)
 *		__array(	char,	next_comm,	TASK_COMM_LEN	)
 *		__field(	pid_t,	next_pid			)
 *		__field(	int,	next_prio			)
 *	),
 *
 *	*
 *	* Assign the entry into the trace record, by embedding
 *	* a full C statement block into TP_fast_assign(). You
 *	* can refer to the trace record as '__entry' -
 *	* otherwise you can put arbitrary C code in here.
 *	*
 *	* Note: this C code will execute every time a trace event
 *	* happens, on an active tracepoint.
 *	*
 *
 *	TP_fast_assign(
 *		memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN);
 *		__entry->prev_pid	= prev->pid;
 *		__entry->prev_prio	= prev->prio;
 *		memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN);
 *		__entry->next_pid	= next->pid;
 *		__entry->next_prio	= next->prio;
 *	),
 *
 *	*
 *	* Formatted output of a trace record via TP_printk().
 *	* This is how the tracepoint will appear under ftrace
 *	* plugins that make use of this tracepoint.
 *	*
 *	* (raw-binary tracing wont actually perform this step.)
 *	*
 *
 *	TP_printk("task %s:%d [%d] ==> %s:%d [%d]",
 *		__entry->prev_comm, __entry->prev_pid, __entry->prev_prio,
 *		__entry->next_comm, __entry->next_pid, __entry->next_prio),
 *
 * );
 *
 * This macro construct is thus used for the regular printk format
 * tracing setup, it is used to construct a function pointer based
 * tracepoint callback (this is used by programmatic plugins and
 * can also by used by generic instrumentation like SystemTap), and
 * it is also used to expose a structured trace record in
 * /sys/kernel/debug/tracing/events/.
 *
 * A set of (un)registration functions can be passed to the variant
 * TRACE_EVENT_FN to perform any (un)registration work.
 */

#define DECLARE_EVENT_CLASS(name, proto, args, tstruct, assign, print)
#define DEFINE_EVENT(template, name, proto, args)		\
	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
#define DEFINE_EVENT_FN(template, name, proto, args, reg, unreg)\
	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
#define DEFINE_EVENT_PRINT(template, name, proto, args, print)	\
	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
#define DEFINE_EVENT_CONDITION(template, name, proto,		\
			       args, cond)			\
	DECLARE_TRACE_CONDITION(name, PARAMS(proto),		\
				PARAMS(args), PARAMS(cond))

#define TRACE_EVENT(name, proto, args, struct, assign, print)	\
	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
#define TRACE_EVENT_FN(name, proto, args, struct,		\
		assign, print, reg, unreg)			\
	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
#define TRACE_EVENT_FN_COND(name, proto, args, cond, struct,		\
		assign, print, reg, unreg)			\
	DECLARE_TRACE_CONDITION(name, PARAMS(proto),	\
			PARAMS(args), PARAMS(cond))
#define TRACE_EVENT_CONDITION(name, proto, args, cond,		\
			      struct, assign, print)		\
	DECLARE_TRACE_CONDITION(name, PARAMS(proto),		\
				PARAMS(args), PARAMS(cond))

#define TRACE_EVENT_FLAGS(event, flag)

#define TRACE_EVENT_PERF_PERM(event, expr...)

#endif /* ifdef TRACE_EVENT (see note above) */
Commit	Line	Data
97e1c18e MD	1	#ifndef _LINUX_TRACEPOINT_H
	2	#define _LINUX_TRACEPOINT_H
	3
	4	/*
	5	* Kernel Tracepoint API.
	6	*
8cd09a59	7	* See Documentation/trace/tracepoints.txt.
97e1c18e	8	*
de7b2973	9	* Copyright (C) 2008-2014 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
97e1c18e MD	10	*
	11	* Heavily inspired from the Linux Kernel Markers.
	12	*
	13	* This file is released under the GPLv2.
	14	* See the file COPYING for more details.
	15	*/
	16
b70e4f05	17	#include <linux/errno.h>
97e1c18e MD	18	#include <linux/types.h>
97e1c18e MD	19	#include <linux/rcupdate.h>
bd2a634d	20	#include <linux/tracepoint-defs.h>
97e1c18e MD	21
	22	struct module;
	23	struct tracepoint;
de7b2973	24	struct notifier_block;
97e1c18e	25
0c564a53 SRRH	26	struct trace_enum_map {
	27	const char *system;
	28	const char *enum_string;
	29	unsigned long enum_value;
	30	};
	31
7904b5c4 SRRH	32	#define TRACEPOINT_DEFAULT_PRIO 10
7904b5c4 SRRH	33
38516ab5	34	extern int
de7b2973 MD	35	tracepoint_probe_register(struct tracepoint tp, void probe, void *data);
de7b2973 MD	36	extern int
7904b5c4 SRRH	37	tracepoint_probe_register_prio(struct tracepoint tp, void probe, void *data,
	38	int prio);
	39	extern int
de7b2973 MD	40	tracepoint_probe_unregister(struct tracepoint tp, void probe, void *data);
	41	extern void
	42	for_each_kernel_tracepoint(void (fct)(struct tracepoint tp, void *priv),
	43	void *priv);
2e26ca71	44
b75ef8b4 MD	45	#ifdef CONFIG_MODULES
	46	struct tp_module {
	47	struct list_head list;
eb7d035c	48	struct module *mod;
b75ef8b4	49	};
de7b2973	50
45ab2813	51	bool trace_module_has_bad_taint(struct module *mod);
de7b2973 MD	52	extern int register_tracepoint_module_notifier(struct notifier_block *nb);
de7b2973 MD	53	extern int unregister_tracepoint_module_notifier(struct notifier_block *nb);
45ab2813 SRRH	54	#else
	55	static inline bool trace_module_has_bad_taint(struct module *mod)
	56	{
	57	return false;
	58	}
de7b2973 MD	59	static inline
	60	int register_tracepoint_module_notifier(struct notifier_block *nb)
	61	{
	62	return 0;
	63	}
	64	static inline
	65	int unregister_tracepoint_module_notifier(struct notifier_block *nb)
	66	{
	67	return 0;
	68	}
b75ef8b4 MD	69	#endif /* CONFIG_MODULES */
b75ef8b4 MD	70
2e26ca71 SR	71	/*
	72	* tracepoint_synchronize_unregister must be called between the last tracepoint
	73	* probe unregistration and the end of module exit to make sure there is no
	74	* caller executing a probe when it is freed.
	75	*/
	76	static inline void tracepoint_synchronize_unregister(void)
	77	{
	78	synchronize_sched();
	79	}
	80
b725dfea MD	81	#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS
	82	extern void syscall_regfunc(void);
	83	extern void syscall_unregfunc(void);
	84	#endif /* CONFIG_HAVE_SYSCALL_TRACEPOINTS */
	85
2e26ca71 SR	86	#define PARAMS(args...) args
2e26ca71 SR	87
0c564a53 SRRH	88	#define TRACE_DEFINE_ENUM(x)
0c564a53 SRRH	89
2e26ca71 SR	90	#endif /* _LINUX_TRACEPOINT_H */
	91
	92	/*
	93	* Note: we keep the TRACE_EVENT and DECLARE_TRACE outside the include
	94	* file ifdef protection.
	95	* This is due to the way trace events work. If a file includes two
	96	* trace event headers under one "CREATE_TRACE_POINTS" the first include
	97	* will override the TRACE_EVENT and break the second include.
	98	*/
	99
ea20d929 SR	100	#ifndef DECLARE_TRACE
ea20d929 SR	101
2939b046	102	#define TP_PROTO(args...) args
8cd09a59	103	#define TP_ARGS(args...) args
287050d3	104	#define TP_CONDITION(args...) args
97e1c18e	105
c63b7682 TS	106	/*
	107	* Individual subsystem my have a separate configuration to
	108	* enable their tracepoints. By default, this file will create
	109	* the tracepoints if CONFIG_TRACEPOINT is defined. If a subsystem
	110	* wants to be able to disable its tracepoints from being created
	111	* it can define NOTRACE before including the tracepoint headers.
	112	*/
	113	#if defined(CONFIG_TRACEPOINTS) && !defined(NOTRACE)
	114	#define TRACEPOINTS_ENABLED
	115	#endif
	116
	117	#ifdef TRACEPOINTS_ENABLED
97e1c18e MD	118
	119	/*
	120	* it_func[0] is never NULL because there is at least one element in the array
	121	* when the array itself is non NULL.
38516ab5 SR	122	*
	123	* Note, the proto and args passed in includes "__data" as the first parameter.
	124	* The reason for this is to handle the "void" prototype. If a tracepoint
	125	* has a "void" prototype, then it is invalid to declare a function
	126	* as "(void *, void)". The DECLARE_TRACE_NOARGS() will pass in just
	127	* "void data", where as the DECLARE_TRACE() will pass in "void data, proto".
97e1c18e	128	*/
2fbb90db	129	#define __DO_TRACE(tp, proto, args, cond, prercu, postrcu) \
97e1c18e	130	do { \
38516ab5 SR	131	struct tracepoint_func *it_func_ptr; \
	132	void *it_func; \
	133	void *__data; \
97e1c18e	134	\
287050d3 SR	135	if (!(cond)) \
287050d3 SR	136	return; \
2fbb90db	137	prercu; \
da7b3eab	138	rcu_read_lock_sched_notrace(); \
38516ab5 SR	139	it_func_ptr = rcu_dereference_sched((tp)->funcs); \
38516ab5 SR	140	if (it_func_ptr) { \
97e1c18e	141	do { \
38516ab5 SR	142	it_func = (it_func_ptr)->func; \
	143	__data = (it_func_ptr)->data; \
	144	((void(*)(proto))(it_func))(args); \
	145	} while ((++it_func_ptr)->func); \
97e1c18e	146	} \
da7b3eab	147	rcu_read_unlock_sched_notrace(); \
2fbb90db	148	postrcu; \
97e1c18e MD	149	} while (0)
97e1c18e MD	150
7ece55a4 JT	151	#ifndef MODULE
	152	#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args) \
	153	static inline void trace_##name##_rcuidle(proto) \
	154	{ \
	155	if (static_key_false(&__tracepoint_##name.key)) \
	156	__DO_TRACE(&__tracepoint_##name, \
	157	TP_PROTO(data_proto), \
	158	TP_ARGS(data_args), \
	159	TP_CONDITION(cond), \
7c9906ca PM	160	rcu_irq_enter_irqson(), \
7c9906ca PM	161	rcu_irq_exit_irqson()); \
7ece55a4 JT	162	}
	163	#else
	164	#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)
	165	#endif
	166
97e1c18e MD	167	/*
	168	* Make sure the alignment of the structure in the __tracepoints section will
	169	* not add unwanted padding between the beginning of the section and the
	170	* structure. Force alignment to the same alignment as the section start.
3a630178 DH	171	*
3a630178 DH	172	* When lockdep is enabled, we make sure to always do the RCU portions of
a15920be MD	173	* the tracepoint code, regardless of whether tracing is on. However,
	174	* don't check if the condition is false, due to interaction with idle
	175	* instrumentation. This lets us find RCU issues triggered with tracepoints
	176	* even when this tracepoint is off. This code has no purpose other than
	177	* poking RCU a bit.
97e1c18e	178	*/
2fbb90db	179	#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
7e066fb8	180	extern struct tracepoint __tracepoint_##name; \
97e1c18e MD	181	static inline void trace_##name(proto) \
97e1c18e MD	182	{ \
c5905afb	183	if (static_key_false(&__tracepoint_##name.key)) \
97e1c18e	184	__DO_TRACE(&__tracepoint_##name, \
38516ab5	185	TP_PROTO(data_proto), \
287050d3	186	TP_ARGS(data_args), \
2fbb90db	187	TP_CONDITION(cond),,); \
a05d59a5	188	if (IS_ENABLED(CONFIG_LOCKDEP) && (cond)) { \
3a630178 DH	189	rcu_read_lock_sched_notrace(); \
	190	rcu_dereference_sched(__tracepoint_##name.funcs);\
	191	rcu_read_unlock_sched_notrace(); \
	192	} \
2fbb90db	193	} \
7ece55a4 JT	194	__DECLARE_TRACE_RCU(name, PARAMS(proto), PARAMS(args), \
7ece55a4 JT	195	PARAMS(cond), PARAMS(data_proto), PARAMS(data_args)) \
38516ab5 SR	196	static inline int \
38516ab5 SR	197	register_trace_##name(void (probe)(data_proto), void data) \
97e1c18e	198	{ \
de7b2973 MD	199	return tracepoint_probe_register(&__tracepoint_##name, \
de7b2973 MD	200	(void *)probe, data); \
97e1c18e	201	} \
38516ab5	202	static inline int \
7904b5c4 SRRH	203	register_trace_prio_##name(void (probe)(data_proto), void data,\
	204	int prio) \
	205	{ \
	206	return tracepoint_probe_register_prio(&__tracepoint_##name, \
	207	(void *)probe, data, prio); \
	208	} \
	209	static inline int \
38516ab5	210	unregister_trace_##name(void (probe)(data_proto), void data) \
97e1c18e	211	{ \
de7b2973 MD	212	return tracepoint_probe_unregister(&__tracepoint_##name,\
de7b2973 MD	213	(void *)probe, data); \
53da59aa	214	} \
38516ab5 SR	215	static inline void \
38516ab5 SR	216	check_trace_callback_type_##name(void (*cb)(data_proto)) \
53da59aa	217	{ \
7c65bbc7 SRRH	218	} \
	219	static inline bool \
	220	trace_##name##_enabled(void) \
	221	{ \
	222	return static_key_false(&__tracepoint_##name.key); \
97e1c18e MD	223	}
97e1c18e MD	224
65498646 MD	225	/*
	226	* We have no guarantee that gcc and the linker won't up-align the tracepoint
	227	* structures, so we create an array of pointers that will be used for iteration
	228	* on the tracepoints.
	229	*/
d430d3d7 JB	230	#define DEFINE_TRACE_FN(name, reg, unreg) \
	231	static const char __tpstrtab_##name[] \
	232	__attribute__((section("__tracepoints_strings"))) = #name; \
	233	struct tracepoint __tracepoint_##name \
	234	__attribute__((section("__tracepoints"))) = \
c5905afb	235	{ __tpstrtab_##name, STATIC_KEY_INIT_FALSE, reg, unreg, NULL };\
d430d3d7 JB	236	static struct tracepoint * const __tracepoint_ptr_##name __used \
d430d3d7 JB	237	__attribute__((section("__tracepoints_ptrs"))) = \
65498646	238	&__tracepoint_##name;
97419875 JS	239
	240	#define DEFINE_TRACE(name) \
	241	DEFINE_TRACE_FN(name, NULL, NULL);
7e066fb8 MD	242
	243	#define EXPORT_TRACEPOINT_SYMBOL_GPL(name) \
	244	EXPORT_SYMBOL_GPL(__tracepoint_##name)
	245	#define EXPORT_TRACEPOINT_SYMBOL(name) \
	246	EXPORT_SYMBOL(__tracepoint_##name)
	247
c63b7682	248	#else /* !TRACEPOINTS_ENABLED */
2fbb90db	249	#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
97e1c18e MD	250	static inline void trace_##name(proto) \
97e1c18e MD	251	{ } \
2fbb90db SR	252	static inline void trace_##name##_rcuidle(proto) \
2fbb90db SR	253	{ } \
38516ab5 SR	254	static inline int \
	255	register_trace_##name(void (*probe)(data_proto), \
	256	void *data) \
97e1c18e MD	257	{ \
	258	return -ENOSYS; \
	259	} \
38516ab5 SR	260	static inline int \
	261	unregister_trace_##name(void (*probe)(data_proto), \
	262	void *data) \
c420970e MD	263	{ \
c420970e MD	264	return -ENOSYS; \
53da59aa	265	} \
38516ab5	266	static inline void check_trace_callback_type_##name(void (*cb)(data_proto)) \
53da59aa	267	{ \
7c65bbc7 SRRH	268	} \
	269	static inline bool \
	270	trace_##name##_enabled(void) \
	271	{ \
	272	return false; \
c420970e	273	}
97e1c18e	274
97419875	275	#define DEFINE_TRACE_FN(name, reg, unreg)
7e066fb8 MD	276	#define DEFINE_TRACE(name)
	277	#define EXPORT_TRACEPOINT_SYMBOL_GPL(name)
	278	#define EXPORT_TRACEPOINT_SYMBOL(name)
	279
c63b7682	280	#endif /* TRACEPOINTS_ENABLED */
38516ab5	281
3c49b52b SR	282	#ifdef CONFIG_TRACING
	283	/**
	284	* tracepoint_string - register constant persistent string to trace system
	285	* @str - a constant persistent string that will be referenced in tracepoints
	286	*
	287	* If constant strings are being used in tracepoints, it is faster and
	288	* more efficient to just save the pointer to the string and reference
	289	* that with a printf "%s" instead of saving the string in the ring buffer
	290	* and wasting space and time.
	291	*
	292	* The problem with the above approach is that userspace tools that read
	293	* the binary output of the trace buffers do not have access to the string.
	294	* Instead they just show the address of the string which is not very
	295	* useful to users.
	296	*
	297	* With tracepoint_string(), the string will be registered to the tracing
	298	* system and exported to userspace via the debugfs/tracing/printk_formats
	299	* file that maps the string address to the string text. This way userspace
	300	* tools that read the binary buffers have a way to map the pointers to
	301	* the ASCII strings they represent.
	302	*
	303	* The @str used must be a constant string and persistent as it would not
	304	* make sense to show a string that no longer exists. But it is still fine
	305	* to be used with modules, because when modules are unloaded, if they
	306	* had tracepoints, the ring buffers are cleared too. As long as the string
	307	* does not change during the life of the module, it is fine to use
	308	* tracepoint_string() within a module.
	309	*/
	310	#define tracepoint_string(str) \
	311	({ \
	312	static const char *___tp_str __tracepoint_string = str; \
	313	___tp_str; \
	314	})
	315	#define __tracepoint_string __attribute__((section("__tracepoint_str")))
	316	#else
	317	/*
	318	* tracepoint_string() is used to save the string address for userspace
	319	* tracing tools. When tracing isn't configured, there's no need to save
	320	* anything.
	321	*/
	322	# define tracepoint_string(str) str
	323	# define __tracepoint_string
	324	#endif
	325
38516ab5 SR	326	/*
	327	* The need for the DECLARE_TRACE_NOARGS() is to handle the prototype
	328	* (void). "void" is a special value in a function prototype and can
	329	* not be combined with other arguments. Since the DECLARE_TRACE()
	330	* macro adds a data element at the beginning of the prototype,
	331	* we need a way to differentiate "(void *data, proto)" from
	332	* "(void *data, void)". The second prototype is invalid.
	333	*
	334	* DECLARE_TRACE_NOARGS() passes "void" as the tracepoint prototype
	335	* and "void *__data" as the callback prototype.
	336	*
	337	* DECLARE_TRACE() passes "proto" as the tracepoint protoype and
	338	* "void *__data, proto" as the callback prototype.
	339	*/
	340	#define DECLARE_TRACE_NOARGS(name) \
287050d3	341	__DECLARE_TRACE(name, void, , 1, void *__data, __data)
38516ab5 SR	342
38516ab5 SR	343	#define DECLARE_TRACE(name, proto, args) \
287050d3	344	__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), 1, \
38516ab5 SR	345	PARAMS(void *__data, proto), \
	346	PARAMS(__data, args))
	347
287050d3 SR	348	#define DECLARE_TRACE_CONDITION(name, proto, args, cond) \
	349	__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), PARAMS(cond), \
	350	PARAMS(void *__data, proto), \
	351	PARAMS(__data, args))
	352
1ed0c597 FW	353	#define TRACE_EVENT_FLAGS(event, flag)
1ed0c597 FW	354
d5b5f391 PZ	355	#define TRACE_EVENT_PERF_PERM(event, expr...)
d5b5f391 PZ	356
ea20d929	357	#endif /* DECLARE_TRACE */
97e1c18e	358
ea20d929	359	#ifndef TRACE_EVENT
823f9124 SR	360	/*
	361	* For use with the TRACE_EVENT macro:
	362	*
	363	* We define a tracepoint, its arguments, its printk format
2621bca8	364	* and its 'fast binary record' layout.
823f9124 SR	365	*
	366	* Firstly, name your tracepoint via TRACE_EVENT(name : the
	367	* 'subsystem_event' notation is fine.
	368	*
	369	* Think about this whole construct as the
	370	* 'trace_sched_switch() function' from now on.
	371	*
	372	*
	373	* TRACE_EVENT(sched_switch,
	374	*
	375	* *
	376	* * A function has a regular function arguments
	377	* * prototype, declare it via TP_PROTO():
	378	* *
	379	*
ef18012b SR	380	* TP_PROTO(struct rq rq, struct task_struct prev,
ef18012b SR	381	* struct task_struct *next),
823f9124 SR	382	*
	383	* *
	384	* * Define the call signature of the 'function'.
	385	* * (Design sidenote: we use this instead of a
	386	* * TP_PROTO1/TP_PROTO2/TP_PROTO3 ugliness.)
	387	* *
	388	*
ef18012b	389	* TP_ARGS(rq, prev, next),
823f9124 SR	390	*
	391	* *
	392	* * Fast binary tracing: define the trace record via
	393	* * TP_STRUCT__entry(). You can think about it like a
	394	* * regular C structure local variable definition.
	395	* *
	396	* * This is how the trace record is structured and will
	397	* * be saved into the ring buffer. These are the fields
	398	* * that will be exposed to user-space in
156f5a78	399	* * /sys/kernel/debug/tracing/events/<*>/format.
823f9124 SR	400	* *
	401	* * The declared 'local variable' is called '__entry'
	402	* *
	403	* * __field(pid_t, prev_prid) is equivalent to a standard declariton:
	404	* *
	405	* * pid_t prev_pid;
	406	* *
	407	* * __array(char, prev_comm, TASK_COMM_LEN) is equivalent to:
	408	* *
	409	* * char prev_comm[TASK_COMM_LEN];
	410	* *
	411	*
	412	* TP_STRUCT__entry(
	413	* __array( char, prev_comm, TASK_COMM_LEN )
	414	* __field( pid_t, prev_pid )
	415	* __field( int, prev_prio )
	416	* __array( char, next_comm, TASK_COMM_LEN )
	417	* __field( pid_t, next_pid )
	418	* __field( int, next_prio )
	419	* ),
	420	*
	421	* *
	422	* * Assign the entry into the trace record, by embedding
	423	* * a full C statement block into TP_fast_assign(). You
	424	* * can refer to the trace record as '__entry' -
	425	* * otherwise you can put arbitrary C code in here.
	426	* *
	427	* * Note: this C code will execute every time a trace event
	428	* * happens, on an active tracepoint.
	429	* *
	430	*
ef18012b SR	431	* TP_fast_assign(
	432	* memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN);
	433	* __entry->prev_pid = prev->pid;
	434	* __entry->prev_prio = prev->prio;
823f9124 SR	435	* memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN);
823f9124 SR	436	* __entry->next_pid = next->pid;
ef18012b	437	* __entry->next_prio = next->prio;
ec6e7c3a	438	* ),
823f9124 SR	439	*
	440	* *
	441	* * Formatted output of a trace record via TP_printk().
	442	* * This is how the tracepoint will appear under ftrace
	443	* * plugins that make use of this tracepoint.
	444	* *
	445	* * (raw-binary tracing wont actually perform this step.)
	446	* *
	447	*
	448	* TP_printk("task %s:%d [%d] ==> %s:%d [%d]",
	449	* __entry->prev_comm, __entry->prev_pid, __entry->prev_prio,
	450	* __entry->next_comm, __entry->next_pid, __entry->next_prio),
	451	*
	452	* );
	453	*
	454	* This macro construct is thus used for the regular printk format
	455	* tracing setup, it is used to construct a function pointer based
	456	* tracepoint callback (this is used by programmatic plugins and
	457	* can also by used by generic instrumentation like SystemTap), and
	458	* it is also used to expose a structured trace record in
156f5a78	459	* /sys/kernel/debug/tracing/events/.
97419875 JS	460	*
	461	* A set of (un)registration functions can be passed to the variant
	462	* TRACE_EVENT_FN to perform any (un)registration work.
823f9124 SR	463	*/
823f9124 SR	464
091ad365	465	#define DECLARE_EVENT_CLASS(name, proto, args, tstruct, assign, print)
ff038f5c SR	466	#define DEFINE_EVENT(template, name, proto, args) \
ff038f5c SR	467	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
f5abaa1b SR	468	#define DEFINE_EVENT_FN(template, name, proto, args, reg, unreg)\
f5abaa1b SR	469	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
e5bc9721 SR	470	#define DEFINE_EVENT_PRINT(template, name, proto, args, print) \
e5bc9721 SR	471	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
287050d3 SR	472	#define DEFINE_EVENT_CONDITION(template, name, proto, \
	473	args, cond) \
	474	DECLARE_TRACE_CONDITION(name, PARAMS(proto), \
	475	PARAMS(args), PARAMS(cond))
ff038f5c	476
30a8fecc	477	#define TRACE_EVENT(name, proto, args, struct, assign, print) \
da4d0302	478	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
97419875 JS	479	#define TRACE_EVENT_FN(name, proto, args, struct, \
	480	assign, print, reg, unreg) \
	481	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
2701121b DK	482	#define TRACE_EVENT_FN_COND(name, proto, args, cond, struct, \
	483	assign, print, reg, unreg) \
	484	DECLARE_TRACE_CONDITION(name, PARAMS(proto), \
	485	PARAMS(args), PARAMS(cond))
287050d3 SR	486	#define TRACE_EVENT_CONDITION(name, proto, args, cond, \
	487	struct, assign, print) \
	488	DECLARE_TRACE_CONDITION(name, PARAMS(proto), \
	489	PARAMS(args), PARAMS(cond))
7cb2e3ee	490
1ed0c597 FW	491	#define TRACE_EVENT_FLAGS(event, flag)
1ed0c597 FW	492
d5b5f391 PZ	493	#define TRACE_EVENT_PERF_PERM(event, expr...)
d5b5f391 PZ	494
7cb2e3ee	495	#endif /* ifdef TRACE_EVENT (see note above) */