[mirror_ubuntu-artful-kernel.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/static_key.h>

struct module;
struct tracepoint;
struct notifier_block;

struct tracepoint_func {
	void *func;
	void *data;
};

struct tracepoint {
	const char *name;		/* Tracepoint name */
	struct static_key key;
	void (*regfunc)(void);
	void (*unregfunc)(void);
	struct tracepoint_func __rcu *funcs;
};

extern int
tracepoint_probe_register(struct tracepoint *tp, void *probe, void *data);
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

#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

#ifdef CONFIG_TRACEPOINTS

/*
 * 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(),			\
				rcu_irq_exit());			\
	}
#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 or we match the
 * condition.  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						\
	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 /* !CONFIG_TRACEPOINTS */
#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 /* CONFIG_TRACEPOINTS */

#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_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>
c5905afb	20	#include <linux/static_key.h>
97e1c18e MD	21
	22	struct module;
	23	struct tracepoint;
de7b2973	24	struct notifier_block;
97e1c18e	25
38516ab5 SR	26	struct tracepoint_func {
	27	void *func;
	28	void *data;
	29	};
	30
97e1c18e MD	31	struct tracepoint {
97e1c18e MD	32	const char name; / Tracepoint name */
c5905afb	33	struct static_key key;
97419875 JS	34	void (*regfunc)(void);
97419875 JS	35	void (*unregfunc)(void);
bd1c8b22	36	struct tracepoint_func __rcu *funcs;
65498646	37	};
97e1c18e	38
38516ab5	39	extern int
de7b2973 MD	40	tracepoint_probe_register(struct tracepoint tp, void probe, void *data);
	41	extern int
	42	tracepoint_probe_unregister(struct tracepoint tp, void probe, void *data);
	43	extern void
	44	for_each_kernel_tracepoint(void (fct)(struct tracepoint tp, void *priv),
	45	void *priv);
2e26ca71	46
b75ef8b4 MD	47	#ifdef CONFIG_MODULES
	48	struct tp_module {
	49	struct list_head list;
eb7d035c	50	struct module *mod;
b75ef8b4	51	};
de7b2973	52
45ab2813	53	bool trace_module_has_bad_taint(struct module *mod);
de7b2973 MD	54	extern int register_tracepoint_module_notifier(struct notifier_block *nb);
de7b2973 MD	55	extern int unregister_tracepoint_module_notifier(struct notifier_block *nb);
45ab2813 SRRH	56	#else
	57	static inline bool trace_module_has_bad_taint(struct module *mod)
	58	{
	59	return false;
	60	}
de7b2973 MD	61	static inline
	62	int register_tracepoint_module_notifier(struct notifier_block *nb)
	63	{
	64	return 0;
	65	}
	66	static inline
	67	int unregister_tracepoint_module_notifier(struct notifier_block *nb)
	68	{
	69	return 0;
	70	}
b75ef8b4 MD	71	#endif /* CONFIG_MODULES */
b75ef8b4 MD	72
2e26ca71 SR	73	/*
	74	* tracepoint_synchronize_unregister must be called between the last tracepoint
	75	* probe unregistration and the end of module exit to make sure there is no
	76	* caller executing a probe when it is freed.
	77	*/
	78	static inline void tracepoint_synchronize_unregister(void)
	79	{
	80	synchronize_sched();
	81	}
	82
b725dfea MD	83	#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS
	84	extern void syscall_regfunc(void);
	85	extern void syscall_unregfunc(void);
	86	#endif /* CONFIG_HAVE_SYSCALL_TRACEPOINTS */
	87
2e26ca71 SR	88	#define PARAMS(args...) args
2e26ca71 SR	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 MD	105
	106	#ifdef CONFIG_TRACEPOINTS
	107
	108	/*
	109	* it_func[0] is never NULL because there is at least one element in the array
	110	* when the array itself is non NULL.
38516ab5 SR	111	*
	112	* Note, the proto and args passed in includes "__data" as the first parameter.
	113	* The reason for this is to handle the "void" prototype. If a tracepoint
	114	* has a "void" prototype, then it is invalid to declare a function
	115	* as "(void *, void)". The DECLARE_TRACE_NOARGS() will pass in just
	116	* "void data", where as the DECLARE_TRACE() will pass in "void data, proto".
97e1c18e	117	*/
2fbb90db	118	#define __DO_TRACE(tp, proto, args, cond, prercu, postrcu) \
97e1c18e	119	do { \
38516ab5 SR	120	struct tracepoint_func *it_func_ptr; \
	121	void *it_func; \
	122	void *__data; \
97e1c18e	123	\
287050d3 SR	124	if (!(cond)) \
287050d3 SR	125	return; \
2fbb90db	126	prercu; \
da7b3eab	127	rcu_read_lock_sched_notrace(); \
38516ab5 SR	128	it_func_ptr = rcu_dereference_sched((tp)->funcs); \
38516ab5 SR	129	if (it_func_ptr) { \
97e1c18e	130	do { \
38516ab5 SR	131	it_func = (it_func_ptr)->func; \
	132	__data = (it_func_ptr)->data; \
	133	((void(*)(proto))(it_func))(args); \
	134	} while ((++it_func_ptr)->func); \
97e1c18e	135	} \
da7b3eab	136	rcu_read_unlock_sched_notrace(); \
2fbb90db	137	postrcu; \
97e1c18e MD	138	} while (0)
97e1c18e MD	139
7ece55a4 JT	140	#ifndef MODULE
	141	#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args) \
	142	static inline void trace_##name##_rcuidle(proto) \
	143	{ \
	144	if (static_key_false(&__tracepoint_##name.key)) \
	145	__DO_TRACE(&__tracepoint_##name, \
	146	TP_PROTO(data_proto), \
	147	TP_ARGS(data_args), \
	148	TP_CONDITION(cond), \
d6284099 PM	149	rcu_irq_enter(), \
d6284099 PM	150	rcu_irq_exit()); \
7ece55a4 JT	151	}
	152	#else
	153	#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)
	154	#endif
	155
97e1c18e MD	156	/*
	157	* Make sure the alignment of the structure in the __tracepoints section will
	158	* not add unwanted padding between the beginning of the section and the
	159	* structure. Force alignment to the same alignment as the section start.
3a630178 DH	160	*
	161	* When lockdep is enabled, we make sure to always do the RCU portions of
	162	* the tracepoint code, regardless of whether tracing is on or we match the
	163	* condition. This lets us find RCU issues triggered with tracepoints even
	164	* when this tracepoint is off. This code has no purpose other than poking
	165	* RCU a bit.
97e1c18e	166	*/
2fbb90db	167	#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
7e066fb8	168	extern struct tracepoint __tracepoint_##name; \
97e1c18e MD	169	static inline void trace_##name(proto) \
97e1c18e MD	170	{ \
c5905afb	171	if (static_key_false(&__tracepoint_##name.key)) \
97e1c18e	172	__DO_TRACE(&__tracepoint_##name, \
38516ab5	173	TP_PROTO(data_proto), \
287050d3	174	TP_ARGS(data_args), \
2fbb90db	175	TP_CONDITION(cond),,); \
a05d59a5	176	if (IS_ENABLED(CONFIG_LOCKDEP) && (cond)) { \
3a630178 DH	177	rcu_read_lock_sched_notrace(); \
	178	rcu_dereference_sched(__tracepoint_##name.funcs);\
	179	rcu_read_unlock_sched_notrace(); \
	180	} \
2fbb90db	181	} \
7ece55a4 JT	182	__DECLARE_TRACE_RCU(name, PARAMS(proto), PARAMS(args), \
7ece55a4 JT	183	PARAMS(cond), PARAMS(data_proto), PARAMS(data_args)) \
38516ab5 SR	184	static inline int \
38516ab5 SR	185	register_trace_##name(void (probe)(data_proto), void data) \
97e1c18e	186	{ \
de7b2973 MD	187	return tracepoint_probe_register(&__tracepoint_##name, \
de7b2973 MD	188	(void *)probe, data); \
97e1c18e	189	} \
38516ab5 SR	190	static inline int \
38516ab5 SR	191	unregister_trace_##name(void (probe)(data_proto), void data) \
97e1c18e	192	{ \
de7b2973 MD	193	return tracepoint_probe_unregister(&__tracepoint_##name,\
de7b2973 MD	194	(void *)probe, data); \
53da59aa	195	} \
38516ab5 SR	196	static inline void \
38516ab5 SR	197	check_trace_callback_type_##name(void (*cb)(data_proto)) \
53da59aa	198	{ \
7c65bbc7 SRRH	199	} \
	200	static inline bool \
	201	trace_##name##_enabled(void) \
	202	{ \
	203	return static_key_false(&__tracepoint_##name.key); \
97e1c18e MD	204	}
97e1c18e MD	205
65498646 MD	206	/*
	207	* We have no guarantee that gcc and the linker won't up-align the tracepoint
	208	* structures, so we create an array of pointers that will be used for iteration
	209	* on the tracepoints.
	210	*/
d430d3d7 JB	211	#define DEFINE_TRACE_FN(name, reg, unreg) \
	212	static const char __tpstrtab_##name[] \
	213	__attribute__((section("__tracepoints_strings"))) = #name; \
	214	struct tracepoint __tracepoint_##name \
	215	__attribute__((section("__tracepoints"))) = \
c5905afb	216	{ __tpstrtab_##name, STATIC_KEY_INIT_FALSE, reg, unreg, NULL };\
d430d3d7 JB	217	static struct tracepoint * const __tracepoint_ptr_##name __used \
d430d3d7 JB	218	__attribute__((section("__tracepoints_ptrs"))) = \
65498646	219	&__tracepoint_##name;
97419875 JS	220
	221	#define DEFINE_TRACE(name) \
	222	DEFINE_TRACE_FN(name, NULL, NULL);
7e066fb8 MD	223
	224	#define EXPORT_TRACEPOINT_SYMBOL_GPL(name) \
	225	EXPORT_SYMBOL_GPL(__tracepoint_##name)
	226	#define EXPORT_TRACEPOINT_SYMBOL(name) \
	227	EXPORT_SYMBOL(__tracepoint_##name)
	228
97e1c18e	229	#else /* !CONFIG_TRACEPOINTS */
2fbb90db	230	#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
97e1c18e MD	231	static inline void trace_##name(proto) \
97e1c18e MD	232	{ } \
2fbb90db SR	233	static inline void trace_##name##_rcuidle(proto) \
2fbb90db SR	234	{ } \
38516ab5 SR	235	static inline int \
	236	register_trace_##name(void (*probe)(data_proto), \
	237	void *data) \
97e1c18e MD	238	{ \
	239	return -ENOSYS; \
	240	} \
38516ab5 SR	241	static inline int \
	242	unregister_trace_##name(void (*probe)(data_proto), \
	243	void *data) \
c420970e MD	244	{ \
c420970e MD	245	return -ENOSYS; \
53da59aa	246	} \
38516ab5	247	static inline void check_trace_callback_type_##name(void (*cb)(data_proto)) \
53da59aa	248	{ \
7c65bbc7 SRRH	249	} \
	250	static inline bool \
	251	trace_##name##_enabled(void) \
	252	{ \
	253	return false; \
c420970e	254	}
97e1c18e	255
97419875	256	#define DEFINE_TRACE_FN(name, reg, unreg)
7e066fb8 MD	257	#define DEFINE_TRACE(name)
	258	#define EXPORT_TRACEPOINT_SYMBOL_GPL(name)
	259	#define EXPORT_TRACEPOINT_SYMBOL(name)
	260
97e1c18e	261	#endif /* CONFIG_TRACEPOINTS */
38516ab5	262
3c49b52b SR	263	#ifdef CONFIG_TRACING
	264	/**
	265	* tracepoint_string - register constant persistent string to trace system
	266	* @str - a constant persistent string that will be referenced in tracepoints
	267	*
	268	* If constant strings are being used in tracepoints, it is faster and
	269	* more efficient to just save the pointer to the string and reference
	270	* that with a printf "%s" instead of saving the string in the ring buffer
	271	* and wasting space and time.
	272	*
	273	* The problem with the above approach is that userspace tools that read
	274	* the binary output of the trace buffers do not have access to the string.
	275	* Instead they just show the address of the string which is not very
	276	* useful to users.
	277	*
	278	* With tracepoint_string(), the string will be registered to the tracing
	279	* system and exported to userspace via the debugfs/tracing/printk_formats
	280	* file that maps the string address to the string text. This way userspace
	281	* tools that read the binary buffers have a way to map the pointers to
	282	* the ASCII strings they represent.
	283	*
	284	* The @str used must be a constant string and persistent as it would not
	285	* make sense to show a string that no longer exists. But it is still fine
	286	* to be used with modules, because when modules are unloaded, if they
	287	* had tracepoints, the ring buffers are cleared too. As long as the string
	288	* does not change during the life of the module, it is fine to use
	289	* tracepoint_string() within a module.
	290	*/
	291	#define tracepoint_string(str) \
	292	({ \
	293	static const char *___tp_str __tracepoint_string = str; \
	294	___tp_str; \
	295	})
	296	#define __tracepoint_string __attribute__((section("__tracepoint_str")))
	297	#else
	298	/*
	299	* tracepoint_string() is used to save the string address for userspace
	300	* tracing tools. When tracing isn't configured, there's no need to save
	301	* anything.
	302	*/
	303	# define tracepoint_string(str) str
	304	# define __tracepoint_string
	305	#endif
	306
38516ab5 SR	307	/*
	308	* The need for the DECLARE_TRACE_NOARGS() is to handle the prototype
	309	* (void). "void" is a special value in a function prototype and can
	310	* not be combined with other arguments. Since the DECLARE_TRACE()
	311	* macro adds a data element at the beginning of the prototype,
	312	* we need a way to differentiate "(void *data, proto)" from
	313	* "(void *data, void)". The second prototype is invalid.
	314	*
	315	* DECLARE_TRACE_NOARGS() passes "void" as the tracepoint prototype
	316	* and "void *__data" as the callback prototype.
	317	*
	318	* DECLARE_TRACE() passes "proto" as the tracepoint protoype and
	319	* "void *__data, proto" as the callback prototype.
	320	*/
	321	#define DECLARE_TRACE_NOARGS(name) \
287050d3	322	__DECLARE_TRACE(name, void, , 1, void *__data, __data)
38516ab5 SR	323
38516ab5 SR	324	#define DECLARE_TRACE(name, proto, args) \
287050d3	325	__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), 1, \
38516ab5 SR	326	PARAMS(void *__data, proto), \
	327	PARAMS(__data, args))
	328
287050d3 SR	329	#define DECLARE_TRACE_CONDITION(name, proto, args, cond) \
	330	__DECLARE_TRACE(name, PARAMS(proto), PARAMS(args), PARAMS(cond), \
	331	PARAMS(void *__data, proto), \
	332	PARAMS(__data, args))
	333
1ed0c597 FW	334	#define TRACE_EVENT_FLAGS(event, flag)
1ed0c597 FW	335
d5b5f391 PZ	336	#define TRACE_EVENT_PERF_PERM(event, expr...)
d5b5f391 PZ	337
ea20d929	338	#endif /* DECLARE_TRACE */
97e1c18e	339
ea20d929	340	#ifndef TRACE_EVENT
823f9124 SR	341	/*
	342	* For use with the TRACE_EVENT macro:
	343	*
	344	* We define a tracepoint, its arguments, its printk format
2621bca8	345	* and its 'fast binary record' layout.
823f9124 SR	346	*
	347	* Firstly, name your tracepoint via TRACE_EVENT(name : the
	348	* 'subsystem_event' notation is fine.
	349	*
	350	* Think about this whole construct as the
	351	* 'trace_sched_switch() function' from now on.
	352	*
	353	*
	354	* TRACE_EVENT(sched_switch,
	355	*
	356	* *
	357	* * A function has a regular function arguments
	358	* * prototype, declare it via TP_PROTO():
	359	* *
	360	*
ef18012b SR	361	* TP_PROTO(struct rq rq, struct task_struct prev,
ef18012b SR	362	* struct task_struct *next),
823f9124 SR	363	*
	364	* *
	365	* * Define the call signature of the 'function'.
	366	* * (Design sidenote: we use this instead of a
	367	* * TP_PROTO1/TP_PROTO2/TP_PROTO3 ugliness.)
	368	* *
	369	*
ef18012b	370	* TP_ARGS(rq, prev, next),
823f9124 SR	371	*
	372	* *
	373	* * Fast binary tracing: define the trace record via
	374	* * TP_STRUCT__entry(). You can think about it like a
	375	* * regular C structure local variable definition.
	376	* *
	377	* * This is how the trace record is structured and will
	378	* * be saved into the ring buffer. These are the fields
	379	* * that will be exposed to user-space in
156f5a78	380	* * /sys/kernel/debug/tracing/events/<*>/format.
823f9124 SR	381	* *
	382	* * The declared 'local variable' is called '__entry'
	383	* *
	384	* * __field(pid_t, prev_prid) is equivalent to a standard declariton:
	385	* *
	386	* * pid_t prev_pid;
	387	* *
	388	* * __array(char, prev_comm, TASK_COMM_LEN) is equivalent to:
	389	* *
	390	* * char prev_comm[TASK_COMM_LEN];
	391	* *
	392	*
	393	* TP_STRUCT__entry(
	394	* __array( char, prev_comm, TASK_COMM_LEN )
	395	* __field( pid_t, prev_pid )
	396	* __field( int, prev_prio )
	397	* __array( char, next_comm, TASK_COMM_LEN )
	398	* __field( pid_t, next_pid )
	399	* __field( int, next_prio )
	400	* ),
	401	*
	402	* *
	403	* * Assign the entry into the trace record, by embedding
	404	* * a full C statement block into TP_fast_assign(). You
	405	* * can refer to the trace record as '__entry' -
	406	* * otherwise you can put arbitrary C code in here.
	407	* *
	408	* * Note: this C code will execute every time a trace event
	409	* * happens, on an active tracepoint.
	410	* *
	411	*
ef18012b SR	412	* TP_fast_assign(
	413	* memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN);
	414	* __entry->prev_pid = prev->pid;
	415	* __entry->prev_prio = prev->prio;
823f9124 SR	416	* memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN);
823f9124 SR	417	* __entry->next_pid = next->pid;
ef18012b	418	* __entry->next_prio = next->prio;
ec6e7c3a	419	* ),
823f9124 SR	420	*
	421	* *
	422	* * Formatted output of a trace record via TP_printk().
	423	* * This is how the tracepoint will appear under ftrace
	424	* * plugins that make use of this tracepoint.
	425	* *
	426	* * (raw-binary tracing wont actually perform this step.)
	427	* *
	428	*
	429	* TP_printk("task %s:%d [%d] ==> %s:%d [%d]",
	430	* __entry->prev_comm, __entry->prev_pid, __entry->prev_prio,
	431	* __entry->next_comm, __entry->next_pid, __entry->next_prio),
	432	*
	433	* );
	434	*
	435	* This macro construct is thus used for the regular printk format
	436	* tracing setup, it is used to construct a function pointer based
	437	* tracepoint callback (this is used by programmatic plugins and
	438	* can also by used by generic instrumentation like SystemTap), and
	439	* it is also used to expose a structured trace record in
156f5a78	440	* /sys/kernel/debug/tracing/events/.
97419875 JS	441	*
	442	* A set of (un)registration functions can be passed to the variant
	443	* TRACE_EVENT_FN to perform any (un)registration work.
823f9124 SR	444	*/
823f9124 SR	445
091ad365	446	#define DECLARE_EVENT_CLASS(name, proto, args, tstruct, assign, print)
ff038f5c SR	447	#define DEFINE_EVENT(template, name, proto, args) \
ff038f5c SR	448	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
f5abaa1b SR	449	#define DEFINE_EVENT_FN(template, name, proto, args, reg, unreg)\
f5abaa1b SR	450	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
e5bc9721 SR	451	#define DEFINE_EVENT_PRINT(template, name, proto, args, print) \
e5bc9721 SR	452	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
287050d3 SR	453	#define DEFINE_EVENT_CONDITION(template, name, proto, \
	454	args, cond) \
	455	DECLARE_TRACE_CONDITION(name, PARAMS(proto), \
	456	PARAMS(args), PARAMS(cond))
ff038f5c	457
30a8fecc	458	#define TRACE_EVENT(name, proto, args, struct, assign, print) \
da4d0302	459	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
97419875 JS	460	#define TRACE_EVENT_FN(name, proto, args, struct, \
	461	assign, print, reg, unreg) \
	462	DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
287050d3 SR	463	#define TRACE_EVENT_CONDITION(name, proto, args, cond, \
	464	struct, assign, print) \
	465	DECLARE_TRACE_CONDITION(name, PARAMS(proto), \
	466	PARAMS(args), PARAMS(cond))
7cb2e3ee	467
1ed0c597 FW	468	#define TRACE_EVENT_FLAGS(event, flag)
1ed0c597 FW	469
d5b5f391 PZ	470	#define TRACE_EVENT_PERF_PERM(event, expr...)
d5b5f391 PZ	471
7cb2e3ee	472	#endif /* ifdef TRACE_EVENT (see note above) */